EPA/600/9-89/072
August 1989
THIRD
INTERNATIONAL CONFERENCE
ON
NEW FRONTIERS FOR HAZARDOUS WASTE MANAGEMENT
Proceedings
September 10-13, 1989
Pittsburgh, Pennsylvania
Sponsored by
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
United Nations Environmental Programme
Paris, France
World Federation of Engineering Organizations
Pasadena, California
American Academy of Environmental Engineers
Annapolis, Maryland
NUS Corporation
Pittsburgh, Pennsylvania
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
For sale by the Superintendent of Documents', U.S. Government
Printing Office, Washington, D.C. 20402
-------
NOTICE
These Proceedings have been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review policies
and approved for presentation and publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for
use.
-------
FOREWORD
The managing of hazardous waste has proved itself to be of vast world
concern. Because of this, the United States Environmental Protection
Agency (USEPA) has developed a keen interest in working with environmental
specialists from all nations. We view face-to-face conferences with our
peers from many lands to be the most effective method of learning the prob-
lems and solutions so important to us all. The papers given at the
sessions of this meeting will, we are certain, bring greater understanding
of what we must accomplish to return this, our earth, to as close to its
primal being as possible. -- At USEPA we feel if we, as world citizens,
fail in this mission we will leave a reminder for future generations of
our dereliction as they suffer the consequences. We must not fail and so
we come to this international gathering to learn from the work being done
in other lands and to share our own findings.
As one of your hosts at the Third International Conference on New
Frontiers for Hazardous Waste Management USEPA1s Risk Reduction Engi-
neering Laboratory assists in providing an authoritative and defensible
engineering basis for assessing and solving the problems posed by hazardous
waste in the environment. Its products support the policies, programs, and
regulations of USEPA, the permitting and other responsibilities of State
and local governments, and the needs of both large and small businesses in
handling their wastes responsibly and economically.
These Proceedings present the papers from this Third International
Conference. USEPA, United Nations Environmental Programme, World Feder-
ation of Engineering Organizations, American Academy of Environmental
Engineers, and NUS Corporation have cosponsored this conference in order
to summarize important new technological developments and concepts with
broad international application.
E. Timothy Oppelt
Di rector
Risk Reduction Engineering Laboratory
n 7
-------
ABSTRACT
The Third International Conference on New Frontiers for Hazardous
Haste Management was held at Pittsburgh, Pennsylvania, September 10-13,
1989. The purpose of this conference was to examine the state of tech-
nology for the disposal of hazardous waste. Emphasis was placed on the
presentation of papers that summarized important new technological devel-
opments and concepts with broad international application.
Sessions were held in the areas of : (1) Thermal Treatment, (2)
Physical/Chemical Treatment, (3) Biological Treatment, (4) Land Disposal,
(5) Solidification/Stabilization, (6) Waste Minimization, and (7) Waste
Management.
These Proceedings are a compilation of the speaker's papers. Where
material for the entire work of a presenter was not available for primary
publication, copies of the full paper may be obtained in the Conference
lobby or later, by contacting NUS Corporation at their Pittsurgh address.
Due to the press of time and distance, manuscripts for five presen-
tations were not included in these Proceedings. To obtain copies please
enquire at the Conference registration desk or, following the conference,
contact NUS Corporation at their Pittsburgh Address.
A topical index is located at the back of the book. We hope you
find it useful.
This Conference was sponsored by United States Environmental Protec-
tion Agency, United Nations Environmental Programme, World Federation of
Engineering Organizations, American Academy of Environmental Engineers,
and NUS Corporation,
-------
CONTENTS
Notice 1 i
Foreword i ii
Abst ract i v
Acknowledgements xi
Index 595
Paper Number (The index, at the back of this volume, will refer to the individual
Paper Numbers indicated below for each paper presented.)
SESSION 1 - PHYSICAL/CHEMICAL TREATMENT
1. Evaluation of Treatment Technologies for Contaminated Soil and Debris
Richard Lauch, U.S. Environmental Protection Agency; Barbara B.
Locke, Majid Dosani, Steve Giti-Pour, Catherine D. Chambers, PEI
Associates, Inc.; and Ed Alperin, Arie Groen, IT Corporation .... 1
2. Pre-Treatment of Hazardous Waste
Weine Wiqvist, Kemiavfall in Skane AB 13
3. Hydraulic Jett Mixing — Versatile Tool for Hazardous Waste Treatment
John R. Ackerman, Hazleton Environmental Products, Inc 17
4. Demonstration of Technologies to Remove Contamination from Groundwater
Kent M. Hodgson and LaPriel Garrett, Westinghouse
Hanford Company 26
5. Soil Decontamination With Extraksol^M
Jean Paquin and Diana Mourato, Sanivan Group 35
6. Organic Waste Treatment With Organically Modified Clays
Jeffery C. Evans, Stephen E. Pancoski, Bucknell University; and
George Alther, Bentec, Inc
48
7. Technologies Applicable for the Remediation of Contaminated soil at
Superfund Radiation Sites
Ramjee Raghavan, George Wolf, Foster Wheeler
Enviresponse, Inc.; and Darlene Williams, U.S. Environmental
Protecti on Agency 59
SESSION 2 - LAND DISPOSAL
8. Advanced Technlogies for Pollutant Detection, Monitoring, and
Remediation in Ground Water
R.A. Klopp, Terra Technologies; J.F. Haasbeek, P.B. Bedient,
Rice University; and A.A. Biehle, Consultant 67
-------
9. Use of Abandoned Coal/Lignite Open Pits for Waste Disposal in
Selected European Countries
Jacek S. Libicki, POLTEGOR ...................................... 76
10. Desiccation and Permiability of Soil Bentonite Materials
Raj P. Khera, Hemendra Moradia, and Mahendraratnam Thill iyar,
New Jersey Institute of Technology .............................. 84
11
12
Design of Clay Liners to Minimize Shrinkage Cracking
Miguel Picornell and Mohd Zaifuddin Idris, University of
Texas at El Paso ................................................ 92
Field Studies on the Hydrological Performance of Multilayered
Landfill Caps
Stefan Melchior and Gunter Miehlich, Universitat Hamburg
13. Design and Construction of the C2-Landfill Maasvl akte Rotterdam
H.L. Sijberden, Rotterdam Pub! ic Works
100
108
14
15
Bedrock Neutralization Study for the Bruin Lagoon Superfund Site
Gerard M. Petulunas, Duane R. Lenhardt, and James E. Neice,
GAI Consultants, Inc ............................................ 116
SESSION 3 - SOLIDIFICATION
Modeling Chemical and Physical Processes in Leaching Solidified
Wastes
Bill Bachelor, Texas A&M University ............................. 123
16. Solidification of Filter Ashes from Solid Waste Incinerators
Peter Friedli, Geotechnik; and Paul H. Brunner, Swill Federal
Institute for Water Resources and Water Pollution Control ....... 132
17. Evaluation of Stabilization-Solidification Techniques
Rene Goubier, Agence National e pouf la Recuperation et
1 'El imination des Dechets ....................................... 143
18. In Situ Stabilization/Solidification of PCB-Contaminated Soil
Mary K. Stinson, U.S. Environmental Protection Agency ........... 151
19. Applications of Geopolymer Technology to Waste Stabilization
Douglas C. Comrie, John H. Patterson, and Douglas J. Ritcey,
D. Comrie Consulting Ltd ........................................ 161
20. Investigation of Stabilizing Arsenic-Bearing Soils and Wastes Using
Cement Casting and Clay Pel leti zing/Sintering Technologies
John J. Trepanowski, David D. Brayack, NUS Corporation; and
Jeffery A. Pike, U.S. Environmental Protection Agency ........... 166
21. Evaluation of the Soliditech Solidification/Stabilization Technology
Walter E. Grube, Jr., U.S. Environmental Protection Agency ...... 176
-------
SESSION 4 - BIOLOGICAL
22.
23.
24.
25.
26.
The Development of Screening Protocols to Test the Efficacy of
Bioremediation Technologies
John A. Glaser, U.S. Environmental Protection Agency 189
Fungal, Biotrap for Retrieval of Heavy Metals from Industrial
Wastewaters
Theodore C. Crusberg, Pamela Weathers, and Ellen Baker,
Worcester Polytechnic Institute 196
Composting of Explosives and Propel!ant Contaminated Sediments
Richard T. Williams, P. Scott Ziegenfuss, Roy F. Weston, Inc.;
and Gregory B. Mohrman, Wayne E. Sisk, U.S. Army
A New Biotechnology for Recovering Heavy Metal Ions from Wastewater
Dennis W. Darnal and Alice Gabel, Bio-recovery Systems, Inc. ..
204
217
27,
Hazardous Waste Management in Research Laboratories
George Sundstrom, Agricultural Research Service 226
SESSION 5 - WASTE MINIMIZATION
Waste Reduction -- What is it? How to do it?
Roger L. Price, Univ. of Pittsburgh 234
28. Paint Removal Strategies Effective in Reducing Waste Volumes
and Risks
Thomas F. Stanczyk, Recra Environmental, Inc 243
29.
30.
Treatment and Recovery of Heavy Metals from Incinerator Ashes
I.A. Legiec, C.A. Hayes, and D.S. Kosson, Rutgers - The State
Univ 253
Government-Provided Technical Assistance for Hazardous Waste
Minimization
Robert Ludwig, Jim Potter, David Hartley, Kim Wilhelm,
California Department of Health Services; and Lisa Brown, U.S.
Environmental Protection Agency 262
SESSION 6 - THERMAL TREATMENT
31.
32.
33.
Rotary Kiln Incineration Systems: Operating Techniques for
Improved Performance
Joseph Santoleri, Four Nines, Inc
The Use of Oxygen in Hazardous Waste Incineration — A State-of-
the-Art Review
Min-Da Ho and Maynard G. Ding, Union Carbide Industrial
Gases, Inc ,
269
285
Flue Gas Cleaning by Wet Scrubbing and Condensation in a Special
Waste Incineration Plant
J.-D. Harbell, P. Luxenberg, and D. Ramke, Gesellshaft zur
Beseitigung von Sondermull in Bayern 306
vn
-------
34. Evaluation of Mechanisms of PIC Formation in Laboratory Experiments:
Implications for PIC Formation and Control Strategies in Full-Scale
Incineration Systems
Philip H. Taylor, Barry Del linger, Debra A. Tirey, University
of Dayton; and C.C. Lee, U.S. Environmental Protection Agency .
SESSION 7 - WASTE MANAGEMENT
314
35. Cost Effective Remediation Through Value Engineering
Patrick F. O'Hara, Kenneth J. Bird, and William C. Smith,
Paul C. Rizzo Associates, Inc 323
36. The Importance of Exposure Pathways in Toxic Substance Control:
A Case Study for TCE and Related Chemicals
Y.C. Yeh and W.E. Kastenberg, Univ. of California at
Los Angel es 331
37. International Perspectives of Cleanup Standards for
Contaminated Land
Robert L. Siegrist, Institute for Georesources and Pollution
Research 348
38. The Status of Hazardous Waste Management in Taiwan, R.O.C.
Larry L.G. Chen, Republic of China 360
39. Chemical Waste Management in Hong Kong
M.J. Stokoe, R.W. Jordan, and R. Tong, Hong Kong Government 367
SESSION 8 - BIOLOGICAL
40. Aerobic Mineralization of Organic Contaminants Bound on Soil Fines
Robert C. Ahlert, David S. Kosson, Rutgers - The State Univ.;
and John E. Brugger, U.S. Environmental Protection Agency 384
41. Environmental Fate Mechanisms Influencing Biological Degradation of
Coal-Tar Derived Polynuclear Aromatic Hydrocarbons in Soil Systems
John R. Smith, David V. Nakles, Donald F. Sherman, Remediation
Technologies, Inc.; Edward F. Neuhauser, Niagara Mohawk Power
Corporation; and Raymond C. Loehr, David Erickson, Univ. of
Texas at Austin 397
42. Design Considerations for Fixed-Film, Aerobic, Microbiological
Degradation of Hazardous Waste
Marleen A. Troy and Wesley 0. Pipes, Drexel Univ 406
SESSION 9 - PHYSICAL/CHEMICAL
43. Applicability of Steam Stripping to Organics Removal From
Wastewater Streams
Benjamin L. Blaney, U.S. Environmental Protection Agency 415
44. SITE Program Demonstration of the CF Systems Inc. Organics
Extraction Unit
Richard Valentinetti, U.S. Environmental Protection Agency 425
VI 1 1
-------
45. Analysis of Vapor Extraction Data from Applications in Europe
Dieter Killer and Horst Gudemann, HARRESS Geotechnics, Inc 434
46. In-Situ Bioremediation of Cyanide
Robert C. Weber, Gregory Smith, Joseph Aiken, Richard Woodward,
and David Ramsden, ENSR Consulting and Engineering 442
SESSION 10 - PHYSICAL/CHEMICAL
47. Evaluation of Alternatives for Impounded High Salt Wastes Contam-
inated with Organic and Inorganic Pollutants
.Robert D. Fox and Victor Kalcevic, International Technology
Corporation 451
48. KPEG Application from the Laboratory to Guam
Alfred Kornel, Charles J. Rogers, and Harold L. Sparks, U.S.
Environmental Protection Agency 460
49. Testing Natural Zeolites for Use in Remediating a Superfund Site
Robert L. Hoye, P.E.I. Associates; and Jonathan G. Herrmann,
Walter E. Grube, Jr., U.S. Environmental Protection Agency 468
50. Treatment of Water Reactive Wastes
John Parker, Lanstar Wimpey Waste 479
51. Use of Innovative Freezing Technique for In-Situ Treatment of
Contaminated Soils
Olufemi A. Ayorinde, Lawrence B. Perry, and Iskandar K. Iskandar,
U.S. Army cold Regions Research and Engineering Laboratory 487
52.
53.
54.
SESSION 11 - WASTE MINIMIZATION
Enhancing Liquid-Liquid and
grating Alternating Current
Water Control Systems
Patrick E. Ryan, Electro-Pure
Stanczyk, Recra Environmental
Sol id-Liquid Phase
Electrocoagul ators
Separation by Inte-
with Processing and
Systems,
Inc. ..
Inc.; and Thomas F.
Machine Coolant Maintenance Leading to Waste Reduction
Barb Loida, Donna Peterson, and Terry Foecke, Univ.
Mi nnesota
of
Waste Gypsum - It's Utilization and Environmental Impacts
Ryszard Szpadt, Technical University of Wroclaw; Zdzislaw
Augustyn, Wroclaw Geological Enterprise; and Wladyslaw
Grysiewicz, Research Center "Hydro-Mech" Kowary
55. Obstacles and Issues in Source Reduction of Chlorinated Solvents-
Solvent Cleaning Applications
Azita Yazdani, Source Reduction Research Partnership
56. Use of EPA's Synthetic Soil Matrix (SSM) in the Evaluation and
Development of Innovative Soil Treatment Technologies
Richard P. Traver, U.S. Environmental Protection Agency; and
M. Pat Esposito, Bruck, Hartmann & Esposito, Inc
499
509
518
530
539
IX
-------
SESSION 12 - THERMAL TREATMENT
57. Pilot-Scale Incineration Testing of an Oxygen-Enhanced Combustion
System
Larry R. Water!and, Johannes W. Lee, Acurex Corporation; and
Laurel J. Staley, U.S. Environmental Protection Agency 547
58. The Partitioning of Metals in Rotary Kiln Incineration
Gregory J. Carroll, Robert C. Thurnau, Robert E. Mournighan,
U.S. Environmental Protection Agency; and Larry R. Water!and,
Johannes W. Lee, Donald J. Fournier, Acurex Corporation 555
59. Treatment of RCRA Hazardous/Radioactive Mixed Waste
M.E. Redmon, M.J. Williams; and S.D. Liedle, Bectel
National, Inc „ 554
60. Operating Experiences of the EPA Mobile Incineration System With
Various Feed Materials
James P. Stumbar, Robert H. Sawyer, Gopa! D. Gupta, Foster
Wheeler Enviresponse, Inc.: and Joyce M. Perdek, Frank Freestone,
U.S. Environmental Protection Agency 572
61. State of the Art Assessment and Engineering Evaluation of Medical
Waste Thermal Treatment
R.G. Barton, G.R. Hassel, W.S. Lanier, and W.R. Seeker, Energy
and Environmental Research Corporation „ 585
-------
ACKNOWLEDGEMENTS
Organizing Board
Allen Cywin
NUS Corporation
Arlington, Virginia
William C. Anderson
American Academy of
Environmental Engineers
Annapolis, Maryland
Jacqueline Aloisi de Larderel, Director
Industry and Environment Office
United Nations Environmental Programme
Paris, France
Ronald D. Hill
U.S. Environmental
Agency
Cincinnati, Ohio
Clyde J. Dial
U.S. Environmental
Agency
Cincinnati, Ohio
Protection
Protection
William J. Carroll, V.P.
Engineering and Environment
Committee
World Organization of
Engineering Organizations
Pasadena, California
Dr. Robert C. Ahl ert
Purdue Univ.
Dr. Michael D. Aitken
The Univ. of Toledo
Dr. James Alleman
Purdue Univ.
Dr. Gary F. Bennett
Univ. of Toledo
Mr. James Bridges
USEPA
Dr. Carl A. Brunner
USEPA
Mr. Richard Conway
Union Carbide Corp.
Dr. Richard A. Dobbs
USEPA
Mr. Brian Flynn
ERM Southwest
Organizing Committee
Lynne M. Casper
Director
NUS Corporation
Pittsburgh, "Pennsylvania
Support Committee
Dr. Raynond A. Freeman
Monsanto Company
Dr. T. Michael Gilliam
Martin Marietta Energy
Systems, Inc.
Dr. Sidney A. Hannah
USEPA
Mr. John Hernandez
New Mexico State Univ.
Mr. H. Lanier Hickman, Jr.
Government Refuse
Collection & Disposal Assn.
Dr. Robert L. Irvine
Univ. of Notre Dame
Dr. Charles F. Kul pa
Univ. of Notre Dame
Mr. Jack Lindsey
USEPA
Dr. Stephen F. Pedersen
Comprehensive Saftey
Compliance
Mr. Robert Peters
Argonne National
Laboratories
Mr. Jerry Roberts
Univ. of Cincinnati
Mr. Jerry M. Schroy
Monsanto Company
Dr. Charles A. Sorber
Univ. of Pittsburgh
Mr. Mark J. Stutsman
USEPA
Mr. Marty Tittlebaum
Louisiana State Univ.
Mr. H. Paul Warner
USEPA
XI
-------
-------
EVALUATION OF TREATMENT TECHNOLOGIES FOR
CONTAMINATED SOIL AND DEBRIS
Richard Lauch
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Barbara B. Locke, Majid Dosani,
Steve Giti-Pour, and Catherine D. Chambers
PEI Associates, Inc.
Cincinnati, Ohio 45246
and
Ed A!perin and Arie Groen
IT Corporation
312 Directors Drive
Knoxville, Tennessee 37923
ABSTRACT
The performance of bench-scale treatment technologies on Superfund site
soil samples was evaluated. The data were required to assist EPA in the
study of alternatives for treating Superfund soils. Soils from three Super-
fund sites were selected for treatment evaluations. Three treatment technol-
ogies were evaluated: 1) chemical treatment (KPEG), 2) physical treatment
(soil washing), and 3) low-temperature thermal desorption. An earlier study
evaluated these three technologies on surrogate soils and also included
incineration and stabilization. A brief comparison of results from actual
Superfund soils with the surrogate soil will also be given. The following is
a brief description of the three treatment technologies that will be dis-
cussed.
KPEG treatment of contaminated material involves mixing the waste with
potassium hydroxide and polyethylene glycol, heating the waste/reagent mix-
ture to 100° to 180°C for 1 to 5 hours, decanting excess reagent, washing the
soil with water, and neutralizing and discharging the cleaned soil. The
process has been successfully demonstrated for treatment of soil containing
chlorinated biphenyls, dioxins, and furans. Decontamination is achieved
through chemical dehalogenation of the aryl halide to form water-soluble
reaction products.
-------
Soil washing involves contacting the waste to be treated with a wash so-
lution of chelate or anionic surfactant. The soil/reagent mixture is agitat-
ed for 15 to 30 minutes. Following agitation, the soil/reagent mixture is
wet-sieved for particle size separation. Following chemical analysis, sieve
fractions are either disposed of in a hazardous waste landfill or returned to
the site as clean. Soil washing has been demonstrated to work for both
inorganic and organic contaminants.
Thermal desorption involves heating organic-contaminated soils in a fur-
nace to prescribed temperatures up to 550°F. As heating occurs, contaminants
are released from the soil as gases that are purged from the system. The
cleaned soils can then either be put back on site or disposed of in a Re-
source Conservation and Recovery Act (RCRA) landfill.
INTRODUCTION AND PURPOSE
In addition to addressing future
land disposal of specific listed
wastes, the RCRA land disposal re-
strictions also address the disposal
of soil and debris from response ac-
tions under the Comprehensive Envi-
ronmental Response Compensation and
Liability Act (CERCLA). Sections
3004(d)(3) and (e)(3) of RCRA state
that the soil/debris waste material
resulting from a Superfund-financed
response action or an enforcement
authority response action implemented
under Sections 104 and 106 of CERCLA,
respectively, will be subject to the
land ban. Because Superfund soil/de-
bris waste often differs significant-
ly from other types of hazardous
waste, the EPA is developing specific
RCRA Section 3004(m) standards or
levels that apply to the treatment of
these wastes. These standards will
be developed through the evaluation
of best demonstrated and available
technologies (BOAT). In the future,
Superfund wastes in compliance with
these regulations may be deposited in
land disposal units; wastes exceeding
these levels will be banned from land
disposal unless a variance is issued.
The data obtained in this study
were required to assist EPA in the
study of alternatives for treating
Superfund soils. Under Phase I of
this research program, which was
conducted from April to November
1987, a surrogate soil containing a
wide range of chemical contaminants
typically occurring at Superfund
sites was prepared for use in bench-
or pilot-scale performance evalua-
tions of five available treatment
technologies: 1) physical treatment
(soil washing), 2) chemical treatment
(KPEG), 3) thermal desorption, 4) in-
cineration, and 5) stabilization/so-
lidification. Under Phase II of this
research program, which was conducted
from April to September 1988, contam-
inated soils from several Superfund
sites were used to reevaluate (at
bench scale) three of these treatment
technologies: 1) chemical treatment
utilizing a potassium polyethylene
glycol reagent (KPEG), 2) physical
treatment (soil washing), ard 3) low-
temperature thermal desorption. Sta-
bilization/solidification was also
conducted in Phase II; the stabiliza-
tion results will be submitted in a
separate report. Incineration was
not evaluated in Phase II because of
the expense involved in pilot-scale
evaluation of incineration and the
plethora of existing data on incin-
eration of hazardous waste.
This report covers Phase II of
EPA's research program for testing
-------
Superfund soils and compares the
Phase I and Phase II efforts. The
bench-scale technology evaluations of
KPEG, soil washing, and thermal de-
sorption were essentially the same in
Phases I and II. The use of actual
contaminated soils in Phase II,
however, produced different removal
efficiencies for many of the contami-
nants than was observed during the
Phase I investigations on surrogate
soils.
APPROACH
Soil Characteristics
Soils from three Superfund sites
were selected for these tests:
Syncon Resins, Berlin-Farro, and Old
Mill. Soils from Berlin-Farro and
Syncon Resins were used for the KPEG
and soil-washing processes, and soils
from Berlin-Farro and Old Mill were
used for the low-temperature desorp-
tion process. Table 1 gives the
grain size characteristics for the
three soils that were used.
Chemical Treatment—KPEG
During 1978, a new chemical rea-
gent was synthesized and effectively
used to dechlorinate PCB oils at the
Franklin Research Center. Since that
time, a series of reagents have been
prepared from potassium hydroxide and
polyethylene glycolates (KPEG) to
degrade the polychlorinated biphenyls
(PCB's), polychlorinated dibenzodiox-
ins (PCOD's), and polychlorinated
dibenzofurans (PCDF's) contained in
soils and in other matrices such as
waste herbicides and waste oils.
A typical KPEG process for
treating contaminated soil involves
mixing the soil with potassium hy-
droxide and polyethylene glycol (ap-
proximate molecular weight of 400),
heating the soil/reagent mixture, to
100° to 180°C, allowing the reaction
to proceed for 1 to 5 hours, decant-
ing the excess reagent, washing the
soil two or three times with water,
and discharging the clean soil.
Figure 1 represents the bench-scale
reaction vessel used in this study.
Source
Table 1. Physical analysis - candidate test soils.
Weight percentage
Coarse sand Fine sand Silt Clay
(>0.5 mm) (0.05-0.5 mm) (0.002-0.05 mm) (<0.002 mm)
Syncon Resins3
Berlin-Farro
Old Mill0
47.8
29.8
4.2
3.3
28.2
37.2
29.1
47.0
35.2
36.1
40.3
35.6
15.1
15.0
33.4
34.1
22.6
19.6
8.0
8.2
27.2
26.5
8.9
7.6
a The soil is primarily alluvial sand, silt, clays, and detritus (land-
fill soil).
The soil is glacial till.
c The soil is a glacial silty clay that also has sand, gravel, and
boulders.
-------
VARIABLE-SPEED
O ) SIIH MOTOR
(50 • 7SO rpm)
THERMOMETER
AND ANGLED ADAPTER
HEATING
MANTLE
VARIABLE
TRANSFORMER
Figure 1 . KPEG reaction vessel.
-------
The reaction of potassium hydroxide
with polyethylene glycol forms an
alkoxide, which in turn reacts ini-
tially with one of the chlorine atoms
on the aryl ring to produce an ether
and potassium chloride salt. In some
KPEG reagent formulations, dimethyl
sulfcxide (DMSO) is added as a cosol-
vent to enhance reaction rate kinet-
ics by improving rates of extraction
of aryl halide wastes into the alkox-
ide phase.
The important feature offered by
the polyethylene glycol-mediated de-
halogenation reaction is a controlled
process that occurs at relatively low
temperatures (70° to 180°C). In
contrast, more conventional dehalo-
genation processes using only solid
caustic generally require substan-
tially higher temperatures, and they
are ofter. violent (or even uncon-
trolled) processes.
Physical Treatment—Soil Washing
Soil washing is a physical
treatment process in which soil par-
ticles are washed with an aqueous
solution to separate the fine soil
fraction (<2 mm) from the coarse
fraction (>2 mm). The steps involved
in soil washing are excavation of the
contaminated soil, breakup of the
soil agglomerates, separation of fine
particles from coarse particles by
physical cleaning and scrubbing with
washing fluids, and treatment of the
fines and spent wash solution. After
the soil-washing process, the clean
coarse fraction (>2 mm) can usually
be deposited back on site without
further treatment, which reduces the
volume of the original contaminated
soil. Contaminated fine particles
can then be stabilized, incinerated,
or subjected to chemical treatment.
If additives are used in the tap
water, recovery and reuse of the
washing fluid are essential for an
economical soil-washing treatment
process.
Successful soil washing of
contaminated soils is based on the
following assumptions:
1) Fine particles (silt, clay, and
humic material) adsorb a signif-
icant fraction of the contami-
nants in the soil.
2) The contaminated fine particles
attach to coarse grains of the
soil by electrostatic, chemical,
hydrodynamic, and Van der Waals
forces.
3) Physical washing of the sand/
gravel/rock fraction effectively
removes the fine sand, silt,
clay, and colloidal-sized mate-
rials from the coarse material.
4) The removal of fine particles
results in the removal of a sig-
nificant portion of the contami-
nants that exist in that frac-
tion.
Most of the soil-washing experi-
ments conducted to date have shown
that metals can be effectively re-
moved from the >2-mm fraction by use
of plain water or addition of a che-
late solution. Organic contaminant
removal has been enhanced by the use
of surfactant solutions. If the con-
taminants are chemically bound to the
soil particles, however, surfactants
may not be able to remove them from
the particles and mobilize them in
the water (Dietz et al. 1986). Nev-
ertheless, washing of soils contain-
ing hydrophobic organics has met with
limited success.
The physical nature of the soils
is also an important consideration in
-------
an evaluation of soil washing as a
possible treatment technology. Gen-
erally, soils high in clay and humic
content are difficult to treat be-
cause of their characteristically
high cation exchange capacity and
high specific surface area. These
experiments entailed the use of three
aqueous wash solutions:
1) Tap water for removing water-
soluble compounds.
2) Chelate wash for removing metal-
lic compounds. A solution was
prepared containing a 3:1 molar
ratio of ethylenediaminetetra-
acetic acid (EDTA) to total
hazardous metals present in the
soil.
3) Surfactant wash for removing or-
ganic compounds. A 0.5 percent
(by weight) solution was pre-
pared of industrial formula Tide
(manufactured by Procter &
Gamble).
Low-Temperature Thermal Desorption
Low-temperature thermal desorp-
tion is based on vapor pressure for
removal of organic compounds. The
thermal desorption process takes ad-
vantage of thermal driving forces to
remove organic contamination while
avoiding typical incineration
processing conditions that are expen-
sive or have negative public percep-
tion. Because thermal desorption is
conducted at lower operating tempera-
tures, it offers significant fuel
savings over high-temperature incin-
eration. The heat required for ther-
mal desorption can be provided by
indirect heating of the soils as
opposed to direct-fired heating of
solids in an incineration process.
This greatly reduces the quantity of
off-gases that must be cleaned prior
to discharge. This design aspect not
only reduces the cost of subsequent
air pollution control, but also
facilitates the design of a closed
system with no visible plume. Thus,
treatment of contaminated soils by
thermal desorption can be more cost-
effective on low-level organically
contaminated soils and may be more
readily accepted by the public.
The experimental approach was
thermal treatment in static trays in
an electric oven for specified peri-
ods of time. The removal of organic
constituents was measured after
treatment at two different test tem-
peratures, 350° and 550°F, for 30
minutes. The effectiveness of treat-
ment was measured by analysis of the
soil before and after thermal desorp-
tion. Figure 2 is a schematic of the
thermal desorber used in this study.
PROBLEMS ENCOUNTERED
Several problems were associated
with the semivolatile analyses, and
these data should be used with cau-
tion. The hold times were exceeded
for analysis of the semivolatile or-
ganic extracts. Furthermore, the de-
tection limit for many of the samples
was in excess of what was necessary
to make valuable conclusions concern-
ing the treatment technologies' ef-
fectiveness. Finally, for pyrene,
the only target compound subject to
QA/QC analyses, the QC limits for ac-
curacy and precision were exceeded.
RESULTS
Chemical Treatment--KPEG
The metals results indicate that
none of the metals was removed by
KPEG. KPEG is not designed to remove
-------
INTERIOR OF
OVEN CHAMBER
OVEN INDICATOR
THERMOCOUPLE
PURGE GAS
TEST THERMOCOUPLE
SOIL THERMOCOUPLE
GAS EXIT AT DOOR SEAL
Figure 2. Schematic of interior of static tray test oven with the tray inserted.
-------
metals; however, small reductions in
some metals were noted and it is
likely that these reductions resulted
from the acid wash after KPEG treat-
ment.
The percentage reductions of
volatile organic contaminants from
the Berlin-Farro soil treated by KPEG
ranged from 28 to greater than 93
percent (Table 2). The reduction
data for semivolatiles (Table 3) show
removals of 59 to 97 percent. The
semivolatile data are suspect, how-
ever, because an extended delay oc-
curred in the analyses and because of
high method detection limits in the
analytical parameters.
The pesticides data (Tables 4
and 5) indicate almost complete
removal (98 percent or higher) of all
pesticides by the KPEG treatment for
both Berlin-Farro and Syncon soils.
The percentage reductions in TCDF,
PeCDF, and HxCDF concentrations from
the Syncon Resins soil treated with
KPEG ranged from greater than 57 to
greater than 94 (Table 6). The KPEG
process was specifically designed to
degrade chlorinated organics such as
PCB, PCDD, and PCDF and other contam-
inants (e.g., pesticides and herbi-
cides). The substantial reduction in
pesticides and some chlorinated di-
benzofurans obtained during this
study further supports the effective-
ness of KPEG in removing these com-
pounds from contaminated soils.
Physical Treatment—Soil Washing
The results from the evaluation
of soil washing show that generally,
most metals were effectively removed
from the coarse fraction (greater
than 2 mm in diameter) of the contam-
inated soils. Table 7 shows that
water alone removed arsenic and lead
in an average of 92 percent from the
Syncon Resins site soil, whereas cop-
per, vanadium, and zinc were reduced
by an average of 64 percent. The
addition of a chelate slightly in-
creased the removal efficiencies for
arsenic and nickel; the addition of a
surfactant slightly increased the re-
moval of antimony, arsenic, and vana-
dium. Pyrene was effectively reduced
with a water wash (by 72 percent for
the Syncon Resins soil). The addi-
tion of a chelate increased the
removal only slightly (pyrene was
reduced by 80 percent). In general,
the addition of a chelate or surfac-
tant did not significantly enhance
metallic or organic contamination
removal over use of a plain tap water
wash. Other organics were in the
soil, but analytical results were too
poor or detection limits too high to
indicate whether removal of organics
was taking place.
Low-Temperature Thermal Desorption
The results of the thermal de-
sorption tests show that this process
is effective in removing organic
contaminants from Superfund soils.
Treatment results for Berlin-Farrc
and Old Mill soils are given in
Tables 8 and 9, respectively. Vola-
tiles ranged from 70 to 96 percent
removal at 350°F and from 63 to 99
percent removal at 550°F for both
soils. Semivolatiles ranged from 55
to 95 percent removal at 550°F for
both soils. In the opinion of the
authors, the lower ranges, such as 72
and 70 percent removal for toluene
and xylenes, respectively, at Old
Mill would be higher if it were not
for poor analytical results because
94 and 96 percent removals were
respectively obtained for these same
two compounds at Berlin-Farro. Fur-
thermore, Berlin-Farro had a higher
-------
Table 2. Results of KPEG treatment of Berlin-Farro soils—volatile organic
compounds.
Parameter
Acetone
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
Untreated soil ,
ppb
1200.0
109. 5C
500. Oc
580. Oc
225. Oc
Treated soil ,
ppb
633b
<33.0
<33.0
215b
161
Percent
reduction
47.3
>69.9
>93.4
62.9
28.4
a Results presented are averages of quadruplicate runs.
Detected in blank.
c Estimated value.
Table 3. Results of KPEG treatment of Berlin-Farro soil—semivolatile or-
ganic compounds
Parameter
Hexachlorocyclopentadiene
Hexachlorobenzene
Pentachlorobenzene
Untreated
soil , ppb
38,667
88,667.
8,800°
Treated
soil , ppb
<4800
<2600
<3565
Percent
reduction
>87.6
>97.1
>59.5
Results presented are averages of quadruplicate runs.
Estimated value.
Table 4. Results of KPEG treatment of Berlin-Farro soils—pesticides.'
Parameter
Untreated
soil, ppb
Treated
soil, ppb
Percent
reduction
2,4-D
Heptachlor epoxide
7163
16.1
169
0.113
97.6
99.3
Results presented are averages of quadruplicate runs.
-------
Table 5. Results of KPEG treatment of Syncon Resins soil—pesticides.'
Parameter
Untreated
soil, ppb
Treated
soil, ppb
Results presented are averages of quadruplicate runs.
Percent
reduction
DDT
ODD
DDE
86.7
23.7
15.7
<0.016
<0.016
<0.047
>99.9
>99.9
99.7
Table 6. Results of KPEG treatment of Syncon Resins soil—furans.'
Parameter
TCDF
PeCDF
HxCDF
Untreated
soil, ppb
8.03
22.9
8.77
Treated
soil, ppb
<1.71
<1.46
<3.69
Percent
reduction
>78.7
>93.6
>57.9
Results presented are averages of quadruplicate runs.
Table 7. Soil-washing results: Syncon Resins soil.'
(Soil particles greater than 2 mm in diameter)
Treated soil3
Contaminant
Metal
Antimony
Arsenic
Barium
Copper
Lead
Nickel
Vanadium
Zinc
Organic
Pyrene, ppb
Untreat-
ed soil
0.064
1.95
0.661
0.145
0.244
0.047
0.044
0.910
7880
Water
wash
0.060
0.144
0.394
0.056
0.020
<0.027
<0.016
0.305
2209
Percent
reduc-
tion
6.3
92.6
40.4
61.4
91.8
>42.6
>63.6
66.5
72.0
Che! ate
wash
0.058
0.096
0.486
0.072
0.030
<0.026
<0.015
0.434
<1552
Percent
reduc-
tion
9.4
95.1
26.5
50.3
87.7
>44.7
>65.9
52.3
>80.3
Surfactant
wash
0.052
0.014
0.518
0.337
8.95
0.041
<0.014
0.493
<2500
Percent
reduc-
tion
18.8
99.3
21.6
NRb
NR
12.8
>68.2
45.8
>68.3
a
Values in ppm, except where noted
b NR = Not reported.
10
-------
Table 8. Summary of organic results of thermal desorption of Berlin-Farro
soil (average).
Parameter
Untreated Percent Percent
soil, 350°F, re- 550°F, re- .
yg/kg yg/kg duction yg/kg duction0
Volatile organics
2-Butanone
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
290
147
280
483
387
343L
<23
<23
19
<23
(18)
>84
>92
96
>94
80 •
<25
3
27
<25
72
>83
99
94
Semivolatile organics
Hexachlorobutadiene
Hexachlorocyclopentadiene
Pentachl orobenzene
Hexachlorobenzcne
1,900C
46,000
10,200
105,000
430C
3,050
15,100
250,000
77
93
(48)
(138)
<3,300
<3,300
2,500
47,000
NCd
93, ,
75
55
a Reduction reported as percent change from initial concentration, not cor-
rected by moisture. Increases in concentration are shown in parentheses.
Compound detected in the laboratory blank.
c Estimated values are presented by the CLP laboratory.
NC = Not calculated. The method detection limit of the analysis was
exceeded by the laboratory; reduction cannot be calculated.
Table 9. Summary of organic results of thermal desorption of Old Mill soil
(average).
Parameter
Untreated
soil,
yg/kg
Percent
350°F, re-
Percent
550°F, re- .
yg/kg duction yg/kg duction
Volatile organics
Trichloroethene
Tetrachloroethene
Toluene
Xylenes (total)
Semivolatile organics
Aroclor 1260
2400
362
152
950
2000
173
35
43
285
3000
93
90
72
70
<25
<25
57
48
>93
63
95
(50)D ND (100) >95
a Reduction reported as percent change from initial concentration, not cor-
rected by moisture.
Increases in concentration are shown in parentheses.
c ND = Not detected (method detection limit).
11
-------
percentage of fine silt and clay,
which should theoretically be harder
to treat.
The percent reduction of semi-
volatile organic compounds from
Berlin-Farro soil was slightly lower
than for the volatile compounds.
This is logical because of the higher
vapor pressure of the volatiles. It
is also noted that some of the semi-
volatiles show an increase in contam-
inant (values in parentheses) after
thermal treatment. This is possible
because moisture in the soil can
evaporate while vaporization of the
contaminant is insignificant. This
could also be due to poor analytical
results. More complete results on
thermal desorption are given in Ref-
erence 2.
COMPARISON OF SUPERFUND SOILS (PHASE
II) WITH SYNTHETIC SOILS (PHASE I)
Comparing these results on ac-
tual Superfund soils with the syn-
thetic soils showed that the same
trend in contaminant removals was
accomplished on both types of soils.
The synthetic soils gave a little
higher percent removal than the
actual Superfund soils. This is
logical because the synthetic soils
were spiked and tested with very
little time for the contaminants to
sorb deeply into the soil. The
actual Superfund soils weathered for
long periods; therefore, contaminant
removal would be more difficult.
ACKNOWLEDGMENTS
The authors thank Robert Thurnau
and Mary Ann Curran, both with the
U.S. EPA, RREL, Cincinnati, Ohio, for
initiating the project and serving as
work assignment managers. We also
thank Benjamin Blaney, Chief, Hazard-
ous Waste Treatment Branch, and
Jonathan Herrmann, Chief, Treatment
Technology Section, also with the
U.S. EPA, RREL, Cincinnati, Ohio, for
their technical assistance on the
project.
REFERENCES
1. Dietz, D. H., et al. 1986.
Cleaning Contaminated Excavated
Soil Using Extraction Agents
(Draft). Prepared for the U.S.
Environmental Protection Agency,
Hazardous Waste Engineering
Research Laboratory, by Foster
Wheeler Corporation under Con-
tract No. 68-03-3255.
2. Lauch, R. P., et al. 1989.
Low-Temperature Thermal Desorp-
tion for Treatment of Contami-
nated Soils, Phase II Results.
Proceedings Fifteenth Annual
Research Symposium,
Ohio, April 1989.
Cincinnati
DISCLAIMER
The work described in this paper was funded by the U.S. Environmental
Protection Agency. However, the contents do not necessarily reflect the
views of the Agency and no official endorsement should be inferred.
12
-------
PRE-TREATMENT OF HAZARDOUS WASTE
Weine Wiqvist
KemiavfaU in Skane AB
S-211 24 Malmo, Sweden
ABSTRACT
The pre-treatment of waste is effected by a new method at KEMIAVFALL's reception and holding
station in Malmo, Sweden. The purpose of the pre-treatment is to convert solid and pasty waste into
pumpable waste. The types of waste most immediately applicable are lubrication greases, bituminous
products, paints and glue waste. The solid and pasty waste are converted into emulsified form together
with waste oil or solvent waste in specially designed grinding equipment. The resulting product, a so-
called dispersion, can thereafter be conveyed further for final incineration. The advantages of the
system are numerous. Since waste in drums is converted into a pumpable state for further transport in
tank carriers, stockholding and transport of drums is reduced. Well-defined mixes can be produced by
suitable mixture and dispersion. The final incineration at SAKAB, with its rotary kiln unit, or other
authorized recipient plants can be better controlled and governed, at the same time that the capacity is
utilized to a greater extent.
INTRODUCTION
KEMIAVFALL is a company which deals
with the collection, transport, pre-treatment and
storage of hazardous waste. KEMIAVFALL is a
sub-division of SYS AV (Southwest Scania Waste,
Inc.), which in turn is owned by a number of
municipalities in southern Sweden. KEMIAV-
FALL operates at cost.
The Swedish model for handling hazardous
waste divides the responsibility among industry
(i.e., the waste producer), the municipalities and
the Swedish State. The waste producer is express-
ly responsible and obliged to declare and packa-
ge his waste and to deliver it to the municipality
for further transport. He is also totally responsi-
ble for the contents of the waste and as the waste
producer, he shall also assume all the expenses
for the treatment. The handling of hazardous
waste is prescribed in a special regulation.
Based on the Public Cleansing Act, the muni-
cipality is responsible for collecting and transpor-
ting hazardous waste, thereby creating coordina-
ted and controlled handling. It should be noted
that this naturally does not imply that the munici-
pality treats waste itself. It can very well be done
by hiring contractors. In order to acquire a
sufficiently large base and the proper competen-
ce, collaboration among the municipalities is
imperative, as in the case of KEMIAVFALL,
which belongs within a commonly owned corpo-
ration.
The Swedish State is responsible for the final
treatment of waste in part through its own waste
corporation SAKAB, in part through other com-
panies which have been given special license.
A proposal for new legislation that would
place full responsibility for transportation and
treatment on the municipalities has recently been
made.
The State through its Environmental Protec-
tion Board also provides definitions of what is
considered hazardous waste. It should be noted
that at the present, no international convention
and definition of what constitutes hazardous waste
exists, a fact which at times can make information
in these issues difficult to understand and even
misleading.
PURPOSE
One of the biggest problems at the present
time in Sweden, as in many other countries, is that
the capacity for the final treatment of hazardous
waste is too limited. This is because industry's
13
-------
internal treatment of its own waste, for both
technical and public opinion reasons, has not
developed according to plan. At the same time,
waste, which previously was handled more or
less illegally, is now being taken care of more and
more; this has led to an increased volume of
hazardous waste. Industrial measures aimed at
decreasing the production of hazardous waste
material have on the whole proved to be of limited
value so far, and have partially been counterba-
lanced by the increased total industrial produc-
tion. All in all, this means that greater and greater
amounts of hazardous waste cannot be treated.
Traditionally, much waste is treated in oil
drums, at least hi Sweden. This means that at the
final destruction, the drums are often burned
whole with their waste substances in for examp-
le the rotary kiln unit operated by the government
corporation SAKAB.
Our hypothesis is that through pre-treatment
of certain types of waste aimed at converting solid
and pasty waste to pumpable waste, the capacity
utilization in traditional rotary kiln units could be
increased. In addition, a fraction of the waste after
more pre-treatment could be treated and incinera-
ted, for example as a specific fuel for the cement
industry.
Principle Scheme
Waste
[Dispe
rse
I Rotary ki
kiln
Further treatment
| Specified fuel[•» [Cement kUn]
APPROACH
The following description and results concern
the first step: a relatively simple pre-treatment
followed by incineration in a rotary kiln. The
report is based on practical experience from exi-
sting installations. In connection with this, a second
step requiring further pre-treatment and the pos-
sibility of preparing a specified fuel is also repor-
ted on. Developmental work is in progress in this
area, but full-scale production is not operational
as yet.
Types of waste
Those types of waste which are applicable for
pre-treatment in this context are those solid and
pasty ones suitable for thermal treatment: certain
types of waste oil, bituminous products, solvent
waste, lubrication greases, paint, glue, and varn-
ish, for example, which are usually handled in
drums.
Technique
The necessary technical equipment consists
of a newly developed "dispersion module". The
drums are handled and emptied into a tank, where
the contents are mixed with so-called dispersion
solution, usually light oil and solvents. The waste
is pumped and decomposed in a special aggregate
until an emulcified ("dispersion") form is achie-
ved, thereby the name. The ratio of dispersion
solution to waste is c.l:5.
The capacity for creating a dispersion is deci-
ded first of all by the speed with which the drums
can be emptied. The dispersion aggregate per se
can normally treat several tens of cubic meters per
hour. The rate for emptying the drums so far is c.
20 drums at a time, equalling 4 cubic meters/hr.
Flow Scheme
Waste classification
Analysis
Storage
Drum handling
- opening
- emptying
- cleaning
Dispension
- with oil/solments
Storage
Transportation
Incineration
14
-------
Specific details
The main purpose of pre-treatment is to con-
vert solid waste to a pumpable condition. This
facilitates further transport and treatment. How-
ever, this is not enough. The final mix must fur-
thermore meet certain previously determined cri-
teria regarding energy content, heavy metal con-
tent, etc. Not until then can the dispersion waste
be treated as a special product and be transported
directly into and incinerated in, for example, a
rotary kiln.
From the above it is clear that great exactness
must be followed when taking samples or choo-
sing the waste which is to be dispersed. The
necessary knowledge does not always have to be
collected by sample taking and laboratory analy-
sis but can equally well be obtained through a
closer record of the origin of the waste, he type of
production, etc. Thus, by an intelligent applica-
tion of chemical knowledge, enough information
can be obtained to prevent the inclusion of certain
particularly polluted types of waste. Knowledge
of what is to be mixed and dispersed thus con-
cerns not only the amount of heavy metals, halo-
gens like chlorine and bromine, etc., in the waste
in question, but also a chemical/technical judge-
ment of the risks involved in joint storage effects,
in the form of polymerization or other undesira-
ble results.
The following items indicate the demands
made on the homogenized waste:
Total halogens max. 1 %, of which bromine 0.1 %
Chromium max. 1000 mg/1
Lead max. 2000mg/l
Total alkali metals max. 500 mg/1
Cadmium trace
Quicksilver trace
PCB max. 50 mg/kg
Particles max. 10 mm
Heat value 16-20 MJ/kg
Pumpability Pump time when flowing in 3" pipe
at least 10 cubic meters/hr
Storage/transport
The dispersion waste is temporarily stored in
tanks and thereafter transported by tank carriers
to the final treatment plant. This waste will gra-
dually settle and therefore in certain cases, circu-
lar pumping must occur during temporary stora-
ge; in any case, the length of time from the
dispersion to the final delivery must be kept as
short as possible. Even transport by rail can be
possible.
Treatment of left-over drums
The drums which are emptied are provisional-
ly cleaned. In the cases when the drums are reused
within the plant, for example for packing cans,
etc., no additional cleaning is required. The
remaining drums are compressed and deposited
as hazardous waste. In order to recover scrapped
drums, further chemical/physical cleansing must
take place.
Installations
Dispersion aggregates are now operative in
three places in Sweden, one in direct connection
with the SAKAB complex, one outside of Stock-
holm, where the plant is run by a private contrac-
tor and is located c. 250 km. from SAKAB, and
one at the KEMIA VFALL plant in Malmo, c. 500
km. from SAKAB.
Final treatment
Up to now, all experience is based on the final
treatment by thermal incineration in rotary kiln
units. Here the dispersion waste is treated like
sludge and with its medium-good heat value is
laid like a base load in the system. In this way both
internally and externally produced dispersion are
received and incinerated without any particular
delay. From the point of view of incineration, itis
obvious thatthehomogenized, pumpable waste is
preferable to the drums. The capacity in the plant
is increased at the same time that the incineration
effectiveness can be kept high. This latter advan-
tage means lower discharge and the former, a
more cost-effective handling.
PROBLEMS ENCOUNTERED
Operational experiences taken from the three
existing installations cover a relatively short period
of time: The oldest plant has been in operation for
c. two years and the other two, c. one year.
15
-------
The following problems have been observed:
* Difficulty in choosing drums whose con-
tents will be dispersed
* Spillage when emptying the drums
* Long emptying time for "tough" waste
* Somewhat unpleasant work environment.
The main problem stems from the preparation
of the dispersion. Transport and final treatment
appear to have few problems.
RESULTS
Experiences to date have shown that pre-
treating by dispersion provides:
- More effective transportation, by replacing
regional and long distance drum transport
with tank carrier transport, providing a greater
effective transport capacity.
- More effective incineration, when the whole
drum, which must itself also be warmed up
and contains solid and half-solid waste, is
replaced by incineration of a liquid material,
which on the whole means higher capacity,
better incineration characteristics and there
with lower discharge.
- More effective resource utilization of drums.
which can be reused or recovered as scrap
metal after cleaning.
- Lower total costs for both transport and final
incineration, including naturally the extra
costs for the pre-treatment.
FUTURE DEVELOMPENT
Up to now all experience has been based on
final incineration in rotary kilns. As implied in the
introduction, one of Sweden's biggest problems
in this area is the crying need for sufficient final
treatment capacity. Due to this, on-going re-
search aims at developing a second pre-treatment
step: through filtering and other efforts, separa-
ting permanent pigments and particles from paint
waste for example, and adjusting the heating
value to a specified fuel product with prescribed
energy values, heavy metal content, etc. Such a
reworked waste, while maintaining requirements
for an environmentally positive incineration,
should be able to be finally destroyed in, for
example, cement kilns or certain other industrial
heating units. In this way, one gains a considera-
bly higher potential for a diversified final utiliza-
tion. It should be observed however that such
handling to the greatest extent possible must be
coordinated with an extensive check of both the
mixing equipment and the capacity and environ-
ment values at the final treatment plant. This is
necessary, as an inadequate handling of the dis-
persion and pre-treated waste can lead to very
unpleasant consequences. This is easily under-
stood, since it can obviously be very tempting to
include certain pumpable waste which has very
high destruction costs, for example PCB waste,
when making the dispersion.
In this area of development with second-stage
pre-treatment, a considerably part of the disper-
sion material - 30-40% - will be separated as
solid material. In order to achieve the capacity-
heightening effect, it is obviously not a good idea
to have the final incineration in a rotary kiln unit,
as the waste is comprised chiefly of inorganic and
solid materials. The goal here is to isolate this
waste for deposition through some type of solidi-
fying process.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
16
-------
HYDRAULIC JETT MIXING - VERSATILE TOOL FOR HAZARDOUS WASTE TREATMENT
John R. Ackerman, P.E., DEE
Hazleton Environmental Products, Inc.
225 North Cedar Street
Hazleton, PA 18201
ABSTRACT
Many of the problems presented in most hazardous waste treatment pro-
cesses are the result of being unable to sufficiently mix the waste stream
with the solid, liquid or gas reactants required to provide treatment. This
paper details the development of a mixing system that through the action of
hydraulic jetts formed by a nozzle ring mixer configuration produces an
extremely turbulent mixing action by the formation of a hydraulic venturi.
The action of this venturi results in the aspiration of considerable volumes
of air through the bore of the unit. Experimentation has proven that in addi-
tion to the oxygen transfer capabilities first evidenced by the unit, suffi-
cient volumes of air are drawn through the unit to effect stripping of
entrained and dissolved gases and volatile organic compounds (VOCs).
The development of the hydraulic jett mixing concept has to this point
led to four separate applications in waste treatment processes. They are met-
als removal from solution through pH adjustment and oxidation, more efficient
reagent mixing, high efficiency dry reagent mixing directly into a waste
stream and air stripping of gases and VOCs from solution. This versatile
piece of equipment attains its treatment efficiency because it was designed to
enhance every aspect of mixing producing a high efficiency mixer. Extremely
turbulent flow regimes are produced not only within the bore of the unit, but
also within the jett streams as they leave the nozzles as evidenced by the
streams' high Reynold's numbers. Additionally, the hydraulic flow path design
reduces losses through the mixer to allow for high efficiency mixing at low
heads.
Patents #4,474,477 & #4,761,077, with
Other Patents Pending) can mix slur-
ry-laden waste streams as well as
clear water streams.
INTRODUCTION
Efficient mixing of reactants
into a waste stream has always been a
problem in that there has been no
mixer capable of combining all
the elements of enhanced mixing into
a single piece of equipment. Through
the development of a mixing system
for the mining industry to treat acid
mine water containing heavy metals, a
versatile new hydraulic jetting
static mixer has been developed that
has no moving parts and a clean bore
with no internal components. As a
result, this patented unit (U.S.
The main goal of the'development
of the hydraulic jett mixer was to
reduce the size of the tankage
required for an acid mine drainage
(AMD) treatment plant through devel-
opment of a static mixing device that
could coincidentally aerate' the
treatment flow. This process equip-
ment being developed would simulta-
neously adjust the pH and oxidize the
metals allowing formation of the
hydroxide sludges required for sedi-
17
-------
mentation and removal of the metals
from the treatment stream. In
effect, the device eliminates two
reaction tanks, the neutralization/
mixing tank and the aeration tank.
Further refinement of the deliv-
ery system allows many dry solids,
such as hydrated lime or soda ash, to
be directly injected into the treat-
ment stream. As was found during the
initial testing of the hydraulic mix-
ing system, dry hydrated lime could
be directly mixed into the acid
treatment stream for pH adjustment.
Not only could sufficient pH adjust-
ment be attained, proving the concept
of rapid direct injection of dry rea-
gents, but further testing proved the
units could produce milk of lime
solutions extremely quickly. This
allows batch tanks for existing lime
solution injection style treatment
systems to be downsized.
The oxidation capabilities of
the hydraulic jett, as evidenced by
its precipitation of high concentra-
tions of metals from solution, led to
investigation of the unit's capabili-
ties to strip dissolved gases and
volatile organic compounds from solu-
tion. Most easily stripped are dis-
solved and entrained gases, such as
carbon dioxide, methane or radon.
Stripping of the volatile
organic compounds with the hydraulic
jett is somewhat harder due to their
lower Henry's Numbers. For example,
tetrachloroethylene (PCE) only has a
removal rate of 75-78% on a single
pass, however, multiple passes, five
recycles, achieve greater than 98%
removal efficiencies with a fourth
generation design. While more passes
may be required to achieve sufficient
removal rates, this is overcome by
the size of the unit, less than 6'
tall, and its use of aspirated air
for stripping.
As with the other hydraulic jet-
ting applications, intimate mixing of
the gas stream with the organic chem-
ical-laden water provides the driving
force to remove the organics from
solution.
PURPOSE
The treatment process for acid
mine drainage (AMD) is straight for-
ward, raise the pH of the liquid,
then aerate to precipitate the metals
from solution. Traditionally, a
flash mixer in a separate tank
receives a lime slurry to mix with
the incoming acid water flow. The
thirty second to five minute reten-
tion period insures that a neutral-
ized effluent is discharged into the
aeration tank where the metals, which
have changed state, can be oxidized
and form a hydroxide precipitate
during the additional twenty to
thirty minute retention period.
(3,5,8)
Development of an enhanced mix-
ing system would produce a more effi-
cient and cost effective treatment
process. The basic parameter
selected for enhancement was the gen-
eration of additional interface sur-
face area. The other design consid-
eration was to produce a mixer that
eliminated the large tankage volume
generally devoted to the mixing and
aeration portions of the treatment
process.
The initial research & develop-
ment of what came to be the hydraulic
jett started with the premise that
the mixer would be of a static design
and would utilize the velocity gener-
ated by passing the flow through
nozzles specially designed to greatly
enhance the turbulence within the
contacting chamber of the mixer.
(1,6,7)
18
-------
APPROACH - ACID MINE DRAINAGE (AMD)
ply was used.
With a mixer designed to produce
neutralization of an acid stream
along with the capability to inject
air to effect oxidation of the metals
present within the acid stream, the
next step was to field test the
equipment to insure that it could be
scaled up to process flows in excess
of 500 gpm.
The original field testing was
divided into two components: the
first consisted of testing at an acid
water flow of 100 gpm, while the sec-
ond was at 800 gpm. The trial used a
side stream from an acid mine water
source in West Virginia and dis-
charged into the holding basin prior
to the existing acid mine water
treatment plant.
The trial set-up included stain-
less steel submersible pumps, both
liquid and dry solid delivery systems
and 2" and 8" hydraulic jett units.
The test started with the 2"
hydraulic jett operating at 100 gpm
through the 1/8" nozzles. (Figure 1)
The milk of lime slurry was injected
through the throat of the unit to
contact the acid water stream within
the bore. A blower supplied low
pressure air for oxidation of the
metals after neutralization of the
acidity.
Upon completion of the testing
with a lime slurry as the neutraliz-
ing agent, the next step was to det-
ermine if the hydraulic jett mixer
could take a dry solid feed and
effect the neutralization reaction
within the short mixing period
available. To accomplish this a con-
verted "rock duster" was used to
pneumatically convey powdered
hydrated lime into the bore through
the throat of the mixer. As with the
previous test a low pressure air sup-
The next stage of the testing
was to operate at a 800 gpm flow and
to attempt to duplicate the results
obtained with the testing at 100 gpm
flows. In this portion of the test-
ing an 8" hydraulic jett mixer (Fig-
ure 2) was used. This unit used 3/8"
nozzles. Both lime slurry and dry
lime injection was performed along
with injection of liquid sodium
hydroxide (NaOH).
PROBLEMS ENCOUNTERED - AMD
Injection of the lime slurry
into the mixer produced no problems
even at high feed rates. However,
pneumatic conveying of a dry lime
suspension resulted in a calcium car-
bonate build-up created by back-
splashing of the jetts. This
build-up occurred in both size units
and steps were taken to eliminate the
problem by relocating the air
entrance to the mixer. Further
design changes resulted in the back
of the mixer bore being left com-
pletely open allowing an air flow to
be aspirated into the bore preventing
backsplashing of the liquid.
RESULTS - AMD
The 2" hydraulic jett mixer was
designed to pass 100 gpm through 72
nozzles placed in the wall of the
center bore. Two additional cham-
bers, each containing an additional
48 nozzles, were located before and
after the main injection ring. These
chambers were for injecting air or
additional flows into the bore of the
mixer.
With the lime slurry addition to
the acid water flow, the hydraulic
jett mixer was able to neutralize the
acidity present in the process stream
and raised the pH from 3.5 through to
near 12 at successive levels of test—
19
-------
Lime
Water
Figure 1: First Generation Hydraulic Jett Mixer
Figure 2: Acid Mine Drainage
Hydraulic Jett Mixer
Figure 3: Air Stripping
Hydraulic Jett Mixer
Water
Water
ing. This was especially impressive
in that the total residence time
within the mixer was 0.2 seconds
prior to discharge. The stochiomet-
ric lime additions produced results
tnat closely followed the titration
curve for the mine water being
tested. Additionally, the air volume
injected was sufficient to oxidize
all the iron present in the mine
water. Iron concentrations in excess
of 500 mg/1 (80% ferrous, 20% ferric)
were present in the flow and wert
precipitated in the form of ferric
hydroxide to pH levels of 10.5.
The next step of the testing of
the 2" unit was to pneumatically
convey powdered hydrated lime into
the bcre of the unit to mix directly
with the acid water. As with the
lime slurry, each level of lime addi-
tion drove the pH up in accordance to
the titration curve reaching a maxi-
mum pH of 11.3. Again, the neutral-
ization reaction was complete as the
20
-------
liquid left the unit as evidenced by
no further pH rise. Oxidation and
precipitation of the metals was also
driven to completion upon sufficient
increase in pH.
The next series of tests, which
incorporated the design changes
developed from the testing of the 2"
mixer, was to determine the scale-up
capabilities of the hydraulic jett
concept. An 8" mixer passing 800 gpm
using the same nozzle configuration
as the 2" unit, except for the nozzle
diameter, was tested.
The major difference between the
units (Figures 1 & 2) was to open the
back of the 8" unit's bore. As was
described in the Problems Encountered
section, the need to eliminate the
backsplashing led to the discovery
that large volumes of air are aspi-
rated into the bore by the venturi
action of the hydraulic jetts. The
AMD units incorporated this design
change throughout the remaining test-
ing.
The 8" unit produced the same
results as the 2" unit in terms of
neutralization and oxidation effi-
ciency for lime slurry injection, dry
hydrated lime feed and with this
series of tests, liquid caustic.
However, in this series of
tests, the air used for oxidizing the
metals was not obtained by low pres-
sure injection, but through air aspi-
rated co-current with the water flow.
The oxygen mass transfer mechanisms
produced are so efficient that the
dissolved oxygen concentration was
driven from 1 to 2 mg/1 to saturation
levels of over 11 mg/1 with just the
single pass through the hydraulic
jett mixer. The mine water tempera-
ture during the trial was in the mid
50's F. (1,4,7)
As a result of this series of
tests the hydraulic jett mixing sys-
tem has been commercially applied in
acid mine drainage treatment plants
in West Virginia and Ohio. These
plants have produced considerable
savings as the cost of treatment has
dropped from $0.54 per thousand gal-
lons for conventional lime and aera-
tion treatment to $0.17 per thousand
gallons with the hydraulic jett sys-
tem. (4)
APPROACH - SLURRYING SOLIDS
From review of the results of
the hydraulic jett mixer's AMD tests,
it was felt that lime could be slur-
ried into a water stream using the
hydraulic jetts. A series of tests
were performed to verify the poten-
tial application.
Initially, to test the premise,
the 2" hydraulic jett mixer used in
the AMD testing along with a further
modified "rockduster", were placed at
a mine site in West Virginia to pro-
duce a "milk of lime" slurry from dry
hydrated lime.
A second test was performed at a
New York State electric utility where
a specially designed hydraulic jett
unit was installed to reduce the time
required to slurry a tanker truck
load of hydrated lime. Subsequent
tests at the same site have been per-
formed aimed at producing higher
slurry concentrations and in batching
slurry production. (2,7)
PROBLEMS ENCOUNTERED
SOLIDS
SLURRYING
Two problems were encountered
during the development phase of the
hydraulic jett for the application of
producing slurries/solutions, pri-
marily milk of lime, from hydrated
lime. They were 1) balancing the
aspirated and conveying air volumes,
and 2) insuring the wetting of all
21
-------
the dry solids entering the unit.
The first problem presented
itself when the air volumes being
used to convey the dry hydrated lime
from its silo exceeded the air volume
the hydraulic jett unit could aspi-
rate and resulted in blowback of lime
solids from the bore. The problem
was solved by balancing the veloci-
ties of the two air streams to insure
a constant inward motion of the air
flow.
The second problem occurred when
milk of lime slurries in excess of
101 solids by weight were being made.
Wetting of all the solids needed to
be insured to attain complete mixing.
This problem was solved by reconfig-
uring the jett nozzles.
RESULTS - SLURRYING SOLIDS
The production of lime slurries
with the hydraulic jett mixing system
is a simple process in which powdered
hydrated lime is pneumatically con-
veyed or dropped into the back bore
of the jett unit. (Figure 2) It is
contacted with a set volume of water,
where it is initially wetted and then
slurried by successive jetted flows
within the bore of the unit.
The hydraulic jett mixers have
been used to make "milk of lime"
slurries in concentrations ranging
from 5% to 25% solids by weight. The
main advantages in using this type of
mixer are the time factor and dust-
less operation of the units.
Testing at a New York State
utility produced a system in which
the time taken to make up the 750
gallon milk of lime "day tank" was
dropped from approximately four hours
to less than ten minutes, while eli-
minating the lime dust problem inher-
ent in their traditional lime slurry
production.
The production of lime slurries
using the hydraulic jetts is being
commercially applied with significant
customer cost savings in operator
time, maintenance time and the more
efficient mixing allows greater effi-
ciency in the lime usage. (2)
APPROACH - AIR STRIPPING
Review of the results of the
development testing for the acid mine
water treatment process showed that
an extremely efficient aeration
device had been produced. As placing
a gas (oxygen) into solution was
easily accomplished, the premise was
forwarded that the same hydraulic
jett principle could just as easily
remove gases from solution.
The same mixing action that pro-
duced the large surface areas needed
for mixing when combined with the
large air flows aspirated through the
unit would act similarly to existing
air stripping technology, such as,
packed, counter-current air stripping
towers.
A new test unit was made to
attempt to remove PCE and radon from
a groundwater source. Further modi-
fications (Figure 3) of the hydraulic
jett design were made to increase the
air handling capacity of the unit.
A total of three tests were per-
formed with the air stripper version
of the hydraulic jett mixer. One
removing PCE and radon from ground-
water, one removing PCE from a spring
source and the third removing methane
and hydrogen sulfide from a ground-
water source. All these tests were
on existing potable water supplies.
PROBLEMS ENCOUNTERED - AIR STRIPPING
The premise that air stripping
with the hydraulic jett would be as
22
-------
easy as aerating with the device,
unfortunately did not hold. The
first tests, while showing promise of
results to come, did not provide
superior removal efficiencies on a
single pass through the unit for
volatile organic chemicals, however,
dissolved radon & methane gases were
readily removed from solution.
Modifications of the hydraulic
jett design to increase the air
volume entrained by a single unit
were made and the units were
retested.
RESULTS - AIR STRIPPING
As previously described three
major tests removing PCE, radon and
methane were performed in 1987-88 to
verify the removal efficiencies of
the hydraulic jett when used to air
strip.
The first test was performed at
a well site in northeastern Pennsyl-
vania contaminated with tetrachlo-
roethylene (PCE) and radon. A trial
was set up for operation of a
2-1/2" Twin hydraulic jett unit
operating on a single pass of the
flow from the well pump. Discharge
of the flow into a receiving tank
allowed an 18" free space above the
water for volatile separation from
solution. The unit used produced a
volumetric ratio of air to water of
17:1.
As the results listed in
Tables 1 & 2 show the hydraulic jett
unit used for this first test handled
PCE concentrations ranging from 10 to
15 ppb in a flow of 120 gpm on a
single pass at removal efficiencies
ranging from 55 to 60% for tempera-
tures over 20 F. There was a signi-
ficant drop in removal efficiency
with the advent of colder ( 20 F)
weather. Average removal efficien-
cies were near 40% for that part of
the testing. This represented about
a 36% drop-off in removal efficiency
for the colder weather.
Table 1: PCE Removal Single Pass
Sample Source Result Reduction
12/11
12/12
12/19
12/20
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
15 ppb
6 ppb
12 ppb
5 ppb
10 ppb
6 ppb
11 ppb
7 ppb
60%
58%
40%
36%
Radon levels in the water stream
ranged from 1500 to 1900 pCi/L. An
average 90% removal rate was attained
for 20 F operation with only a drop-
off to 85.5% for cold weather oper-
ation.
Table 2: Radon Removal Single Pass
Sample Source Result Reduction
12/11
12/12
12/19
12/20
Radon
Influent
Effluent
Influent
Effluent
Influent
Effluent
. Influent
Effluent
1910
180
1675
188
1518
213
1752
303
levels measured
+/-100
+/-
+/-
+/-
+/-
+/-
+/-
+/-
in
45
85
40
82
40
80
40
91%
89%
88%
83%
picocuries
per liter (pCi/L).
The results from the first test
showed that improvements could be
made to the unit to enhance the
removal efficiencies by increasing
the air volumes and recycling the
flow through the unit. Table 3 shows
the results of a recycle test on PCE
removal from a contaminated spring in
central Pennsylvania.
As indicated the PCE levels
could be dropped considerably below
the MCL of 5 ppb with three passes
through the hydraulic jett with
23
-------
approximately 99% removal with five
passes. This unit had an increased
air flow that produced a volumetric
air to water ratio of 28:1.
Table 3: PCE Removal Recycle
Sample Result Reduction
Pennsylvania American
are appreciated.
Water Company
Influent
Recycle 1
Recycle 2
Recycle 3
Recycle 4
Recycle 5
39.1 ppb
12.3 ppb
5.7 ppb
2.0 ppb
1.0 ppb
0.4 ppb
73.7%
85.4%
94.8%
97.3%
98.9%
As a result of these tests the
hydraulic jett air stripping process
has been issued an innovative permit
by the DER in Pennsylvania for use in
volatile organic compound removal for
potable water treatment.
One other on-going test is in
New York State in which a methane and
hydrogen sulfide contaminated well
is being successfully treated for
potable use. In addition to produc-
ing a palatable water, the unit is
also oxidizing the iron present in
the flow eliminating the previously
required potassium permanganate addi-
tion. This unit is in operation on
an emergency permit from the New York
Department of Health.
The hydraulic jett has proven to
be a versatile piece of equipment
that has considerable potential in
the treatment of contaminated water
sources. Its capabilities are accom-
plished through the high turbulence
generated within the jetted nozzles
which produces extremely large sur-
face areas for contacting the water
with the aspirated air stream.
ACKNOWLEDGEMENTS
Analytical services and support
received from Dr. Brian Dempsey, Penn
State University, Mr. Jack Mitchell,
Lemont Water and Mr. Paul Zielinski,
REFERENCES
1. Coudriet, Lawrence, Personal Cor-
respondence with Mr. Coudriet,
Techniflo Systems, Wexford, PA
2. Galgon, R.A., Personal Conversa-
tions with Mr. Galgon, Hazleton
Environmental Products, Inc.,
Hazleton, PA
3. Holland, Charles T., James L.
Corsaro, Douglas J. Ladish, 1968,
"Factors in the Design of an Acid
Mine Drainage Plant", In: Second
Symposium on Coal Mine Drainage
Research, May, Mellon Institute,
Pittsburgh, PA, pp. 274-290
4. Kolbash, Ronald L., PhD, 1988,
"New Lower Cost Method for Treat-
ing Acid Mine Drainage", A Case
History - Martinka Mine #1", In:
49th Annual Meeting International
Water Conference, October, Pitts-
burgh, PA
5. Selmeczi, Joseph G., 1972, "Design
of Oxidation Systems for Mine
Water Discharges", In: Fourth Sym-
posium on Coal Mine Drainage
Research, April, Mellon Institute,
Pittsburgh, PA, pp. 307-330
6. Smith, William, Personal Corre-
spondence with Mr. Smith, Strato-
Flo, Inc., Cannonsburg, PA
7. Werner, Roy. Personal Correspon-
dence with Mr. Werner, Barrett,
Haentjens & Company of Pittsburgh,
Lawrence, PA
8. Wilmoth, Roger C., Robert B.
Scott, 1970, "Neutralization of
High Ferric Iron Acid Mine Drain-
age", In: Third Symposium on Coal
Mine Drainage, May, Mellon Insti-
tute, Pittsburgh, PA, pp. 66-90
24
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
25
-------
DEMONSTRATION OF TECHNOLOGIES TO REMOVE CONTAMINATION
FROM GROUNDWATER
Kent M. Hodgson and LaPriel Garrett
Westinghouse Hanford Company
Richland, WA 99352
ABSTRACT
The Westinghouse Hanford Company has been testing various technologies
for decontaminating groundwaters and liquid effluents. The results of pre-
liminary testing of three technologies are reported. The technologies are
iron coprecipitation/filtration, supported liquid membranes, and reverse
osmosis. The processes were tested to determine their capability to remove
uranium, chromium, nitrates, and technetium. All processes removed contamin-
ants to less than maximum contaminant limits. The secondary waste volumes
were estimated for each process. The supported liquid membranes secondary
waste volume was the smallest, followed by iron coprecipitation, and the
largest volume was created by the reverse osmosis process.
INTRODUCTION
The Hanford Site at Richland,
Washington, is operated for the
U.S. Department of Energy (DOE) by
the Westinghouse Hanford Company
(Westinghouse Hanford). Until re-
cently, an acceptable method for
disposing of water containing small
amounts of contaminants was to dis-
charge the water to underground
cribs. Forty years of operation at
the Hanford site has resulted in the
contamination of some of the
groundwaters with such compounds as
uranium, chromium, technetium, and
nitrates. The current goal of
Westinghouse Hanford and the DOE is
to cease disposal to the soil column
and to investigate the removal of
the contaminants from the ground-
waters. The diverse nature of the
processes at the Hanford Site has
created groundwaters at different
locations with different contami-
nants, thus increasing the difficul-
ty of implementing remedial actions
in a cost-effective manner.
Westinghouse Hanford has a
program to demonstrate and develop
technologies that will remove con-
taminants from groundwaters and to
determine the volume of secondary
waste that will be generated so that
it can be minimized. Reverse os-
mosis and coprecipitation with iron
are two water treatment technologies
that are being tested with ground-
water samples to demonstrate the
purity of water that can be pro-
duced. Initial testing is also
designed to identify potential
problems and to provide information
to help evaluate these processes.
In addition, a new, emerging
technology, supported liquid mem-
26
-------
branes (SLM), is being developed.
The SLMs have the potential to be a
very powerful tool in restoring
groundwaters to their uncontaminated
state.
A primary criterion for a
potential treatment process is that
it produce a relatively small amount
of secondary waste. One purpose of
this testing was to determine the
secondary waste volume generated by
each of these processes.
The capability of the processes
that were tested to remove contam-
inants from groundwater will be
evaluated by comparing final
contaminant concentrations to the
maximum contaminant limits (MCL).
Iron Coprecipitation/Filtration
Iron coprecipitation is a pro-
cess that is used in the Uranium
Mill Tailing Remedial Action (UMTRA)
program to remove radium, uranium,
and other contaminants from the
surface runoff wastes generated
during remedial action (5). It is
also used at the Oak Ridge Y-12
Plant to remove uranium from
nitrate-containing wastes (2).
Iron is added to the stream and
then precipitated with the contami-
nants when the pH of the solution is
raised by the addition of lime or
sodium hydroxide.
Once the precipitation has
occurred, the contaminant-containing
solids must be separated from the
water. This can be done using micro-
filtration as at the UMTRA site at
Lakeview, Oregon, or by settling as
used at the Oak Ridge Y-12 Plant.
Coprecipitation is a process that
removes metal ions, however, it will
not remove nitrate ions, which are
a serious contamination problem in
some of the groundwaters at Hanford.
Supported Liquid Membranes
The SLM technology has been
investigated on a laboratory scale
for the past 20 years and on a very
limited pilot scale for the past 10
years. There are two types of
liquid membranes: emulsion mem-
branes and SLM. Only tests of SLM
have been performed at Westinghouse
Hanford.
An SLM is an organic phase that
is held by capillary forces within
the pores of a microporous poly-
meric membrane. A detailed descrip-
tion of SLM can be found in Danesi
et al. (3).
SLM have several advantages
over conventional solvent extraction
and water treatment technologies as
fol1ows:
o Very low solvent requirements
o Improved selectivity
o Reduced cross contamination
o Recovery of species present
in low concentrations
o Low secondary waste volumes
o Nitrate removal.
The SLM development at Westing-
house Hanford is a joint effort with
the Argonne National Laboratory
(ANL) and has focused on the problem
of removing uranium and nitrate from
groundwater. An extraction system
for removing uranium has been iden-
tified and tested at ANL (1, 2); the
results of a demonstration of this
system using a 2.2 square meter
(mz) SLM module and a sample of
contaminated groundwater are repor-
ted here.
Reverse Osmosis
Reverse osmosis is a separation
technology used for very difficult
separations (i.e., salt from water)
27
-------
and can produce highly purified
water. However, the rejection of
salts is a function of the salt
concentration in the feed. In order
for reverse osmosis to compete as a
process for cleaning groundwater, it
must produce a small secondary
waste. This means that the feed
will become very concentrated in
dissolved solids, and the percent
rejected may decrease. The goal of
these tests is to reduce the con-
taminants to below MCL. These
levels are very low; therefore, a
small decrease in rejection of a
contaminant may cause that contam-
inant to exceed the drinking water
standards in the permeate.
The combined disadvantage and
advantage of reverse osmosis is that
it removes all of the ions present.
This is a disadvantage because the
secondary waste volume is increased
by ions, such as sodium and calcium,
that do not need to be removed. It
is an advantage because ions, such
as nitrate, are also removed. Most
of the contaminated groundwaters
contain nitrate in excess of the
MCL.
Secondary Waste
The volume of secondary waste
that is generated in the cleanup of
a groundwater or plant effluent is
very important to the economic
viability of a process. This is
especially true when the secondary
waste must be treated as a hazardous
waste or mixed hazardous/radioactive
waste and disposed of in accordance
with applicable State and Federal
laws. Therefore, the testing of
these processes is not only to
evaluate their capability to reduce
contaminants to MCL, but to help
estimate the amount of secondary
waste that will be generated during
processing.
The secondary waste volume for
each process that is tested will be
one of the process characteristics
used to determine the process with
the best capability to economically
decontaminate groundwater.
PURPOSE
Development is needed to provide
engineering and management with the
information necessary to select
processes that will decontaminate
groundwater while generating a
minimum amount of secondary waste.
The diverse nature of the con-
taminated groundwaters and liquid
effluents make the testing of
various process options necessary
to be able to select the process or
processes with the capabilities to
reduce the contaminants to the
required levels. The levels used
for evaluation were the MCL. The
MCL for the contaminants of concern
in this study are given in Table 1.
Table 1. Maximum Contaminant
Limits (4).
Contaminant
Maximum
contaminant limits
Technetium-99
Uranium
Chromium
Nitrate
900 pCi/L
10 ppb
50 ppb
45 ppm
APPROACH
The three technologies descri-
bed above were tested using con-
taminated groundwater. The scope of
the testing was limited to demon-
strating the capabilities of the
technologies and to providing the
information necessary to allow the
estimation of the secondary waste
volumes.
28
-------
Iron Coorecicitation/Filtration
The iron coprecipitation/fni-
tration tests were performed on
three contaminated groundwaters.
The groundwaters and the contami-
nant levels are given in Table 2.
Table 2. Contaminant Levels in
Groundwater Used in Testing.
~ NT=CyaT"
Ground- U Cr Tc trate nide
water (ppb)(ppb)(pCi/L) (ppm) (ppm)
No. 1 3460 -- 786 38
No. 2 140 190 -- 273
No. 3 -- -- 4300 368 2.2
Samples of these groundwaters
were placed in 55-gallon drums and
transported to the laboratory for
testing. The apparatus used for
testing consisted of a feed tank, a
filtrate tank, a pump, and a cross-
flow filter that was a 1-inch tub-
ular polyfluorotetraethylene micro-
filtration membrane 3.7 meters (m)
long. Several tests were conducted
with each groundwater to try to
identify the conditions under which
maximum decontamination could be
realized.
Supported Liquid Membrane
The supported liquid membrane
process was tested with groundwater
number one. Two separate tests were
conducted using an apparatus that
contained two membrane modules.
Each membrane module contained 2600
microporous hollow fibers. The
hollow fibers were made of micro-
porous polypropylene and had an
inside diameter of 0.6 millimeters
(mm). The internal surface area of
each module was 2.2 m2. One module
was used to remove uranium from the
groundwater and the other module was
used to remove nitrate and tech-
net i urn.
The SLM process for removing
uranium from groundwater was devel-
oped at ANL (1, 2) for Westinghouse
Hanford. This process uses an
extractant that is 0.1 molar (M)
bis(2,4,4-trimethylpentyl)phosphinic
acid (H[DTMPePA]) in n-dodecane.
The H[DTMPePA] is the active reagent
of the commercially available Cyanex
272j. The H[DTMPePA] was selected
because of the high distribution
ratio for uranium and its selec-
tivity for uranium over calcium and
iron. The strip solution used in
the uranium removal process is 0.1M
1-hydroxyethyl-1,1 diphosphonic acid
(HEDPA).
The SLM process for removing
nitrate and technetium uses 0.1M
primene JM-T2 in n-dodecane as the
extractant. The primene JM-T is a
product of the Rohm and Haas Com-
pany. The strip solution used was
0.5M sodium hydroxide (NaOH).
In the first test, 50 gallons of
water were recirculated through the
uranium removal module. In the
second test, the treated groundwater
from the first test was recirculated
through the nitrate/technetium
removal module.
Reverse Osmosis
The feed to the reverse osmosis
test was groundwater from the same
location as that used in the sup-
ported liquid membrane testing
(1) Registered Trademark American
Cyanamid Company.
(2) Registered Trademark Rohm and
Haas Company.
29
-------
(groundwater 1). The reverse
osmosis system consisted of two feed
tanks, a pump, a reverse osmosis
membrane, assorted pressure gauges,
flowmeters, a conductivity sensor, a
temperature sensor, and valves. The
membrane that was used in this test
was a Filmtec FT-30, constructed of
a thin film composite polyamide. It
is a spiral wound cartridge and the
dimensions are 1x16 centimeters (cm)
with 2.1 m2 of membrane area. The
feed was to be concentrated until
precipitation occurred or until the
water flux dropped to 60% of the
original value.
PROBLEMS ENCOUNTERED
The determination of contam-
inant concentrations in the ground-
water samples after substantial
amounts of the contaminants had been
removed was a problem. This is true
for both the radioactive and non-
radioactive contaminants. One way
this was countered was to take large
samples and to concentrate them
prior to analysis.
RESULTS
Iron Coorecipitation/Filtration
Uranium, chromium, and cyanide
were successfully removed from the
three groundwaters.
The results of these tests
(summarized in Table 3) indicate
that uranium and chromium can be
reduced to below the MCL in
groundwater. However, the
treatment scheme for each groun-
dwater will be different. In treat-
ing groundwater 2, sodium
metabisulfite was added to the water
prior to precipitation to reduce the
hexavalent chromium to trivalent
chromium. A treatment with sodium
hypochlorite was used with ground-
water 3 to destroy the cyanide.
In these tests, one contaminant
was targeted for removal from each
stream and in general was reduced to
below the MCL for that contaminant.
However, in several cases there were
other contaminants in the stream
that were not reduced to below their
MCL. For example, the chromium in
groundwater 2 was removed, but the
uranium present was not reduced to
less than the MCL. Through further
testing, it should be possible to
modify the process to remove most of
the contaminants. The contaminant
present in most of the contaminated
groundwaters that cannot be removed
by this process is nitrate.
Supported Liquid Membranes
The goal of runs with ground-
water 1 was to reduce the uranium,
nitrate, and technetium to less than
their MCL. The MCL are given in
Table 1. In the first run, the
uranium was removed using 0.1M
bis(2,4,4-trimethylpentyl)phosphinic
acid (H[DTMPePA]) in dodecane. The
results of these tests are given in
Table 4. The uranium concentration
in the groundwater was reduced from
3460 ppb to less than the MCL of
10 ppb in less than twelve hours.
The technetium, nitrate, sulfate,
and chloride concentrations did not
change during the run.
The second run performed was to
remove nitrate and technetium from
the groundwater that had been
treated in the first run to remove
uranium. The membrane was 0.1M
primene JM-T in dodecane and the
strip solution was 0.5M NaOH. The
results of this test are also
presented in Table 4. The nitrate
was reduced to half its original
concentration in less than 20 hours,
while the technetium was reduced to
less than half of its original
concentration in less than 4 hours.
30
-------
Table 3. Summary of Coprecipitation/Filtration Test Results.
Contaminant
Uranium
Chromi urn
Cyanide
Percent removal
88.6-99.9
97.9-99.5
100
Final
concentration
ranae
1-7 ppb
1-4 ppb
not detectable
Maximum
contaminant
limit
10 ppb
50 ppb
N/A
Table 4. Uranium, Nitrate, and Technetium Removal Demonstration Results.
Time (Hr.)
0
4
8
12
16
20
24
36
48
strip 48
a Run #1
b Run #2
Uranium3
(DDb)
3,460
257
29.9
7.3
2.1
1.4
1.3
0.97
1.31
31,200
Technetium^
(oCi/U
786
361
139
51
18
9
8
4
2
11,600
Nitrateb
(oom)
38.5
37.0
31.0
24.7
20.9
18.1
15.3
10.2
42.0
358.0
Sulfateb
(pom)
1,120
1,110
1,020
974
848
810
716
920
793
8,420
Chloride5
(ppm)
14.6
14.3
13.6
13.7
12.8
11.8
12.7
10.6
9.4
78.8
It appears that the technetium was
extracted faster than the nitrate.
The pH of the groundwater was
monitored during the run and when it
increased to 2.3, additional
sulfuric acid was added to ensure
that sufficient hydrogen ions were
present to transport the nitrate, as
nitric acid, across the membrane.
The data show that sulfate and
chloride were also transported
across the membrane.
A third run in which the
modules were connected in series to
remove the uranium, nitrate, and
technetium simultaneously was con-
ducted. The data for this run are
not presented here. However, it is
important to point out that the
strip solutions used in the third
run were the same ones used in
the first and second run. That
is, the uranium from approximate-
ly 100 gallons of groundwater was
concentrated into 4 gallons, and
the nitrate and technetium from
100 gallons was also concentrated
into 4 gallons.
Reverse Osmosis
The feed for the reverse osmo-
sis test was 770 gallons of ground-
water 1. The concentrate stream was
recycled through the membrane and
the permeate was collected and
sampled. After the concentrate
31
-------
stream had been processed through
the membrane nine times, a precipi-
tate formed that was determined to
be calcium carbonate. Hydrochloric
acid was added to dissolve the pre-
cipitate and processing was con-
tinued through five more membrane
passes. The acid addition changed
the test parameters considerably so
results will be discussed as before
acid addition and after acid
addition.
The results before acid addi-
tion (Table 5) show that uranium and
technetium were removed to well
below the HCL. The analysis of the
feed water showed 11 to 26 ppm
carbon tetrachloride. None was
detected (ND) in the permeate.
However, there was not a correspond-
ing increase in the concentrate
stream, so the results are not
conclusive. Ninety-two percent of
the nitrate was rejected, which left
a concentration of less than 3 ppm
in the permeate stream.
After processing through nine
passes the volume had been reduced
from 770 to 88 gallons. This was
calculated to be a recovery of
88.6%. The flux during the ninth
pass was 83% of the original flux.
Rejection rates decreased for
all three of the constituents
measured after hydrochloric acid was
added. This was expected since the
salt passage through the membrane is
a function of the concentration
differential across the membrane.
Usually this rate decreases
gradually and makes very little
difference in the overall percent
rejection. However, the addition of
hydrochloric acid changed the con-
centration of ions in the feed by a
considerable amount, and rejection
rates of all the constituents were
affected. This was most noticeable
in the technetium and nitrate
results (compare Table 5 with
Table 6). Concentrations in the
permeate were still below the MCL
in all cases.
After processing through four-
teen passes, the concentrate stream
had been reduced to about 31 gallons
for an overall recovery of 96%.
During processing through the four-
teenth pass, the flux decreased to
60% of the starting flux. At this
time the test was terminated.
Secondary Waste Volumes
One of the important criteria
for judging the desirability of a
treatment system is the volume of
waste generated during the treatment
process. The testing reported here
was exploratory and was not exten-
sive enough to optimize process
parameters to minimize the secondary
waste volume. However, to aid in
the evaluation of this technology,
an attempt has been made to esti-
mate, based on the testing results,
the volume of secondary waste. The
estimated secondary waste volumes
are presented in Table 7.
32
-------
Table 5. Summary of Reverse Osmosis Test Results Before Acid Addition
(Nine Passes).
Contaminant
Uranium
Technetium
Carbon tetrachloride
Nitrate
Permeate
concentration
4.9 ppb
57.3 pCi/L
ND
2.6 ppm
Percent
re.iection
99.9
95.3
--
92.1
MCL
10 ppb
900 pCi/L
5 ppb
45 ppm
Table 6. Summary of Reverse Osmosis Test Results After Processing
Through Fourteen Passes.
Contaminant
Uranium
Technetium
Carbon tetrachloride
Nitrate
Permeate
concentration
7.8 ppb
192.7 pCi/L
ND
5.6 ppm
Percent
re.iection
99.8
83.9
--
82.0
MCL
10 ppb
900 pCi/L
5 ppb
45 ppm
Table 7. Comparison of Secondary Waste Volumes for Iron Coprecipitation,
Supported Liquid Membranes, and Reverse Osmosis.
Process
Iron coprecipitation
Supported liquid membranes
Reverse osmosis
Secondary waste volume
0.4 gallons/1000 gallons of feed
0.1 gallons/1000 gallons of feed
114 gallons/1000 gallons before acid
was added
40 gallons/1000 gallons with acid
addition
33
-------
ACKNOWLEDGEMENTS
The support and help of Dr.
E. P. Horwitz and Dr. R. Chiarizia
of ANL and Dr. Eric Tiepel of
Resource Technologies Group are
appreciated. The untiring efforts
of Mr. D. E. Gana are also acknow-
ledged.
REFERENCES
1. Chiarizia, R., Horwitz, E. P.,
"Extraction of Uranium (IV)
with Cyanex-272, to be publis-
hed in Solvent Extraction and
Ion Exchange.
2. Chiarizia, R., Horwitz, E. P.,
"Study of Uranium Removal from
Groundwater by Supported Liquid
Membrane," to be published in
Solvent Extraction and Ion
Exchange.
3.
4.
5.
Danesi, P. R., Horwitz,
E. P., Chiarizia, R., and
Vandegrift, G. F., 1981,
"Mass Transfer Rate Through
Liquid Membranes: Interfa-
cial Chemical Reactions and
Simultaneous
Controlling
Separation
Technology.
Diffusion as
Permeability
Factors,"
Science and
16(2), pp 201-211.
EPA, "Water Pollution
Control; National Primary
Drinking Water Regulations;
Radionuclides; Advance
Notice of Proposed
Fuelmaking," Federal
Register. September 30,
1986.
Tran, T. V., "Advanced
Membrane Filtration Process
Treats Industrial Waste
Water Efficiently," Chemical
Engineering Progress. March
1985, pp. 29-33.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
34
-------
SOIL DECpNTAMiNATION WITH EXTRAKSOL™
Jean Paquin and Diana Mourato
SANIVAN GROUP
Montreal, Canada
H1K4E4
ABSTRACT
Polychlorinated biphenyls, polyaromatic hydrocarbons, oils, pentachlorophenols have been
succesfully extracted from clay-bearing soil, Fuller's earth, oily sludge, activated carbon and
gravel by the one ton per hour Extraksol™ unit.
The Extraksol™ process is a mobile decontamination technology which treats unconsolidated
materials by solvent extraction. Treatment with Extraksol™ involves material washing,
drying and solvent regeneration. Contaminant removal is achieved through
desorption/dissolution mechanisms. The treated material is dry and acceptable to be
reinstalled in its original location.
The process provides a fast, efficient and versatile alternative for decontamination of soil and
sludge. The organic contaminants extracted from the matrix are transferred to the extraction
fluids. These are thereafter concentrated in the residues of distillation after solvent
regeneration. Removal and concentration of the contaminants ensures an important waste
volume reduction.
This paper presents the process's operational principles and the steps involved in Extraksol's
development with results of the pilot tests and full-scale demonstrations.
INTRODUCTION
Extraksol™ is a mobile decontamination
process which extracts organic
contaminants from unconsolidated materials
such as soil, sludge, gravel, etc. Polar and
non-polar contaminants are extracted from
the soils by solvent washing through
desorption / dissolution mechanisms.
Treatment with Extraksol involves material
washing and drying, which generates a
decontaminated, dry soil, ready to be
returned to its original location.
Extraksol™ has been developed and
commercialized by the Sanivan Group to
complement its hazardous waste treatment
technologies. Initially developed to extract
polychlorinated biphenyls (PCBs) from
contaminated soil, Extraksol™ has since
demonstrated to be an efficient sorution to
recycle sand, gravel, mixed soil and sludge
contaminated with oils, greases,
polyaromatic hydrocarbons (PAHs),
chlorinated organics such as
pentachlorophenols (PCPs) and other
common organic pollutants.
35
-------
Extraksol™ has been designed as a closed
system (no gaseous or liquid discharges)
which considerably reduces the volume of
contaminants. The solvents used are
non-chlorinated, non-toxic and non-
persistent. Solvent regeneration is an
integral part of Extraksol™ and is carried
in parallel to the extraction activities.
Solvent regeneration allows to associate the
benefits of soil decontamination to the
advantages of contaminant volume reduction.
Treatment of soil by extraction of the
contaminants with a non-toxic,
non-persistent solvent, is an elegant
approach which provides an environmental
recommendable solution; the contaminants
are removed and contained without need to
destroy the soil. Furthermore, since soil
drying is part of the process, the treated,
dry soil can often be returned to its
original location after mixing with top-soil
for revegetation.
Treatment of contaminated soil by solvent
extraction has been first applied on an
industrial scale in the Netherlands in 1983
and 1985. Large, fixed extraction plants
were built to solve the problems of
hydrocarbons and cyanides. Plants such as
the Heijmans Milieutechniek B.V., HWZ
Bodemsanering B.V., Ecotechniek B.V.,
Heidemij Milieutechniek B.V. and Mosmans
Milieutechniek B.V. combined the techniques
of coagulation/flocculation, sedimentation,
flotation, high pressure jets, thermal
washing and froth flotation, to
decontaminate 10 to 30 tons of soil per
hour (1, 2).
In North America, the tendency is to develop
flexible, mobile systems which can treat
the materials on-site. Available techniques
employ organic solvents (both as low
pressure liquids and critical fluids),
amines, supercritical carbon dioxide as
extraction fluids (3, 4, 5).
This paper describes the Extraksol™
process and presents the objectives and
results of the pilot-scale work and of the
various demonstrations conducted with the
one ton per hour Extraksol™ unit on PCB
containing soil as well as on other organic
contaminants and matrices.
PURPOSE
In Eastern Canada, there is no alternative
for PCB contaminated soil other than
containment and securisation. Incineration
of PCBs is not yet an accepted solution
whereas biodegradation techniques are not
available.
Solutions exist for other priority
contaminants, but these are often not
applied due to high treatment costs, NIMBY
effects, extensive treatment periods, local
regulations, etc. Often non-PCB sites will
be rehabilitated by landfilling or
encapsulating the wastes with or without
previous fixation. To reduce costs,
clean-up is often carried-out through
solutions which are not really
environmentally acceptable.
Based on these facts, the Sanivan Group had
identified, in 1987, a pressing need for
the development of an efficient, mobile,
non- destructive, environmentally
acceptable technology to treat
PCB-contaminated soil and other
non-consolidated materials.
Market studies had also revealed the need
for a non-specific technology which would
be flexible enough to extract a large
number of organic contaminants from a
wide range of solid matrices (soil, sand,
gravel, sludge, stones, etc.)
APPROACH
The Extraksol™ process was developed
through 5 distinct steps;
36
-------
1 -Laboratory-scale tests were conducted to
develop the process's chemistry and
better understand the kinetics of the
extraction.
2 -These were followed by pilot-scale tests
conducted on PCB-contaminated soil.
These tests validated the results obtained
in the smaller scale but also helped to
better define the range of application of
the process as well as the most desirable
operating parameters. The influence of
the matrix on the extraction efficiency
was also evaluated.
3 -A one ton per hour Extraksol™ unit was
constucted from the conclusions derived
in the pilot scale tests. The unit was
constructed by ChemVac, a sister
company, part of the Sanivan Group.
4 _A series of tests and demonstrations were
conducted on the unit's capability to
decontaminate PCB containing soil.
Troubleshooting of the unit and process
upgrading was also performed during this
period.
5 -After having demonstrated the unit's
capacity to extract PCBs from soil, tests
were done to establish the application
range and limits of the system. A series
of demonstrations were conducted to
evaluate the capacity of Extraksol™ to
extract a variety .of contaminants from
different matrices.
c) effect of contact time between the
extraction fluid and the soil;
d) effect of different solvent ratios on
treatment efficiency.
2 ) Description of the tests
a) 5 liters of homogenized soil were mixed
for 10 min at 30 rpm with 15 liters of the
extraction fluid. Mixing was carried
within a cement-mixer;
b) After agitation, both liquid and solid
phases were transferred into a settling
chamber. After separation of both phases,
solvent and soil samples were taken for
analysis. The solvent was discarded and the
soil transferred to the cement-mixer for a
second extraction cycle;
c) Steps a and b were repeated as required.
The number of extraction cycles applied on
each sample was computed. After each
extraction cycle the soil and solvent were
analyzed for PCBs and oils & greases.
d) After extraction, 0.5 L of the treated
soil were weighed and transferred into a
sealed, 2 L neck glass bottle. Evaporation
of the solvent was carried at a vacuum
equivalent to 20 inches of mercury while
rotating the flask at 10 rpm on a rotary
evaporator.
Description of the pilot-scale work
1) Objectives
a) establish Extraksol's operating
parameters; type of solvent, type of soil that
can that can be treated, type of mixing
required and appropriate number of
extraction cycles;
b) verify the efficiency of the process to
extract low concentrations of PCBs from soil
heavily contaminated with oils and greases;
Description of the full-scale work
1) Process description
A schematic diagram of the one ton per
hour Extraksol™ unit is presented in
Figure 1, whereas the process's flow
sheets are presented in Figures 2 and 3.
The process consists of the following
phases:
37
-------
SOIL WASHING
After introducing 8 or 10 drums of soil (or
other contaminated solid material) within
the extraction vessel, the extractor is
purged and the washing cycle is initiated.
During the soil washing phase, clean solvent
is continuously pumped and withdrawn into
and from the extractor, creating a dynamic
system (Figure 2). To further enhance
soil-solvent contact, the extractor is slowly
rotated on its axis. At later stages of
treatment, used solvent is recycled.
The circulating fluid migrates through the
soil and dissolves/desorbs the contaminants
present in the matrix according to the
affinity of the solvent for the contaminant.
The contaminant is transferred from the
soil matrix to the solvent and is carried out
of the extractor with the fluid. The used
solvent is transferred to a storage tank.
The extraction phase is continued until the
soil is thought to be decontaminated. Visual
analysis of the solvent sampled at the
extractor's outlet provides an indication of
the state of decontamination. The washing
time varies from 1 hour to 2 hours
depending on the type and concentration of
contaminant within the soil to be
decontaminated.
After the soil washing cycles are
terminated, the solvent is withdrawn from
the extractor and transferred to the
contaminated solvent tank.
SOLVENT REGENERATION PHASE
Solvent regeneration is conducted in
parallel to the soil washing phase.
Regeneration is carried by distillation and
takes advantage of the differences in
evaporation temperatures of the extraction
fluids and the contaminants. The
contaminants are concentrated and
recovered as residues of distillation.
The solvent regeneration phase is an
integral part of the Extraksol™ process
which considerably reduces the cost of
decontamination. Since the solvent is
regenerated to its original composition, it
can be reused without limitation and
without the need of additives.
SOIL DRYING PHASE
After treatment with Extraksol™, the soil
is dry and can often be relocated in its
original location.
Soil drying consists of removing the
residual solvent from the soil in a closed
system operation (Figure 3) .
The drying phase extends from 1 hour to 1
1/2 hours depending on the type of matrix
to be treated.
2) Description of the treatments
The demonstrations were conducted
according to Extraksol's operational
procedures. Further details on the
demonstrations are presented in the result
tables (Tables 2 to 5).
PROBLEMS ENCOUNTERED
Extraksol™ was conceived, designed,
constructed and is now being operated by
the Sanivan Group. The system is unique
and was developed without background
information to be used as reference.
Troubleshooting of the process was tedious
and improvements to the unit had to be
carried out in a stepwise fashion as
technical problems arouse whilst
operating the full-scale unit. Major
process improvements have been added to
the full-scale unit during the last year, to
improve its treatment's efficiency and to
widen the types of matrices and
contaminants treatable by Extraksol™.
38
-------
RESULTS
2)Results of the
demonstrations
full-scale
1)Results of the pilot-scale tests
The results generated from the pilot-scale
tests described above are presented in Table
1. The conclusions from these tests are the
following:
a) Although the equipment and procedures
used could not provide optimal extraction
conditions, the percentages of removal
achieved in most tests were higher than
80% with some results higher than 95%.
b) From the tests, solvent mixture #2 has
better removal efficiency for both PCB and
O&G than mixtures #1, #3 or #4. Solvents
#3 and #4 are efficent for O&G extraction,
but do not extract PCBs in a similarly
efficient fashion. This improved efficiency
may be explained by the decrease in mass
transfer resistance induced by the water
adsorbed onto the soil particles. The slight
polarity of solvent #2 would overcome such
a water interference.
c) The PCB concentration in the solvent
decreases after each extraction cycle and
tends to be asymptiotic after 3 extraction
cycles (refer to Figure 4).
d) The data presented show that there is no
apparent benefit in increasing the contact
period between soil and solvent for more
than 10 minutes.
e) The removal of PCB is not hindered by
the large concentration of oil and grease in
the soil. Also, a good extraction efficiency is
obtained even at low PCB concentrations.
f)The extraction efficiency and the
mechanisms of extraction appear to be
comparable in both clays and sands. The
granulometry of these materials could
however interfere in a full-scale
treatment.
The results of the full-scale
demonstrations are summarized in Tables
2, 3, 4 and 5.
a) PCB extraction
Full-scale tests confirmed the conclusions
derived from the pilot tests, ie. solvent
mixture #2 is more efficient for both PCB
extraction (Table 2) and oil and greases
removal (Table 3). Again, the slight
polarity of solvent #2 would allow
penetration into the water layer with
subsequent dissolution/ desorption of
adsorbed contaminants. This mechanism
becomes important when decontaminating
surface active soils such as clay or clayey
soil.
The differences in efficiencies observed
between solvents #1 and #2 for PCB
extraction in mixed soil was confirmed by
the concentration of contaminant within
the solvents with time (Figure 5). Figure
5, also shows that the rate of contaminant
removal is considerably reduced as the
contaminant concentration in the soil is
reduced.
This behaviour could be accounted for
slower diffusion rates induced by lower
concentration differentials between the
soil and solvent, which indicates that an
equilibrium between the soil and solvent
has been reached. Furthermore, after the
free contaminants have been flushed with
the solvent, different extraction
mechanisms become responsible for soil
decontamination. Extraction of adsorbed or
complexed contaminants involves slower
mechanisms where partition coefficients
and contaminant affinities become the
driving force for soil decontamination.
39
-------
b) Oils and Greases Extraction
!) Mixed soil
The first 3 results presented in Table 3
refer to mixed soil contaminated with both
PCBs and transformer oils. Extraction of
the two contaminant sources was done
simultaneously. Both PCB and O&G
extraction performance were good with
solvent #2, with removal rates greater
than 90%.
ii) Refinery sludge and soil
The Extraksol™ process was shown to be
very efficient to extract hydrocarbons from
clayey soil and sludge originating from
refinery sites. These materials are often
oily and sticky with O&G concentrations
exceeding 5% (50,000 ppm). The system
has the capacity to declassify the sludge but
also to have the treated sludge be
relocalized as "soil". After removal of the
contaminants, the sludge is a dry, powdery
material which can not be distinguished
from the refinery's clayey soil.
Extraksol™ is a batch process which offers
the flexibiity to treat different types of
materials with a large range of
granulometries and fabrics. The batch
operation also provides a large flexibility of
operation; no process adjustement is needed
to select the number of extraction cycles
required for a particular waste to reach a
specific residual contaminant concentration.
For example, the residual O&G
concentration aimed when treating the
refinery sludge was of 5,000 ppm (Table
3). If required, the sludge could have been
further treated, but the costs of
decontamination would be raised with
increasing number of extraction cycles.
iii) Fuller's earth
Extraksol™ has succesfully decontaminated
and regenerated more than 30 tons of
Fuller's earth. This diatomic earth is
currently used as absorbant and filtration
media to purify oils. After use, the Fuller's
earth are very oily with O&G concentrations
exceeding 30%. Removal efficiencies
achieved were in the order of 99% after 2
hours of washing.
iv) Porous gravel and stones
Extraction of hydrocarbons from porous
materials is feasible with Extraksol™, but
the removal efficiency is lower than with
non-porous materials. The surface area of
the contaminated matrix appears to
interfere with solvent extraction.
With removal percentages ranging from
60% to 80%, solvent circulation does not
appear to be the most efficient treatment
for porous materials. It is believed that
contaminants' extraction could have been
improved if the gravel would have soaked
in the solvent for a given time, before
initiating solvent circulation.
c) PAH extraction
Extraksol™ was shown to be an efficient
solution to decontaminate soil and sludge
bearing polyaromatic hydrocarbons
(PAH). With extraction efficiencies
ranging between 80% and 95% and
residual PAH concentrations as low as 10
ppm (Table 4), the process offers an
acceptable solution for emerging PAH
problems.
d) POP extraction
Extraksol™ has the capacity to extract a
large number of chlorinated and
non-chlorinated organic contaminants
from unconsolidated solids. Although many
contaminants such as pesticides, dioxins
and furans remain to be tested, treatment
of pentachlorophenols (PCP) by
Extraksol™ was a success. Even porous
gravel which has a tendency to be difficult
to treat for oils (refer to Table 3), was
decontaminated to non-detectable levels of
PCP.
Treatment of activated carbons have shown
that these surface active materials have a
tendency to be more difficult to
decontaminate. Nevertheless, an extraction
efficiency equivalent to 89% was obtained
40
-------
solvent #2 (Table 5). After
treatment, the activated carbons retained
their physical characteristic and could be
re-used as a pre-treatment carbon.
Regeneration of carbons by solvent washing
is therefore possible with Extraksol™,
reducing the cost of disposal and of new
material replacement.
CONCLUSIONS
Enough work has been done with the one ton
per hour Extraksol™ unit to confirm that
the process can effectively extract oils,
PCB, PAH, and PCP from soil and generate a
"clean", dry soil, which once mixed with
top-soil for revegetation can be returned to
its original location.
Decontamination of sand, mixed soils clayey
soil, sludge, gravel, activated carbon and
stones have demonstrated that the process is
very flexible and can be adapted to solve a
range of environmental problems.
The flexibility and mobility of the 1
ton/hour unit and its 3 day set-up period
makes is a perfect process to be used on
small projects with a maximum of 300 tons
of material to be treated.
The Sanivan Group is planning to build a
larger Extraksol unit which will have the
capacity to treat 6 to 8 tons per hour. This
mobile unit will be designed on the same
principles of operation as the smaller unit.
This larger unit will be operated by 2
• operators as the smaller unit.
The cost of treatment with the larger unit
will range from $100 to $200 per ton of
material to be treated, depending on the type
of matrix, the contaminants' concentration
and residual contaminant concentrations.
REFERENCES
1. Assink, J. W. Extractive Methods for
Decontamination; a General Survey and
Review of Operational Treatment
Installations. In Contaminated Soil; J.
W. Assink, W. J. van den Brink (eds.)
Martinus Nijoff Publ. Dordrecht.
1986.
2. Assink, J. W. Extractive Methods for
Soil Decontamination; Operational
Treatment Installations in the
Netherlands. Proceedings second Int'l
Conf. on New Frontiers for Hazardous
Waste Management. Pittsburgh, USA.
Sept. 1987.
3. Dooley, K. M. et al. Supercritical Fluid
Extraction and Catalytic Oxidation of
Toxics from Soils. Proceedings second
Int'l Conf. on New Frontiers for
Hazardous Waste Management.
Pittsburgh, USA. Sept. 1987.
4. Irvin, T. R. et al. Supercritical
Extraction of Contaminants from
Water and Soil with Toxicological
Validation. Proceedings second Int'l
Conf. on New Frontiers for Hazardous
Waste Management. Pittsburgh, USA.
Sept. 1987.
5. Moses, J. and Abrishamian R. Case
Study: SITE Program puts critical
fluid solvent Extraction to the Test.
Hazardous Waste Mgt. Mag. Jan-Feb.
30-32. 1988.
41
-------
c
13
H
03
v.
•*-«
X
LLJ
=3
O
JZ
CD
a.
c
o
c
o
CD
o
o
"•J2
CO
S
o
o
CO
o
1_
3
O)
iZ
O
CO
i—
X
LLJ
DO
CD
CO
2
D)
CO
a.
E
CO O
o cs
CD O
r
o
H—
03
Q
i
-------
Figure 2, Schematic of the Extraksol™ Process
Extraction Mode
-a-
soil within
extractor
Soil Washing Phase
contaminated
solvent
tank
Solvent Regeneration Phase
distiller
+ condenser
Figure 3, Schematic of the Extraksol™ Process
Drying Mode
hot
inert
gas
&
steam
treated soil
within
extractor
inert gaseous
gas + solvent
cold inert gas
, ,
inert gas
•f
liquid solvent
liquid
solvent
solvent
tank
43
-------
FIGURE 4, TYPICAL CONCENTRATIONS OF PCB IN THE SOLVENT
PILOT-SCALE WORK
250
Cone. PCB 200
in solvent
(ppm) 15°
100
50
0
soils' initial PCB cone. = 600 ppm
soils' final PCB cone. = 6.3 ppm
3 4
Extraction Cycles
FIGURE 5, TYPICAL CONCENTRATIONS OF PCB IN THE SOLVENT
FULL-SCALE WORK
70 r-
60
Cone. PCB 50
in solvent
(ppm)
40
' 30
20
1 0
0
1
soils' initial PCB cone. = 150 ppm
soils' final PCB cone. = 14 ppm
234
Extraction Cycles
44
-------
TABLE 1, SUMMARY OF RESULTS OBTAINED WITH EXTRAKSOL'S PILOT SCALE UNIT
TYPE OF SOIL
Clay
Clay
Sand
Mixed Soil
Mixed Soil
Mixed Soil
Mixed Soil
TYPE OF FLUID
# 1
# 2
# 1
# 1
# 2
#3
#4
REMOVAL OF OILS AND GREASES
Initial
Concent
(ppm)
1.970
8,040
470
13,890
14,400
34,300
21,540
Final
Concent
(ppm)
847
590
220
2,270
1.210
1,440
1,880
% Removal
(%)
56.9
92.7
53.2
87.7
92.0
96.0
91.0
REMOVAL OF PCB
Initial
Concent
(ppm)
7,925
2,055
600
3.6
5.3
5.2
4.8
Final
Concent
(ppm)
2.080
48.8
6.3
0.69
0.70
1.0
1.1
% Removal
(%)
73.8
97.6
98.9
89.0
87.0
81.0
77.0
TABLE 2, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR,
EXTRAKSOL UNIT - PCB REMOVAL
TYPE OF
SOIL
clay-bearing
clay-bearing
clay-bearing
TYPE OF
FLUID
# 2
# 1
#2
Initial PCB
Concent.
(ppm)
150
163
54
Final PCB
Concent
(ppm)
1 4
28
4.4
% Removal
(%)
91.0
82.0
92.0
45
-------
TABLE 3, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR
EXTRAKSOL UNIT - O&G REMOVAL
TYPE OF
SOIL
clay-bearing
clay-bearing
clay-bearing
refinery clayey soil
refinery oily sludge
refinery oily sludge
refinery oily sludge
refinery oily sludge
Fuller's earth
Fuller's earth
Fuller's earth
Fuller's earth
pulp & paper porous gravels
pulp & paper porous gravels
TYPE OF
FLUID
#2
# 1
# 2
#2
#2
# 2
# 2
#2
# 2
#2
# 2
# 2
# 2
# 2
Initial O&G
Concent.
(ppm)
1 ,801
1,789
600
15,000
49,000
72,000
73,000
70,000
366,000
447,000
313,000
332,000
10,000
1 ,040
Final O&G
Concent.
(ppm)
1 82
1 66
80
800
4,200
2,000
4,800
340
4,200
5,500
3,700
4,000
3,690
207
% Removal
(%)
90
82
92
95
91
97
93
99
99
99
99
99
63
80
TABLE 4, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR
EXTRAKSOL UNIT - PAH REMOVAL
TYPE OF
SOIL
refinery clayey soil
refinery oily sludge
refinery oily sludge
refinery oily sludge
refinery oily sludge
TYPE OF
FLUID
# 2
# 2
# 2
# 2
# 2
Initial PAH
Concent.
(ppm)
332
81
240
150
1,739
Final PAH
Concent.
(ppm)
55
1 6
1 0
1 9
1 30
% Removal
(%)
83
81
96
87
92
46
-------
TABLE 5, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR
EXTRAKSOL UNIT - PCP REMOVAL
TYPE OF
WASTE
porous gravel
porous gravel
porous stones
activated carbon
TYPE OF
SOLVENT
#2
#2
# 2
#2
Initial PCP
Concent.
(ppm)
8.2
81.4
38.5
744
Final PCP
Concent.
(ppm)
<0.82
<0.21
19.5
83
% Removal
(%)
>90
>99.7
50
89 '
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
47
-------
ORGANIC WASTE TREATMENT WITH ORGANICALLY MODIFIED CLAYS
Jeffrey C. Evans, Ph.D., P.E.
Associate Professor of Civil Engineering
Stephen E. Pancoski
Research Associate
BUCKNELL UNIVERSITY
Lewisburg, Pennsylvania 17837
and
George Alther
President
BENTEC, INC.
Ferndale, Michigan 48220
ABSTRACT
A relatively new technology for the retention and adsorption
of organic pollutants involves the use of, organically modified
clays. The accessible surfaces within the crystalline structure
of clay minerals are chemically modified with organic derivatives
such as alkylammonium ions. This clay modification imparts an
organophilic character to the clay. The clay surface is thus
rendered suitable to adsorb organic molecules.
Clays such as bentonite have been used for many years as
pond and landfill liners because of their low permeability to
water. The low permeability of these clays has been shown to be
affected adversely by fluids containing organics. Organically
modified clays allow an extension of clay barrier technology into
organic systems. As a result of the affinity between organically
modified clays and organic pollutants, applications for their use
in waste treatment and remediation have evolved. These
applications include:
1) Waste stabilization - organically modified clay is mixed
with organic wastes and then cementing agents to produce a
solidified matrix, resulting in reduced Teachability of the
organics from the stabilized matrix.
2) Water treatment - organically modified clay is used for
treatment of ground and surface water to remove organic
constituents within the waste stream.
3) Spill control - organically modified clay can be distributed
on water or soil surfaces to sorb organic liquids as necessary
for spill control.
4) Tank farm liners - organically modified clay can be used in
liner systems for fuel oil storage tanks.
5) Hazardous waste liner systems - organically modified clay
can be used as a barrier layer component within liner systems for
hazardous waste storage and disposal sites.
48
-------
INTRODUCTION
The use of organically
modified clays in hazardous
waste management applications
offers a significant new and
untapped potential. These
.clays-may be used in the stabi-
lization of organic wastes and
organically contaminated soils,
for waste water treatment, for
oil spill control, for liner
systems beneath fuel oil stor-
age tanks, and as a component
within liner systems of haz-
ardous waste storage treatment
and disposal facilities.
Organically modified clays
(organophi1ic clays) may be
employed in each of these sys-
tems to adsorb organic waste
constituents, enhancing the
performance of these applica-
tions. This paper first
describes the nature of
organophi lie clays, and then
discusses their application in
each of the five areas.
ORGANICALLY MODIFIED CLAYS
The production of organi-
cally modified clays begins
with the use of a natural clay
mineral. The clay minerals
most commonly used in this
process are smectites
(montmori1lonite and hectorite)
and attapulgite (palygorskite).
Since the structure of each of
the base clays differs, the
performance of the modified
clays will likewise vary.
Detailed descriptions of these
clays is found elsewhere (6,
11)
Organic Modification of Clays
The. investigation of clay-
organic interactions began over
50 years ago. An early study
reacted organic bases and their
salts with montmori11onitic
clays and presented evidence
that an ion exchange reaction
had occurred (6). Similar
early experiments using organic
chemicals in montmori11onitic
clay demonstrated that the
exchangeable inorganic cations
could be replaced by organic
cations, and that uncharged
polar compounds could enter the
inner layer region without the
release of cations (13). It
was also found that bentonite,
after reaction with certain
organic compounds, gains the
properties of swelling and
dispersing in organic fluids
(8). These studies describe
the clay-organic interactions
which impart the organophilic
characteristics upon the modi-
fied clay. Since that time,
these interactions have proven
to be effective and, in many
cases9 commercially viable in
transforming a naturally
hydrophilic clay into an
organophilic clay.
A number of chemical
interactions between the clay
and organic compound were iden-
tified in the organic modifica-
tion of clays. The primary
reactions which occur are
adsorption, intercalation, and
cation exchange. Additional
reactions include ion exchange,
an ion exchange, protonation of
organic molecules at the clay
surface, hemisalt formation,
ion-dipole coordination, hydro-
gen bonding, pi-bond ing,
entropy effects, and covalent
bonding (10). It is beyond the
scope of this paper to discuss
in detail these reactions,
which influence the organic
modification of clays.
To produce an organically
modified clay, an unmodified
clay mineral is reacted with an
organic compound. In this
49
-------
process, a cat ionic surfactant,
such as quaternary ammonium,
replaces the exchangeable inor-
ganic sodium, calcium and/or
magnesium ions on the nega-
tively charged surface of the
clay. In this reaction, the
clay's nature is converted from
a hydrophilic to an
organophilic condition. Reac-
tion of the clay with the
appropriate organic cation will
result in a modified clay which
will swell and disperse in the
presence of a variety of
organic liquids.
The organic compounds most
commonly used to modify clays
are quaternary ammonium salts.
A quaternary ammonium salt is a
form of an organic nitrogen
compound in which the molecular
structure includes a central
nitrogen atom joined to four
organic groups along with an
acid radical. They are all
considered cationic, surface
active coordination compounds
and tend to be adsorbed on sur-
faces, thus the term surfac-
tants. The most commonly
employed types of quaternary
ammonium compounds used to mod-
ify clays are dimethyl ammo-
nium, methyl benzyl ammonium,
dibenzyl methyl ammonium, and
benzyl dimethyl ammonium quats.
Manufacture
of
Organically
Modified Clays
Organically modified clays
are manufactured using either a
dry process or a wet process.
In a wet process, the unmodi-
fied clay is mixed with water,
forming a slurry. The result-
ing slurry is centrifuged to
remove inert, non-clay miner-
als. The supernatant, which
contains ultra-pure clay, is
then reacted with the specified
organic compound. The mixture
is filtered, dried and ground.
In the dry process, limited
amounts of water are first
added to the unmodified clay.
The clay is then reacted with
the organic compound in a
mixer, pug mill or extrusion
device. Finally, the reacted
material is dried and ground.
Since centrifugation is not
performed in the dry process,
some impurities still exist in
the finished clay product.
Additional detail regarding the
clay manufacture can be found
elsewhere (1, 5, 12}.
Adsorption by Organophilic
Clays
Organically modified clays
are suitable media for the
adsorption of soluble organic
compounds from dilute aqueous
solutions. This adsorption
occurs -through electro-
static/hydrogen bonding forces
at the hydrophilic sites, and
by van der Waals forces at the
organophilic sites of the
organophilic clay. The follow-
ing factors affect the adsorp-
tion of organic compounds from
dilute aqueous solutions by
organophilic clays: a) the
nature of the adsorption sites,
b) the nature of the organic
molecules to be adsorbed, c)
spatial considerations, d)
thermodynamic quantities, and
e) solubility of the adsorbate
in the solvent (4). Portions
of the organically modified
clay surface which were not
modified during the organic
modification process are still
hydrophilic. As a result,
adsorption by electro-
static/hydrogen bonding with
the hydrophilic portion of the
adsorbate molecule occurs at
these hydrophilic clay sites.
On the remainder of the clay
surface, which is organophilic,
van der Waals bonding occurs.
Since adsorption occurs at both
50
-------
types of sites on the clay sur-
face, a balance between the
organophilic and hydrophilic
clay sites optimizes the
adsorption capacity of the clay
(4).
LABORATORY TESTING PROGRAM
The adsorption capacity of
a number of commercially avail-
able organically modified clays
was quantified utilizing the
free swell, or sedimentation,
test. In each of these tests,
50 milliliters (ml) of the test
fluid were poured into a 100-mT
graduated cylinder and 2.5
grams of the clay were then
rained into the fluid. The
clay typically settled to the
bottom of the graduated cylin-
der. The free swell volume was
recorded after 24 hours. The
laboratory test procedure for
this free swell test was modi-
fied from laboratory procedures
previously developed (7). The
major modification from the
original procedure was the
reduction in the quantity of
clay and test fluid utilized.
These investigations parallel a
study published earlier in
which a single organically
modified clay type was evalu-
ated (8, 9). This swell volume
is only an indicator of the
clay's adsorption capacity, and
may not adequately quantify the
performance of the clay in the
applications discussed herein.
The clays and their manu-
facturers are listed in Table 1
along with pertinent informa-
tion regarding the manufacture
of the organophilic clays. The
organic compounds reacted with
the base clays are quaternary
ammonium salts, primarily
dimethyl di(hydrogenated tal-
low) ammonium chlorides. Tal-
low is an animal fat with each
organic molecule containing 16
to 18 carbon atoms.
Ten test fluids were used
in these studies: acetic acid,
acetone, aniline, carbon tetra-
chloride, deionized water,
diesel fuel, hexane, kerosene,
unleaded gasoline, and xylene.
All of the fluids employed in
the study were added in a con-
centrated form with each of the
clays studied. The results of
the free swell tests conducted
for these studies, along with
the density of each clay, are
summarized in Table 2.
DISCUSSION OF
RESULTS
LABORATORY TEST
Examination of the average
free swell volume enables a
comparison between the various
organically'modified clays for
a wide range of organic fluids.
Viewing the data in this way,
it is revealed that all of the
organically modified clays have
an average free swell volume
between 11 and 35 ml. The wet
process clays with the highest
average free swell volumes are
Baragel 3000 (34.3 ml),
Benathix 1-4-1 (32.1 ml) and
Tixogel SP (31.6 ml). The dry
process clays with the highest
average free swell volumes are
PC-1 (23.8 ml) and TS-55 (23.1
ml). The two unmodified clays,
bentonite and attapulgite, have
the lowest average free swell
volumes, indicating they did
not swell appreciably in the
presence of the concentrated
organic test fluids.
In summary, it is believed
that these tests provide a use-
ful way to rapidly and quanti-
tatively compare the expected
performance of organically
modified clays. Note that clay
cost must be evaluated in
comparison to performance in
51
-------
selecting a particular clay for
a particular application.
APPLICATIONS
With the ability of the
organically modified clays to
adsorb organics, a number of
applications for organically
modified clays in pollution
control and hazardous waste
site remediation have been
identified. These applications
range from those fully devel-
oped and in the marketplace to
those recently proposed. The
following section discusses
several applications for organ-
ically modified clays.
Waste Stabilization
The remediation of organi-
cally contaminated soils and
wastes employing stabilization
and solidification techniques
has become increasingly
widespread. The stabilization
process is designed to maximize
shear strength and minimize the
rate of leaching of hazardous
constituents from the stabi-
lized matrix into the environ-
ment. Conventional stabiliza-
tion techniques, such as cement
and flyash, are usually limited
to inorganic, metal-bearing
wastes. For organic wastes,
modified clays have been
employed to adsorb organic con-
stituents. Preliminary labora-
tory data indicate that
organophilic clays are effec-
tive as stabilization agents
(12).
When used in conjunction
with conventional cement-based
or pozzolanic additives, organ-
ically modified clays are
effective in reducing the
mobility of organic con-
stituents from the stabilized
matrix. Reductions in the
mobility of organics has been
demonstrated and is the subject
of several patents (2, 3).
Organically modified clays are
first mixed with the waste to
adsorb the organic con-
stituents. In this manner, the
organics are chemically bound
within the organoclay, thereby
reducing the organic interfer-
ence with the normal cement
reactions and. lattice forma-
tion. The clay, with organic
contaminants bound within the
clay structure, is then
macroencapsulated in a cementi-
tious matrix formed by a cement
or pozzolan. This technique,
which utilizes an organophilic
clay in conjunction with a
cement product, is employed by
the Silicate Technology Corpo-
ration of Scottsdale, Arizona
for the stabilization of
organic hazardous wastes. In
this technique, the leaching
potential of the organic con-
stituents is decreased through
the use of organophilic clays
as compared to techniques using
only cement or pozzolan in the
stabilization process.
Water Treatment
Organically modified clays
are used to treat organically
contaminated waste water. In
this process, the aqueous solu-
tion is filtered through organ-
ically modified clay and the
organic contaminants are
adsorbed by the clay. Unlike
activated carbon, which adsorbs
organic contaminants through
surface related phenomena,
organically modified clays
swell as the organic contami-
nants are sorbed into the clay
structure. Thus, the organic
molecules of the contaminant
preferably partition into the
organic phase of the organoclay
instead of the aqueous phase
52
-------
(2). Volatile organics are
poorly adsorbed, whereas oils
and greases are readily
adsorbed as a result of their
differing partition coeffi-
cients.
Organically modified clays
are typically used along with
other water treatment technolo-
gies. For example, when used
in a treatment system upstream
of activated carbon, the carbon
life is greatly extended as a
result of the removal of high
molecular weight organics. Two
commercially available prod-
ucts, Calgon's Klensorb 100TM
and Electrum's Organosorb, con-
tain an organically modified
clay to remove organics from a
waste stream. The organophilic
clay in these products is used
with an anthracite filter media
to provide an effective column
filtration medium in-water and
waste water treatment. The
combination of this mixed media
and granular activated carbon
adsorption facilitates the
removal of a broad range of
both soluble and insoluble com-
pounds. Shown on Figure 1, are
the influent and effluent oil
concentrations for an oily
steam condensate filtered with
a mixture of Organosorb and
anthracite filter media. The
treatment effectiveness is
demonstrated on Figure 1 and an
annual savings of about
$150,000 per year was projected
for a 500 to 600 gpm system
14). In another application,
the organically modified clay
removes non-volatiles prior to
air stripping, thereby increas-
ing the efficiency of the
system.
Spill Control
The organophilic nature of
organically modified clays make
them suited for use in spill
control applications. As
demonstrated in the laboratory
studies, these clays will
either float on or sink to the
bottom of an aqueous solution,
depending on the nature of the
organic modification. For
example, an oil spill on water
can be sorbed by the organoclay
and held within the clay crys-
talline structure for cleanup
and disposal. For a spill
where the materials sink, the
clays could likewise be
selected to sink through the
water and sorb the spilled
fluid.
Tank Liners
Fuel oil storage tanks are
typically surrounded with a
liner and berm system to
contain the • fuel oil should a
leak occur. As a result of the
impervious nature of the liners
used in these systems, they
also contain precipitation. As
an alternative to conventional
liner systems, it is proposed
that organically modified clays
be used. An organically modi-
fied clay liner would permit
precipitation to flow downward
through the liner without accu-
mulating in the containment
system. In the event of a tank
leak, the clay will swell in
the presence of the fuel and
form an impervious barrier
layer. This would prevent
migration of contaminants into
the subsurface. It may also be
possible to use organically
modified clays as secondary
containment barriers be-
neath/around underground stor-
age tanks in a similar manner.
Landfill Liners
The technology for land-
fill liners has been evolving
in recent years to include
53
-------
multimedia barrier layers to
optimize the environmental
protection. Presently, liner
systems include both geomem-
brane barrier layers and natu-
ral clay barriers. The
greatest environmental concern
at present is that both geomem-
branes and natural clays may be
subject to degradation in the
presence of organic contami-
nants. Further, organic con-
taminants may migrate through
these materials in response to
chemical diffusion gradients.
It is proposed that the liner
system include a barrier layer
composed of an organically
modified clay. In this way,
the multimedia liner system
would have superior performance
in the presence of organic con-
taminants. The sorptive capac-
ity of these materials would
significantly reduce the rate
of organic contaminant trans-
port across the liner system.
ACKNOWLEDGEMENTS
These investigations are
part of a research project
funded by the Ben Franklin
Partnership of Pennsylvania,
the Sun Refining and Marketing
Company, and the Earth
Technology Corporation. The
authors appreciate the coopera-
tion of the clay suppliers.
Appreciation is also extended
to Kate Toner, Holly Borcherdt
and Chris Bailey who provided
laboratory assistance.
REFERENCES
I. Alther, G.R., Evans, J.C.,
and Pancoski, S.E.,
"Organically Modified
Clays for Stabilization of
Organic Hazardous Wastes,"
Proceedings of Superfund
2.
4.
5.
6.
7.
8.
'88, pp. 440-445, Nov.,
TW8.
Beall , Gary W., "Method of
Immobilizing Organic
Contaminants to Form Non-
Flowable Matrix There-
from,
U.S.
Patent Number
4,650,590, March 17, 1987.
Seal! , Gary W., "Method of
Removing Organic Contami-
nants from Aqueous
Compositions," U.S. Patent
Number 4,549,966, October
29, 1985.
Cowan, C.T. and White, D.,
"Adsorption
Complexes," _
Minerals, Proceedings
by Organo-Clay
Clays and Clay
the Ninth
Conference on
Clay Minerals,
1960, ed. Ada
pp. 459-467,
Academy
N a t i o n a'l
1962.
Evans, J
S.E.,
Modified
of
National
Clays and
Oct. 5-8,
Swineford ,
National
of Sciences
Research Council,
Transportation
C. and Pancoski,
"Organically
Clays," Preprint:
Research
Board
Annual
1989.
Clay
Second
Book
Meeting , Jan.,
Grim, Ralph E.,
Mineralogy,
Edition, McGraw-Hil1
Company, 1968.
Hettiaratchi, J.P.A., and
Hrudey, S.E., "Influence
of Contaminant Organic-
Water Mixtures on
Shrinkage of Impermeable
Clay Soils with Regard to
Hazardous Waste Landfill
Liners," Hazardous Waste &
Hazardous Materials, Vol .
4, Number 4, pp. 377-388,
1987.
Jordan, John W.,
"OrganophiVic Bentonites.
I Swelling in Organic
Liquids," The Journal of
Phys ical
&
Col loidal
Chemistry,
Number 2",
Feb., 1949.
Volume
PP
294-306
54
-------
9.
10,
11
Jordan, John W.,
"Alteration of the
Properties of Bentonite by
Reaction with Amines,"
Mineralogical Magazine and
Journalofthe
lineralogicaI
Society,
Volume 28, Number 205, pp.
598-605, June 1949.
Mortland, M.M., "Clay-
Organic Complexes and
Interactions," Advances in
Agronomy, Volume22,
Academic Press, Inc., pp.
75-117, 1970.
Newman, A.C.D., and Brown,
G., "The Chemical
Constitution of Clays," in
Chemistry of Clays and
Clay Minerals, ed. A.C.D.
Newman, pp. 1-128, John
12.
13.
14.
Wiley & Sons, New York,
1987.
Pancoski, Stephen E.,
"Stabilization of
Petroleum Sludge,"
Master's Thesis, Bucknell
University, June, 1989.
Raussel1-Colom, J.A. and
Serratosa, J.M.,
"Reactions of Clays with
Organic Substances," in
Chemistry of Clays and
Clay Minerals^ed
A.C.D.
pp. 371-422, John
Sons, New York,
Newman,
Wiley &
1987.
Electrum, Inc.,
"Organosorb Cleanup of
Oily Steam Condensate in a
Petrochemical Plant,"
Company Literature,
Fairfield, Kentucky.
14
13
12
11
10
9
8-
7
6-
5-
4-
3
2
H
Detection Umlf*
I
6
"o
I
G
1
§
o
>50 >5O
x x
x x
>5.0
>5.O
I AM VI
Acceptable Effluent QunlH-y y II x
-i—i—i—i—r—T—f r •r' ii i '"i i1 "i" ....
5 10 15 20 25 3O 35 4O 45 50 55 6O 65 70 75 80 85 9O 95 1OO
Days
Figure 1. Influent and Effluent Concentrations (from Electrum, Inc.)
55
-------
I
>— i
>>
a
O
-
n
S
CO
.23
3
se
O
m
tr
S
UNLEADED
t/1
UJ
O
cc
Ul
u
UJ
*
o
Ul
ri
lu
o
u
CO
UJ
t3
GASOLINE
_i
u_
UJ
z
£
Ul
UJ
o
DC
UJ
Z
X
3:
UJ
o
%
0
UJ
z
UJ
*
o
o
DC
IU
^
J
1—
UJ
o
u
^»
i
^
G
fmtt
E
•^
^
e
~
^
E
^
E
^
E
^
S
i
Ol
eo
NO
O
o
o
N.
O
o
CO
in
o
l1*-
in
in
CM
in
o
Si
a
CO
4-*
CO
O
o
0
in
o
in
o
in
•J-
0
in
^T
o
vf
g
o
(U
I Bent on
r>-
o
o
o
KS
0
CM
in
«-
o
0
in
o
o>
o
00
0
>o
SO
o
-,
lAttatot
K,
tn
o
^
0
CM
in
CO
0
o
N.
0
0
o
in
CO
CO
in
O
t—
SB
|Attator
CO
o
c>
o
o
CM
O
CO
CM
O
CO
o
o
tn
in
*-
o
*-
tn
CO
CO
CO
g
O
CO
K,
o
o
o>
0
*
CO
°
o
o
in
0
ro
o
CO
CO
f^.
tn
0
0
!
CO
in
W
O
to
O
"
0
0
in
o
CM
o
o
Kl
in
r-
o
0
3
0
o_
4-*
CO
tj
03
in
a
o
•«*
0
CO
CO
CM
CM
O
o
o
CM
0
**
O
CO
0
•»*
0
0
*-J
X
(0
0
o
0
CO
o
o
CM
0
ro
CO
K.
CM
o
0
o
CO
0
*-
o
CO
o
NO
CM
in
0
z
r>
0
o
o
CM
0
03
o
00
o
o
in
03
0
<0
o
0
in
0
c
0)
0;
•o
£
Kl
in
CM
o
00
V,
0
-
03
o
0
5
0
o
CM
0
CM
o
0
0
03
0
Kl
o
o
o
^
in
0
o
03
w
o
in
CM
0
in
CM
0
a
o
K
O
0
0
CO
0
o
CO
CO
CO
o
ro
0
u
^
0
0
CO
CM
0
CM
CM
CO
CM
ro
o
0
o
CO
0
CM
O
O
O
l^
2
0
1—
CM
0
8
O
N-
o
R
CO
tn
ro
o
0
0
o
0
tn
o
«-
o
o
o
tn
0
o
1
3
,-
0
0
CM
0
CM
O
CM
O
T
0
o
c-
0
o
1^.
o
*n
^
0
in
in
(n
„
0
in
CM
o
a
0
in
CO
K)
O
0
o
CM
in
ro
o
o
N.
CM
-a-
0
0
0)
O)
o
X
h.
0
tn
o
o
CM
0
O
CM
CO
0
ro
o
0
0
*-
0
fM
O
*-
o
*o
o
0
CO
o
UJ
D)
O
X
CM
0
ro
o
CM
0
a
o
~
tn
o
en
0
.-
o
o
N.
ro
tn
0
m
en
o
X
o
0
o
o
(M
0
K
CO
tn
o
o
CO
NO
0
CO
o
*-
o
J^
£
0
Ul
o.
1
X
0
0
0
tn
o
o
tn
o
o
tn
o
3
O
O
in
Kl
o
tn
o
o
t—
ro
0
ex
V)
X
ro
'CM
0
CO
in
0
o
in
o
tn
o
fc
o
0
o
Kl
0
•O
O
O
00
(M
•**•
0
UJ
H-
Itixoge
CM
0
in
Kl
o
o
o
N
o
£
o
0
o
CM
0
in
o
o
0
o
in
0
a.
en
X
CM
0
Kl
O
0
Kl
0
°
O
o
o
CM
0
ro
tn
CO
0
NO
in
0
ro
8
Bentone
,-
0
03
O
S
O
CM
in
0
—
o
0
o
r-
0
0
o
o
CO
in
CM
0
X
JC
1
o
ca
56
-------
3
CO
(2
H
o
ji
rt
IN
UJ
U.
CJ
s
o
fV
UJ
Z>
o
ID
Z
X
_J
CJ
0)
c
o
5
g
5
Of
o
w
to
u
4-*
ra
u
8
00
•D
(0
Ol
c
UJ
apulgite
4-»
4-*
C
o
0)
c
o
c
O)
CO
o
•o
o
CO
z
g
c
Ot
CO
01
iS
0
_c
§
'c
i
1
«0
"8
to
g
cn
o
1
-a
g
0
fi
8.
4-*
4-f
t.
0
4-1
C
CO
-3
O
4-»
4-*
4-*
.,-
o
-C
s
1
o
ro
•8
to
g
cn
o
1
•a
1
0
0)
a.
4-*
4-*
a
c
CO
»—
o
Ol
."E
0
JC
1
'E
(0
I
(0
"8
ogenal
.£:
xi
I
o
V
L_
O
1
u
4-*
C
CO
o
0
CO
4-1
(0
cr
£
.—
S
3
O
4->
C
o»
1
•a
J!
a
x
.
C
o
c
Ot
CO
L
a
T3
O
09
Z
o
4-*
1
CO
V
'i
to
m
01
a
o
o_
0)
C
o
c
0)
CO
«
u
t_
>.
CJ
t
?
a.
o
(0
CJ
4J
3
§
'E
to
i
(0
"8
ro
g
cn
o
j=
•D
8
Q
0)
C
O
c
Ot
CO
01
L.
CJ
fc
3
co
O
o
0
4-*
to
CT
E
•—
a
o
4-*
1
c
1
£
|
•o
c
-C
0>
X
u
c
o
c
Of
CO
a
-a
b
09
Z
t—
3C
O>
•^
0
J=
I
'E
1
S
to
•8
to
g
Cn
o
•R
•a
g
0
0)
4-*
*E
c
01
3
L
o
ca
Aberdeen
^-
01
..1
0
-C
s
1
3
O
tO
"8
(0
g
en
o
•R
3
g
0
0»
C
O
g
CO
L
a
c
CO
T
•S
0
.c
s
1
3
O
(0
"8
CO
g
cn
o
*
-o
0
0
c
o
c
CO
L
a
4-*
c
CO
*7"
*J
p
5
'c
1
8
to
"S
ogenal
"R
XI
g
0
OJ
c
o
c
CO
L.
a
a,
L.
O
D
O
c
0)
a
CO
*f
-S
0
j;
I
ro
|
to
•8
ogena
1
X)
g
0
0>
C
O
g
CO
t_
o
s
4-*
c
m
CJ
-S
0
x:
S
*E
o
1
ro
15
to
g
cn
o
*
•D
"x.
g
0
g
C
Ol
CO
t_
o
c
CO
1
t—
4->
$
§
'E
0}
3
O
to
15
(0
g
cn
o
"R
3
g
0
s
a
jj
L
o
XI
e
CO
z
01
g
4-*
§
jat w/ polymer (elastomer)
CT
E
•—
to
o
4-1
1
c
O)
1
TO
JS
jC
X
4»
g
c
Ol
CO
L
o
T3
0
t-
co
z
in
CO
o>
13
L.
0
u
§
to
1
4-*
1
g
cn
|
XI
1
_c
4>
g
C.
Ol
CO
Ol
DE
2.
to
to
"R
CO
CO
X
—
0
.c
1
'E
(0
3
O
(0
"8
ro
C
V
cn
o
*
•D
g
O
uf
g
C
CO
s
£
0}
m
T?
fs*
UJ
0)
cn
x
—
o
.c
s
'E
(O
3
O
ro
"8
1
O)
O
*
•o
g
ca
Of
g
c
Ol
CO
o
s
SL
ro
CO
-J,
13
CO
cn
x
Uoride
u
§
i
I
"8
4-*
g
cn
I
10
1
Ol
X
g
c
o
CO
s
>.
to
4-*
ro
"S
UJ
o.
01
cn
x
Ol
o
0
f=
—
i
8
"8
4-*
§>
|
XI
I
4-*
Ol
X
g
c
01
CO
s
a
to
(0
x.
o_
0)
cn
X
0)
.^-
o
j:
§
'E
CO
3
O
to
"8
Ol
cn
o
"S
Xt
1
O
O>
g
C
0)
CO
s
SL
a
4-*
a
x.
UJ
f-
01
cn
x
,~
o
-s
s
'E
a
3
O
a
"8
Ol
cn
o
"R
XI
"x
I
Ol
g
C
0)
CO
s
51
ro
4-*
CO
-j.
a.
(U
x
Uoride
u
g
i
|
4-t
1
C
0)
cn
|
1
x:
4-*
Ol
X
Ol
g
CO
S
Z
a
re
"S
ra
i
s
4-*
•8
4-t
cn
|
"°
1
4-*
Ol
X
Ol
c_
0
u
01
s
to
u
g
z
rv
01
c
0
c
Ol
1
o
"to
"8
s
Ol
g1
"R
XI
.1
a
a>
g
c
Ol
CO
s
ro
g
z
01
g
c
at
ca
1
3
O
to
"8
Ol
•R
XI
g
O
Ol
0
u
31
ro
g
z
s
01
c
0
c
cc
ro
3
O
to
"S
cn
•R
xi
g
a
a>
g
c
(U
CO
s
ro
g
z
o
o
CD
ro
01
en
t_
to
m
I
|
4-*
g
O)
-C
•D
M
1
J=
Ol
X
c
g
C
EA
S
a
g
z
£5
01
g
C
01
CO
I
a
|
"8
4-*
g
•§.
g"
g
O
g
c
CO
s
ro
g
z
ru
c
§
c
0>
ca
4-*
5
CT
|
1
g
cn
•R
•a
N
1
JZ
Ol
X
0
u
0>
3=
s
a
g
z
10
0
(/]
I
c
a
CO
S
a
8
1
g
cn
•R
M
">•
4-*
g
g
c
Ol
CO
s
to
g
z
X
c
o
ca
57
-------
Disclaimer
Ihe work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
58
-------
TECHNOLOGIES APPLICABLE FOR THE REMEDIATION OF
CONTAMINATED SOIL AT SUPERFUND RADIATION SITES
Ramjee Raghavan, George Wolf
Foster Wheeler Enviresponse, Inc.
GSA Raritan Depot
Woodbridge Ave.
Edison, NJ 08837
Darlene Williams
Releases Control Branch
USEPA RREL
GSA Raritan Depot
Woodbridge Ave.
Edison, NJ 08837
ABSTRACT
This paper identifies technologies that may be
stabilizing radioactive contamination at Superfund
containing radioactive material. The radioactive
these sites consist primarily of waste from radium,
processing. Twenty-five existing Superfund sites
radionuclides.
useful in removing or
hazardous waste sites
materials at some of
uranium, and thorium
are known to contain
Sites contaminated with radioactive material pose a unique problem
because, unlike organic wastes, radioactive contaminants cannot be
destroyed by physical or chemical means; they can only decay at their
natural rate. Alteration of the radioactive decay process thereby changing
the fundamental hazard is not possible. Several technologies have poten-
tial for removing or stabilizing radioactive material at Superfund sites.
These fall into the categories of disposal, on-site treatment, chemical
extraction, physical separation, and soil washing. Applicability of these
technologies is controlled by site-specific factors, and their feasibility
must be determined on this basis.
INTRODUCTION
The United States Environmental
Protection Agency (USEPA) has iden-
tified 25 Superfund sites in the
country that are radioactively con-
taminated. They are
1. The radioactive
many Superfund sites
ucts of uranium,
radium processing in
listed in Table
materials at
are by-prod-
thorium, and
the form of
tailings, contaminated buildings
-------
TABLE 1. SUPERFUND SITES CONTAINING RADIOACTIVE CONTAMINANTS
Site
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Name City/County State/EPA Region
Shpack/ALI
Haywood Chemical Co./
Sears Property
U.S. Radium Corp.
U. R. Grace & Co.
Montclair, West Orange,
Glen Ridge Radium Site
Lodi Municipal Well
Lansdowne Property
Haxey Flats Nuclear
Disposal Site
West Chicago Sewage
Treatment Plant
Reed-Keppler Park
Kerr-McGee Off -Site
Properties
Kerr-McGee Kress Creek/
West Branch of Dupage
River
The Komestake Mining Co.
United Nuclear Corp.
Weldon Spring Quarry
Monticello Radioactivity-
Contaminated Properties
Denver Radium Superfund
Sites
Lincoln Park
U.S. DOE Rocky Flats
Plant
lira van Uranium Project
Teledyne Wah Chang
Kanford 200-area (USDOE)
Hanford 300-area (USDOE)
Hanford 100-area (USDOE)
Norton/Attleboro
Maywood/Bergen Co.
Orange, Essex Co.
Wayne/Passaic Co.
Essex Co.
Essex Co.
Lodi, Bergen Co.
Lansdowne
Fleming City/Hillsboro
West Chicago
West Chicago
West Chicago
West Chicago
Cibola Co.
Church Rock
St. Charles City
Monticello
San Juan, Co.
Denver
Canon City
Golden
Mont rose City/Uravan
Albany
Benton, CO
Benton, CO
Benton, CO
MA/ 1
MJ/II
NJ/II
NJ/II
NJ/II
NJ/II
NJ/II
PA/I I I
KY/IV
IL/V
IL/V
IL/V
IL/V
NM/VI
NM/VI
MO/VI I
UT/VI 1 1
CO/VIII
CO/VIII
CO/VIII
CO/VI 1 1
OR/X
WA/X
WA/X
WA/X
Acres
31
42
1
6.5
127.0
wells
1.9
25.0
25.0
0.25
--
--
245
170
220
--
40
900
6,550
900
--
--
--
--
Cu yds
270,000
10,000
120,000
350,000
--
2,000
178,000
40,000
, 20,000
61,000
--
16,500,000
4,700,000
780,000
182,000
106,000
1,900,000
--
10,000,000
--
1,000,000,000
27,000,000
4,300,000,000
-------
and equipment, and stream sedi-
ments. These sites, located across
the United States, vary greatly in
size and may involve radiation
exposure to people who reside on
and around them.
These sites if not remediated
pose a potential threat to human
life and the environment. Possible
effects on human health include the
increased risk of cancer and
increased risk of genetic damage
that may cause inheritable defects
in future generations.
PURPOSE
Sites contaminated with radio-
active material pose a unique prob-
lem because unlike organic waste,
radioactive materials cannot be
destroyed by physical or chemical
means; they only decay at their
natural rate. The purpose of this
paper is to review potential soil
remediation technologies that can
reduce the mobility or volume of
the contaminated material, such
that treatability studies for reme-
diation of these sites can be
identified.
APPROACH
Technologies that have potential
to remediate structures (buildings)
and groundwater are of interest at
some Superfund radiation sites, but
these are beyond the scope of this
paper. The soil remediation technol-
ogies discussed in this paper fall
into the categories of disposal,
on-site treatment, chemical extrac-
tion, physical separation, and soil
washing. Applicability of these
technologies to Superfund radiation
sites is controlled by site-
specific factors; therefore, their
usefulness must be determined on a
site-by-site basis.
On-Site Disposal: Capping
This concept involves covering
the contaminated site with a bar-
rier sufficiently thick and imper-
meable to minimize the diffusion of
radon gas (1). Barrier materials
can be either natural low-permeabil-
ity soils (e.g., clay) or synthetic
membrane liners, or both.
Application: Appropriate for
large discrete contaminated areas,
or several smaller areas that are
close together.
- Advantages: Low cost, easily
applied, well-known, and a proven
technology.
- Limitation: Limits further use
of the site. The cap must be main-
tained as long as the contaminant
exists at the site. Also, horizon-
tal migration of the radioactive
material in groundwater could still
occur.
- Experiences: Exists for radio-
active contaminated soils and
tailings (1,2). -
Cost: $13-200 per mj(3). Low
cost for cap only, high cost for
excavation, transportation, legal
assistance, and cap.
On-Site
Walls
Disposal: Vertical Barrier
Vertical barrier walls may be
installed around the contaminated
zone to help confine the material
and any contaminated groundwater
that might otherwise flow from the
site. The barrier walls, which
might be in the form of slurry
walls or grout curtains (4), would
have to reach down to an imperme-
able natural horizontal barrier,
such as a clay zone, in order to be
61
-------
effective in impeding groundwater
flow.
- Application: Could be considered
for large discrete waste material
or around several smaller areas
that are close together.
- Advantages: Simple to install,
and applicable to a variety of soil
conditions.
- Limitation: Restricts further use
of the site, possible deterioration
of the barrier walls resulting from
the chemicals contained in the
waste, would not stop vertical con-
tamination to groundwater below.
- Experiences: Exists for hazardous
wastes and not for radioactive
wastes (4).
- Experiences: Exists for radio-
actively contaminated soils (3),
per nr of
Cost: $33-377
vertical face (4).
Off-Site Disposal:
Land Encapsulation
Land encapsulation has been the
disposal method most used to this
point in time for low-level radio-
active waste materials. Land encap-
sulation can be as simple as exca-
vating the contaminated material
and, without further treatment,
hauling it to a secure site.
Application: Appropriate for
wastes that have not been treated,
as well as for radionuclides extrac-
ted from a soil or other type of
matrix.
- Advantages: Low cost, proven,
workable technology for the dis-
posal of low-level radioactive
wastes.
- Limitation: Finding a site is
politically and socially difficult.
Transportation of large volumes
also carries certain costs and
risks. Longevity is a consideration
not fully addressed by this dis-
posal method.
Cost: $276-895
contaminated soil (3).
per nr of
Off-Site Disposal: Land Spreading
This technology involves excava-
tion of the contaminated material,
transporting it to a suitable site,
and spreading it on unused land,
assuring that radioactivity levels
approach the natural background
level for these materials when the
operation is completed.
- Application: Appropriate for dry,
granular tailings and soils with
very low level radioactivity.
- Advantages: Simple and rela-
tively inexpensive.
- Limitation: Selecting a site is
both politically and socially diffi-
cult. Also, it could contribute to
a non-point source pollution
problem.
- Experiences: Very limited.
- Cost: Not available.
Off-Site Disposal:
Underground Mine Disposal
Underground mine disposal could
provide secure and remote contain-
ment for radioactive waste. The
radioactive waste could be excava-
ted and transported without treat-
ment to the mine site, pretreated
for volume reduction, or solidified
to facilitate transport and place-
ment, thus reducing associated
costs. Movement of radionuclides
into groundwater must be investi-
gated and prevented.
Application: Appropriate for
variety of radionuclides and matrix
types.
- Advantages: Would provide a very
secure and remote containment.
62
-------
- Limitation: Expensive. Transpor-
tation costs and associated risks
need to be researched further.
- Experiences: Very limited (1).
- Cost: $399-942 nr of contamin-
ated soil (1).
Off-Site Disposal: Ocean Disposal
A possible alternative to land-
based disposal options is ocean dis-
posal. This alternative should only
be considered for tailings and
other radioactive soils that are
free of other hazardous waste,
because of potential danger to
marine biota.
- Application: Appropriate for low-
level radioactive waste.
- Advantages: Offers extreme isola-
tion of low-level radioactive
waste.
Limitation: Stringent permit
requirements.
- Experiences: Exists for radio-
active wastes (1).
- Cost: $332-400
taminated soil (2).
per m3 of con-
On-Site Treatment:
Stabilization/Solidification
This method immobilizes radio-
nuclides by trapping them in an
impervious matrix (4). The solidi-
fication agent (i.e., Portland
cement, silica grout, or chemical
grout) can either be injected in
situ, or the waste can be excava-
ted, mixed, and returned.
- Application: Can be applied to
buried and/or capped material.
- Advantages: Solidification may
be able to reduce the release of
radon and associated radioactivity
to acceptable levels. Solidifica-
tion also may make it easier to
transport and dispose of the waste
material off site.
- Limitation: Long-term effects
are not known. There can be unde-
sired reaction between the addi-
tives and other types of hazardous
waste.
- Experiences: hxists for hazard-
ous wastes and not for radioactive
wastes (4). .
- Cost: $44-328 per nr of contam-
inated soil (4).
On-Site Treatment: Vitrification
This technology immobilizes
radioactive contaminants by trap-
ping them in an impervious matrix.
The in situ process melts the waste
materials between two or more elec-
trodes, using a large amount of
electricity. The melted material
then cools to a glassy mass in
which the radionuclides are
trapped.
- Application: Applicable for low-
level radioactive wastes.
- Advantages: Minimal site prepara-
tion required.
- Limitation: Many substances vola-
tilize, requiring gas collection
system. Radium may volatilize,
therefore extra precautions are
required.
- Experiences: Limited (1).
- Cost: $161-600 per m6 of contam-
inated soil (3).
Chemical Extraction
The various applicable chemical
extraction techniques include
extraction with inorganic salts,
mineral acids, and complexing
agents.
Radioactive contaminants can be
extracted by thoroughly mixing soil
and mill tailings with different
chemical solutions. The clean
coarse solids are separated from
the extractant solution by physical
63
-------
methods. The radioactive material
is removed from the extractant solu-
tion by ion exchange, coprecipita-
tion, or membrane filtration.
- Application: Various applica-
tions depending on reagents used.
- Advantages: High percentage of
radium, thorium, and uranium
removal possible depending on
reagents used.
- Limitation: The resulting chem-
ically leached material may create
a harmful waste stream. Reagents
can be expensive and may require
corrosion resistant materials.
- Experiences: Limited laboratory-
and bench-scale testing. Some
extraction of radioactive material
from ores. (1,5-10)
- Cost: $66-199 per m3 of contam-
inated soil.
Physical Separation
The radioactive contaminants in
soils and tailings in many cases
are associated with the finer frac-
tions (1,11). Thus, size separation
may be used to reduce the volume of
concentrated material for disposal,
leaving a cleaner fraction. Physi-
cal separation may be used with
extraction to further reduce contam-
inant volume. Four physical separa-
tion technologies that can be used
are screening, both wet and dry;
classification; flotation; and
gravity concentration. (12-14)
- Application: Applicable for a var-
iety of soils depending on method
used.
- Advantages: Can be simple and
inexpensive method.
- Limitation: Some soils may be
hard to separate. Some methods have
low capacity.
- Experiences: Mature technologies
with extensive use in industry and
ore processing.
- Cost: Equipment costs for four
technologies are:
0 Screening--$4.4-ll per Kg/hr
capacity
0 Classification-^. 22-1.1 per
Kg/h capacity
0 Flotation--$10-52 per L/s
capacity
0 Gravity Concentration--$4.1-
4.73 per L/s capacity
Soil Washing
Soil washing uses a combination
of physical separation and chemical
extraction technologies. Contamina-
ted soil or tailings are mixed with
water and/or extraction reagents.
The clean coarse particle sizes are
separated from the liquid contain-
ing the fines and radioactive mate-
rial by a combination of physical
separation methods. The radioactive
material would then be extracted
from the liquid by standard water
treatment processes such as filtra-
tion, carbon treatment, ion
exchange, chemical treatment, and
membrane separation.
- Application: Depends on chemical
reagents used.
- Advantages: High percentage of
contaminant removal is possible.
Recycling of reagents is possible.
- Limitation: Expensive reagents.
Chemically leached material may
create a harmful waste stream.
- Experiences: Only laboratory-
and bench-scale testing /I).
- Cost: $66-132 per m3 of contam-
inated soil.
RESULTS
- The 25 Superfund sites have radio-
logically contaminated soil spread
over a total of 9,500 acres and
have several contaminated ground-
water wells.
64
-------
Alteration of the radioactive
decay process thus changing the fun-
damental hazard is not possible.
- Any choice of remediation technol-
ogies for radioactive waste at
Superfund sites is site specific.
Extensive site soil characteriza-
tion studies would be required
prior to development and applica-
tion of most of the technologies.
- Since none of the chemical extrac-
tion -and physical separation tech-
nologies have been used in a radio-
active site remediation situation,
their application must be
approached cautiously. The same
holds true for solidification and
stabilization processes. Histori-
cally, only land encapsulation has
been used to remediate similar
sites; ocean disposal has been used
for low-level radioactive wastes.
- Various remediation technologies
have potential to reduce the volume
of the contaminated waste with an
associated increase in concentra-
tion of the radioactive material.
- Every site remediation involving
radioactive materials must involve
a final environmentally safe dis-
posal site for the radioactive
materials.
- Even if it proves feasible at a
particular site to lower the concen-
tration to some acceptable level of
radionuclides in the soil by physi-
cal separation and/or chemical
extraction, the "clean" fraction is
likely to contain traces of radio-
nuclides. Therefore, adequate atten-
tion must be given to whether the
"clean" fraction may be returned to
the original site or an unrestric-
ted location or must be sent to a
disposal site.
ACKNOWLEDGMENTS
The authors wish to express their
gratitude for the contributions of
F. Freestone, P.Shapiro, R.Hartley,
W. Gunter, and G. Snodgrass of the
U.S. Environmental Protection
Agency; and G. Gupta of Foster
Wheeler Enviresponse, Inc.
REFERENCES
1..U. S. Environmental Protection
Agency. Technological
Approaches to the Cleanup of
Radiologically Contaminated
Superfund Sites, USEPA. Report
EPA/540/2-88/002, August 1988.
2. Camp, Dresser & McKee et al.
Draft. Final Feasibility Study
for the Montclair/West Orange
and Glen Ridge, New Jersey,
Radium Sites, Volume 1. USEPA
Contract 68-01-6939, 1985.
3. U.S. Department of Energy. Long
Term Management of the Existing
Radioactive Wastes and Residues
at the Niagara Falls Storage
Site, DOE/EIS-0109D, Washington
DC, 1984.
4. U.S. Environmental Protection
Agency. Handbook -- Remedial
Action at Waste Disposal Sites
(Revised). EPA-625/6-806,Hazard-
ous Waste Engineering Research
Laboratory, Cincinnati, Ohio,
1985.
5. Borrowman, S. R., and P. T.
Brooks. Radium Removal from
Uranium Ores and Mill Tailings.
RI-8099, U.S. Bureau of Mines,
Salt Lake City Research Center,
Salt Lake City, Utah, 1975.
-------
6. Clark, D. A. State of the Art:
Uranium Mining, Milling and
Refining Industry. USEPA/60/
2-74-038m 1974.
7. Landa, E. R. Leaching of Radio-
nuclides from Uranium Ore and
Mill Tailings. Uranium,
1:53-64, 1982.
8. Organization for Economic Coop-
eration and Development (OECD).
Uranium Extraction Technology -
Current Practice and New Devel-
opment in Use Processing. OECD,
Paris, 1983.
9. Ryon, A.D., F.J.Hurst, and F.G.
Seeley. Nitric Acid Leaching of
Radium and Other Significant
Radioncuclides from Uranium
Ores and Tailings. ORNL/TM-
5944, Oak Ridge National Labor-
atory, Oak Ridge, Tennessee,
1977.
10. Taskayev, A.I., V.Ya.Ovchenkov,
R.M. Altkaskhin, and I.I.
Shuktomova. Effect of pH and
Liquid Phase Cation Composition
on the Extraction of 226 Ra
from Soils. Pochvovedeniye,
12:46-50, 1976.
11. Raicevic, D. Decontamination of
Elliot Lake Uranium Tailing.
CIM Bulletin, 1970.
12. Garnett, John, et al. Initial
Testing of Pilot Plant Scale
Equipment for Soil Decontamina-
tion. U.S. Dept. of Energy,
RFP 3022, 1980.
13. Kelly, E. G., and D. J.
Spottiswood. Introduction to
Mineral Processing. John Wiley,
New York, 1982.
14. Wills, B. A. Mineral Processing
Technology. Pergamon Press,
New York, 1985.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
66
-------
Advanced Technologies for Pollutant Detection, Monitoring, and Remediation in
Ground Water
R. A. Kloppi
J. F. Haasbeek2
P. B. Bedient3
A. A. Biehle4
Abstract
Advances in technology for detection, monitoring, and remediation of hazardous
waste constituents both in soil and ground water are rapidly changing. Conventional
drilling and sampling for pollutant detection is giving way to sophisticated in situ
technologies. One of the most promising is the use of cone penetration testing. The
cone penetration test has been useful in the siting of waste disposal facilities and in the
design of remedial action alternatives. Various types of cone penetrometers have been
used to conduct a number of in situ tests where accurate information on soil stratigraphy
and variability is essential for the consideration and evaluation of various hazardous
waste disposal technologies.
Introduction
During site investigations, the
owner/operator is interested in obtaining
data toward the ultimate goal of ground
water remediation. The data of primary
concern are the vertical and lateral extent
of the contaminants and their respective
concentrations, as well as subsurface
geological data. Conventional sources
available for the gathering of this data
include soil borings, installation of
monitoring wells, geophysical methods,
and aquifer testing.
One method of collecting site data is
the cone penetrometer test (CRT). This
method can be used to define such factors
as sand geometry, hydraulic
conductivities, detailed stratigraphy, and
other soil properties. Once these data are
collected, the remediation options can be
examined to determine which is best
suited for the project at hand.
Equipment and test procedures
The cone penetration test involves
pushing a cone-shaped instrument into
the soil and measuring its resistance to
penetration. The two basic types of
instruments are mechanical and electrical
cone penetrometers.
The electronic cone penetrometer is
shown in Figure 1. As it penetrates,
sensitive strain gauges transmit electronic
measurements of resistance to
penetration of the cone tip and friction
sleeve to an automatic data acquisition
system. Measurements of the lateral drift
of the cone are also obtained during the
sounding via a built-in inclinometer. The
test was recently standardized by the
American Society of Testing and Materials
(ASTM, D-3441 79). An alternate design
which incorporates a piezoelement in the
cone tip is often used for environmental
studies. The piezoelement has the ability
to measure the pore water pressures in
the soil that are developed during
penetration. Ground water sampling
cones (Figure 2) are also capable of
monitoring pore pressure and are
discussed in greater detail below.
1 Principle, Terra Technologies, Houston, Texas
2 Graduate Student, Rice University, Houston, Texas
3 Professor, Rice University, Houston, Texas
4 Consultant, Houston, Texas
67
-------
Digital output of soil measurements are
printed continuously during the sounding.
All data are stored on a magnetic medium
for future processing, and a graphical
presentation of the data is immediately
available for in-field stratigraphic
correlation and evaluation. This feature
allows testing to be concentrated in critical
areas. Field plots consist of cone tip
resistance, sleeve friction resistance,
friction ratio, pore pressure, and
differential pore pressure ratio. The
interpretive data are compatible with most
personal computer systems for further
data reduction, including stratigraphic
cross sections and contouring.
Ground water sampling cones are
used to obtain selected samples of in situ
gasses and liquids, retaining volatile
components. Samples are encapsulated
for easy and accurate lab analysis,
minimizing external contamination.
The equipment is ideal for
determination of the vertical and lateral
extent of contamination. Tracer tests are
often conducted where the system is used
for the controlled injection of a tracer fluid.
The spread of the fluid can then be
observed by repeated sampling at various
distances from the point of injection
(Torstensson, 1984; Torstensson, et al.,
1986). The major advantages of the
electronic cone penetrometer over
conventional sampling and testing are
speed, continuous data measurements,
economy, and reliability.
A typical field program consists of an
initial site survey with the cone
penetrometer, followed by a limited
number of soil borings. One of the
disadvantages of CRT equipment is the
lack of soil samples for testing. Although
near-surface samples are easily
obtainable via a sampling adapter, its
depth capability is limited. The cone
penetrometer will, however, define the
optimum depth and location for samples to
be taken, thereby reducing the high cost
normally associated with soil borings.
With the incorporation of the
piezoelement into the system, the cone
penetrometer has become an accepted,
cost effective tool for environmental site
assessments. Additional information on
the use of piezocone data including
discussions on corrections for pore
pressure effects and normalization of data
can be found in Robertson, et al., 1986.
Use of the Cone Penetrometer: Case
History
A large truck mounted electronic cone
penetrometer was used to collect data at a
site in the southwestern United States.
Based on limited data from a few
monitoring wells at the site, an extensive
survey of the subsurface was conducted
using the CRT equipment. Data from the
survey included detailed stratigraphy, soil
types and properties, estimates of
hydraulic parameters such as
transmissivities and potentiometric
surfaces, and the vertical and lateral
extent of contamination. These data were
then used to simulate contaminant
infiltration and transport at the site, and to
begin design of a remedial strategy for the
aquifer. The objectives of the modeling
effort were to simulate the extent of the
contaminant plume, and to examine the
effectiveness of various remedial
schemes.
The CRT data showed the stratigraphy
at the site to consist of an upper clay layer
extending 30 feet (ft) below the ground
surface. This clay was underlain by a
sandy aquifer approximately 40 ft thick,
which was in turn underlain by another
clay layer of undetermined thickness.
Transport Concepts - Upper Zone
The upper clay at the site was
relatively tight, with hydraulic
conductivities averaging 10"6 centimeters
per second (cm/sec), and ranging from
68
-------
10'5 to 10'9 cm/sec. The rates of vertical
migration were governed mainly by rainfall
and infiltration processes within the
unsaturated and saturated zones of the
upper clay unit, and were computed
analytically. The average water table
elevation in the clay was approximately 5
ft msl, whereas the average elevation of
the piezometric surface in the sand below
was approximately 3 ft msl. Darcy's
equation was used to calculate velocity
and travel time of contaminants through
the upper clay to the lower sand. The
computations showed that for average
hydraulic conductivities at the site,
contaminants may require from 22 to 220
years to reach the sandy aquifer.
Once the contaminants in the clay
travel to the base of the clay layer, they
enter the sand, in which horizontal
transport dominates.
Transport Concepts - Lower Zone
The sandy aquifer at the site
comprised the lower zone for transport.
The general direction of flow in this zone
was horizontally to the west. Velocities in
the sand, as indicated by observed
perchlorethylene (PCE) values, appeared
to be relatively low. Measured values
indicated that the contaminants had
migrated about 300 ft from the source
area. In modeling this site, several
transport mechanisms which may also
affect the movement of the contaminants
through the sands were not considered
including adsorption, chemical reactions,
and biodegradation. The exclusion of
these mechanisms normally produces a
conservative result.
Description of the USGS Ground Water
Model
One of the most widely used two-
dimensional ground water transport
models is by Konikow and Bredehoeft
(Konikow, 1978). The USGS method of
characteristics (MOC) model simulates
solute transport in flowing ground water
and can be applied to a wide range of
problem types involving steady-state or
transient flow. The model computes
concentration changes over a grid with
time caused by advection and mixing
(dispersion) from fluid sources. The
model allows the solute to adsorb or
degrade linearly, and assumes that the
gradients of fluid density, viscosity, and
temperature do not effect the velocity
distribution. However, the aquifer may be
heterogeneous and/or anisotropic with
variable pump rates or head values.
The program uses an iterative
alternating direction implicit (ADI)
procedure to solve a finite-difference
approximation to the ground water flow
equation. After the head distribution is
computed, the velocity of the ground water
flow is computed for each node using an
explicit finite-difference form of Darcy's
equation. The model then uses a particle-
tracking procedure to represent advective
transport and a two-step explicit
procedure to solve a finite-difference
equation that describes the effects of
hydrodynamic dispersion, fluid sources
and sinks, and divergence of velocity.
Input parameters for the model
include: transmissivity, porosity, storage
coefficient, longitudinal and transverse
dispersivity, grid system and overall
geometry of boundary conditions,
locations of sources and sinks, wells,
boundary conditions, and initial conditions
for heads and background concentrations.
The model output can include 2-D
head distributions, the x and y velocity
distributions, the drawdown depths, the
concentration values at each grid location,
and detailed breakthrough curves at
selected observation wells. In order to
analyze the model output, a graphical
array package was used. This package,
Biograph, is being developed at Rice
University (Newell, 1988) as part of a
Decision Support System for ground water
69
-------
modeling. It displays numerical data over
the model grid as patterns varying in
darkness: darker patterns represent
higher values. The program also allows
the user to graph values versus time
(breakthrough curves), or over a cross
section across the grid. Many of the
figures in this paper are printed output
from this package.
Initial Model Setup
A 19 by 25 grid of cells 50 ft on a side
was set up and located in an east-west
direction. The source area is located
toward the eastern end of the grid. The
overall length and width of the grid were
1250 ft and 950 ft, respectively, giving a
simulation area of approximately 27.3
acres. The grid setup is shown in Figure
o.
The hydraulic conductivity used in the
base run simulation was 0.0052 cm/sec,
based on aquifer tests and CRT data at
the site, and was assumed to be constant
and isotropic throughout the aquifer. The
longitudinal and transverse dispersivity
values used in the base run were selected
from the literature and set to 10 ft and 1 ft,
respectively. An early investigation of the
effects of variable thickness showed little
effect, so a constant sand thickness of 40
ft was used. Finally, the background
concentrations in the region were
assumed to be negligible.
The source area was simulated in the
model using injection wells in cells
designated as source areas. The source
cells used are shown in Figure 3. The
leakage rate from the source into the
aquifer was estimated using Darcy's law,
and the extent of observed PCE. The final
estimate was 0.0006 cubic feet per
second (cfs), or 0.27 gallons per minute
(GPM). The contaminant concentrations
were calculated as a percentage of the
source concentration, which was taken to
be approximately 66,000 micrograms per
liter (p.g/1), based on observations in a
monitoring well near the source and
solubility data for the chemicals involved.
The hydraulic conditions in the aquifer
were assumed to be at steady state in all
of the simulations presented here. This
assumption greatly reduces run times, and
seldom produces any effect on model
results. The validity of this approach was
verified using the model in a hydraulically
transient mode.
The period of time since the
contaminants first reached the sand layer
was difficult to determine. A simulation
time of 5 years was chosen based on
records of source activity, and on initial
model results which indicated that
approximately 5 years of contaminant
movement were required to match
observed data points.
Sensitivity Analysis
The model was run using the
parameters described above to produce a
base run to be used as a base line for
comparison in the sensitivity analysis.
Three additional runs were made in
order to investigate the sensitivity of the
contaminant plume to changes in model
parameters. The results of these runs are
compared to the base run in Table 1. The
table shows that the model is most
sensitive to the source leakage rate, and
shows little response to changes in other
parameters.
Model Calibration
The goals of the hydraulic calibration
effort were to simulate the general
direction and velocity of the ground water
in the area of interest. Water level
contours, plume extent, and the geologic
nature of the aquifer system were all used
to develop a satisfactory "match" between
reality and simulation.
The final calibration was achieved
using constant head cells around the grid
perimeter, and three additional cells
representing local peaks and depressions
70
-------
(Figure 3). The head values from the
model output were entered into a
contouring package, and the resulting
contours were overlaid on the contour
map created from observed head data.
The final match presented a close
correlation, as shown in Figure 4.
The observed data for PCE were
chosen as the data to which the
contaminant transport results of the model
would be compared. This was only a
preliminary calibration process, as
normally at least 6 to 10 observed data
points are required within the plume to
calibrate a 2-dimensional model.
Measurements from one monitoring well
and two CRT points were used for
comparison. Assuming a source
concentration of 66,000 u,g/l, the observed
PCE concentrations as percentages of the
source concentration in three model cells
would be:
Cell
7,9
10,13
8,13
% of source concentration
0.333
0.355
0.0167
The results of the final model run are
shown in Figure 5. The concentrations in
the above three cells are 0.5%, 0.8%, and
0% of source concentration, respectively.
The run shows good areal agreement, and
is satisfactory based on the available data.
The parameters used in this final run
were:
Hydraulic Conductivity
Sand Thickness
Leakage rate
Time
Longitudinal Dispersivity
0.00052 cm/sec
40ft
0.0006 cfs
5 years
5tt
In addition, the source area in the model
was slightly reduced. The leakage of
contaminants into cells (8,8) (9,8) and
(9,10) was eliminated, effectively reducing
the source area by 7500 square feet.
Remedial Design Criteria
Based on the CRT data collected at
the site, several remedial techniques were
found to be infeasible because of the
depth of the contamination. These
included a collector trench for contaminant
removal, and slurry walls or sheet piling
for plume containment. It was determined
that the installation of a recovery well field
was the only economically feasible option.
A recovery field may include the
installation of extraction and injection
wells: extraction wells to reverse the
gradient, and injection wells to increase
the gradient and expedite the remediation
time. The number of wells and the
pumping rates required for a 5-year
cleanup operation were investigated by
numerical modeling.
Remedial Modeling - Basic Scenario
As an initial design of the extraction
well field, three pumping wells were
simulated approximately 75 ft down-
gradient of the source area, spaced 150 ft
apart. These locations in the model grid
are shown in Figure 3. The wells were
operated for 5 years. The results of
operating these wells at various pumping
rates are shown in Table 2. A longitudinal
cross section of the plume through time
for the run labeled "Low Rate 2" is shown
in Figure 6. The table shows that the 3-
well scheme is probably inadequate for a
5-year cleanup operation.
Additional Wells
Due to the lack of response from the 3-
well system, four additional wells were
placed in grid cells (8,9) (10,7) (12,5) and
(11,8) and the pumping rate in each of the
wells was increased to 3.5 GPM, which
produced drawdowns of up to 20 ft. The
source area was assumed to have been
removed. After 5 years of pumping, the
maximum concentration in the grid was
reduced to 0.1%, and the plume area was
less than 0.3 acres. A longitudinal cross
71
-------
section of the plume through time is
shown in Figure 7. The plume after 5
years is shown in Figure 8.
Conclusions
After reviewing all available
hydrogeology, CRT and monitoring well
data 'for the site, it was possible to
successfully model the migration of
contaminants in the sand. Although the
calibration of the model was based on
limited data, the results of the remedial
design effort are indicative of the scale of
the remedial systems to be examined. In
addition, the modeling effort at the site has
exposed several areas where data were
inadequate, and has shown which design
options must be investigated further.
The availability of the CPT equipment
has proven extremely useful in the
investigation of this site, and can be used
to collect the additional data required in
order to complete the design of the
remediation scheme. In order to fully
utilize the power of the CPT method,
communication must be increased
between modeling and data collection
stages. After the initial site survey, the
investigation should iterate between
modeling and data collection steps, with
each step directing the efforts of the next.
This combination of two powerful
technologies can provide cost savings
through increased productivity and
efficiency, and is a concept which applies
in all engineering disciplines.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
References
1. ASTM, 1979, "Standard Method for
Deep, Quasi-Static Cone and
Friction-Cone Penetration Tests of
Soil", Designation D3441-79, pp.
550-557.
2. Konikow, L. F., and J. D.
Bredehoeft, 1978, "Computer Model
of Two-Dimensional Solute Transport
and Dispersion in Ground Water", U.
S. Geological Survey Techniques of
Water Resources Investigation, Book
7, Chapter C2.
3. Newell, C. J., P. B. Bedient, and J. F.
Haasbeek, 1988, "Oasis: A Graphical
Hypertext Decision Support System
for Ground Water Contaminant
Modeling", Submitted for Publication
4. Robertson, P. K., R. G. Campanella,
D. Gillespie, and J. Greig, 1986, "Use
of Piezometer Cone Data", ASCE
Specialty Conference on the Use of
In-Situ Tests in Geotechnical
Engineering, pp. 1263-1280.
5. Torstensson, B., 1984, "A New
System for Ground Water
Monitoring", Publication of BAT
Envitech, Inc., pp. 131-138.
6. Torstensson, B., and A. M. Petsonk,
1986, "A Hermitically Isolated
Sampling Method for Ground Water
Investigations", ASTM Symposium
on Field Methods for Ground Water
Contamination Studies and their
Standardization, Cocoa Beach,
Florida.
72
-------
Run Parameter Changed (units) Maximum Plume
Number (base run value) Concentration Area
(modified value) (%) (Acres)
1 Hydraulic Conductivity (cm/sec)
0.0052 85 4.4
0.00052 100 4.8
2 Source Leakage Rate (cfs)
0.0006
0.0001
3 Longitudinal Dispersivity (ft)
10
1
85 4.4
27 2.9
85 4.4
90 3.0
Run Source Pumping Maximum Plume
Description Removed Rate Concentration Area
(GPM) (%) (Acres)
No Action 1 No
No Action 2 Yes
0 100 7.5
0 100 4.4
High Rate No 4.5 76.4 3.2
Low Rate 1 No
Low Rate 2 Yes
• 9
*•" '' ' i±£t/////£ZY^
1 99.8 4.6
1 89.4 4.3
4 ;
- ' '1
^u* *u^r^. £^gfry^ / «
Cmk«l faitt (It «»•)
* Fnciio* tfem «l JO c*1
Figure 1: Electric Friction Cone (after ASTM)
73
-------
OUlCft C0-M.lt! *O«fTI«
Figure 2: Ground Water Sampling Probe
(courtesy of BAT Envitech)
North
1
5
10
15
20
25
1
5 10 15 19
:£
•
i
iiliii*
u
«
'FT
*
.....••
i-:'1
:->•
E
••3.
S-:-:
:
Leaend
J
'fZ\
.&•
*$
»5
::vi
s:;
300
iS:
:¥;:;
*•*
:S:
;¥::
H:
S?
S
*B
r;g
:s;
•:%
:;|^:
«S
•:;;:
S*
;$;
S
-------
300 feet
Figure 5: Calibrated Plume
100
Direction of Flow
t = 0 years
t = 2.5 years
,t = 5 years
0 East - West Distance (ft) 60°
Figure 7: Cross Section through
Plume - 7-Well Pumping Scenario
100
Direction of Flow
t = 0 years
t = 2 years
t = 5 years
0 600
East - West Distance (ft)
Figure 6: Cross Section through
Plume - 3-Well Pumping Scenario
(run description: Low Rate 2)
10
15
20
25
10
15
19
300 feet
Figure 8: 7-Well Pumping
Scenario - Plume after 5 Years
75
-------
USE OF ABANDONED COAL/LIGNITE OPEN PITS
FOR WASTE DISPOSAL IN SELECTED EUROPEAN COUNTRIES
Jacek S. Libicki
Central Research and Design Institute for Surface Mining
POLTEGOR
Powstancow SI. 95, 53-332 Wroclaw, Poland
ABSTRACT
The use of abandoned coal/lignite pits as disposal sites for solid waste
appears to be a reasonable approach to a difficult problem, especially if they
are located close to the waste source. However, a potential for groundwater
and soil pollution exists. This issue was discussed by a "Group of Experts on
Opencast Mining of the UN Economic Commission for Europe" because most of the
sites are operated by mining companies. This paper contains the major topics
of discussion including the significance of the problem, legal aspects,
characteristics of the open pits, waste intended for disposal, investigations
required to obtain a disposal permit, disposal techniques, protection
measures, monitoring environmental impacts, and research trends. A few
countries are used as examples.
INTRODUCTION
Abandoned open pits from the
extraction of minerals such as
coal/lignite, sand/gravel, and clay,
are attractive potential waste dis-
posal sites. Their advantages are
that an excavated area is available,
no new land would be disturbed, and
the waste can be placed below the
original land level (landscape may
even be improved), and the risk of
fugitive dust is minimized. The
disadvantages are that the waste may
be in contact with the groundwater,
or leachate from the waste may reach
the groundwater and thus result in
its pollution. The use of abandoned
coal/lignite open pits for disposal
of different types of waste was dis-
cussed in the "Group of Experts on
Opencast Mining of the UN Economic
Commission for Europe." The aim of
the discussion was to compare legis-
lation and regulations, current pro-
cedures, environmental impacts that
have been found, protection measures
in use, and future trends. The major
contributions to the discussion were
from the United Kingdom (UK), Poland,
Czechoslovakia, and the German Demo-
cratic Republic (GDR).
LEGISLATION AND REGULATIONS
Both in Eastern and Western
European countries, use of abandoned
open pits for waste disposal is sub-
jected to different laws and regula-
tions. In the UK the most important
one is the Disposal of Poisonous
Waste Act (1972). This Act makes it
an offense punishable by heavy
penalties, to deposit on land any
poisonous, noxious, or polluting
substances which can give rise to an
76
-------
environmental hazard or cause danger
to persons and animals. The Act
gives responsibilities to waste manu-
facturers and disposal operators,
especially regarding chemical compo-
sition of the disposal and its loca-
tion as well. The other one is Con-
trol of Pollution Act (1974) which
sets out arrangements for disposing
of controlled waste, and duties of
disposal authorities, and regulations
under which the waste may be deposi-
ted. The subordinary to this Act,
Regulations of 1976, detail the
licensing of Waste Disposal (1976).
According to this last one the dis-
posing of waste, other than mining
and quarry waste, require an applica-
tion with listed waste volume, detail
characteristics, type of containers,
and their number, size and descrip-
tion. Moreover, the name of the per-
son/employer who brought the waste to
the disposal must be specified. The
disposal of less hazardous waste such
as colliery waste, fly ash, and
domestic waste are the subject of the
Town and Country Act (1971) and the
Town and Country Planning Development
Order (1979). According to the above
provisions, permission for disposal
is required, and mining authorities
must consult with appropriate water
authorities on all planning and
applications. Finally in the UK, the
Council of European Communities
Directive for Groundwater Protection
Against Pollution caused by Dangerous
Substances (80/86) EC) has to be
observed.
In Poland disposal into open
pits is subjected to several laws
such as: Environmental Protection
Law (1980), Land Management Law
(1985), Farmland and Forest Protec-
tion Law (1982), Water Law (1974) and
finally the Mining Law (1978). Each
of these laws is supplemented by many
detailed regulations edited by the
responsible ministries and local
authorities. There are eight major
steps in obtaining a permit.
1. A Feasibility Study that shall
contain the characteristics of waste,
characteristics of disposal site
(geology, hydrogeology, surface
water, geotechnics, dimensions, loca-
tion, etc.), method of disposal,
forecast of impact on the environ-
ment, protection measures, monitor-
ing, and further research. This
study is conducted by a consulting
company.
2. An application of the waste dis-
posing company to the Province
Authorities for approval of a
location. Ask whether a given waste
can be stored in a given place, and
what conditions must be met (the
Feasibility Study must be enclosed).
3. The Application is reviewed by
various departments of the province
administration (Protection of Envi-
ronment, Health, Regional Develop-
ment. Transportation, Agriculture,
etc.) and county/community admini-
stration. The Province and Local
Councils (elected bodies) also review
the application. The province
authorities give their opinion on the
site and list their requirements, and
if necessary, they request further
research or collection of informa-
tion.
4. A General Technical Plan which
discusses in more detail the subjects
in the Feasibility Study, responds to
the recommendations of the Province
and Local Authorities, and presents
results of any additional research
that may have been performed.
5. Final application to the Province
Authorities for the permit (the Gen-
eral Technical Plan is enclosed) is
made.
6. The Authorities review the appli-
cation using the procedure similar to
the one given above (3.), and approve
the Realization Plan with comments
and recommendations.
77
-------
7. Detail designs of the site are
made, e.g., sealing, road construc-
tion, electrical power supply, and
water management.
8. Disposal takes place with peri-
odic inspection of the Province
Authorities. They check to see that
their recommendations and conditions
are being met.
In the GDR Law, wastes are
divided into three groups: toxic,
polluting, and low or non-toxic ones.
Only the third category is allowed to
be disposed in open pits. This is
regulated in general by the Envi-
ronmental Policy Act, Water, Act,
Matter Act, Mining Act, Toxicity Law,
and in details by the Regulations.
In the GDR, the laws require a study
of waste reuse (recycling) prior to
disposal application. The permission
is always issued for a fixed period
of time, fixed volume of wastes and
disposal procedures.
In all the above countries per-
mission is granted or confirmed by
local authorities, and it includes
the conditions, stipulations, and
responsibilities of the company, even
if there is more than one user of the
site. In Eastern European countries
laws and regulations are not as com-
mon as in Western Europe. In several
European countries (Greece, Turkey,
Yugoslavia, and others) there are no
regulations.
Characteristics of the Open Pits used
for Disposal
The open pits designated for
waste disposal in Europe can be
divided into two general groups. The
first group are the open pits after
bituminous coal extraction and will
be represented here by examples in
the UK.
Typical void dimensions and capa-
cities for waste disposal are:
1. 400 x 400 m x 15 m (depth),
1,1 mill. m3 capacity
2. 450 x 425 m x 56 m (depth),
3,0 mill. m3 capacity
3. 1000 x 350 m x 80 m (depth),
15,0 mill. m3 capacity.
The surrounding rocks are mostly
shales, sandstones, and limestones.
The second group are the open pits
after lignite extraction, and here
two types can be distinguished. The
first ones are the final voids with
an area of 50 - 800 hectares, depth
20 - 60 m and capacity up to 30 mill.
m . The second ones are deep open
pits which can have an area of 1000
hectares, depth 200 m and capacity up
to 500 mill. m3. The rocks surround-
ing lignite open pits consist mostly
of sands and clays, both occurring in
the slopes, and in the bottom of the
void intended for waste disposal.
Very often the aquifers are in poten-
tial contact with waste and are wide-
spread and must be protected by law.
Therefore, in all countries with
environmental laws, obligatory waste
disposal in these pits is restricted
to the types of waste, as well as the
disposal procedures and protection
measures.
Characteristics of Waste Disposed in
Open Pits
Due to the large size of aban-
doned open pits, reliable sealing
would be very difficult and expen-
sive; therefore the hazardous wastes
of high toxicity are mostly banned
from these sites. The most common
use is the disposal of coal refuse
(Poland, UK), fly ash from coal-fired
power plants (most of the countries),
domestic refuse (many countries),
discards (Poland, UK, Czechoslovakia,
construction/demolition materials
(UK, Czechoslovakia), and wastes from
briquetting factories (GDR).
78
-------
Some examples of these wastes
are presented below. The coal refuse
has Teachable IDS 500 mg/kg, Cl 50
mg/kg, S04 40 mg/kg, Phenols 0.06
tng/kg, CN 0.005 mg/kg, Zn 0.18 mg/kg,
Cu 0.04 mg/kg, Pb 0.04 mg/kg, Mo 0.03
mg/kg, Cr 0.067 mg/kg, As 0.002
mg/kg, Sr 0.08 mg/kg, Hg 0.01 mg/kg,
Cd 0.05 mg/kg.
The ash from the lignite-fired
power plant contains Teachable TDS up
to 1000 mg/kg, and respectively Cl
100 mg/kg, SO, 2000 mg/k§» Cu 0.75
mg/kg, Cr 0.93 mg/kg, Pb 0.8 mg/kg,
Ni 1.0 mg/kg, Zn 5.2 mg/kg, Co 1.8
mg/kg, Cd 0.18 mg/kg, and Mo 0.75
mg/kg. These figures refer to the
fresh ash. After disposal, the
Teachable heavy metals change from
easily Teachable oxides to less
TeachabTe carbonates. In some pTaces
where metalTurgicaT discards are
stored in open pits, high content of
heavy metaTs, for example, Cr up to
10 mg/kg have been found. The dome-
stic household wastes contain plas-
tic, cans, paper, food, ash, etc.,
and may contain smaTT amounts of
hazardous waste disposed of with the
househoTd waste, for example; pesti-
cides, soTvents, and batteries. To
concTude; in European countries the
hazardous toxic wastes are not dis-
posed of in the open pits. OnTy Tess
toxic wastes are aTTowed; however
Targe voTumes of these wastes (up to
a few million tons per year in one
operation) can be very hazardous,
even if toxicity per one ton of waste
is relatively low.
Waste and Site Investigation
In all countries waste must be
investigated prior to obtaining per-
mission for its disposal. Results of
the investigation have to be included
in the application submitted by the
company responsible for the disposal.
In the waste investigation, special
emphasis is put on their future reuse
(GDR), to ascertain its risk of
toxicity, and the categorization of
substances containing harmful
metalloids and metals that could have
effect on groundwater (all coun-
tries). In Poland, for example,
there are no precise regulations
unequivocally defining the waste
investigation techniques. The plan-
ning organization recommends an
investigation in relation to the
given site and disposal method. The
investigation of precise chemical
composition is not recommended, but
investigation of waste Teachability
under the conditions within the
disposal area are recommended. Tests
under full saturation and/or only
periodical leaching by rain water,
are recommended. It is possible to
obtain quick results from glass
columns (diameter of 10 cm and 1 m
long, and intensive leaching for 72
or 96 hours. More precise results
can be obtained in a lysimeters study
(e.g., lysimeters of a diameter of
1 m and 3 - 4 m high), which are
leached in the natural rain condi-
tions during the whole year. The
chemical analysis of leachates shall
cover alT possible components that
can be observed, incTuding heavy
metaTs and organic compounds. All
possible toxic compounds are Tisted
on the speciaT official Tist. Basic
site investigation is undertaken
before open pit mining operations
commence, to determine the hydro-
geoTogicaT conditions in the vicinity
and they are usuaTTy continued during
mining operations. During these
operations other experiences aTso are
gained, which can be used in the dis-
posaT site studies. Some additional
site investigation is aTso done prior
to disposaT if the information from
the mining period is not adequate.
In Poland the following data are rec-
ommended for site evaluation.
1. Characteristics of aquifer having
contact with waste disposal (thick-
ness, area! extent, hydraulic
conductivity, permeability and
79
-------
specific yield).
2. Distribution of hydrodynamic
heads,
3. Chemical composition of natural
groundwater examined during the whole
year prior to disposal (seasonal
changes in the aquifers close to the
surface are significant),
4. Lithology of aquifers in the
aspect of absorption and ion
exchange,
5. Detailed description of the open
pit with respect to permeability of
its bottom and slopes,
6. Identification of climatic condi-
tions, especially water balance, and
rain water infiltration rate.
7. Present and future use of ground-
water in the disposal area.
All the above data regarding the
waste and open pit characteristics
must be included in the application
for disposal permission. In GDR, a
special official certificate is to be
prepared for the locations.
Disposal Techniques
Deep mine colliery waste and
discards are handled by train cars or
trucks and spread with use of bull-
dozers. Ash from coal-fired power
plants is handled as slurry by
pipelines, as dry, by special tight
train cars or trucks (then spread
with water jet) and as semi-dry by
belt conveyors (spread with
spreaders). Domestic waste is
contained in plastic bags (UK).
Industrial waste, other than building
rubble, that can cause pollution of
groundwater is contained in sealed
steel drums, sometimes pvc.
Dry wastes are always disposed
of by "layer tipping." The thickness
of layers doesn't exceed 1 m (UK) and
about 5 - 10 m (Poland). Layers are
compacted and sealed if necessary.
In cases where spontaneous combustion
is expected, the layers of waste are
always sealed with clay. A very
special technique of ash disposal in
an active open pit has been applied
in Poland. In the Belchatow mine-
mouth operation, 38 million tons of
lignite mined from one open pit is
fired in the nearby power plant of
4320 MW capacity. About 4 million
tons of fly ash containing Teachable
sulphates and heavy metals has to be
disposed of within the internal dump
of a still active open pit. The open
pit is 200 m deep and both of its
slopes and bottom are in 70 percent
sand. The groundwater table is a
widespread aquifer (both in over-
burden and under the mined lignite
seam) and has been drawn down by 200
m for mining operations. It is kept
below the pit bottom. Mine water is
discharged, after treatment, to a
protected river. After mining is
completed, the groundwater will again
saturate the overburden and ash
disposal. Thus the aquifer must be
protected forever. Sealing of such
huge open pits (dumping of about 130
million m of overburden plus 4 mil-
lion tons of ash per year) is impos-
sible. An extensive study of the
problem showed that the waste should
not be slurry-dumped due to the
excess of polluted water, and its
possible leaching to the aquifer; and
not dry-ash dumped due to the dust
that would be created. The solution
was a mixing of ash with 30 percent
water in special screw mixers. It
provides the consistency of a wet
non-dusting substance. During about
30 minutes of handling on belt con-
veyors, the consistency becomes that
of a slightly cemented globule. Two
thirds of the ash is mixed in the
proportion of 1.5 with clayey (70
percent) overburden. When this over-
burden is disposed of on special
selected benches of the internal dump
80
-------
^ underlain mostly by clays) it forms
an almost impermeable structure.
This way groundwater is protected now
and for the future.
Protection Measures
Sealing of open pits is a major
groundwater protection measure. Two
types of sealing are in use. The
first one is to cover the bottom and
slopes of the open pit with a clay
blanket. In the UK the clay blanket
is about 1 m thick to prevent both
the ingress of groundwater and the
aquifer pollution. On completion of
waste disposal the surface is sealed
with about 1 m of clay blanket that
is covered with subsoil, topsoil, and
revegetated. In the GDR the thick-
ness of a clay blanket must be at
least 2m and it must have
permeability of k = 10"8 m/sec. On
the slopes it must be at least 1 m
above the highest natural groundwater
table. In Poland there are no
regulations about the sealing layer
thickness, but in the plan its
efficiency must be proved by hydro-
geological and geomechanical calcula-
tions. In one of the completed lig-
nite open pits (40 m deep) the sandy
bottom was sealed with 5m of clay,
but in this case clay was easily
available from a nearby active open
pit. The second type of sealing used
in Poland is the vertical cutoff
slurry wall constructed down to a
continuous impermeable strata. The
strata must be continuous, at least 2
m thick, and have a permeability of
at least 10"7 m/sec. Its presence
must be proven by site examination.
The depth of slurry cutoff walls can
be up to 25 m with a thickness of 0.5
- 1.0 m, and a required permeability
of 10"8 m/sec. An example of this is
a disposal site of waste from a coal
desulphurization plant located in the
vicinity of the aquifer used as a
potable water supply for a large
city. For neutralization of some
kinds of wastes the injection of
approved chemicals is in use. A suc-
cessful case in the UK is the injec-
tion sealing of deep colliery waste
that was polluting the surrounding
water. The operation was run in two
stages. To a depth of up to 7 meters
a 2.1 lime/cement slurry was injected
under pressure through boreholes
drilled in 2 m pattern. Deeper por-
tions, below 7 m were sealed by
injection of pure lime milk in a
dosage of 5 tons per hole. This
procedure improved the quality of the
discharge to acceptable limits.
Surface water impounded within the
waste disposal is always under
control. In case of hydraulic
disposals the polluted water is in a
closed circuit. The excess of
polluted surface water is pumped to
surface lagoons and treated before
discharge to natural streams. This
water must meet the required
standards in all countries. For
example in the UK it is TDS below 75
mg/dm3, zinc below 0.5 mg/dm3,
mineral, oils and hydrocarbons below
5 mg/dm3 and pH between 6 and 9.
In all countries the surface of
the filled abandoned open pit must be
sealed and revegetated. In one coun-
try (GDR) the abandoned open pits
cannot be used for waste disposal if
they are closer than:
- 2000 m to facilities of public
health and food production and,
- 1000 m to residential areas, rec-
reation and sports facilities.
Beside the technical measures dis-
cussed above, in some countries,
e.g., Poland, the company intending
to dispose waste in an abandoned open
pit must pay to the local community
10 USD/T for the waste of category I
(the most hazardous) $2.5/T for the
waste of category II, $0.50/T for the
category III (fly ash) and $0.25
USD/T for the category IV (coal
refuse).
81
-------
Monitoring
Monitoring of groundwater, sur-
face water and sometimes air is
required by law. It is effected
mostly by the waste disposal operator
under the supervision of state or
local environmental authorities.
Groundwater is monitored in
piezometric tubes. For sealed dis-
posal sites in Poland the monitoring
wells are located around the dis-
posal :
- 20 m up gradient of the disposal
site
- 30 m from the disposal site in an
intermediate zone
- 50 m from the disposal down
gradient.
In the case of unsealed dispo-
sals which can produce some ground-
water pollution, the monitoring wells
are recommended to be situated downs-
tream 1-4 radial lines:
1 st well 50 - 100 m from the edge
2nd well 100 - 300 m from the edge
3rd well 400 - 700 m from the edge
4th well 800 - 1500 m from the edge
Sampling is recommended once per
month to once per three months. The
same frequency of monitoring is
recommended for surface water out-
f1ows.
Air and plants are monitored,
but there are no detailed regula-
tions. Requirements of local author-
ities determine the type and fre-
quency of monitoring.
Environmental Impacts
1. Air Pollution
It has been found that air is
polluted at a distance of up to 500 m
from disposal sites. The permitted
standards for dust fall
(250T/km2/year) and admissible
concentration of suspended dust in
air (instantaneous value 0.5 mg/m3
and 0.022 mg/m3 - average during the
entire year) have been exceeded at
that distance.
2. Groundwater Pollution
Groundwater pollution around a
coal waste disposal site located in
an abandoned open pit is given in the
table below:
Desig-
nation
PH
Con-
ducti
vity
TDS
Cl
so4
Na
K
Ca
Mg
Mn
Fe
total
NNH,
P04
CM
Phe-
nols
Al
Zn
Cu
Pk
Cr
As
Sr
Mg
Cd
Ho
B
Unit
us/cmr
mg/dnr
»
it
"
it
ii
ii
»
ii
n
it
n
«
n
n
n
n
n
11
n
«
ii
n
11
Average
concen-
tration
prior to
disposal
6.66
247.1
169.2
15.08
54.1
7.84
2.77
16.26
4.95
0.24
4.60
0.43
0.014
0.0049
0.0034
0.16
0.360
0.023
0.0165
0.0064
0.0168
0.130
0.630
0.0024
0.0148
0.032
Average
concen-
tration
after
disposal
6.25
460.72
329.13
40.84
117.98
33.50
5.51
34.11
10.23
0.266
3.7433
1.22
0.0244
0.0059
0.0036
0.181
0.1672
0.0102
0.246
0.0056
0.0274
0.1472
0.6294
0.0037
0.0083
0.0685
Maximum
concen-
tration
after
disposal
6.88
801.0
550.07
72.73
209.89
81.99
11.31
53.60
17.39
0.79
8.75
2.47
0.053
0.0172
0.0066
0.444
0.497
0.0313
0.047
0.075
0.057
0.216
1.300
0.0058
0.024
0.095
3. Surface Water Pollution
Surface water pollution is illustrated by the
examples given below of leachates from a coal waste
disposal:
Disposal No. 1
mg/dm
Effluent
at the slope
foot
TDS 26999
Cl 15200
SO, 1277
Ca 534
Mg 380
Disposal No. 2
mg/dm
25238
12125
99
820
500
82
-------
Surface
water flow
TDS
Cl
SO,
Ca
Hg
5100
520
900
120
220
2756
1046
670
no data
no data
5. Soils and Plants Contamination
So far, soil and plants contami-
nation by disposal of wastes in open
pits below the terrain surface has
not been investigated enough to draw
any reasonable conclusions. The
observations, analyses, and measure-
ments carried out far above terrain
surface show contamination of soil
and plants. It is a different quan-
titative phenomenon and conclusions
resulting from it cannot be applied
to waste disposal below the surface
(in open pit).
All of the above figures and
facts show that even non-hazardous
waste (in US categorization) can be
harmful to the environment, espe-
cially when disposed of in large vol-
umes, and in open pits where the con-
tact with groundwater is easy. They
prove also that abandoned open pits
must be sealed before disposal of
waste if contact of waste with
groundwater and surface water is pos-
sible, and groundwater quality is
under protection.
Research Needs
Research should be focused on:
- increase of reuse of mining and
metallurgical waste to diminish
their volume and make them less
hazardous to the environment,
- different leaching procedures for
more accurate evaluation of waste
toxicity,
- improvement of open pits explora
tion,
- chemical treatment of waste (for
example, by injection) for
neutralization and reduction of
permeability,
- and sealing efficiency.
ACKNOWLEDGEMENTS
We appreciate the use of the
national reports provided by ECE,
Group of Experts on Opencast Mining,
and especially from the UK, GDR,
Czechoslovakia, and Poland.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
83
-------
DESICCATION AND PERMEABILITY OF SOIL BENTONITE MATERIALS
Raj P. Khera
Hemendra Moradia and Mahendraratnam Thilliyar
Department of Civil and Environmental Engineering
New Jersey Institute of Technology
Newark, NJ 07102
ABSTRACT
A series of experiments were conducted to study the effect of
drying on the backfill materials constituted from bentonite and
other soil components. To simulate the field conditions, one half
of a specimen was subject to cycles of drying and wetting while
the other half was kept dry. Solutions of aniline, phenol and
hydrochloric acid were used as the liquids. Material containing
10 percent bentonite showed the most cracking. The largest cracks
were formed with phenol and the smallest with water. Hydrochloric
acid showed no cracks. The cracks were smaller when the soil
mixture consisted of 5 percent bentonite, 10 percent kaolinite,
and 85 percent sand. The cracks were further reduced in size and
density when the proportion of kaolinite was increased to 20
percent.
All the cracks essentially closed when the specimen was subjected
to permeability tests. Permeability tests on the cracked speci-
mens did show some what higher values for the hydraulic conduc-
tivity. The permeability ratio of cracked to uncracked specimens
ranged from about 2 to 25.
INTRODUCTION
Seepage has been success-
fully controlled in many field
problems by providing a soil-
bentonite backfill in slurry
walls. Such walls are effective
in controlling seepage at un-
contaminated sites. A typical
backfill mix consists of sodium
bentonite, clay, and coarse
grained materials. The mix is
designed to achieve a perme-
ability of 10 ' cm/sec or less.
In recent years slurry walls
are being extensively used for
containment of waste sites. A
literature survey shows that
the effect of desiccation on
the integrity of slurry walls
still remains untouched. During
its service life, the portion
of a slurry wall below the
lowest ground water level
remains in a saturated state
while the portion above
undergoes drying and wetting
cycles as a result of ground
water fluctuation and
precipitation. The degree to
which the change in permeabil-
ity in a particular zone occurs
may be affected by the number
of drying and wetting cycles,
amount of precipitation, and
volume of leachate generated.
PURPOSE
The purpose of this research
was to conduct a series of
experiments to study the effect
of drying on the backfill
84
-------
materials constituted from
bentonite and other soil
components. The study was also
to provide information on the
size and distribution of drying
cracks as affected by water and
various organic and inorganic
chemicals
APPROACH
Test specimens were pre-
pared by one dimensional con-
solidation to a maximum pres-
sure of 50 kPa (kilo Pascal).
The consolidated specimens were
70 mm in diameter and 30 to
40 mm in height. For each soil
mix three specimens were pre-
pared for testing with each of
the chemicals. One of the sam-
ples was maintained in a satu-
rated state (S1) and subjected
to permeability tests. The sec-
ond specimen was allowed to dry
in air and afterward placed in
an oven at 40°C for 48 hours
(S2). The third sample was sub-
jected to wet and dry cycles
(S3). After the desiccation
studies these specimens were
used for permeability tests
(1).
Desiccation Study
The proportion of various
soi1-bentonite mixes are shown
in Table 1. The chemicals and
their concentrations are shown
in Table 2.
Table 1. Design Mix.
Symbol Bentonite Kaolinite Sand
(*) (*) (%)
M11 10 90
M5 5 10 85
M6 5 20 75
Table 2. Chemicals
Symbol Liquid Concentration
ppm or N
L2 Aniline 10000
L5 Phenol 10000
L9 HC1 1.0
Bentonite was thoroughly
mixed with the given liquid to
form a slurry and was allowed
to activate for 24 hours. After
that the slurry was mixed with
other soi1 components and
poured into a cylindrical
plastic mould. Each end of the
specimen was covered with a
filter paper and a porous stone
before consolidation.
Tests on dried backfill
materials (2) showed that they
had lower permeability than the
nondried specimens. Since uni-
form drying does not represent
field conditions, the method of
drying was modified. The plas-
tic mould for the desiccation
tests consisted of two
longitudinal halves held
together by clamps. After
completion of the consolidation
the specimen along with its
mold was laid on its side. The
upper half of the mould was
removed and the sample was
allowed to dry out either in
air at room temperature or in
an oven at 40° C. At the end of
the drying period liquid was
added to the tray so that half
the sample was submerged. The
two ends of the specimen were
kept covered with semicircular
filter papers and porous stones
to prevent any possible
sloughing. The time for
completion of one dry and one
wet cycle ranged from 5 to 8
days or more in case of air
drying. The wet and dry cycles
were repeated as desired.
85
-------
The sample subjected to
this form of drying was consid-
ered to represent field condi-
tions for a slurry wall. Fi-
nally the second half of the
cylinder was also removed and
the sample was mounted in a
triaxial cell for the perme-
ability measurements using the
flow pump method.
Permeab i1i ty Measurements
The permeability measuring
equipment consisted of a stain-
less steel flexible wall perme-
ability cell and air pressure
system. Pore pressure parameter
B= 0.95 was used for satura-
tion. Permeability was detei—
mined using a constant rate of
flow (3).
PROBLEMS ENCOUNTERED
In some instances the soil
adhered to the consolidation
mold. The adhesion was overcome
by smearing the mold with sili-
con grease. Because of the low
governing back pressure for the
flow pump, both vacuum and back
pressure were utilized for sam-
ple saturation. Considerable
volume of flow was required to
realize saturation.
RESULTS
Fig. 1 shows the state of
the specimen from mix M11 con-
taining 10 percent sodium ben-
ton ite and 90 percent sand af-
ter the third drying cycle with
water. The first drying cycle
showed only hairline cracks.
After the second cycle the
cracks were more clearly no-
ticeable. But after the third
cycle the cracks had become
fairly wide. The largest crack
developed horizontally along
the wet-dry interface. Most of
smaller cracks intersected the
larger crack. During the wet-
ting cycles the cracks appeared
to close. But they opened up
with greater intensity during
the drying cycle.
Tests with phenol showed
minor cracks at dry-wet inter-
face with the very first dry
cycle (Fig. 2). Cracks in the
horizontal direction became
more extensive at the end of
the second dry cycle. As shown
in Fig. 3 the cracks are also
visible along the side of the
sample indicating that they
extended from one face to the
other. As indicated in Fig. 4
with aniline the cracks were
smaller even after the third
cycle. No cracks were found
with hydrochloric acid, only
the color of the sample had
changed.
There was only negligible
cracking for mix M5 with water
though some sloughing was ob-
served in the zone which was
subjected to frequent wetting.
Only the sloughing appeared to
increase with the number of cy-
cles. As indicated in Fig. 5
phenol produced cracks. They
were visible on both faces of
the specimen. However, the
cracks shown in the Fig. 5 are
for the fifth cycle. When com-
pared with Fig. 2 which shows
cracks for only the second cy-
cle, these cracks are much
smaller. Fig. 6 hows the effect
of aniline during the fifth cy-
cle. The cracks are thin, hori-
zontal and at the dry-wet in-
terface. The lower section
which underwent wetting and
drying shows some cracks too.
The cracks are still smaller as
compared with mix M11 which had
86
-------
undergone only three dry-wet
cycles . The difference in the
cracking behavior may be at-
tributed to the total percent-
ages and the types of clays
present in the two mixes. M11
contained 10 percent bentonite
while M6 had 5 percent ben-
tonite and 10 percent kaolin-
ite. The reaction with hy-
drochloric acid was similar to
that with M11.
There was not much differ-
ence in response of M6 which
contained 25 percent clay as
compared with M5.
In most of the cases the
location and the pattern of the
cracks were typical. Deep and
wide cracks were observed at
the wet-dry interface. The
size, number, and the density
of the cracks increased with
increasing number of cycles.
The organic chemicals showed
larger cracks than water. Most
of the time the cracks appeared
to close upon wetting, but they
became more severe as the wet-
ting was followed by drying. No
cracks could be detected with
hydrochloric acid.
PERMEABILITY TESTS
The results of permeabil-
ity tests are shown in Table 3.
Table 3. Results of Permeabil-
ity Tests.
M6S1L1
M6S2L1
M6S3L1
wet
1.5x10~9
1,6x10~9 dry
1.7x10 9 eye
water
water
water
No.
M11S1L1
M11S2L1
M11S3L1
M11S1L2
M11S3L2
3x10~s
6x10-1°
6x10~9
2x10~9
5x10~8
k,cm/s state
wet
dry
eye
wet
eye
permeant
water
water
water
ani1ine
ani1ine
For comparison purposes,
the permeability of the soil
samples were measured in wet,
dry, and cycled state for some
of the samples. For M11 the
lowest permeability was for the
specimen which was dried uni-
formly. Since drying reduces
the void ratio and the subse-
quent swelling during satura-
tion does not restore the orig-
inal volume of the specimen,
this should be expected. When
the permeability of the speci-
men which was subjected to dry-
wet cycles is compared with
uniformly dried specimen it is
one order of magnitude larger.
Compared with the wet specimen
the permeability of the cycled
specimen is only twice as
large. Considering the large
and continuous cracks exhibited
in Fig. 1 and Fig. 2, the in-
crease in permeability appears
inconsistent. A photograph of
the same specimen (Fig. 1) af-
ter the permeability test is
shown in Fig. 7. Note that none
of the cracks that were visible
in Fig. 1 are discernible. The
combination of swelling during
saturation and the confining
pressure resulted in nearly
complete healing of the speci-
men. Had the cracks been com-
pletely closed the coefficient
of permeability would be of the
same order of magnitude as the
dry specimen.
With aniline the cycled
specimen shows a greater in-
crease in permeability over the
wet specimen (25 times larger)
when compared with water (2
times larger).
87
-------
There was little differ-
ence in the permeability for M6
with water regardless of the
state of the test specimen.
Recall that the cracks with
water were only marginal for
this soil mix. No data is
available on mix M6 with
chemicals.
CONCLUSIONS
Based on the work done,
the following conclusions may
be drawn
1. For bentonite-sand mix the
major cracks developed at
the dry-wet interface. With
the increasing number of
cycles the density and the
size of the cracks in-
creased. During wetting cy-
cles they appear to close.
2. The organic chemicals showed
much larger cracks than wa-
ter. Among the organics
phenol displayed larger
cracks than aniline.
3. The increased percentage of
clay helped in reduction of
the crack formation.
4. The permeability of the cy-
cled samples was generally
higher than the wet samples.
The difference was more for
the organics than for water.
5. The swelling of the cycled
specimens during saturation
combined with cell confining
pressure resulted in closing
of most of the cracks.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
ACKNOWLEDGEMENTS
This research was funded
under Grant SITE 13, by the
Hazardous Substance Management
Research Center, a National
Science Foundation Indus-
try/University Cooperative
Center, and a New Jersey
Commission on Science and
Technology, Advanced Technology
Center, and the Department of
Civil and Environmental
Engineering, at the New Jersey
Institute of Technology,
Newark, NJ. Their support is
gratefully acknowledged.
REFERENCES
1. Khera, Raj. P., Thilliyar,
M., and Moradia, H.,
"Cracking of Backfill
Materials In Soil Ben-
tonite Walls", Report
SITE-13, NSF Industry/
University Cooperative
Research Center, NJ Comm.
on Sci & Tech. Adv. Tech.
Center, NJIT, Newark, NJ,
Nov. 1988.
2. Khera, Raj P., Wu, Y. H.,
and Umer, M. K.,
"Durability of Slurry Cut-
Off Walls around the Haz-
ardous Waste Sites,"
Progress Report SITE-8,
NSF Industry/University
Cooperative Research Cen-
ter, NJIT, Newark, NJ,
Sept. 1986.
3. Olsen, H. W., Nichols, R.
W., and Rice, T. L., "Low
Gradient Permeability
Measurements in a Triaxial
System," Geotechnique, V.
35, No. 2, 1985, pp. 145-
157.
88
-------
Figure 1. M11-S3-L1
after 3 dry-wet
cycles with water
Figure 2. M11-S3-L5
after 3 dry-wet
cycles with phenol
Figure 3. Side view
of M11-S3-L5
» $ / ^*_ - '5.V
v J* . . 1 «, .
89
-------
Figure 4. M11-S3-L2 after 3 dry-wet cycles with aniline
Figure 5. M5-S3-L5 after 5 dry-wet cycles with phenol
90
-------
Figure 6. M5-S3-L5 after 5 dry-wet cycles with aniline
Figure 7. M11-S3-L1 after permeability test,
91
-------
DESIGN OF CLAY LINERS TO MINIMIZE SHRINKAGE CRACKING
Miguel Picornell and Mohd Zaifuddin Idris
Department of Civil Engineering
University of Texas at El Paso
El Paso, TX 79968-0516
ABSTRACT
Clay liners have been used to store hazardous liquid wastes or as back-
up to plastic liners. The design of the clay liner mix is normally based
on achieving a sufficiently low coefficient of permeability. This has been
sometimes approached by mixing native soils with bentonite. Although the
bentonite addition does reduce the permeability of the mix it also causes
an increase of the shrinkage potential of the liner mix. This study has
been focussed on the feasibility of reducing the shrinkage potential of the
liner mix, while maintaining a sufficiently small permeability coefficient.
The variables considered are the clay mineral used in the mix and the
addition of coarse aggregate to form a stiff framework of grains that would
resist the volume change imposed on the liner. The study also addressed
the selection of the best gradation of coarse aggregate that would minimize
the shrinkage potential.
For these purposes, specimens prepared by mixing bentonite or kaolinite
with three alternative gradations of coarse aggregate were compacted under
optimum conditions. The shrinkage strains experienced upon air drying of
the specimens were measured, duplicates of these specimens were subject to
permeability tests, and to a cycle of consecutive swell and shrink. The
formation of cracks in the specimen subjected to the swell/shrink cycle was
investigated by visual observation of the split specimens after staining
the crack walls with methylene blue.
The results of this study indicate that the addition of inert clays or
coarse aggregate to a liner mix are feasible means to reduce the shrinkage
potential of the mix while maintaining a reasonably low permeability. The
best gradation of coarse aggregate was found to be the gradation of the
theoretical Fuller's curve for the maximum size of aggregate being used.
INTRODUCTION
The containment of hazardous
liquid wastes is one of the most
troublesome and certainly most
urgent problems facing the
industrial community. Existing
regulations prohibit the disposal of
free liquids unless the landfill has
an adequate liner. The availability
of. suitable clayey soils in many
regions of the United States makes
the use of clay liners an
economically attractive solution.
Furthermore, clay liners have and
can always be used as back-up for
92
-------
synthetic liners.
Clay liners have been normally
built using native soils mixed with
bentonite and compacted to form the
liner. The bentonite is used to
reduce the permeability of the liner
mix to some small value. At
present, the design of the liner mix
is based on achieving a material
with the desired permeability
coefficient. However, the clay
particles of bentonite are highly
susceptible to changes in moisture
content, to capillary forces, and to
physico-chemical interactions with
the liquid waste. Due to these
effects, the properties and the
structure of a compacted liner can
change with time, resulting in a
reduction of the effectiveness of
the liner as a barrier to contain
waste.
The deterioration of the liner
can be the result of volumetric
changes imposed by climatic wet-dry,
or freeze-thaw cycles (1), or
volumetric changes associated with
the interaction of clay particles
with the liquid waste (2).
The designer has practically no
control over the magnitude of the
changes that will be imposed on the
liner; that is, the degree of
desiccation the liner will undergo,
or the concentration of the stored
liquid waste. However, the designer
can influence the formation and the
extent of shrinkage cracks by
manipulating the composition of the
liner mix, in order to reduce the
shrinkage potential of the compacted
liner to a minimum.
PURPOSE
The primary goal of this study
was to evaluate the feasibility of
choosing liner mixes that minimize
the shrinkage potential of the liner
while the permeability is kept at
some reasonably small value, such as
1 x 10 cm/sec (3).
The most obvious variable.
available to reduce shrinkage of the
liner mix is the type of clay
mineral used to mix with the soil.
The shrink/swell behavior of
compacted clayey soils has been
related to the predominant clay
mineral (4). The study has focussed
on comparing the permeability and
shrinkage behavior of liner mixes
prepared with commercially available
bentonite and kaolinite.
The inclusion of large sizes of
inert ballast has been shown (5) to
.decrease the shrinkage potential of
the soil mix. At the same time, the
inclusion of large grain sizes is
known to decrease the permeability
of the mix (6), provided that the
large size particles do not dictate
an increase in the pore volume of
the soil mix. The second goal of
this study has been to evaluate the
effect of large particle aggregates
on the permeability and shrinkage
potential of the soil mix and the
selection of a best gradation that
would maximize the improvement of
both properties.
APPROACH
Materials
The specimens tested were
prepared by mixing clay with coarse
aggregate. The coarse aggregate was
obtained by sieving crushed gravel
and sand into five size intervals.
The effect of the gradation of the
coarse aggregate was evaluated by
preparing and testing triplicate
specimens with three different
gradations. The specimens of Batch
1 were prepared with the gradation
of Fuller's curve for the maximum
size aggregate of 3/4 in. The
93
-------
specimens of Batch 2 and Batch 3
were prepared with gradations on
either side of Fuller's curve. The
coarse aggregate compositions of the
three batches are presented in Table
JL •
Table 1. Coarse Aggregate
Percentile Composition
of the Three Batches
Particle Batch Batch Batch
Size Fraction 123
3/4 "-3/8" 31.37
3/8 "-No. 4 21.17
No.4-No.10 18.62
No.lO-No. 40 18.79
No. 40-N0.200 10.05
46.94 22.99
18.37 17.24
14.29 20.69
13.27 22.99
7.14 16.09
One set of batches was prepared
by mixing the coarse aggregate with
pure sodium-base Wyoming bentonite.
A^second set of batches was prepared
with a local low shrinkage clay
"kaolinite" that exhibits a liquid
limit of 35.5 and a plastic limit of
20.1. Hydrometer analyses indicated
that the kaolinite contained 30% of
clay-size particles, 55% of silt-
size particles, and 15% of fine
sand. Although no mineralogical
identification was attempted, the
Skempton's activity coefficient of
0.5 is typical of a kaolinite (7).
Compaction Tests
For each trial mix, a compaction
test was performed to determine the
maximum dry unit weight and the
optimum moisture content. All
specimens for subsequent testing
were prepared by compacting the mix
at these optimum conditions.
A standard Proctor compaction
test in a six-inch mold was
performed for each trial mix. The
mold was filled in three lifts, and
the amounts of coarse aggregate and
clay for each lift were weighed,
mixed, and wetted separately to
reduce the risk of segregation.
Shrinkage Measurements
Specimens of each trial mix were
compacted, extruded from the mold,
and left to air dry in a 30%
relative humidity controlled room.
The _ shrinkage of the specimen was
monitored with a dial gage measuring
the axial deformation, and a caliper
was used to measure the changes in
diameter. The drying process was
prolonged until the dial gage
measurements indicated that the
volume changes had stopped.
Permeability Tests
The permeability of the trial
mixes was determined on specimens
compacted inside a section of a six
inch PVC pipe. The specimen was
then allowed to saturate by ponding
permeating fluid in the pipe for
about a month. Then the pipe was
capped and connected to a standing
burette to perform a series of
falling head tests. The permeating
fluid used was a 0.01 N calcium
chloride solution. The hydraulic
gradients during the tests ranged
from 6 to 8. The tests were
prolonged until a nearly constant
coefficient of permeability was
obtained in consecutive
measurements. This period of time
ranged from. 3 to 4 weeks.
Consecutive Swell/Shrink Behavior
Specimens of the trial mixes
were subjected first to swelling and
then to a shrinking cycle. The
swelling and shrinkage strains were
monitored and the cycles were
prolonged until the deformations had
stopped. The comparison of the
swelling and shrinkage strains
allowed an indication of whether the
swelling process imposed permanent
changes on the compacted specimen.
After the shrinkage cycle, the
specimens were stained with a
solution of methylene blue deposited
inside the shrinkage cracks with a
94
-------
burette. After staining and ait-
drying, the specimen was split and
the length of penetration of the
stains through the specimen or the
contact specimen-mold was recorded.
PROBLEMS ENCOUNTERED
One problem encountered in the
measurement of shrinkage strains
upon air drying the specimens was
the axial deformation of the
specimen due to its own weight. To
assess the importance of this
effect, concurrent measurements of
the diametral deformation of the
specimen were recorded. The
deformations due to its own weight
were very important in the specimens
subject to swelling. This imposed
the need to perform the swelling
phase on specimens inside molds.
RESULTS
Specimens of each batch were
prepared by mixing the fixed coarse
aggregate gradations with increasing
percentages of clay. A total of 20
trial mixes with bentonite and 28
with kaolinite were tested.
However, some of the trial mixes did
reach the desired coefficient of
permeability, and further testing of
the trial mix was cancelled. Due to
space constraints, the results
presented in this paper include only
the trial mixes that were subject to
all the tests described. The
complete set of all the tests
performed have been presented
elsewhere (8). A summary of the
tests results on bentonite trial
mixes is presented in Table 2.
Table 3 summarizes the test results
on kaolinite trial mixes.
The maximum dry unit weights,
obtained in the compaction tests,
decreased for increasing clay
content. This trend was very
apparent in the bentonite trial
mixes, but was less apparent for the
kaolinite trial mixes shown in Table
3. However, the whole set of
compaction test results on kaolinite
mixes showed a slight increase of
the maximum dry unit weight at low
clay contents and, then, a general
decrease towards larger clay
contents.
The maximum dry unit weights
obtained on Batch 1 specimens with
bentonite were clearly larger than
for the other batches. This
indicated that Fuller's _curve
provides the more compact specimens;
that is, it allows packing the
maximum amount of clay and aggregate
into the mix. The dry unit weights
of kaolinite mixes were consistently
larger than for bentonite mixes.
This effect is related to the lower
affinity for water of the kaolinite
particles, "thus, allowing a more
compact specimen.
The permeability of bentonite
specimens with small clay contents
was very large. To reach an
acceptable permeability coefficient,
it was necessary to mix in about 6%
clay for the gradation of Batches 1
and 2, and about 8% for the
gradation of Batch 3. The addition
of more clay did not reduce
significantly the permeability of
the mix. The differences in
permeability shown in Table 2 are
not significant and are within the
accuracy of the determination.
The permeability of kaolinite
specimens was larger than for the
bentonite specimens. However, all
batches showed a clear pattern of
decreasing permeability with clay
content, and with sufficiently large
clay contents, they reached
comparable permeabilities to the
bentonite specimens. The change
from high permeabilities to
acceptable permeabilities for a
liner was more gradual for the
kaolinite specimens than for the
95
-------
bentonite specimens. This suggests
that defects of mixing of the soils
in the field could have a much less
drastic effect when kaolinites are
used instead of bentonites. The
permeability tests on kaolinite
specimens of Batch 1 achieved an
acceptable (3) permeability with
smaller clay percentages than
Batches 2 and 3. This indicated
that the gradation of Fuller's curve
provided the best liner mix. The
permeabilities of the specimens of
Batch 3, corresponding to the more
sandy gradation, were the largest
and never reached an acceptable
permeability even for the largest
clay contents used.
The specimens left to air dry in
a controlled environment experienced
axial strains consistently larger
than the transversal strains. This
is believed to be the result of
axial deformations of the specimen
due to its own weight. Both sets of
measurements indicated that the
specimens decreased in length and in
diameter during the drying process.
For all batches and clay types,
Table 2. Summary of Test Results on Bentonite Trial Mixes
Batch Percent Maximum Stable
No. Clay Dry Unit Permeability
( % ) Weight ( cm/min . )
(pcf)
1
1
1
1
1
2
2
2
2
2
2
2
2
3
3
3
3
3
6.05
7.92
9.79
11.4
13.1
5.77
7.96
9.93
10.9
12.5
14.0
15.5
16.9
8.0
10.0
12.0
14.0
16.0
128
126
126
126
126
124
123
121
119
118
118
116
116
123
121
120
117
116,
.0
.7
.2
.1
.2
.4
.0
.0
.1
.8
.3
.0
.3
.1
.2
.9
.5
.2
5.25
4.75
5.25
1.10
7.90
8.05
4.05
3.90
2.45
7.20
5.80
4.90
4.70
6.25
6.50
1.05
9.45
5.80
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
10- '
10~7
10~7
io-6
10~7
io-7
io-7
10~7
io-7
io-7
io-7
10~7
io-7
io-7
io-7
10"6
io-7
io-7
Strains Upon Strai
Air Drying Swell/
Axial Trans.
(%) (%)
0.48
0.90
0.81
1.12
2.22
0.50
0.20
0.74
0.30
0.65
0.13
0.74
2.75
1.29
1.59
0.73
1.15
1.10
0.20
0.37
0.32
0.37
0.50
0.18
0.18
0.32
0.33
0.27
0.47
0.20
0.45
0.40
0.72
0.43
0.62
0.77
Swell
(%)
6.99
11.05
12.57
11.81
13.91
5.90
14.63
13.52
14.69
17.53
18.04
17.08
16.12
9.00
10.16
13.37
17.41
17.99
.ns in Crack
'Shrink Continuity
Shrink
(%) (%)a
4.30
5.99
6.27
7.75
11.61
2.47
6.52
5.17
8.50
10.03
10.00
11.86
6.62
6.02
5.96
7.57
8.73
9.39
35.0
100
100
100
100
100
100
100
100
100
100
100
100
100
'100
100
100
100
Indicates the length of the specimen stained with methylene blue as a
percentage of the total length of the specimen.
96
-------
these measurements indicated that
the shrinkage potential of the mix
increases with increasing clay
contents. These results illustrate
that the skeleton of aggregate was
not able to completely prevent the
shrinkage of the specimens. In
order to achieve this goal, these
results suggest that higher
compactive efforts than the standard
Proctor might be necessary. For
comparable clay contents, the
kaolinite specimens experienced
smaller strains than the bentonite
specimens. Thus, the use of
kaolinite can help to reduce the
risk or the extent of cracking of a
liner.
The swelling strains of all
three batches of bentonite specimens
increased with increasing clay
content. By way of contrast, the
swelling strains of kaolinite
specimens did not exhibit this
trend. The shrinkage strains,
experienced by the specimens after
the swelling phase were consistently
lower for both clays and for all the
batches. This suggests that the
swelling process caused some
irreversible changes on the
specimen, thus implying that
successive cycles of wetting and
drying of a liner can induce a
progressive change of the compacted
soil structure. Although, both clay
types exhibited these structural
changes, both the swelling and
shrinking strains experienced by the
kaolinite specimens were smaller
than those experienced by the
bentonite specimens. Therefore, a
liner built with kaolinite would be
less affected by successive wet/dry
cycles. The swelling and shrinking
Table 3. Summary of Test Results on Kaolinite Trial Mixes
Batch
NO.
1
1
1
2
2
2
2
3
3
3
3
3
Fine Maximum Stable
Grained Dry Unit Permeability
Fraction Weight (cm/min.)
(%) (pcf)
15.30
17.00
18.70
11.93
13.20
14.42
16.15
8.50
10.20
11.90
13.60
17.00
133.5
133.4
133.4
137.3
137.3
135.5
139.5
133.5
133.5
134.2
134.2
136.1
9.50
7.00
3.50
3.09
1.13
1.05
1.50
3.60
1.29
4.53
3.48
1.20
X
X
X
X
X
X
X
X
X
X
X
X
io-b
io-6
io-6
io-2
io-3
io-4
io-6
io-2
io-2
o
10 z
IO-3
10"3
Strains Upon
Air Drying
Axial Trans.
(%) (%)
0.07
1.08
1.20
0.53
0.24
0.58
0.35
0.31
0.21
0.26
0.50
0.76
0.57
0.72
1.02
0.28
0.15
0.52
1.05
0.33
0.32
0.33
0.44
1.00
Strains in Crack
Swell/Shrink Continuity
Swell Shrink
(%) (%) (%)
8.65
7.59
8.38
7.76
13.98
5.95
6.64
2.57
6.47
5.16
4.61
7.91
3
3
3
4
6
2
2
1
2
2
3
3
.90
.58
.69
.21
.73
.35
.83
.02
.96
.53
.07
.11
9.5
100
100
21.0
31.6
26.3
10.0
19.0
18.2
100
28.6
47.6
a
indicates the length of methylene blue penetration at the mold-specimen
contact as a percentage of the total length of the specimen.
97
-------
strains experienced by the specimens
of Batch 1 were smaller than those
of Batches 2 and 3, thus indicating
that the gradation affects the
potential changes induced by wet/dry
cycles, and that a liner with a
gradation adjusted to Fuller's curve
will be less prone to a progressive
deterioration by successive wet/dry
cycles.
After the swell/shrink cycle,
the bentonite specimens exhibited
very apparent cracks uniformly
distributed through the exposed
surface of the specimen. By way of
contrast, the kaolinite specimens
did not crack, with only three
exceptions of specimens showing a
small crack at the contact between
the specimen and the mold, when the
methylene blue solution was
deposited in the cracks of (almost
all) the bentonite specimens, it
moved through the crack very fast
and appeared at the bottom of the
specimen, thus indicating that the
shrinkage cracks extended through
the length of the specimen. From
the observation of the split
specimens, it was determined that
the methylene blue had moved through
the soil mass in some specimens and
inBothers, after moving through the
soil mass a fraction of the length
of the specimen, it had moved to the
contact mold-soil. The methylene
blue solution did not penetrate
through the soil mass of the
kaolinite specimens in any
measurable amount; the only observed
path of penetration was at the mold-
soil contact. These observations on
the cracks of the specimens after
the swell/shrink cycle are another
indication that a liner built with
kaolinites would be much less
susceptible to be damaged by wet/dry
cycles.
The larger cracks developed
through the bentonite specimens are
attributed to the extrusion of the
bentonite out of the pores of the
aggregate during the swelling phase.
It was observed that, during the
swelling phase, a soft layer of clay
particles with a consistency of a
gel was forming on the top of the
bentonite specimens. This layer was
not observed in the kaolinite
specimens.
In summary, the results of the
compactions tests indicate that the
use of kaolinite, and adjusting the
gradation of the liner mix to
Fuller's curve, can provide a more
dense liner; the permeability tests
results indicate that it is also
possible to achieve acceptable
permeabilities for the liner when
using inactive clays such as
kaolinites; the shrinkage strain
measurements indicate that a liner
built with kaolinites will
experience smaller strains upon
drying and thus, will be less likely
of cracking; the strains recorded
during the swell/shrink cycles
indicate that a liner built with
kaolinite will experience smaller
progressive damage and this will be
further reduced if the gradation of
the mix is adjusted to Fuller's
curve. These considerations
indicate that adjusting the
gradation of the liner mix to
Fuller's curve and using inactive
clays such as kaolinites are
feasible means to reduce the risk of
liner damage by shrinkage cracking,
while still achieving acceptable
permeability coefficients.
ACKNOWLEDGEMENTS
This research was supported by
the Engineering Foundation Grant RI-
A-86-2. This support is gratefully
acknowledged. Qusai s. El-Jurf and
Syed Hussaini helped in performing
the work reported. Their good will
and enthusiasm is also acknowledged.
98
-------
REFERENCES
1. Brewer, R., 196 4, "Fabric and
Mineral Analysis of Soils,"
John Wiley and Sons, Inc., New
York.
2. Anderson, D.C., 1981, "Organic
Leachate Effects on the
Permeability of Clay Soils,"
M.Sc. Thesis, Texas A&M
University.
3. Brown, K.W., and D.C. Anderson,
1983, "Effect of Organic
Solvents on the Permeability of
Clay Soils," Report No. EPA -
600/2-83-016.
4. Yong, R.N., and B.P. Warkentin,
1975, "Soil Properties and
Behavior," Elsevier, New York.
5. Dixon, D.A. M.N. Gray, and A.W.
Thomas, 1985, "A Study of the
Compaction Properties of
Potential Clay-Sand Buffer
Mixtures for Use in Nuclear
Fuel Waste Disposal," Eng.
Geology, Vol. 21, Elsevier,
Amsterdam, pp. 247-255.
6. Johnson, E.E., 1972, "Ground
Water and Wells," Johnson Div.
Inc., Universal Oil Products
Co., St. Paul, Minnesota, pp.
41-43.
7. Mitchell, J.K., 1976,
"Fundamentals of Soil
Behavior", John Wiley and Sons,
Inc., New York.
8. Idris, M.Z., 1988, "Design of
Clay Liners to Minimize the
Effect of Shrinkage Cracks on
the Permeability of the Liner,"
M.Sc. Thesis, University of
Texas at El Paso.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
99
-------
FIELD STUDIES ON THE HYDROLOGICAL PERFORMANCE OF MULTILAYERED LANDFILL CAPS
Stefan Melchior and Gunter Miehlich
Institut fUr Bodenkunde der Universitat Hamburg
Allende-Platz 2, D-2000 Hamburg 13
Federal Republic of Germany
ABSTRACT
Along with the remedial action on the waste disposal site Georgswerder
in Hamburg (FRG), a research and development program with regard to the
hydrological performance of various covering systems is being carried out.
For this purpose, six large-scale test fields with a size of 500 ma each
have been built. All fields are designed as multilayered systems with a
combination of topsoil, drainage layer and barrier system. Compacted clay
liners, combined systems of flexible membrane liners on top of compacted
clay layers and a combined system of compacted clay on top of a capillary
barrier system are tested as barrier system variants on two different
slopes. On these test fields, all discharges (surface run-off, interflow in
topsoil and drainage layer and the leakage through the barrier systems), as
well as soil hydrological and meteorological parameters, are measured.
The tests focus on the comparative evaluation of the long-term effec-
tiveness of the different covering systems and on the physics of water
movement within the different barrier systems. The measurements were
started in the'beginning of 1988. The first results show that up to now all
systems perform better than required by regulations- The course of the
discharges, collected below the barrier systems show a peak during the
first months. This is interpreted as pore water outflow of the compacted
clay layers due to consolidation. The best results are achieved by the
extended capillary barrier systems, where even the pore water flow is
drained off laterally so that there is no infiltration of water into the
disposed waste.
INTRODUCTION
In the Federal Republic of
Germany (FRG), about 45,000 aban-
doned waste disposal sites are
suspected to cause contamination to
the environment (1). About 2,400 of
those sites are concentrated in the
Hamburg industrial district. About
140 of those sites are suspected to
have a high toxic potential. In most
cases, the use of groundwater as
drinking or irrigation water is
threatened and require remedial
action. Very often, highly toxic
industrial waste has been mixed with
municipal waste. Clean-up strate-
100
-------
gies, therefore, have to deal with
large amounts of waste with a
variable composition. In all those
cases where excavation or in situ
treatment are not available due to
political, financial or technologi-
cal limitations, the only remaining
and at least temporarily efficient
strategy is to isolate and drain
the waste. Covering systems play a
central role in those concepts.
Covering systems have to meet
different requirements: (1) the
system itself has to prevent in-
filtration of rainfall into the
waste, enable a controlled gas
collection, and often must be suit-
able for recultivation. In the long
run, a cap must be resistant against
loads and stresses like subsidence,
erosion, biopenetration, desicca-
tion and clogging of the drain
system. (2) Evaluating not only the
efficiency of the system itself but
also the contamination risk of the
whole site, the demand is that the
covering system must be built as
fast as possible. The more compli-
cated it is to plan and to build a
cap, the more rainfall will infil-
trate the waste. (3) In general, a
covering system must be cost-effec-
tive so that it enables the protec-
tion of as many sites as fast as
possible with a given budget.
PURPOSE
Landfill caps usually are
rather complex and expensive due to
the requirements they have to meet
as a system. In most cases, they are
multilayered systems with a combina-
tion of topsoil, drainage layer and
one or more barrier layers. A lot of
information concerning the percola-
tion of compacted clay liners have
been derived from laboratory experi-
ments (summary in 2). The differen-
ces between laboratory data and the
often much higher hydraulic conduc-
tivity observed in the field are
well known (3). But even though a
number of caps have beeii planned
and constructed during the recent
period, only a few studies deal
with the hydrological performance of
different caps in the field (4 - 8).
Almost nothing is known on the
long-term effectiveness of covering
systems. Field data on the water
balance of different systems and on
the physics of water movement within
the barrier layers are needed to
better understand and predict the
efficiency of planned systems, to
enable an improved control of the
factors which determine the ef-
ficiency, and to optimize the design
of covering systems in general.
APPROACH
Along with the remedial action
on the former waste disposal site
Georgswerder in Hamburg, one of the
biggest sites in Europe, six test
fields have been built for this
study. Each test field is 10 m wide
and 50 m long in slope direction
(Fig. 1). The fields are integrated
in the landfill cover, but flow from
the high slope and lateral flow into
the test fields are prevented by a
combination of flexible membrane
liners (HOPE) and clay seals.
Therefore, each test field has its
own water balance. Gravel-filled
underdrains consisting of welded
HOPE liners are installed below the
barrier systems. They enable the
collection and direct measurement of
the barrier system leakage rates.
The test fields are located in two
areas with different slope (F-fields
with 4 %, S-fields with 20 % slope).
The slope and design of the barrier
system are varied in such a way that
the influence of these factors on
the water balance can be determined
101
-------
SOIL HYDROLOGICAL
MEASURING UNITS
HDPE-LINER
CLAY SEAL
TOPSOIL (75cm)
UPPER DRAINAGE
LAYER (25cm)
CLAY BARRIER (60cm)
UNDERDRAIN(20cm)
FLEXIBLE MEMBRANE
LINER (HOPE)
GASDRAIN (35cm)
FORMER COVER
•WASTE
Figure 1. Schematic view on a test field.
separately (Fig. 2). All fields are
designed as multilayer-systems with
a combination of topsoil, drainage
and barrier layers. Due to the
comparative testing concept, the
topsoil and the drainage layer of
all fields are designed alike (75 cm
of sandy loam respectively, 25 cm of
fine gravel, separated by a geo-
fabric to protect the drainage
layer). The barrier systems vary as
follows:
- A compacted clay liner on F1 and
S1 (three lifts of glacial till,
60 cm thick).
- A combined barrier system composed
of a welded flexible membrane
liner (HOPE) on top of a compacted
clay liner (F2 and S2, the clay
liner has the same design as on F1
and SI). The quality of the HOPE
liners has been thoroughly con-
trolled during the construction.
Therefore, they are assumed to be
watertight during the first years
of the study. During that time,
these fields serve as control
plots concerning the percolation
of the barrier system and for that
reason allow the precise detection
of pore water discharge due to
consolidation of the clay liners
102
-------
after the installation of the
upper drainage layer and topsoil.
Standard design of the Georgswer-
der cover (F3). Like F2 and S2,
HDPE liners, however, not welded
but installed overlapping in slope
direction on top of a compacted
clay liner.
Combined barrier system consisting
of a compacted clay layer (40 cm
thick), lying above a capillary
barrier consisting of the capilla-
ry layer (60 cm, fine-grained
sand) on top of the capillary
block layer (25 cm, coarse sand/-
fine-grained gravel, S3). In that
extended capillary barrier design,
the upper clay layer provides a
low intensive infiltration into
the capillary layer. Therefore,
the capillary barrier can operate
under unsaturated conditions using
the wick effect even after heavy
rainstorms or after melting of
snow.
FLAT SLOPE (
F1 F2
F3
i
• STEEP SLOPE (*20X.)
51
52
S3
SCALE in cm
STANDARD DESIGN
GEORGSWERDER COVER
LEGEND:
TOPSOIL
UPPER DRAINAGE
LAYER and
UNDERDRAW
CLAY BARRIER
CAPILLARY LAYER
GASDRAIN
PROTECTIVE LAYER
FORMER COVER
WASTE
ROOT AND RODENT
BLOCK (overlapping)
HDPE-LINER
(welded)
The measuring program covers the
following range:
- The discharges (surface run-off,
interflow in topsoil and drainage
layer as well as leakage through
the barrier system) are collected
and measured individually and with
high temporal resolution.
- In some fields, various soil
hydro!ogical measuring units with
a total of 24 neutron probe tubes
and 531 automatically recording
tensiometers are installed (e.g.,
in Fig. 1) to define moisture
content and hydraulic potential in
high temporal and spatial resolu-
ti on.
- A weather station serves to
measure precipitation, air tem-
perature, humidity, wind velocity,
Figure 2. Layer design of the
fields
test
radiation balance and soil tempe-
rature to determine the water
input by precipitation and all
parameters necessary to compute
evapotranspiration using the
Penman/Monteith equation.
In addition, several soil physical,
chemical and mechanical parameters
are analyzed in the laboratory or
determined by additional experiments
in the field: (e.g., grain-size and
pore-size distribution, saturated
and unsaturated hydraulic conduc-
tivity, bulk density, proctor densi-
ty, plasticity, shear strength, clay
mineralogical composition, inorganic
constituents of the discharges,
103
-------
tracer and infiltrometer experiments
to determine preferential flowpaths
and transit times).
PROBLEMS ENCOUNTERED
Hydrological studies on test
fields in technical scale have to
deal with specific methodological
problems such as boundary effects
and the representation" ty of the
construction methods that are used
(for example, compaction equipment
and quality control). Any kind of
material cutting through liners can
produce undesired preferential
flowpaths. To avoid those boundary
effects as have been shown by Hotzl
and Wohnlich (9), a new boundary
design was developed in cooperation
with the engineering office IGB
(Hamburg, FRG). A combination of
HOPE liners and clay seals above the
barrier systems define the test
field boundary. Therefore, it was
not necessary to cut through the
barrier systems. In addition, it was
possible to construct the clay
liners of the test fields con-
tinuously over all fields with the
same equipment, technique and
quality control similar to the whole
landfill cover.
A large number of technical
problems had to be solved concerning
construction details, measuring
techniques and data collection which
cannot be mentioned here in detail
(e.g., leak-free construction of the
pipes between test fields and
measuring shafts in a way that
allows subsidence; installation of
instruments in the barrier systems
without causing leaks; protection
against lightning). Some new instru-
ments have been developed (e.g.,
pressure transducer tensiometers,
discharge measuring instruments). In
general, the following strategies
have proven to be very important:
(1) a very close quality control
with always at least one person at
the site observing the construction
progress; (2) a redundant measuring
concept concerning instrumentation,
data collection and independent
verification of the data (the
crucial data, for example, are all
measured and registered on two
independent ways, automatically and
manual); (3) a modular setup of the
instrumentation and easy-to-cali-
brate-instruments reduce interrup-
tions of measurements and data
collection due to technical defects.
Due to the bad weather condi-
tions during the construction of the
fields, the three flat fields were
completed later than the steep
fields. This resulted in slightly
different initial soil water con-
tents and germination conditions
for the seeds. Therefore, the spe-
cies composition of the vegetation
is not totally similar on all fields
(yet the type of vegetation, a short
grassland with no shrubs and trees,
is the same).
RESULTS
The research and development
program was started in 1986. It took
one year to plan the test fields and
another year to build them. The
measurements to determine the water
balances of the fields (precipita-
tion, discharges and soil water
contents) started at the end of
1987. Therefore, only preliminary
results can be given.
Looking at the measured dis-
charges, the following phenomena can
be described:
- There is surface run-off under
both slope conditions, though
frequency and intensity are higher
on the S-fields (2.6 % of rainfall
104
-------
on the S-fields, 0.5 % on the F-
fields). Only very little erosion
occurred during the first winter
even though there was almost no
vegetation on the fields in that
time.
Lateral flow within the topsoil
only occurs on the S-fields. It
contributes only 0.8 % to the
total balance.
The widest range of flow inten-
sities can be observed in the
upper drain layers. There is only
little effect of slope concerning
the total volume of flow, but the
flow intensity can be more than
two times higher on the steep
slope.
Fig. 3 shows the amounts of water
that have been collected below the
different barrier systems. The
measurements started during
construction, which ended on the
steep fields in week No. 40 and on
the flat fields in week No. 48. In
Fig. 3a and 3b, the flow rates on
the fields S2 and F2 with a
combination of HOPE and clay
liners are compared with the flow
rates of the respective fields SI
and Fl with just clay liners. Con-
sidering the HOPE liners on S2 and
F2 as impermeable, the black bars
show the pore water flow out of
the compacted clay liners due to
consolidation during and after the
construction of drainage layer and
topsoil. The white bars give the
sum of pore water flow and leakage
through the clay liners of SI and
Fl.
In general, the data show a low
intensity of flow. There is a
significant pore water discharge
during the construction of the
topsoil. The flow decreases in the
following weeks, but one year
later, still exists. No change in
LITER / WEEK
MM / WEEK
*M S2
C3 81: Oh
!: HOPE'Clay Liner I
: Clay Lit.ar I
0.3
0.2
0.1
o • IVrh lllTI'Hll'ITnTmii'irniiiiriiiii'iTmTnTmTnT- o
34 39 44 49 1 6 11 16 21 26 31 36
No. of Week 1987 / 88
LITER / WEEK
MM / WEEK
F2: HDPE'Clay Liner I
Ft; Clay Liner I
0.3
0,1
0 n,, 111, | innnmn'flTOWwmwi'iw WTmiTTi mwnr- o
34 39 44 49 1 6 11 16 21 26 31 36
No. of Week 1987 / 88
LITER / WEEK
MM / WEEK
34 39 44 49 1 6 11 16 21 26 31 36
No. of Week 1987 / 88
Figure 3. Water collected below the
different barrier systems
in liters and millimeters
per week
105
-------
the volumetric water content of
the clay has been observed, but
there still is a pore water
pressure within the systems. The
measured hydraulic gradients
within the clay liners are around
+1.8. Therefore, no steady condi-
tions concerning the percolation
of the clay liners have been
established yet, and values of the
hydraulic conductivity of the clay
barriers on Fl and SI can not be
calculated based on these data. It
can at least be said that there
is percolation through the clay
barriers of these fields in
addition to the pore water dis-
charge because the water volumes
collected on Fl and SI are always
significantly higher than on the
respective fields F2 and S2. The
higher flow rates on the F-fields,
compared with the S-fields are not
resulting from slope, but are
caused by the higher initial
moisture contents of the clay
layers of these fields. Because of
different weather conditions (that
is i.e., more rainfall) during
construction more moisture was
introduced. Nevertheless, it must
be stressed that the total water
volumes collected (pore water
discharge included) on all fields
are much less than the amounts of
water one would expect calculating
the leakage through clay barriers
assuming a hydraulic conductivity
of 1*10-9 m/s and realistic
gradients which have been required
by regulations.
Even better results are achieved
with the extended capillary
barrier system on S3 (Fig. 3c).
Nearly 100% of the water which
infiltrates from the above lying
clay barrier into the capillary
layer moves laterally within the
fine sand and does not pass
through the capillary barrier.
Up to now, all systems are working
satisfactorily. Further measurements
will show how the systems perform on
a long-term basis. On sites where
suitable materials are available
nearby, the extended capillary
barrier might be the best and most
cost-effective type of cap. Further
work will deal with the maximum
drainage length, the influence of
slope, vapour movement and material
properties on those systems. A more
detailed interpretation of the data
of the Georgswerder test fields
showing the water balances, response
times of the different discharges to
rainfall events, and of the spatial
variability of the soil properties
measured during construction is in
progress.
ACKNOWLEDGEMENTS
The studies are supported by
the German Federal Ministry of
Research and Technology (BMFT) and
the City of Hamburg (FHH, Amt fur
Altlastensanierung der Umweltbehor-
de).
REFERENCES
1. Henkel, M.J., 1988, Altlasten
in der Bundesrepublik Deutsch-
land - Eine Zwischenbilanz, In:
Altlasten. Untersuchung,
Sanierung, Finazierung, Brandt,
E. (ed.), Taunusstein, FRG, 25-
35.
2. Grube Jr., W.E., M.H. Roulier
and J.G. Herrmann, 1987,
Implications of Current Soil
Liner Permeability Research
Results, In: Proceedings, 13th
Annual Research Symposium, Land
Disposal, Remedial Action,
Incineration and Treatment of
Hazardous Waste, EPA (ed.),
Cincinnati, OH, 9-25.
106
-------
3. Daniel, D.E., 1984, Predicting
Hydraulic Conductivity of Clay
Liners, J. Geotechnical En-
gineering, 110, 2, 285-300.
4. Andersen, L.J., E. Clausen, R.
Jakobsen and B. Nilsson, 1988,
Two-Year Water Balance Measure-
ments of the Capillary Barrier
Test Field at B^tterup, Den-
mark. Preliminary Results, In:
Proceedings, Impact of Waste
Disposal on Groundwater and
Surface Water, UNESCO Workshop,
Copenhagen, Denmark, 24 p.
5. Cartwright, K., T.H. Larsen,
B.L. Herzog, T.M. Johnson, K.A.
Albrecht, D.L. Moffet, D.A.
Keefer and C.J. Stohr, 1987, A
Study of Trench Covers to
Minimize Infiltration at Waste
Disposal Sites, Final Report,
U.S. Nuclear Regulatory Commis-
sion, NUREG/CR-2478, Vol.2, 122'
p., Washington.
6. Hoeks, 0., A.H. Ryhiner and J.
van Dommelen, 1987, Onderzoek
naar de praktische uitvoerbaar-
heid van bovenafdichting op
afvalstortterreinen, Instituut
voor Cultuurtechniek en Water-
huishouding (ICW) (ed.),
Rapport 21, Wageningen, The
Netherlands, 12 p.
7. Warner, R.C., J.E. Wilson, N.
Peters and W.E. Grube Jr.,
1984, Multiple Soil Layer
Hazardous Waste Landfill Cover:
Design, Construction, Instru-
mentation and Monitoring, In:
Proceedings, 10th Annual
Research Symposium on Land
Disposal of Hazardous Waste,
U.S. EPA (ed.), Cincinnati, OH,
211-221.
8. Wohnlich, S., 1987, Auswirkung
nachtraglicher Grundwasser-
schutzmaBnahmen auf den Wasser-
haushalt von Deponien unter
besonderer Berucksichtigung von
Oberf1achenabd i chtungen,
Dissertation Univeritat Karls-
ruhe, FRG, 269 p.
9. Hotzl, H. and S. Wohnlich,
1987, Experiences with Large-
scale Lysimeters for Deter-
mining the Infiltration Rate,
In: Proceedings, International
Symposium on Groundwater
Monitoring and Management, Vol.
7/II, 1-16, Dresden, GDR.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
107
-------
DESIGN AND CONSTRUCTION OF THE C2-LANDFILL MAASVLAKTE ROTTERDAM
ing. H.L. Sijberden
Rotterdam Public Works
Harbour Engineering Department
Rotterdam, Netherlands
ABSTRACT
By assignment of the AVR Chemie C.V. Rotterdam Public Works started
in 1985 with a design of the C2-Landfill for non-treatable chemical waste
of the so-called middle class (C2). The disposal is under construction
now. The chemical waste will be stored above ground level on the Maas-
vlakte in Rotterdam. The Maasvlakte was created by hydraulic sand filling
of an in part of the Haringvliet estuary. The landfill consists of a con-
crete basin with a volume of 210.000 m3. The dimensions are
50 x 320 x 11 m, the concrete surfaces of the walls are protected with a
HOPE liner. The bottom liner consists of a layer of 100 mm of cast-
asphalt and a crack-bridging membrane which provides a high degree of
safety. A drainage system in and under the basin will be made for trans-
port and control of the leachate.
After the C2-landfill is filled completely it will be covered with a
final layer of two HOPE liners with a bentonite-clay liner in between. On
top of this cover and the concrete basin a sand layer will be applied, so
that a dune is created.
INTRODUCTION
In the Netherlands there are
at present too few possibilities to
dispose of non-treatable chemical
wastes in a sensible, environmen-
tally hygienic manner. The major
part of this type of waste is now
transported to land- fills abroad.
The rest must be stored where it is
produced. In 1984 the Dutch govern-
ment, the municipality of Rotterdam
and eight large industrial companies
established the AVR Chemie C.V. The
storage site for non-treatable
wastes is an essential component of
this company.
Government policy on chemical
waste is based on prevention of the
production of such waste and promo-
tion of its re-use. It is anticipa-
ted that modification of production
methods, new treatment technics and
re-use will lead to a reduction of
the stream of untreatable chemical
wastes. Therefore the C2-landfill is
intended to receive C2-waste for
the coming ten years.
108
-------
PURPOSE
Because storage of C2-chemical
waste at producers' sites involves
an environmental risk, spreading of
such storage activities criss-
cross over the country must be
avoided. Therefore a storage of
nontreatable waste is necessary by
building and operating a secure
landfill at a place that has no
consequences to the environment. To
this purpose the landfill is situa-
ted above ground level.
APPROACH
To build a C2-landfill faci-
lity, a number of permits are re-
quired. When decisions have to be
made on environmental licenses, it
is important to have a good idea of
the consequences of a C2-landfill
to the environment beforehand. The
present Environmental Impact
Statement (EIS) has been drawn up
to aid decision making on:
- only non-treatable chemical waste
of the so-called middle class
(C2);
- compartimentation;
- capacity 210.000 m3;
- filling period about 10 years;
- the waste cannot be driven over
due to high moisture content and
low stability;
- the leachate and percolate see-
page must have been expelled
within a period of 30 years;
- a control and securety system un-
der the basin to monitor ground
water pollution;
- a final minimum distance of
0.50 m between the foot of the
chemical waste and the ground
water;
- a maximum permissible crack width
is fixed at 0,2 mm;
- free rain water penetration in
the C2-waste must be avoided by
movable roofs;
- final cover after filling which
prevents rain water to penetrate
the waste;
- a sand layer will be applied, so
that a dune is created;
- a suitable location.
- dispensation under
Chemical Waste Act;
- the Nuisance Act license;
- the Discharge license in pur-
suance of the Pollution of
Surface Waters Act.
This EIS will be open to in-
spection when the requests for
these licenses are submitted, and
will be discussed during a public
hearing. Only after this, and after
assessment of possible objections
that are submitted, will a decision
be taken about the requests for li-
censes.
Some important features for
making the design of the C2-land-
fill are:
the Dutch DESIGN AND CONSTRUCTION
C2-waste
Non-treatable chemical wastes are
divided into a number of classes,
depending on the quantities in
which they are produced, the treat-
ment possibilities and the envi-
ronmental hygiene considerations.
Chemical wastes which may only be
deposited above ground under spe-
cial regulations and special provi-
sions are considered to belong to
the C2-class.
Examples are:
- metal sludge.
Originates mainly from the treat-
109
-------
ment of wastewater from electro
plating and engraving plants. It
contains various heavy metals and
other components.
- Leather tanning sludge.
Originates from the treatment of
wastewater from leather tanning
plants. Chrome is the main conta-
minant.
- Gas scrubber sludge.
Originates from incineration of,
among others, chemical waste. It
contains various heavy metals.
It is estimated that 25,000 -
30,000 tons of C2-waste per annum
will be deposited in the landfill.
Situation
The landfill is situated on
the Maasvlakte Rotterdam. The
choice of a site on the Maasvlakte,
was mainly based on the compatibi-
lity with the landscape, the future
use that would be made of the area,
the timely availability of a site
and the stable subsoil.
water and waste.
Capacity
The chemical waste will be
stored above ground level in a re-
inforced concrete basin with a vo-
lume of 210.000 m3. This volume
will probably be enough for 5 to
10 years of use. A cross-section is
given in, figure 1.
Compartimentation
The basin is divided in
two main sections separated by a
(vertical) concrete wall. In the
smaller section all wastes contai-
ning more than 2% by weight of
lead, chromium, copper and nickel
will be deposited. This section
(above 15% of the total capacity)
is aimed at waste with a potential
for re-use in the (near) future. In
the main section all other middle-
class waste stream will be disposi-
ted. The sections are presented, in
figure 2 (lay out).
Settlements and ground water levels Design criteria
The calculation of settlements
has been based on soundings carried
out in the area. The total settle-
ment of the lower side of the basin
has been calculated at 0.50 m. The
highest ground water level
(1 x 10 years) during the dumping
phase will be approx. NAP +4.00 m.
A horizontal drainage system round
the C2-landfill will prevent fur-
ther extreme rises in the ground
water.
Taking account of the settle-
ments (0.50 m), sinking of the
ground and rising of the sea level
(0.10 m) and the capillary head of
the ground water (0.10 m), the foot
of the chemical waste should be at
NAP +5.20 m (during initial pe-
riod), in order to achieve a final
minimum distance of 0.50 m between
The basin is constructed of
reinforcement concrete executed in
accordance with the VB 1984
(Regulations for concrete) as wa-
tertight concrete. The materials
used are: blast furnace cement
class A, with a slag content 65%
and reinforcement steel Feb 400 HW
quality with a yield strength of
400 N/mm2. The cover has to be
50 mm minimally. The materials men-
tioned have been chosen to obtain
the greatest possible watertight-
ness of the concrete and the
highest guarantee for the durabili-
ty of the structure (about
150 years).
Reinforced concrete always
shows cracking. For the greatest
possible watertightness it is im-
portant to limit the width of the
110
-------
cracks as much as possible. For
this reason a maximum permissible
crack width of 0.2 mm is taken for
designing the reinforcement. The
calculation and design of the
structure with the aid of the com-
puter program DIANA, is further
based on a safety factor of 1.7 in
relation to the ultimate limit
state (failure). A concrete struc-
ture always shows cracking. This is
the reason why the Authorities that
give the permits take the view that
the waste can be in direct contact
with the ground-water.
In a risk analyses it is sup-
posed that the concrete structure
is not watertight, due to which the
sealing of the basin (HDPE liner
and' cast asphalt) must supply the
watertightness. In practice the
concrete structure can be consi-
dered as watertight (mixture of the
concrete, thickness of the floor
and walls and the crack width re-
quirement). In the construction
phase, measures are taken to allow
the shrinkage of the concrete due
to temperature and hardening pro-
cess to take place freely. If this
is not done, uncontrolled cracking
in the structure could result. The
concrete floor construction, which
is made without expansion joints is
therefore provided with a series of
shrinkage strips. The joints in the
concrete floor will be made later
on, so shrinkage is limited. Cracks
that possibly develop during the
construction phase will be injected
with epoxy-resin. The concrete
walls are made with expansion
joints every 40 m. The joint strip
is made of HDPE.
Innerliner
tinuous connection between liner
and concrete is obtained.
Bottom liner
The general functional re-
quirement for the bottom liner is,
that it must form a "moisture
proof" barrier between the landfill
and the environment for at least
the period of drainage (30 years at
minimum). The bottom liner consists
of a crack-bridging rubber bitumen
membrane and cast-asphalt layer.
The membrane must prevent cracks in
the concrete floor to extend into
the cast-asphalt layer. The asphalt
layer approx. 100 mm thick, forms
the sealing layer. The chemical re-
sistance to the leachate should be
such that the chance of emission in
the dewatering phase, is at an ac-
ceptable level.
Solvents and oil products in
sufficiently high concentrations
attack bitumen. The most extreme
and safest approach is to assume
that the total amount of solvent in
a mixture of solvent is absorbed
into the bitumen. Then 10 mg/1 will
be taken as maximum summed concen-
tration of solvents in both lea-
chate and percolate. Mo account is
taken of the fact that the lea-
chate-percolate is continuously
pumped up and transported to a
storage basin and that evaporation
of the solvent will occur.
It is known from research on
bitumen by the Royal Shell labora-
tory Amsterdam that up to a concen-
tration of 100 mg/1 the mechanical
properties of the bitumen do not
change.
The concrete walls are protec- Drainage system in the basin o
ted with a 2 mm thick HDPE liner.
The HDPE liner is prefabricated in On the asphalt layer a 0.75 m
length of 11 m and 1.40 m widths thick drainlayer is constructed
and welded on vertical HDPE-strips consisting of coarse sand. This
embedded in the concrete. So a con- layer contains a drainage system of
111
-------
HDPE pipes. Valves in these drain
pipes make it possible, that during
the filling phase the leachate-per-
colate and clean rain water are
separated. The leachate-percolate
and the clean water shall be conti-
nuously pumped up and stored in a
reservoir for leachate-percolate
and in a reservoir for clean water.
Drainage and security system under
the basin
The soil under the concrete
basin will be divided in 5 separate
soil compartments by HDPE synthetic
walls. In each section a drainage
system is constructed so that moni-
toring of leakage to the ground wa-
ter is possible. If, due to a pos-
sible leak in the bottom liner the
ground-water under the basin is
polluted, each section can be
drained and sampled separately. By
pumping on this system, the ground-
water level under the structure
will be lower than the surrounding
ground-water level, so the effect
of the pollution will be limited to
an area immediately below the land-
fill. All systems of drainage in,
as well as under the basin ends in
4 vertical shafts. Leachate-perco-
late and clean water are pumped
from these shafts to the reservoirs.
Portal crane
Because the C2-waste due to
high moisture contents and low sta-
bility cannot be driven over by
trucks, the filling is done by a
portal crane with a span of approx.
50 m. This crane runs on rails
along the length of the concrete
basin. A road is constructed on the
soil embankments beside the basin.
The crane is semi-automatic and can
dump the C2-waste in a network of
3 x 3 m2 and the time, the place
and specification of the C2-waste
are registered by computer automa-
tically.
Movable roofs
During the filling phase, the
dumping front must be protected
against penetration of rain water.
Because one of the requirements is
that the leachate-percolate seepage
muste have been expelled within a
period of 30 years, free precipita-
tion on the waste would be contrary
to this requirement. The require-
ment can be met by using movable
roofs. The slope of the waste is
estimated to be about 15° and so
has a length of about 50 m.
Two roofs elements of
22 x 50 m are necessary for the
smaller section and two roofs ele-
ments of 29 x 50 m for the main
section. A rail construction is
laid of both long sides of the con-
crete basin.
Leachate-percolate basin
The leachate-percolate is col-
lected in the leachate basin. This
structure is made by two HDPE li-
ners with a sandlayer in between.
The drain layer is provided with an
observation well which functions as
a leakage detector. To avoid fil-
ling with rain water, a tentlike
roof structure is laid over the
basin. To cope with possible stag-
nation of the leachate transport,
the basin has a volume of 300 m3
equivalent to the amount of lea-
chate and relevant waste water oc-
curring in at least two months.
Clean water basin
Non-polluted water is collec-
ted in the clean water basin. The
basin is sunk in the ground and
consists of a single HDPE sheet. To
reduce the pump capacity of the
pumping station the basin has a
content of 100 m3. The clean water
is discharged into the Mississippi
Harbour by pressure pipe.
112
-------
Monitoring system
CONCLUSIONS
Minimally six observation
wells are to be placed round the
C2-landfill. One by each shaft, as
these are places with a relatively
higher risk, and one at each narrow
side. The leachate basin may be a
place with higher risk potential.
Two observation wells will be
placed near the leachate basin.
Permanent cover
When applying the cover sys-
tem, account is taken of the
settlements occuring in the waste
fill. These settlements will occur
especially in the filling phase.
For this reason the final cover
system will only be applied about
two years after completion of the
filling. In the intermediate period
a temporary liner will be laid. The
final cover consists of two layers
of HDPE liner with a bentonite-clay
liner in between. On top of this
cover and the concrete basin, a
sand layer will be applied so that
a dune is created. The proposed
design incorporates a cover system
with a limited lifetime, so that
regular inspection will be neces-
sary.
By realising the secure land-
fill for non-treatable chemical
waste in the Netherlands, a storage
is possible in a sensible and envi-
ronmental hygienic manner. Espe-
cially for the small producers of
such waste, this site is important.
At the same time a program
must be started for treating chemi-
cal waste so that the stream of
waste will be reduced to acceptable
proportions in future.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
COST AND PLANNING
The cost of investment of the
C2-landfill are about
Dfl. 34.000.000.— and will be
payed by the Dutch government. The
running cost are for the AYR Chemie
C.V.
The construction activities
started in the beginning of 1988.
The landfill will be in operation
at December 1989.
113
-------
\
114
-------
TJ
cd
115
-------
BEDROCK NEUTRALIZATION STUDY FOR THE BRUIN LAGOON SUPERFUND SITE
Gerard M. Patelunas, P.E.
Duane R. Lenhardt, Ph.D.
James E. Niece, P.E.
GAI Consultants, Inc.
Monroeville, PA 15146
ABSTRACT
The Bruin Lagoon site is located in Bruin, Butler County, Pennsylvania,
and is listed as No. 3 on the U.S. Environmental Protection Agencies' National
Priority List. The Lagoon contains waste petroleum tars, sulfuric acid, coal
combustion ash, spent bauxite and other waste materials.
The bedrock neutralization study was conducted to assess the feasibility
of injecting caustic solutions into acid-contaminated bedrock beneath the
lagoon. The site is underlain by a fine to medium grain quartz sandstone
which is contaminated with acid to depths in excess of 30 feet. For this
investigation, Nx-cores were obtained and pressure tests conducted to a depth
of 30 feet below the top of rock.
Leach tests were conducted on contaminated core sections using sodium
hydroxide and sodium carbonate solutions. A total of 12 core sections were
exposed in 3-inch diameter test cylinders and permeated under a positive
pressure of 25 to 50 psi. Measurements of leachate volume, temperature, pH,
and hydraulic conductivity were recorded. Following the tests, the cores were
split, and examined for visual evidence of caustic penetration and crushed to
determine residual strength. The cores were then tested for pH, TOC and
extractable petroleum hydrocarbons.
INTRODUCTION
The Bruin Lagoon Site is located
in Bruin, Butler County, Pennsylvania
and is listed as No. 3 on the U.S.
Environmental Protection Agency's
(EPA) National Priority List. The
lagoon contains waste petroleum tars,
sulfuric acid, coal combustion resi-
due, spent bauxite and other waste
materials associated with an aban-
doned refinery located adjacent to
the site (2). The Record of Decision
(ROD) issued by EPA in September 1987
required the use of a lime slurry to
neutralize acidified bedrock beneath
the lagoon (3). This remedy pre-
sented several problems including the
inherent difficulty of working with a
lime slurry, potential instability to
the overlying muHi-layer cap and
projected long-term ineffectiveness
116
-------
of the treatment. As a result, GAI
Consultants, Inc. (GAI) proposed to
perform a laboratory study to eval-
uate alternative methods of
neutralizing the bedrock.
PURPOSE
The bedrock neutralization study
was conducted to assess the feasi-
bility of caustic solution injection
as an alternative for the treatment
of the acid contaminated bedrock
beneath the lagoon. The program
included field measurements of bed-
rock permeability and laboratory rock
column tests to determine the effec-
tiveness of several alkaline
treatments.
APPROACH
Rock cores obtained at the Bruin
Lagoon indicate that the site is
underlain by a fine to medium grain
quartz sandstone. Cores were ob-
tained to a depth of between 25 to 30
feet below the sludge/rock inter-
face. The upper 5 feet of the rock
section tends to be very broken to
broken, black to dark gray in color,
and impregnated by sludge in both
pores and fractures. At increasing
depth, the rock becomes broken to
massive and grades to a light gray
color with natural, black carbonized
laminae that are oriented at between
15 and 35 degrees. Angular fractures
oriented at between 20 to 90 degrees
typically exhibited iron staining
with black tar-like residue on some
fracture faces to a depth of 20 feet
beneath the rock interface. Measured
pH values of between 1.2 and 1.7 were
recorded in the upper 5 feet of the
core sections with pH values of
between 2.7 and 6.2 recorded at the
bottom of the core sections (1).
Hydraulic pressure tests were
performed in 4 of the 5 core holes.
Measured bulk rock .permeabilities
varied between 1x10"^ and 5xlO~°
cm/sec with higher permeabilities
controlled by rock fracturing. In
general, the effect of pore and frac-
ture clogging by translocated sludge
material is reflected in the lower:
permeabilities (2.2xlO~b to <4.9xlO~6
cm/sec) measured in the upper 5 feet
of the core-hole sections. Similar
permeability values were calculated
for rock core sections permeated in
the rock column tests.
Bedrock neutralization tests
were performed on select core sec-
tions using sodium hydroxide and
sodium carbonate solutions. Tests
were performed on individual core
sections using 0.01, 0.1, 1.0 and 10
normal sodium hydroxide solutions and
0.1 and 1.0 normal sodium carbonate
solutions. A total of twelve (12),
2-inch thick Nx-core sections were
epoxied into 3-inch diameter, sched-
ule 80 test cylinders with threaded
end caps and drainage ports. After
sufficient setting time, the rock
sections were tested for leaks,
sealed and placed in a protective
outer casing, and leached under a
positive pressure (25 to 50 psi)
until 125 milliliters of test solu-
tion (approximately 8 pore volumes)
had passed through each core section.
During testing, measurements of
leachate volume, temperature and pH
were recorded. Following leaching
the core sections were removed from
the test cylinder, split axially and
examined visually for signs of caus-
tic leaching. The core sections were
then tested for unconfined compres-
sive strength and crushed for
chemical analysis. Chemical analyses
performed on the crushed core samples
included pH (SW-9045), total organic
carbon (Walkley-Black) and total
117
-------
recoverable hydrocarbon (EPA 418.1).
Representative sections of untreated
rock core were also tested and ana-
lyzed to define baseline conditions.
PROBLEMS ENCOUNTERED
A 10 N sodium carbonate solution
was used in several column tests, but
failed to permeate the rock samples
due to crystal formation and precipi-
tation on the upper rock surface. No
other significant problems were
encountered in conducting the study.
RESULTS
Column test data for rock
specimens are summarized in Tables 1
and 2. The core sections were per-
meated with a weak base solution
(sodium carbonate) and a strong base
solution (sodium hydroxide). In each
instance, the resulting leachate was
noticeably darkened indicating
leaching of the impregnated sludge
(hydrocarbons) from the rock cores
(Table 2). The effects of contam-
inant removal were particularly
evident in visual comparisons of
axial core fractures with the treated
cores demonstrating a lighter colora-
tion characteristic of the uncon-
taminated quartzitic sandstone.
Leaching of the cores using both test
solutions also produced a significant
reduction in the unconfined compres-
sive strength of the rock (Table 2).
This result may be due to the
selective removal of natural binding
agents and/or interstitial sludge
which may have held sand grains
together in the untreated sample.
During performance of the tests,
leachate volumes were recorded and
subsequently used to calculate rock
permeabilities. Permeabilities of
the sludge impregnated core samples
were found to vary between 6.9xlO~5
and 4.4xlO~° cm/sec in conformance
with field permeability results. In
general, lower permeability valves
were measured for rock samples from
the 409 core series, a finer grain
sandstone, which appeared to be more
heavily impregnated with sludge and
provided low pH values. Measurements
of test solution pH before leaching
and after permeation of the core
sections indicate that the sodium
carbonate solutions (0.1 and 1.0 N)
were readily acidified by the con-
taminated rock samples (Table 2). In
contrast, the sodium hydroxide
solution (0.1 to 10 N) showed no sig-
nificant alteration of effluent pH
except at low solute concentrations
(Table 2).
These results and the corres-
ponding rock pH values before and
after treatment (Table 2) confirm
that use of a weak base would be
ineffective in treating the acidic
bedrock. It was further determined
that use of a 0.01 to 0.1 normal
sodium hydroxide solution would be
effective in treating the rock
beneath the lagoon. Chemical reac-
tions between the acid contaminated
rock and caustic permeate were not
observed during testing, nor was
there any measureable back-pressure
build up within the test cells after
24 hours as a result of gas forma-
tion.
Several alternatives were
considered for neutralizing the
acidic bedrock condition which has
been documented to extend 20 to 30
feet beneath the site. The most
effective option combines pressure
injection of a caustic solution
followed by pressure grouting using a
high pH cement. The injected caustic
solution would neutralize readily
accessible acidity in surrounding
rock pores and fractures, while
118
-------
Table 1* Column Test Readings Bedrock Neutralization Study
Column Test Using NaOH
Core Test
Date
02-29-88
02-29-88
02-29-88
02-29-88
02-29-88
02-29-88
02-29-88
Core Test
02-29-88
02-29-88
02-29-88
02-29-88
Core Test
02-29-88
02-29-88
02-29-88
02-29-88
03-01-88
Core Test
04-12-88
04-12-88
04-12-88
04-12-88
04-12-88
Core Test
03-03-88
03-03-88
03-03-88
03-03-88
03-04-88
No. 109
A:
Wt
Time (min)
1200
1230
1300
1345
1430
1500
1600
No. 109
1100
1200
1230
1300
No. 109
1345
1500
1630
1800
1000
No. 109
900
910
950
1035
1245
No. 409
1100
1300
1400
1500
930
0
30
60
105
150
180
240
B:
0
60
90
120
C:
0
75
165
255
1215
D:
0
10
50
95
225
A:
0
120
180
240
1350
Soln. Cone. = 0.1 N
Pressure Leach Vol Pore Vol
(psi) (ml) (Approx)
25
25
25
30
30
30
30
Soln. Cone.
25
25
25
25
Soln. Cone.
30
30
30
40
40
Soln. Cone.
15
15
15
15
15
Column
Soln. Cone.
20
20
20
30
30
0
10
15
20
60
80
125
= 1.0 N
0
10
80
125
= 10 N
0
10
15
30
125
= 0.01 N
0
25
40
60
125
Test Using
= 0.1 N
0
10
15
20
125
0
0.
1.
1.
4.
5.
8.
0
0.
5.
8.
0
0.
1.
2.
8.
0
1.
2.
4.
8.
7
0
3
0
4
3
7
4
3
7
0
0
3
7
6
0
3
Temp
(°F)
68
68
_
_
63
58
68
_
68
68
68
63
_
63
65
_
_
_
-
PH
_
_
_
13.
13.
13.
13.
13.
12.
_
12.
3.
3.
4.
5.
5
5
7
8
8
0
0
1
7
2
2
K
(cm/sec)
_
1.7 x 10~5
_
_
_
3.2 x ID'5
_
—
6.9 x 10~5
—
4.7 x KT6
_
4.5 x 10~6
_
_
_
5.8 x 10~5
NaCO,
0
0.
1.
1.
8.
7
0
3
3
_
_
_
68
_
—
_
0.
4
.
5.5 x 10~6
_
4.9 x 10'6
(continued)
119
-------
Table 1. (Continued)
Column Test Using NaC03
Date Time
Wt
(min)
Pressure
(psi)
Leach Vol
(ml)
Pore Vol
(Approx)
Temp
K
(cm/sec)
Core Test No. 409 B: Soln. Cone. = 1.0 N
03-03-88
03-03-88
03-03-88
03-03-88
03-03-88
03-04-88
1100
1300
1400
1430
1500
930
0
120
180
210
240
1350
25
25
25
30
30
30
0
20
25
30
35
125
0
1.4
1.7
2.0
2.3
8.3
9.2 x 10
-6
68
4.1
4.4 x 10
-6
Core Test No. 409 C: Soln. Cone. = 1.0 N
03-03-88
03-04-88
03-04-88
03-04-88
03-04-88
03-07-88
03-07-88
03-07-88
03-07-88
1500
1030
1200
1330
1500
1000
1100
1200
1230
0
1170
1260
1350
1440
1440
1500
1560
1590
30
30
30
50
50
50
50
50
50
0
60
68
75
100
100
105
115
125
0
4.0
4.5
5.0
6.7
6.7
7.0
7.7
8.3
_ _
- -
_ _
_ _
_ _
_ _
71 0.9
— —
71 0.5
_
-
_
_
_
_
_
™~ f
6.0 x 10~b
pressure grouting of fractures would
provide a long-term neutralization
capability for remaining acidity at
greater distance from the injection
point.
Pressure injection of the
caustic solution would be implemented
in vertical increments within a net-
work of closely spaced borings (Z 20
feet) followed by pressure grouting
of open boreholes. Additional con-
tainment of the underlying bedrock
contamination could be ensured by a
peripheral grout curtain composed of
acid resistant silicate grout
mixture.
Pressure injection was
determined not to be a cost effective
solution at this site by U.S. EPA.
Since the rock directly beneath the
sludge can be considered almost
impermeable as demonstrated in the
previous section, the effectiveness
of any caustic material placed
directly on top of the rock will be
severely limited. In addition, the
multi-layer cap will be virtually
impermeable to downward migration of
gravitational water. Therefore, a
layer of crushed limestone supple-
mented with agricultural lime was
used in place of the lime slurry
concept in the remedial design. This
procedure will eliminate the problems
associated with the lime slurry while
providing long-term neutralizing
capacity.
120
-------
cn c.
Eg.
•o o
QJ t_
ro 4-J
O w O
o **~ *o
o re QJ
c: cn re
gQJ
r— t-
.— QJ
QJ t—
>,t
cn t-
cn QJ c
TD C ZS
ZJ O t-
C_ «3 in 4-* 4->
4-* o ro i—
re r-
JZ -i- t-
O E QJ
re Q) 4->
QJ jz ro
_l (J E
CJ
QJ O
QJ tn
He
re QJ
c e o
cn o 40 r
CLTD QJ
EC jc
•t- re o
tn re
QJ QJ
cn M i—
O •.- -r-
,— (_ O i
jz ro r—
tn t- Q.
«^ QJ E
5 c re
o *r- tn
r— E
,— OJ
QJ JZ t_
>,•!- O
QJ •«- ro
jr 4J QJ
O O C-
^ 8
r-t CO
!g^
O O
00 CO
CO i-H
121
-------
ACKNOWLEDGEMENTS
This project was performed for
the U.S. Army Corps of Engineers,
Omaha District, Contract
No. DACW45-87-C-0057.
REFERENCES
1. GAI Consultants, Inc., Pre-Design
Activities, Bruin Lagoon Super-
fund Site, for the U.S. Army
Corps of Engineers, Omaha
District, July 1988.
2. Roy F. Weston, Inc., Draft
Remedial Investigation/Feasibil-
ity Study Report for the Bruin
Lagoon Superfund Site, Bruin
Borough, Pennsylvania, prepared
for the U.S. Environmental
Protection Agency, Contract
No. 68-01-6939, June 1986.
3. Seif, James M., Record of
Decision, Remedial Alternative
Selection, Bruin Lagoon Site,
Bruin Borough, Pennsylvania, U.S.
Environmental Protection Agency,
Region III, September 29, 1986.
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
122
-------
MODELING CHEMICAL AND PHYSICAL PROCESSES IN LEACHING SOLIDIFIED WASTES
Bill Batchelor
Civil Engineering Department
Texas A&M University
College Station, Texas 77843
(409) 845-1304
ABSTRACT
Solidification is an important treatment process in hazardous waste
management and will continue to be so until waste minimization and waste
recycle processes are perfected for all hazardous wastes. It is generally
recognized that immobilization of contaminants in solidified wastes occurs
through both physical and chemical mechanisms. Standard techniques for
analyzing contaminant leaching measure only an observed diffusivity that does
not separate chemical and physical factors. The suitability of the standard
data analysis procedure is reviewed and alternative models developed that
describe the separate effects of chemical and physical processes. These
models describe physical transport through a solidified waste matrix according
to Pick's law. Chemical reactions of sorption/desorption and precipitation/
dissolution are described. The observed diffusivity that would be calculated
by ignoring chemical processes is shown to depend on the true effective
diffusivity and coefficients that describe the chemical phenomena.
INTRODUCTION
Solidification is an important
process for treating hazardous wastes
and should become increasingly
important in the future. Many wastes
such as those containing inorganic
contaminants cannot be destroyed by
treatment nor can their production be
avoided. Even wastes that are
primarily organic and can be destroyed
by processes such as incineration or
chemical oxidation, often produce
inorganic residuals that must be
disposed. Until recovery processes
are greatly improved, many wastes that
contain toxic inorganic chemicals will
continue to be disposed on or above
land. Therefore, solidification
processes that immobilize hazardous
components of these wastes are
essential to sound environmental
management of these wastes.
Immobilization by solidification
is generally considered to occur by
two mechanisms. Physical processes
operate by either immobilizing a
contaminant completely by entrapping
it in a solid matrix, or reducing its
mobility, by reducing the area through
with it can migrate. Chemical
processes can reduce mobility by
converting the contaminant to a less
123
-------
mobile, or completely immobile form.
Precipitation is an example of a
chemical process that converts the
contaminant to an immobile solid, and
sorption is an example of a process
that reduces the contaminant's
mobility by fixing it onto the solid
matrix.
Quantification of the degree of
immobilization achieved by
solidification processes is typically
done by conducting leach tests.
Results of these tests are usually
interpreted in terms of a simple
leaching model.1~7 Leaching models
can also be used to predict long term
performance of the solidified waste
material in the environment. As
demands are made on solidification
processes to provide a greater degree
of immobilization for a wider range
of contaminants, models will be
expected to provide more information
about the performance of these
processes in order to aid in
formulating further modifications.
PURPOSE
The purpose of this paper is to
review and evaluate the standard
model used to interpret results of
leaching tests of solidified
hazardous wastes, and to examine its
ability to provide information on the
mechanisms of immobilization.
APPROACH
A model can be thought of as a
set of assumptions and the
mathematical equations that can be
deduced from those assumptions. The
standard model used to interpret
leaching test results is based on the
following assumptions:
1) contaminant is being leached from
an infinitely thick slab;
2) contaminants leach into an
infinitely large bath of fluid;
3) all of the contaminant is mobile
and the concentration of mobile
contaminant is known and is
constant throughout the slab at
the start of the leaching test;
4) the contaminant is transported in
one direction through the slab in
accordance with Pick's law of
diffusion applied with a constant
diffusivity coefficient;
5) the contaminants do not react in
any way.
Applying a material balance to
the slab with these assumptions leads
to the following partial differential
equation, initial condition, and
boundary conditions.
(1)
3 C
a t
C - CtQ for t •= 0, all x
C - 0 for t > 0, and
x - +/- L
where:
C -=
concentration of mobile
contaminant , [ M/L3 ]
~ concentration of
mobile contaminant when
leaching begins,
De = effective diffusivity,
t
x
L
— time since leaching
began, [T]
distance from center of
slab, [L]
distance from center of
slab to edge of slab
(half -thickness) , [L]
The above equations can be solved
to give the concentration of mobile
124
-------
contaminant in the slab at any time.
The internal concentration profiles
are of little use by themselves.
However, they can be integrated to
obtain the amount of material that
has left the slab at any time. This
is normally expressed as the fraction
of contaminant that has leached.
(2)
H
M
t -
4 D
JT L
|0.5 t0.5
where:
-fraction of con-
taminant leached, [ ]
This result can be used to
determine the diffusivity of
contaminant in the solidified waste.
Measured values of Mt/Mo can be
regressed against the square root of
time and the slope obtained. The
diffusivity can then be calculated
directly from the slope. If the
assumption of no chemical reactions
is valid, the diffusivity calculated
in this manner is the effective
diffusivity. If the assumptions upon
which the model based are not known
to be valid, the diffusivity
calculated with this procedure should
be called an observed diffusivity.
This approach uses the term
"effective diffusivity" to refer to
the coefficient that describes the
physical transport of contaminant
through the matrix. The term
"observed diffusivity" refers to the
coefficient that describes the effect
of both physical transport and
chemical reaction.
infinitely thick slab can be
evaluated by comparing estimates of
Mt/Mo calculated with the standard
model with those calculated using a
model that assumes a finite slab
thickness.8 Figure 1 shows the error
associated with assuming the slab is
infinitely thick. Significant
errors are not observed over the
range of Mt/Mo that would typically
be used. Therefore the assumption of
an infinite slab does not limit the
suitability of the standard model for
most applications.
The accuracy of the assumption of
an infinite bath can be evaluated
with analytical solutions of models
that describe how a finite slab
leaches into a finite bath. The
predictions made by these models
depend on the relationship between
the concentration in the slab and the
concentration in the bath that would
exist if the leaching were allowed to
continue forever. This relationship
is affected by the ratio of the
volume of the slab to the volume of
the bath, and by any correction
factor needed to express slab
concentrations on the same basis as
bath concentrations. A typical
correction factor would be applied
when slab concentrations are
expressed on the basis of the total
slab volume and only a fraction of
the volume of the slab actually
contains water. In this case, the
correction factor would be the
porosity of the slab. The combined
effect of these factors is expressed
in a term represented by a.
RESULTS <3> a "
The suitability of the standard
leach model could be limited if any of
its major assumptions are not correct. where: a
The validity of the assumption of an V
V K
s
coefficient, [ ]
volume of bath, [L3]
125
-------
Vs — volume of slab, [L3]
K - ratio of contaminant
concentration in slab
to equivalent concen-
tration in bath, [ ]
Figure 2 shows the extent to
which the assumption of an infinite
bath leads to errors in the standard
model. When a is large (relatively
large bath volume, or small correction
factor), the error is manageable.
When a is small, the error is larger
but it is predictable. The error is
nearly proportional to the fraction
leached. The error in the ultimate
fraction leached that is incurred by
the standard model can be calculated
by knowing that the standard model
predicts (Mt/Mo)inf - 1.0 and the
finite bath model predicts (Mt/Mo)inf
- a/(l + a). The error in the
standard model can then be estimated
at each value of Mt/Mo by assuming it
is linearly related to Mt/Mo. This
result can be used to design leaching
experiments that do not lead to
significant errors when the standard
model is applied to interpret
results.
(4) Relative Error in Mt/M0 (%) -
(Mt/M0)meas (100%/a)
where: (Mt/M0)meas - fraction of
contaminant leached as
measured using the
standard model, [ ]
The most significant error
associated with the assumption of a
constant, known initial contaminant
concentration would likely occur when
a fraction of the contaminant is
completely immobilized by some
physical or chemical process. If
this were not recognized, there would
be an error in calculating the
fraction leached, because the wrong
value for the initial contaminant
mass (M0) would be used. Observed
diffusivity calculated using the
conventional model would be in error
in proportion to the fraction of
contaminant that is mobile (Fm).
(5) (Mt/M0)meas - Fm (Mt/Mo)
- Fm (4 De/ L2)°-« t°-«
(6) Dobs - (Fm)2 De
where: DODS - observed diffusivity,
i.e. that value
calculated using
standard model, [L2/T]
Fm — fraction of contaminant
that is mobile, [ ]
The assumptions that the
contaminants are completely mobile
and that they do not react chemically
are probably the most likely to be
invalid. The effect of chemical
reactions on observed diffusivities
can be evaluated using modifications
of the standard model. Precipitation
and sorption are two of the most
likely chemical reactions to affect
contaminants in solidified wastes.
If precipitation occurs, a
portion of the contaminant will be
converted to the immobile solid form.
As mobile contaminant leaves the
slab, the precipitate will tend to
dissolve to maintain the mobile phase
concentration at the saturation
concentration for that precipitate.
This process can be modeled rather
simply by assuming that the
dissolution reaction is rapid
compared to the diffusion step, and
that a pseudo- steady state
concentration profile is set up in
the slab.9 This model can be solved
to provide the following estimate for
Mt/Mo as a function of time.
126
-------
(7) A.
M
(8) D .
obs
2 D (K + 0.5)
L (K + 1)
•K (K + 0.5) D
p ' e
2 (Kp+ I)2
0.5fc0.5
if D
if K » 1, then D , - 0 .... —
p ' obs 2 (K + 1)
matrix. The relation between
immobile and mobile forms can be
expressed in terms of an isotherm.
By assuming that the exchange between
mobile and immobile forms is rapid
compared to the process of diffusive
transport, the partial differential
equation describing leaching can be
converted to the following.
(10)-
3 C
a t
a x
where: Kp - ratio of mass of cont-
aminant in immobile
precipitated phase to
mass of contaminant in
mobile soluble phase
when leaching begins,
I 1
Since (Kp + 1) - l/Fm, then when
Kp
(9) D
*• F D
m e
obs
The precipitation model predicts
the same dependence of Mt/Mo on time
as does the standard model. However,
the observed diffusivity that would
be calculated using the standard
model is smaller than the effective
diffusivity. The difference is due
to the effect of chemical
immobilization of the contaminant
through precipitation. The model
presented above is able to separate
the chemical and physical factors
that effect leaching behavior.
Another chemical process that
could effect leaching behavior is
sorption/desorption. In this
process, the contaminant is exchanged
between an mobile form in solution
and an immobile form on the solid
where: (dCj/dC) - derivative of
isotherm relating the
concentration of
immobile contaminant
(C^) to the concen-
tration of mobile
contaminant (G), [ ]
If sorption is linear, the
derivative in Equation 10 will be a
constant, and will equal Kp.
Therefore, the observed diffusivity
can be related to the effective
diffusivity by the following.
(11)
obs
1 + K
F D
m e
This demonstrates how the
observed diffusivity can be shown to
be affected by physical factors (De)
and chemical factors (dCj/dC). When
linear sorption occurs, the observed
diffusivity is proportional to the
effective diffusivity. However, when
nonlinear sorption occurs, the
observed diffusivity is a function of
the concentration of mobile
contaminant. The Langmuir isotherm
is an example of a nonlinear isotherm
that often used to describe the
equilibrium relationship between
sorbed and soluble components.
127
-------
(12)
(13)
C. b C
l.max
1 + b C
C. b
l.max
( 1 + b C )5
When nonlinear sorption occurs,
Equation 10 must be solved
numerically to determine how
contaminants leach from the solid.
An orthogonal collocation technique9
has been developed to do this.
Initial results indicate that the
model with sorption/desorption
predicts that the fraction leached
will be proportional to the square
root of time, which is similar to the
results obtained for the model without
reaction. However, the slope of the
plot is substantially less than would
be obtained if sorption did not occur.
It may be possible to make "first-
approximation estimates of observed
diffusivity by assuming that the
concentration of mobile component is
low. With this assumption, the
denominator in Equation 13 approaches
one, and the observed diffusivity is
proportional to the effective
diffusivity. When applied to
preliminary data, this approach
predicted the effective diffusivity
within a factor of two.
The results presented above for
leach models incorporating chemical
reactions shows how these processes
can strongly affect the value of the
observed diffusivity measured in
standard leaching tests. Large
variations in observed diffusivities
would be expected for contaminants
that display different chemical
behavior. Reported values of
observed diffusivities support this
observation. One study reported
observed diffusivities to range from
5x10~9 cm2/s for sodium to 4x10"16
cm2/s for lead.5 Sodium would be
expected to be relatively
nonreactive, while lead would be
likely to precipitate or sorb in
wastes solidified with portland
cement. The effective diffusivities
for sodium and lead should vary by
less than a factor of two,10'11 so
the importance of chemical reactions
is clear in this case.
Based on the analysis of the
standard leach model and alternative
models the following conclusions can
be drawn.
1. Errors in calculating the observed
diffusivity that are associated
with assuming an infinitely thick
slab and an infinite bath volume
can be kept at small levels by
proper design of leach tests.
2. Errors in calculating the observed
diffusivity that are associated
with complete and permanent
immobilization of a fraction of
the contaminant are directly
related to the fraction
immobilized. This fraction is
difficult to measure unless very
long leaching tests can be
conducted.
}. Observed diffusivities depend on
physical and chemical factors.
The effect of the chemical
factors of precipitation and
sorption can be quantified. In
both cases, the models
incorporating chemical processes
predict behavior in agreement
with the form of the standard
model, but with observed
diffusivities lower than
effective diffusivities.
K Simple models for leaching with
linear sorption or with
dissolution of a precipitate show
that the observed diffusivity
128
-------
should be approximately equal to
the effective diffusivity times
the fraction of contaminant that
is mobile. A simple model for
leaching when a fraction of the
contaminant is completely
immobilized predicts that the
observed diffusivity is equal to
the effective diffusivity times
the fraction that is mobile
squared.
5. Leaching results should be
reported as observed diffusiv-
ities, unless it is known that
chemical factors are not
affecting leaching rates.
REFERENCES
1. Hespe, E.D., "Leach Testing of
Immobilized Radioactive Waste Solids,
A Proposal for a Standard Method",
Atomic Energy Review. Vol. 9, No. 1,
pp 195-207, 1971.
2. Godbee, H.W., Joy, D.S.,
"Assessment of the Loss of
Radioactive Isotopes from Waste
Solids to the Environment, Part 1:
Background and Theory", Oak Ridge
National Laboratory, TM 4333, 1974.
3. Johnson, J.C., Lancione, R.L. ,
Sanning, D.E., "Stabilization,
Testing and Disposal of Arsenic
Containing Wastes", in Toxic and
Hazardous Waste Disposal. Vol. 2,
R.J. Pojasek (ed), Ann Arbor
Scientific Publishers, Ann Arbor,
Michigan, 1979.
4. Mahloch, J.L. , "Leach Testing of
Stabilized Industrial Sludges-
Interpretation and Application", in
Toxic and Hazardous Waste Disposal.
Vol. 2, R.J. Pojasek (ed), Ann Arbor
Scientific Publishers, Ann Arbor,
Michigan, 1979.
5. Cote, P.L., Isabel, D.,
"Application of a Dynamic Leaching
Test to Solidified Wastes", in
Hazardous and Industrial Waste
Management and Testing. 3rd
Symposium, ASTM Special Techniques
Publication 851, L.P. Jackson, A.R.
Rohlik, R.A. Conway (eds), 1984.
6. Gilliam, T.M., Dole, L.R.,
"Determining Leach Rates of
Monolithic Waste Forms", U.S.
Department of Energy, DE86 012618,
1986.
7. "Measurement of the Leachability
of Solidified Low-Level Radioactive
Wastes by a Short-term Test
Procedure", ANSI/ANS-16.1-1986 ,
Americal Nuclear Society, La Grange
Park, Illinois, 1986.
8. Crank, J. The Mathematics of
Diffusion. 2nd Edition, Clarendon
Press, Oxford, 1975.
9. Finlayson, B.A., Nonlinear
Analysis in Chemical Engineering.
McGraw-Hill Book Company, New York,
1980.
10. Weast, R.C. Handbook of Chemistry
and Physics. 51st Edition, Chemical
Rubber Company, Cleveland, Ohio,
1970.
11. Perry, R.H. , Chilton, C.H. ,
Chemical Engineers Handbook. 5th
Edition, McGraw-Hill Book Company,
New York, 1973.
ACKNOWLEDGEMENT
This work was funded by a grant from
the Gulf Coast Hazardous Substance
Research Center through the Texas
Engineering Experiment Station.
129
-------
_o
3
o
§
LJ
a?
g
o
01
m
c
i_
o
UJ
7.0
ACCURACY OF INFINITE SLAB SOLUTION
Infinite Both
No Reactions
6.0 • •
5.0-
4.0 - •
3.0-•
2.0--
1.0--
0.0
0.00 0.20 0.40 0.60 0.80
Fraction Contaminant Leached (Mt/Mo)
Figure 1. Accuracy of Infinite Slab Solution
ACCURACY OF INFINITE BATH ASSUMPTION
Finite Slab, No Reactions
volume of liquid/volume of solid = 9
9.0
0.00 0.20 0.40 0.60 0.80
Fraction of Contaminant Leached (Mt/Mo)
Figure 2. Accuracy of Infinite Bath Solution (a - 9)
130
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
131
-------
SOLIDIFICATION OF FILTER ASHES FROM SOLID WASTE INCINERATORS
Peter Friedli1) and Paul H. Brunner2)
x>Geotechnik, CH-8008 Zurich, Switzerland, and
>Dept. of Waste Management and Material Balances,
Swiss Federal Institute for Water Resources and Water Pollu-
tion Control, CH-8600 Diibendorf, Switzerland
ABSTRACT
Residues from flue gas purification of Swiss MSW incinerators
consist primarily of silicates, alkaline and alkaline earth
metals,aluminum compounds, carbonates, chlorides and sulpha-
tes. Since trace metals such
as Zn, Cd or Hg may be leached
r •*- — «-<-^j Akhv^jr *w'*— -i.^-ti\-*j.j.^3^i.
from these residues, treatment of the ash is required in order
to prevent hazardous emissions from ash landfills. The
stabilization of electrostatic precipitator ash, dry scrubber
ash and wet scrubber sludge with various cement based binders
improved the structural quality as well as the leaching
behaviour. The results for leaching and stability from field
tests compared favourably with laboratory experiments. For a
stabilization which retains the hazardous constituents in the
landfill body for long periods of time, it proved to be neces-
sary to equilibrate the ashes with water prior to treatment
with a binder. By appropriate stabilization of the pretreated
ashes, a compressive strength > 1 N/mm2, mass losses after
twelve wet/dry and freeze/thaw cycles of < 10%, and metal lea-
ching < 1 °/oo of total metal present were obtained.
INTRODUCTION
For modern urban areas,
the importance of the incine-
ration of municipal solid
wastes (MSW) increases due to
the scarce space for landfil-
ling. By the combustion pro-
cess, the mass of residues
which has to be disposed of in
landfills can be reduced by
about 80 %. Still, after inci-
neration one fifth of the MSW
remains to be landfilled in
the form of bottom ash, filter
ash or scrubber sludges. MSW
contains considerable amounts
of the national consumption of
many materials. For highly ur-
banized regions, it has been
estimated that as much as 20 %
of the imported cadmium and
lead is contained in the MSW.
During incineration, many ele-
ments are concentrated in the
offgas and removed by the air
pollution control devices.
Therefore, the filter ashes or
132
-------
scrubber sludges act as very
important carriers for certain
elements (Hg, Cdr Pb, Sn, Sb
etc.) from the antroposphere
to the environment, or where
economically possible back
into consumption cycles. It is
of chief importance to control
the quality and the treatment
including the final storage of
the residues of MSW incinera-
tion. This is further emphasi-
zed by the fact that due to
the high content in chlorides,
the filter ashes contain a
large fraction of mobile heavy
metal compounds.
The present paper focuses
on the chemical and physical
immobilization of hazardous
compounds in filter residues
from MSW incinerators. It is a
summary based on solidifica-
tion and leaching experiments
which have been performed by
various Swiss institutions
from 1984 to 1987.
PURPOSE
The objectives for the
treatment of re'sidues from MSW
incineration in Switzerland
are: 1. the transformation of
leachable hazardous constitu-
ents into compounds which are
not mobile under landfilling
conditions for long periods of
time (<103 years', and 2. if
economically feasible, the
recycling of metals by further
concentration and use as a raw
material in smelters. In this
paper, we discuss the stabili-
zation of residues from flue
gas cleaning with cementious
binders. The objectives for
this treatment are:
1. Minimize the leaching rate
by minimizing permeability
and diffusivity, 2. minimize
the leachate concentration of
metals by producing metal spe-
cies of low solubility, and 3.
prevent long term acid corro-
sion by establishing a high
buffering capacity at the pro-
per pH level.
The goal of this paper is
to present composition and be-
haviour of a number of filter
ashes and to discuss various
ash treatments which may
potentially reach the desired
goal of long term im-.
mobilization. The following
questions are addressed:
1. What is the composition
of the filter residues of var-
ious wet and dry cleaning sys-
tems of MSW incinerators?
2. What is the leaching
behaviour of these products?
3. How is the leaching
behaviour and the stability of
the filter ashes improved by
the addition of simple stabi-
lizers such as cement?
4. What general conclu-
sions can be drawn for the
immobilization and solidifica-
tion of filter ashes?
APPROACH
The composition of filter
ashes from five Swiss munic-
ipal solid waste incinerators,
equipped with either electro-
static precipitators (ESP),
spray dryer absorbers, or wet
scrubbers, has been determined
by the usual analytical
methods. The leaching beha-
viour of the untreated filter
ash has been examined by labo-
ratory leaching experiments.
Samples of filter residues
have been solidified with var-
133
-------
ious methods. The treated sam-
ples were tested by physical
and chemical methods in the
laboratory. Three mixtures of
cement based binders and ashes
have been tested for leaching
in field experiments of
various sizes.
element [g/kg] A
B
Si02
Al
Ca
Na
K
Mg
Fe
Cl
S
C
P
Zn
Pb
Cu
Sn
Cr
Ni
Cd
Hg
270
80
60
50
40
18
35
60
30
30
6
25
9
3
4
1
0.2
0.4
0.1
150
50
230
6
12
8
8
120
40
27
3
12
4
0.8
n.d.
0.2
0.1
0.2
0.3
n. d.
4
260
4
3
7
33
5
80
50
0.4
8
3
0.3
2
n.d.
n.d.
0.1
1
Table 1: Mean values of some
constituents of residues from
three flue gas cleaning sy-
stems. Due to the high content
in CaO, the pH value of a
mixture of water and ash A
increases after three days of
mixing from 7 to 11.6. The pH
value of the mixture of ash B
with water increases to about
12. (A: esp ash, B: dry
scrubber (spray absorber) ash,
C: wet scrubber sludge)
Composition and Behaviour of
Filter Ash, Dry Scrubber Ash
and Wet Scrubber Sludge
Filter ash from MSW
incinerators consists prima-
rily of silicate, aluminum,
and alkaline and alkaline
earth metals (cf. table 1).
The prevalent metal species
are mainly silicates, oxides,
chlorides and sulphates, with
carbonates and phosphates as
minor compounds. In the pro-
ducts from dry scrubbing, the
calcium and chloride concen-
trations are enhanced due to
the addition of Ca(OH)2 and
the removal of HC1 from the
offgas. Most other elements
are less concentrated in the
dry scrub-ber product because
of the dilution effect of the
calcium chloride. An exception
is mercury, which appears to
be sorbed by the surface of
the dry scrubbing agent and
hence is partly removed from
the offgas.
The difference between the
esp dust and the wet scrubber
sludge concerns mainly chlo-
ride, phosphorus, aluminum,
and other mobile elements,
which are in part removed in
the aqueous phase. If not
removed separately by ion
exchange or precipitation,
mercury is enriched in the
scrubber sludge. More
important is the fact that
scrubber sludges are generated
in water, which means that
they have gone through such
processes as dissolution,
hydrolysis, neutralization and
precipitation, and that they
are more or less in an
equilibrium state with water.
This is not the case for dry
products, which upon contact
134
-------
with water will go through
many processes. It is
therefore more difficult to
stabilize dry products; in
most cases it is advisable to
thoroughly wet or even wash
the dry products before
solidification.
In contact with water, al-
kaline and other metal oxides
are hydrolysed, resulting in a
significant increase of the pH
value of the aqueous solution
from 6-7 to 9-10. The metal
chlorides present in the ash
are readily dissolved when
leached with water. They will
precipitate as hydroxides when
the pH value increases due to
the contact with water. Thus,
an important fraction of heavy
metals such as cadmium and
lead are leached from the ash
at the first contact with
water. With longer residence
times in water, they will be
retained in the ash/water mix-
ture because the pH value in-
creases to around 10-11. In
the laboratory leaching ex-
periments, which lasted 14
days, no significant differ-
ences in the metal leaching
behaviour of pulverized small
grain ash samples versus
solidified (no binders) sam-
ples were detectable.
The content of organic
substances in the filter ash
ranges from 10-50 g/kg (Brun-
ner et al,1987). Although the
content of certain micropol-
lutants such as PCDD and PCDF,
PAH, chlorinated benzenes and
phenols in the mg/kg range is
well known, there exists no
detailed information on the
nature of the bulk of the or-
ganic carbon in the ash. A
large fraction is easily ex-
tractable with organic sol-
vents. Of all the organic sub-
stances investigated, the
chlorinated phenols are the
most likely to be leached from
a filter ash landfill.
Solidification Experiments
The objective of the expe-
rimental work was to find sim-
ple recipes to solidify and
immobilize products from flue
gas cleaning by the addition
of commercially available bin-
ders, and to test some of the
promising recipes in the
field. In many laboratory ex-
periments, proctor cylinders
(d=56 mm, h=100 mm) of various
mixtures were produced (table
2) . The cylinders were tested
for stability and for leaching
of selected metals and nonme-
tals. The data of the leaching
experiments have been used to
ash Composition of mixture
type %ash %cement %water
A
Bl
B2
C
40
60
30
45
40 HTA
20 HTA
30 HTS
15 HTA
20
20
40
40
Table 3: Solidification reci-
pes for ashes, resulting in
minimum mass losses after 12
wet/dry and freeze/thaw cycles
and a compressive strength of
> 1 N/mm^. A: esp ash, Bl: dry
scrubber ash, B2: dry scrubber
ash washed, C: wet scrubber
sludge; HTA High aluminum
cement plus retarder; HTS
calcium silicate cement (24 %
Si) .
135
-------
5.8m
pSolidified Material
, 1.2m
0.5m
'Leachate
Subbase covered
w/Membrane
L Embankment
Leachate
1— Asphalt (5cm)
covered w/Sand
Fig.la Test Site A. Covered
w/roof. Dry scrubber
ash + slag + binder
(HTA)
calculate a "pseudo" diffusion
coefficient according to the
US EPA Uniform Leaching Proce-
dure.
Table 3 summarizes the re-
cipes and the products which
have been succesfully solidi-
fied. Prior to testing, the
samples were cured for 10 days
at 20 °C and 98 % rel. humi-
dity. The physical testing
included the compressive
strength, the resistance to
freez/thaw cycles as well as
dry/wet cycles. A compressive
strength of > 1 N/mm2 and a
loss of mass < 10 % after 12
freeze/thaw and wet/dry cy-
cles, respectively, was consi-
dered as appropriate for dis-
posal in a final storage site.
The cyclic freeze/thaw
testing is considered to
reflect expansion forces while
the wet/dry tests simulate
shrinking forces on the
treated material.
Fig.lb Test Site Bl and B2.
Bl: Dry scrubber ash +
sand + HTA
B2: Dry scrubber ash + slag +
HTA + Trisilicate
The chemical analysis compri-
sed the measurement of the
concentrations of Cl, Pb, Cu,
Zn, Cd, in a leachate which
was obtained after consecutive
leaching with deionized water
in equilibrium with air, and
in certain cases in equili-
brium with C02•
In the field experiments,
the mixing, handling, the
solidification and the lea-
ching behaviour of the various
filter products was tested.
The laboratory test methods
were compared to the field re-
sults in order to check their
validity. Three experimental
landfills of 30 to 150 m3
volume were constructed, each
containing dry scrubber ash
solidified by different
methods. One of these test
sites was covered by a roof
and run under strictly
controlled and enhanced rai-
ning conditions (cf. fig. la
and Ib).
136
-------
PROBLEMS ENCOUNTERED
Time proportional sampling
of leachates proved to be
insufficient to follow the
leaching behaviour in the
field. In order to collect
leachates from individual rain
events r sampling was done ma-
nually in relation to the rain
events. Also, it was impossi-
ble to obtain a leachate which
actually had penetrated the
landfill body. The collected
leachate was rainwater which
ran along the surface and the
interface to the base liner of
the solidified body. This re-
minds one of the fact that
most laboratory tests are per-
formed with a mass-to-leachate
ratio which is very convenient
for laboratory testing (1:3 to
1:20) but does not reflect the
actual field conditions where
the mass to leachate ratio is
usually much higher (5:1).
The rapid setting, typical
for HTA cements, had to be
counteracted with retarders.
Due to the over-stochiometric
addition of Ca(OH)2 and the
alkalinity of the binder, the
leachate of the dry scrubber
material, was highly alkaline
(pH<12) . Thus the leaching of
Pb as a soluble hydroxo com-
plex is a problem. The addi-
tion of trisilicates (Na2S) up
to 1% of the binder was only
partly effective.
After several months of
exposure to rain and atmos-
phere, effects of weathering
were discerned at the surface
of the landfill body. They
included disintegration of the
surface into grains (site Bl)
and/or squaling due to ettrin-
gite formation (site B2).
These crystallization effects
can be diminished with a sim-
ple pretreatment of the ash
with water. The handling of
the material proved to be dif-
ficult . For rapidly reacting
binders, it is not advisable
to use continuously operating
mixing devices. When batching
with a regular concrete mixer,
the cleaning and maintenance
of the equipment became a se-
rious problem. If possible the
water used for cleaning should
be recycled. The evaporation
from the landfill was enormous
(1 mm/h) for certain weather
conditions .
Highly mobile salts such
as NaCl and CaCl2 were readily
leached from the landfill
body. The high concentration
of chloride (up to 40 g Cl~/l)
may have a negative impact on
the aquatic ecosystem of a
small receiving water. No bin-
ding connection developed bet-
ween the individually compac-
ted layers of the landfill.
RESULTS
The following results were
obtained: It is possible to
stabilize a pretreated (e.g.
washed) filter ash with cement
binders so that it becomes
structurally stable for seve-
ral decades. For a successful
immobilization and solidifica-
tion, it is necessary to equi-
librize filter ash with water
prior to mixing it with bin-
ders. The leaching of metals
and nonmetals from the filter
ash can be much reduced by the
stabilization process. Ne-
vertheless, for a few metals
such as lead and zinc complete
retention in the landfill body
137
-------
residues
ESP ash, dry dry scrubber ash, wet scrubber
and washed dry and washed sludge
binders
aggregates
additives
PC, PCHS,
HTS, HOZ,
Bitumen
slag, klin-
ker, calcium
HTA, HTS, PC
slag, sand
retarders,
PC, HTA, HTS
Table 2: Summary of ashes, binders, aggregates and additives
tested; PC: portland cement, PCHS: PC with high resistivity
for sulfate, HOZ: blast furnace slag, HTA: high aluminum
cement, HTS: high silica cement.
residue type
mixture
ash bind.
results of leaching test
Cd Zn Pb Cl
A B A B A B A
untreated esp ash 100
solidified esp ash 40 402)
solidified mixture 35 182)
12 10 2 11
20 0.01 15 0 15
22 <0.01 15 0.04 14
4 10 52
0.01 15 28
0.5 9 nd
dry scrubber ash and 25% bottom ash
washed and solidi- 32 28^)
fied dry scrubber ash-*-)
40 0.4 11 0.4 10
0.02 13 1
untreated wet 100 -
scrubber sludge4)
solidified wet 50 102)
scrubber sludge^)
40
6 10 89 3 10 nd
1 13 0.5 15 0.03 15 100
Table 4: Leaching tests according to US EPA uniform leaching
procedure; A: % metal cumulatively leached during 11 leaching
cycles; B: - log D?; De = Diffusion coefficient after 11
leaching cycles; *•' leaching with CO2 saturated water; 2) HTA;
3> HTS; 4> neutralization with Ca(OH)2 Range of leachate
concentrations (mg/1): Cd 0.0002-4, Zn*0.02-700, Pb 0.001-6
138
-------
was not achieved by the tech-
nics employed in this work.
The results obtained in the
field compared favorable to
the laboratory experiments.
Structural Stability
Some of the mixtures stu-
died appear to produce stabi-
lization effects with satis-
factory physical properties
(compressive strength >1 N/mm^
and resistant to 12 wet/dry
and freeze/thaw cycles). It
has to be emphasized, that the
laboratory tests applied allow
to qualify the stabilized ma-
terial for short to medium
terms only. Thus, it is not
possible to judge the struc-
tural stability for periods
longer than about one hundred
years. Based on the field ex-
periences, it became obvious
that the solidification of dry
filter ashes from esp, and
much more pronounced from dry
scrubbing processes, contain
large amounts of highly active
materials such as CaO, metal
chlorides and others. The CaO
will react with water and will
subsequently produce a lea-
chate with a high pH value
between 10.5 to 12.5. These
reactions are not completed
during the short stabilization
process of the ash products.
In the landfill, these reacti-
ons will continue, and are of-
ten accompanied with a change
in volume and physical stabi-
lity. In order to overcome
these adverse effects, it pro-
ved to be necessary to wash
the filter ashes first. The
products of such an enhanced
and controlled water contact
are much better suited for
long term stabilization. This
fact might have implications
on the type of flue gas clean-
ing technology to be chosen.
Chemical Immobilization
Table 4 shows the leaching
results of some typical trea-
ted and untreated ash test
specimens. From untreated re-
sidues of esp as well as dry
and wet scrubbers large
amounts of heavy metals and
chlorides are leached when
landfilled [Brunner and Bacc-
ini, 1985, Sawell et al,
1988]. The initial washout be-
haviour is controlled by fast
reactions, while the longterm
stability in the landfill may
be determined even by very
slow reactions like exchange
of constituents of the cement
matrix or crystallization.
The chemistry of the slow
reactions of the stabilized
ashes has not been determined
yet. The short term leaching
is mainly pH controlled and
consists of an initial wash
off of the highly soluble me-
tal chlorides. This may result
in a significant (> 1%) loss
of e.g. cadmium and lead. The
initial loss can be prevented
if the ash is put into contact
with water until the soluble
metals are precipitated as hy-
droxides due to the hydrolysis/
reaction of the CaO with
water. If alkaline binders
such as portland cement are
added to improve the
structural stability of the
ash, the pH value should be
kept below 12 in order to pre-
vent the mobilization of lead,
zinc and other metals due to
the formation of soluble hy-
droxo complexes.
139
-------
A sufficient mixing time
(e.g. > 3 min. per nr* total
mixture) is required for opti-
mum immobilization effects.
Comparison of Laboratory Tests
to Field Tests
In order to assess the be-
haviour of the tested materi-
als, it is indispensable to
know the composition of the
ash mixtures. Laboratory tests
should be directed towards the
understanding of the chemicalr
physical and structural re-
actions which take place dur-
ing the interaction of the
ash, the binders and the en-
vironment (water, air,
biosphere). Standard tests
which yield certain factors
like a "pseudo" diffusion
coefficient or the cumulative
leaching of an element from a
test sample, may be used addi-
tionally to compare various
stabilizing recipes; they are
generally not suited to pre-
dict the longterm behaviour of
reactive ash materials in
landfills. The following sta-
tements have been derived from
the laboratory testing and
were confirmed by the field
tests:
Tests with deionized water
are well suited to investigate
the short term behaviour of
the ashes. The results compare
well with the field experience
of the first few months of ex-
posure to the weathering. The
long term behaviour is better
(but by no means completely)
understood if acidic leachates
such as carbonic acid are used
as leaching agents, e.g. to
titrate the buffering capac-
ity.
Laboratory and field tests
show: Chloride is highly
mobile in the treated and
untreated filter ash and can
hardly be retained in the
landfill body. An exception
may be the formation of
Friedel salts with HTA
Lead and zinc are leached
from the treated and untreated
ash samples if the pH value is
allowed to surpass 12 (cf.
also Sawell et al, 1988) .
There is a decrease in the
concentration of leached me-
tals with increasing leaching
(-time and -volume) in labora-
tory tests as well as the
field experience .
Results from laboratory
tests of solid proctor cylin-
ders are generally similar to
the results of tests of the
pulverized proctor cylinder;
this is explained by the hypo-
thesis that the chemical
immobilization of metals is
more important than the physi-
cal exclusion of water. How-
ever, in the field the solidi-
fication decreases the pene-
tration of water into the
landfill body very much, so
that due to the reduced water
flow the material transport is
highly decreased also, at
least in the first few months.
The future development of the
water flux through the body
has not been assessed yet.
The laboratory testing of
the structural properties did
not allow to predict the field
observations of the deteriora-
tion of the structure of the
landfill body. The water
transport in the field is much
different from the laboratory
140
-------
tests. The difficulties of
handling the ash and how to
solve them are only experien-
ced in the field.
helped us to initialize this
paper. The reviewing by Paul
Bishop (Univ. of Cincinnati)
is highly appreciated.
Outlook
Swiss guidelines for waste
management require long term
safety for all products of wa-
ste treatment. In the case of
landfills ("final storage
sites" [Baccini, 1989]), this
objective has to be reached by
treating wastes before land-
filling,, so that no aftertre-
atment of the effluents of the
landfill is needed even for
long time periods. Residues
from FGC systems show a
relatively narrow composition
range. They are considered a
hazardous waste, which
preferably should be disposed
of separately (treatment and
mono-landfill). Thus, a future
reuse as an ore might be
possible. The stabilization
improves the quality of the
filter ashes considerably, so
that most constituents are re-
tained in the landfill body.
Nevertheless, the longterm
fate of some elements is not
well known yet. It is there-
fore still an open question,
if other kinds of processes
such as high temperature tre-
atment of the filter ash come
closer to the high goal of the
Swiss guidelines.
ACKNOWLEDGMENT S
We are gratefull to Clyde
Dial of the US EPA for the in-
formation regarding previous
EPA studies on the subject.
Ebbe Jons of Niro Atomizer
supplied valuable comments and
REFERENCES
US EPA, 1982, Guide to the
Disposal of Chemically Stabi-
lized and Solidified Wastes,
SW 872, Washington DC.
Baccini P. (Ed), "The
Landfill, Reactor and Final
Storage", Lecture Notes in
Earth Sciences, Vol. 20,
Springer Verlag, Berlin and
New York, 1989.
Brunner P.H. and Baccini
P. , The Generation of
Hazardous Wastes by MSW-
Incineration Calls for New
Concepts in Thermal Waste
Treatment, in: Proc. of the
Second Conference on New
Frontiers for Hazardous Waste
Management, Pittsburgh, Pa.f
EPA/600/9-87/018F, p.343-350,
1987.
BUS, Behandlung und Verfe-
stigung von Riickstanden aus
Kehrichtverbrennungsanlagen,
Schriftenreihe Umweltschutz,
Nr. 62, BUS, Bern (Switzer-
land), 1987.
Sawell S.E., Bridle T.R.
and Constable T.W., NITEP
Phase II - Testing of the
FLAKT Air Pollution Control
Technology at the Quebeq City
Municipal Energy From Waste
Facility, Assessment of Ash
Contaminant Leachability, Pre-
pared for Industrial Programs
Branch, Env. Canada, Environ-
mental Protection Waste Water
Technology Centre, Burlington,
Ontario, 1988.
141
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
142
-------
EVALUATION OF STABILIZATION-SOLIDIFICATION TECHNIQUES
Rene GOUBIER
Agence Nationals pour la Recuperation et I'Elimination
des Dechets - Les Transformeurs
FRANCE
ABSTRACT
Among the techniques applied to treat polluting residues in France for the past ten years
has been the mixing of pollutants with reactive agents in order to "fix" the contaminants and to
give them a solid consistancy The first applications of these stabilization/solidification
processes occured in 1978 in the treatment of oil residues from the AMOCO CADIZ spill. They
have also been used for the treatment of a.mayor dump site for petroleum residues, for the
disposal of mineral sludges of a detoxication plant, and for the rehabilitation of sites
contaminated by various industrial residues, specially acid tars generated by oil refining plants.
Although from the beginning these techniques appeared to be able to transform filthy
lagoons into solid and apparently safe areas, it was necessary to evaluate their efficiency and
to determine the conditions and limits of application.
In the beginning of the eighties. laboratory studies were performed to answer these
questions and more recent tests of the present state of sites treated up to ten years ago have
been carried out. Laboratory evaluation showed that processes using lime as the main
reactive agent are efficient in treating organic waste (specialy acid tars) and that those using
pouzzolanic materials are better for fixing mineral contaminants. However, tests of treated sites
show that although these sites have a satisfactory appearance without any harmful! effect
observable in their environment the performances of the treated material are not always as
well "fixed" as desired This situation may be partly explained by the unability of some
treatment processes to fix some specific contaminants, but it also indicates the lack of an
effective analytical technique and of a clear set of performance requirements at the time of
treatment . Since that time, these points have been improved. However, this emphasize the
necessity for the specification of evaluation tests, especially for the estimation of long term
releases of contaminants from treated material. This has become more important as
stabilization solidification techniques seem to have a wide opportunity of developpement for
the treatment before disposal of waste which are the residues of other treatment (i.e
incineration) or which cannot be treated by other means.
143
-------
INTRODUCTION
In France, two main commercial
stabilization-solidification processes have
been applied : the EIF-Ecologie process
based on the use of lime as main reactive
agent an the TREDI-PETRIFIX process based
on the use of pouzzolanlc material.
In the EIF-Ecologie process, the waste
is excaved and placed in a shallow layer
above the ground at the disposal site and the
reactive agent is spread on the waste and
mixed with it. The treated material is then
disposed and compacted in shallow layers.
In the Petrifix process the waste is
extracted, usually by pumping and
transported to a modular system of mixing
and fed with the adjusted quantity of reactive
agent, the treated material is then disposed
at the site.
PURPOSE
Evaluation of stabilization/solidification
processes is needed for the physical
TABLE 1: Elf ECOLOGIE Process (Units in mg/kg
except for pH and resistivity)
caracteristics of the treated materials as well
as for the capacity to fix the various
contaminants. For the case of the two
processes presented, laboratory evaluation
was carried out in 1980 and 1982 and field
efficiency has been recently evaluated at the
occasion of a study realized within the NATO
CCMS Pilot Study Demonstration of Remedial
Action Technologies for Contaminated Land
and Groundwater.
APPROACH
A - Laboratory studies
1 - Evaluation of EIF-Ecologie process
The study was carried out for the
application of this treatment to petroleum
sludges and to acid tars. It consisted in a
leaching test with agitation of a shallow
cylinder of compacted treated waste during
48 hours in demineralized water (100 g per
liter). The result of the analysis of leachates
are gathered in the following table 1 (figures
given in mg/kg except pH and resistivity).
pH
resistivity /cm
COD
Hexane extract.
Chloroforme extract.
Ag
Cd
Cu
Cr
Fe
Mn
Ni
Pb
Zn
Ca
Acid tar
untreated
1.15
37
372840
9800
39400
0.5
1.4
7.6
6.8
165
3.0
6.0
37.0
450
800
treated
12.15
157
3718
75
396
0.13
0.25
0.2
0.21
0.40
0.14
1.33
3.9
0.62
6350
Petroleum sludge 1
untreated
7.12
2960
220
720
4040
<0.2
<0.2
0.4
<0.2
6.5
1.0
1.0
6.0
4.7
170
treated
12.10
161
4080
200
63.5
<0.1
0.2
0.1
<0.1
0.2
<0.1
1.0
1.2
0.4
4235
144
-------
Then, similar tests were performed to
compare pure water to other solvents :
solution representing athmospheric water
(acid rain), water polluted by residues of
aerobic degradation of organic material and
water polluted by residues of anaerobic
degradation of organic material. No
significant differences were found in the
release of contaminants, except for the last
extractive solution in which the contaminants
released were about two times those
released in other solutions.
2 - Comparative evaluation of
stabilization-solidification process
An other study was realized in 1982
that compared the efficiency of different
treatment methods to stabilizeand solidifie
sludges. The selected residues were
representative of inorganic sludge
neutralized by sodium hydroxide (sludge A),
inorganic sludge neutralized by lime (sludge
B), petroleum sludge (sludge C). The following
table 2 represents the characteristics of the
liquid phase of the untreated sludge (figures in
mg/l except pH and resitvitivy (A cm)
Test of physical characteristics and various
leaching tests were applied to treated
materials : leaching with different solvents :
pure water, acid rain, water contaminated by
municipal waste. There after are presented
some of the mains results of these
investigations applied to the EIF-Ecologie and
TREDI Petrifix processes.
Characteristics of treatment
TABLE 3
TREDI
Petrifix
EIF
Ecolo-
gie
Sludge
(kg)
reagent
(kg)
Sludge
(kg)
reagent
Sludged
33.3
26.7
38
10.45
Sludge
33.3
26.7
35
14.40
Sludge C
46.55
77.20
53.1
27.60
Mechanical properties of treated
materiel
(Bending strength (daN/cm2) and
compressive strength (daN/cm2) 28 and 90
days after treatment)
TABLE 4
SABLt Z
PH
resistivity
COD
chloroform
phenol
Sulfate
Chloride
Nitrate
Sodium
Caldium
Potassium
Cd
Crtot.
Cr6+
Cu
Fe
Mn
Ni
Pb
Zn
LJ
Sludge A
9.36
14
- .
-
-
30006
62166
5900
65780
377
105
0.9
2990
2760
0.9
2.3
<0.2
1.4
2.0
1.8
31
Sludge B
9.42
14
-
-
-
3527
70034
6052
7480
54400
154
2.5
1683
1598
1.5
3.4
0.85
5.5
8.1
4.3
140
Sludge C
12.07
111
6291
2112
90
-
601
52
496
75
2541
<0.12
0.8
<0.24
0.5
4.1
<0.2
<0.6
12.3
1.4
0.9
TREDI
Petrifix
EIF
Ecdo-
gie
days
5.Str.
CStr
B.Str.
D.Str.
Sludg A
28 90
J1.5 -
121 170
5 3
7.7 12
Sludge
28 90
23.5 61
105 320
3.12 19
3 9udge C
28 90
19 12
120 210
Freeze-Thaw durability (10 cvcles)
alterat
Petrifix
during
alterat
yellow
altera
Sludge treated by EIF-Ecologie : no
ion observed. Sludge treated by TREDI
: fissuration of treated inorganic sludge
the third cycle.
Wet - dry durability (10 cvcles)
Sludge treated by EIF-Ecologie : no
ion observed - liquid phase colored in
in the case of inorganic sludge (Cr)
Sludge treated by TREDI Petrifix :
lion during the third cycle (mineral
145
-------
B-Rel study
Within the scope of the NACO CCMS
Pilot Study Demonstration of Remedial Action
Technologies for Contaminated Land and
Groundwater, the evaluation of the present
state of sites treated by EIF-Ecology and TREDI
Petrifix has been carried out. For every
treatment technique, two treated sites have
been evaluated. Their main characteristics
are summarized in the following table 5.
Sampling
For every site, three sampling
trenches were realized and for every trench
three samples of three kilograms of material
were taken : one of the treated material from
the upper layer of the treated section; one of
the treated material from the middle layer,
and one from the ground material located
under the treated zone. The idea was that the
upper sample might be considered as
representative of the treated material in
contact with biosphere conditions : freezing,
leaching by infiltrated water, the middle
sample representative of the average
treated material and lower sample would
give an estimation of possible releases of
contaminants from the treated materiel.
However, it appeared that the
representativity of the upper and lower
sample was not reliable and finely we
considered only the middle sample as
representative.
Tests
- measuring of physical properties
. water content
. permeability
. compressive strength
- leaching tests
In France, at the present time, there
are no specific standardized tests for
evaluation of contaminated material treated
by stabilization-solidification. However,
investigations are now carried out that will
propose such tests within less than one year.
Consequently, it was decided :
a) to perform for every average
sample the present leaching test (called INSA
test) applicable to waste material candidate
for landfilling in special industriel waste
landfills. The main features of this test are :
- extraction solvent : demineralized
water saturated with CO2 and air (pH about 5),
- tested material crushed in parts
smaller tham 4 mm,
-100 g of material mixed with 1 liter of
extraction solvent,
- extraction of solution for analysis
after 16 hours of agitation.
Two successive extractions (1 and 2)
have been carried out (table 7)
TABLE 5
Site
A
B
C
D
Waste
Petroleum waste
Acid tars
Agro-industrial
wastes
Tanning sludges
Quantity
24000t
25000t
7500t
9000t
Contort
HC^netab
HCxacids,
metals
organics
cranium
ammonic
, Yearof
treat
1978
1984
1980
1982
Technique
EIF-ECOLOGIE
EIF-ECOLOGIE
TREDI-PETRIFIX
TOEDI-PETRIFIX
146
-------
b) In addition, in order to take in
account the specific characteristics of
solidification techniques it was decides to
perform the new test which is now prepared to
be later standardized for evaluation of
solidified material.
This test called oedometric pressure
leaching test is based on the use of a
pressure permeater, a section of which is
represented on the following figure 1.
FIGURE 1
1 - Sample
2 - Cylinder of confinement
3 - COntact resin
4 — Lower Baseplate
5 — Upper Baseplate
6 - Porous stone
7 - Filler joint
8 - Stud t>olts
The test is performed first to give an
evaluation of the permeability of the treated
material, then the oedometer is operated for
leaching test. The pressure is adjusted in
order to get a discharge of 0.01 cm3/s (36
cm3/hour). Successive extractions are
carried out and it is possible to add
separatley the extracted quantities of every
contaminant and to represent their variation in
function of the quantity of the liquid
discharged through the sample;
This function is hyperbolic qnd its
interpretation allows the evaluation of the total
quantity of the considered contaminant
which can be extracted if the volume of
extraction liquid or the time of extraction was
infinite. The resulting figure will be considered
as the maximum extractible quantity of the
considered contaminant (mg/kg). For the
present study, five successive extractions
have been performed. In some cases they
were not enough to get a reduction of release
of the considered compound and the
evaluation of the total extractible quantity
was not possible ; such situation is
mentionned In tables by > of the total of the
five extractions.
Results
The following tables 6 to 8 give the
results of tests and analyses of the samples
extracted from sites A, B, C and D.
Physical properties
TABLE 6
Water
content %
Permeabi-
Itym/s
Compr.
strength
daN/cm2
Site A
17
7.910-6
4.7
SteB
24
1.2 ID"5
3.8
SiteC
16
3.4 ID"8
1.1
SiteD
57
3.8 Itf8
1
147
-------
TABLE 7 : LEACHING TEST (INSA TEST- FIGURES IN mg kg
except pH and conductivity )
Sites A and 6. sites c and D
Extract
PH
Conduct ms/m
COD
TOC
Cu
Pb
Cd
Co
. Va
N
HC
S0*=
s-
STEA
1 2
127 127
5.6 5.4
2800 2003
900 490
62 3.7
25 21
04 0.2
<0.5 10.6
>o.n
0.02
0.06
0
1.27
>11.8
SfTEB
12.8
5.8
4760
1690
3.26
>0.12
-
-
-
-
>5.9
>5394
<0.5
: ESTIMATION OF LONG TERM RELEASES
Sites C and D
PH
Conduct.
COD
TOC
NH4-
Fe
Cu
Zn
P
0
Ca
SfTEC
8.5
0.4
160
60
271
5.1
0.06
0.03
>4.5
-
-
SITED
9.2
0.44
2690
1200
131
0.6
-
-
-
<0.05
315
148
-------
Problems encountered
The main problem encountered in
the field study was the Inexistence of reliable
data to characterize the initial residues before
treatment (except for site B). This fact made
impossible any comparison before after
treatment in terms of yield. In connection with
this problem remains the question of the tests
used for the evaluation of treatment
efficiency. We can expect to have a
standardised .test procedure in France within
some months as a result of the research work
which is carried on at the present time, but in
the scope of the present studies the situation
was still not clear. The use of the pressure
lixiviation test which is in preparation
remained still experimental.
Results
If we consider the laboratory studies,
some main results may be mentioned :.
- processes bases on the use of
hydraulic cements (like PETRIFIX) are more
efficient for the fixation of mineral
contaminants than those that use lime as
main reagent (like EIF-ECOLOGIE)
- the fixation of the metallic elements
needs the maintain of a sufficient pH. not too
high, to avoid the release of amphoteric Ions.
Hexavalent chronium appears however
specially difficult to fix
very good mechanical
characteristics can be obtained especially
for processes using hydraulic cements
if we consider the field studies:
- we have already mentioned the
problems resulting of the inexistence of any
reliable characterization of the waste before
treatment. One exception is this of site B'for
which leaching tests (INSA) have been
performed before treatment for some
parameters with the following results of
leachates analysis that can be compared
with the results of similar tests of treated
wastes made in the recent field study (one
extration)
TABLE 9
Parameters
PH
Conductivity
ms.m
CODmg/kg
sulfates
mg/kg
before
treatment
210
4.3
52000
41220
after
treatment
12.5
5.3
3500
10310
- Although the characteristics of the
waste before treatment are not well known, it
appears that the TREDI PETRIFIX treatment has
not well fixed ammonia (sites C and D) and
that release of lead and copper remains
important in the case of EIF-Ecologie
treatment. This last point Is not surprising
because of the amphoteric characteristics of
these metals and of the high pH of the
leachate.
- An other weak point of both
treatment applied to the considered sites is
their relatively for efficiency in terms of
physical characteristics of the treated
products. From the consideration of the
results of the previous laboratory tests in which
this efficiency was much better, it seems that
the studied sites treated some years ago had
lack of an efficient survey during the
treatment operation, it Is very probable that
these projects were realized with the idea to
minimise the expenses.
- Since that time, the applications of
these techniques have been improved : the
requirement of the authorities In charge of
surveying the projects have become more
clear and more stringent and the companies
are able to propose insurance guaranties of
the efficiency of their treatment (evaluated
by tests <^f physical and chemical
characteristics) up to twenty years.
Progress will be made by the
publication of standardized tests for
evaluation of the efficiency of stabilization-
solidification processes. This is especially
Important, not only in the scope of projects of
rehabilitation of contaminated sites but also in
the view of the use of such techniques for the
treatment of ultimate polluting waste, mainly
mineral residues of other treatment
techniques (incineration).
149
-------
REFERENCES
1. F. COLIN - Evaluation du precede EIF-
Ecologie applique 6 une boue
d'hydrocarbures et a un goudron acide
residuaire. Institut de Recherche
Hydrologlques de NANCY - RH-80121 Sept.
I960
2. F. COLIN - Evaluation de la fiabilite des
precedes de fixation des boues utilisees en
France - Etude bibliographique et
documentaire prealable - Institut de
Recherche Hydrologiques de NANCY - RH-81-
25-Avritl981
3, F. COLIN - Evaluation de la fiabilite des
precedes de fication des boues utilisees en
France - Rapport final. Institut de Recherche
Hydrologiques de NANCY - RH 82-185 -
Decembre 1982
4. R. 6OUBIER - Evaluation of solidification
stabilization processes. Procedings from the
November 1988 Bilthoven Meeting of the
NACO/CCMS Pilot Study for Contaminated
Land and Groundwater
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
150
-------
IN SITU STABILIZATION/SOLIDIFICATION
OF PCB-CONTAMINATED SOIL
Mary K. Stinson
Risk Reduction Engineering Laboratory
U. S. Environmental Protection Agency
GSA Raritan Depot
Edison, New Jersey 08837
Stephen Sawyer
Foster Wheeler Enviresponse,
GSA Raritan Depot
Edison, New Jersey 08837
Inc.
ABSTRACT
Under the SITE program, a demonstration has been performed on the
International Waste Technologies' (IWT) in situ stabilization/solidifica-
tion process utilizing the Geo-Con deep-soil-mixing equipment. This was
the first field demonstration of an in situ stabilization/solidification
process. The demonstration occurred in April 1988 at the site of a General
Electric Co. electric service shop in Hialeah, Fla., where the soil con-
biphenyls (PCBs) and localized concentrations of
heavy metal contaminants. The demonstrated process
soil in situ with a cementitious proprietary
tained polychlorinated
volatile organics and
mixed the contaminated
additive, called HWT-20, and water.
The technical criteria used to evaluate the effectiveness of the IWT
process were contaminant mobility measured by leaching and permeability
tests and the potential integrity of solidified soils indicated by measure-
ments of physical and microstructural properties. Performance of the
Geo-Con deep-soil-mixing equipment was also evaluated.
The process appeared to immobilize PCBs. However, due to the very low
PCB concentrations in the leachates, caused in part by the low concentra-
tions of PCBs in the soils, confirmation of PCB immobilization was not
possible. Physical properties were satisfactory except for the freeze/thaw
weathering tests, where considerable degradation of the test specimens
occurred. The microstructural analyses showed the process produced a
dense, homogeneous mass with low porosity, which shows a potential for
long-term durability.
The Geo-Con deep-soil-mixing equipment performed well, with only minor
difficulties encountered, which can be easily corrected. The HWT-20 addi-
tive was well dispersed into the soil, as evidenced by the relatively
uniform change in chemical and physical characteristics of treated versus
untreated soils.
The estimated remediation cost with operation of the 1-auger machine,
used for the demonstration, is $194/ton (SIBO/yd-3). For larger applica-
tions, using Geo-Con's 4-auger machine, costs would be lower.
151
-------
INTRODUCTION
Concern by the public and many
government groups exist over using
landfills for the containment of
hazardous wastes. In response to
the Superfund Amendments and Reauth-
orization Act of 1986 (SARA), the
Office of Research and Development
(ORD) and the Office of Solid Waste
and Emergency Response (OSWER) of
the Environmental Protection Agency
(EPA) have established a formal pro-
gram to accelerate the development,
demonstration, and use of new or
Innovative technologies. This pro-
gram is called Superfund Innovative
Technology Evaluation or SITE (1).
The major objective of the SITE Pro-
gram is to develop reliable cost
and performance information on inno-
vative alternative technologies, so
that they can be adequately consid-
ered in Superfund decision making.
The International Waste Technol-
ogy (IWT) in situ stabilization/
solidification process, utilizing
their proprietary additive HWT-20,
and the injection and deep-soil-
mixing technology of Geo-Con, Inc.
were evaluated (2). IWT claims that
their additive chemically bonds to
the contaminants and creates a hard-
ened, leach resistant, concrete-
like solidified mass when treating
soils containing organics. The
demonstration to evaluate the tech-
nology was performed at a closed
electric service shop in Hialeah,
Fla., contaminated with polychlorin-
ated biphenyls (PCBs). In addition,
one small area contained volatile
organics (VOCs) and low levels of
priority pollutant metals. The
owner is required by the local regu-
latory authorities to remediate the
site for PCBs. The SITE program
demonstration was carried out on
two 10x20-ft test sectors selected
by the owner and known, from prior
soil sampling, to be high in PCBs.
The remediation test was performed
in April 1988 and lasted six days.
Pretreatment soil sampling occurred
in March and posttreatment sampling
in May 1988. The SITE Program eval-
uation analyses were performed inde-
pendently of those required by the
local authorities.
PROGRAM OBJECTIVES
The objectives of this SITE Proj-
ect were to evaluate the IWT/Geo-
Con, Inc. in situ stabilization/
solidification technology in the
following areas:
1. Immobilization of the PCBs in
the soil.
2. Effectiveness, performance, and
reliability of the Geo-Con deep-
soil-mixing equipment used for
the in situ solidification.
3. Potential long-term integrity of
the solidified soils.
4. Degree of soil consolidation
(solidification) produced by the
HWT-20 additive.
5. Costs for applying this technol-
ogy on a commercial scale.
APPROACH
The following technical criteria
were used to evaluate the effective-
ness of the process:
o Mobility of the contaminants --
Sampling was conducted in areas
of high PCBs, the primary contam-
inant targeted for immobiliza-
tion, as well as volatile organ-
ics and heavy metals, with the
analytical emphasis on leaching
characteristics. Three Teachabil-
ity tests were performed: the
Toxicity Characteristics Leach-
ing Procedure (TCLP) and two
152
-------
leach tests that evaluate perfor-
mance of the solidified mass,
MCC-1P and ANS 16.1, both orig-
inally developed for the nuclear
industry. Permeabilities were
also measured before and after
soil treatment. The permeabili-
ties indicate the degree to
which the solidified material
permits the passage of water
through the soil mass, and thus,
the degree of water contact with
the contaminants.
o Durability of the solidified
mass -- Core samples of treated
soils were analyzed to determine
uniformity and long-term endur-
ance potential. The analyses
obtained the following:
-Integrity of the treated soil.
-Unconfined compressive strength
(UCS), which is an indication of
long-term durability.
-Microstructural characteristics
as a source of information on
treated soil porosity, crystal-
line structure, and degree of
mixing.
-Freeze/thaw and wet/dry weather-
ing test data on weight loss;
permeability and UCS tests of
the weathered samples provided
indications of short-term
durability.
The stabilization/solidification
process utilized the deep soil in-
jection and mechanical mixing equip-
ment of Geo-Con,Inc. A batch mixing
system prepared a feed slurry of
approximately 57 wt% solids of HWT-
20 additive. The slurry was pumped
to the injection and mixing auger.
Supplemental water was also fed to
the auger; the quantity of water
was dependent on the moisture con-
tent of the soil, i.e., whether the
sample was from above or below the
water table.
The mixing auger consisted of
one set of cutting blades and two
sets of mixing blades attached to a
vertical shaft rotating at 15 rpm.
Two conduits in the hollow auger
shaft allowed for the injection of
the additive slurry and supplemen-
tal water from the base of the
auger. HWT-20 was injected on the
downstroke, with further mixing
occurring upon auger withdrawal.
The soil columns, drilled to a
diameter of 36 in., were positioned
to provide an overlapping pattern
to insure treatment of the entire
area. About 25% of each soil column
overlapped the other columns.
Two 10x20-ft sectors were
treated, each containing 36 soil
columns. One sector was treated to
a depth of 18 ft and the other to a
depth of 14 ft. The depth of injec-
tion was determined by the need to
treat all soils containing PCBs
above 1 mg/kg. The nominal HWT-20
additive rate used was 15 Ib of dry
additive per 100 Ib of dry soil.
For the actual demonstration, the
additive rates were 0.17 and 0.19
Ib/Tb dry soil for sectors B and C,
respectively.
PROBLEMS ENCOUNTERED
The low quantity of contaminants
along with dilution resulting from
the soil blending operation caused
some difficulties in evaluating the
technology. The PCB concentrations
in the untreated soil, with the ex-
ception of a few points, were below
300 mg/kg, with the largest value
measured being 950mg/kg. After soil
treatment, due to the mixing of the
more highly contaminated soils with
soils of lower PCB concentration
plus some dilution from the addi-
tion of HWT-20 and water, the maxi-
mum treated soil concentration was
153
-------
170 mg/kg, with the remaining sam-
ples 110 mg/kg and below. There-
fore, due to the low mobility of
PCBs, the leachate values were very
near the detection limit of PCBs in
water of 1.0 ug/L. For a few sam-
ples at the end of the analytical
program, the PCB analysis procedure
was modified to allow measuring to
a detection limit of 0.1 ug/L. Even
with the increased sensitivity of
the analytical method, the ability
of the stabilization/solidification
process to immobilize PCBs cannot
be confirmed by this project.
Only 3 out of 34 sample point
locations were found to contain
volatile organics (VOCs) and heavy
metals. This did not provide a suf-
ficient number of points to evalu-
ate immobilization of these contam-
inants. In addition, the soil mix-
ing, upon injection of the HWT-20,
severely reduced the concentration
of the VOCs and metals in the
treated soil, further complicating
the evaluation.
RESULTS
The results are presented in
three parts: immobilization of the
contaminants, durability of the
solidified mass plus supporting
physical tests, and operations of
the deep-soil-mixing equipment of
Geo-Con.
Mobility of Contaminants
Solidification and stabilization
are treatment processes that are
designed to accomplish one or more
of the following (3):
o Improve the handling and physical
characteristics of the waste.
o Decrease the surface area of the
waste across which the transfer
of contaminants can occur.
o Limit the solubility of hazardous
constituents, such as by pH
adjustment.
o Change the chemical form of the
hazardous constituents, such as
by chemical bonding.
Solidification obtains these
results primarily by producing a
monolithic block of treated waste
with high structural integrity.
Stabilization techniques limit the
mobility of waste contaminants or
detoxify them, whether or not the
physical characteristics of the
waste are changed or improved (4).
This is accomplished usually
through the addition of materials,
such as treated organophilic clays
(5) which may provide for chemical
bonding, to ensure that the hazard-
ous constituents are maintained in
their least mobile form.
For each test sector, pretreat-
ment and posttreatment samples were
collected at the same locations.
Seventeen samples were collected in
each sector. The soil was analyzed
for PCBs, and a corresponding TCLP
leach test was performed. There-
fore, for each leachate concentra-
tion measured, the soil concentra-
tion was known. Results of treated
soils could be compared to those of
untreated soils by comparing the
quantity of PCBs in the extract to
that in the solid specimen being
leached--regardless of the fact
that localized soil concentrations
changed significantly as a result
of the soil blending operation.
The maximum PCB concentration
measured was 950 mg/kg of untreated
soil,with most of the samples under
300 mg/kg. The untreated soil TCLP
leachates showed PCB concentrations
up to 13 ug/L. Soil samples with
PCB concentrations below 63 mg/kg
had leachate concentrations below
the detection limit of 1.0 ug/L.
154
-------
All soil samples with more than 300
mg/kg PCBs had measurable leachate
concentration. For the untreated
soil samples with PCB concentration
between 63 mg/kg and 300 mg/kg,
only some leachate samples had
detectable quantities.
After additive injection and mix-
ing, the maximum treated soil concen-
tration was 170 mg/kg, with all the
other samples below 110 mg/kg. All
leachates of treated soil samples
were below 1.0 ug/L. In addition,
seven treated soil leachates from
the higher concentration soils were
analyzed a second time with detec-
tion limits reduced to 0.1 ug/L.
Four of the samples were below the
revised detection limit. Thus, it
appears that the IWT process may
immobilize PCBs,but due to the very
low values measured, absolute con-
firmation was not possible. The
results from the highest PCB concen-
tration soil samples are provided
in Table 1.
Volatile organics, specifically
xylenes,chlorobenzenes,and ethyl ben-
zenes,were found in three untreated
soil samples with a total concentra-
tion up to 1,485 mg/kg. Once the
soil was disturbed by the injection
and mixing operation, the maximum
treated soil total VOCs was reduced
to 41 mg/kg. The total VOCs for the
untreated soil TCLP leachates was
2.5 to 7.9 mg/L and for the treated
soil, 0.32 to 0.61 mg/L. This reduc-
tion in VOC concentrations in the
leachates is equivalent in order of
magnitude to the reduction in VOC
concentrations in the soil. Since
the VOC concentrations were over a
wide range and only three samples
were collected, immobilization of
the VOCs could not be determined.
In addition, IWT claims to have
tailored their additive specifi-
cally for PCB immobilization.
Heavy metals, primarily lead,
copper, chromium, and zinc, were
found in only three untreated soil
samples,with a maximum total metals
concentration of 5,000 mg/kg. After
the remediation operation,the heavy
metals concentrations in the soil
samples were reduced to between 80
and 279 mg/kg. The total metals in
the untreated soil TCLP leachate
ranged from 0.32mg/L to 12.65 mg/L,
and in the treated soil leachate
from 0.12 mg/L to 0.21 mg/L. These
total metal concentrations are
quite low compared to their detec-
tion limits and any likely appli-
cable regulatory levels. Due to the
limited quantity of data and the
low soil concentrations, immobiliza-
tion of heavy metals could not be
determined, although it would be
anticipated to occur, since most
cementitious processes immobilize
heavy metals.
The leachate analyses from leach
tests MCC-1P and ANS 16.1 showed
all analyte concentrations, PCBs
and VOCs, below detection limits.
Thus, these tests provided only
limited information and did not
help in confirming immobilization
of PCBs or VOCs.
The permeabilities of the treated
soils ranged from 10"bto 10~'cm/s.
There was a substantial permeabil-
ity reduction compared to the un-
treated soils, which averaged about
1.8 x 10"z cm/s. Even though the
treated soil permeabilities were
greater than the EPA guideline
value of 1 x 10"' cm/s, which is
targeted for hazardous landfill
liners, the four to five order-of-
magnitude permeability reduction in
treated soil will cause groundwater
to flow around, not through, the
solidified monolith.
In summary, it appears from the
155
-------
Table 1. PCBs in soils and leachates.
Sample Untreated
Treated Treated Soil
Desig- Soil
Soil TCLP Leachate
nation* mg/kg
ug/L
PCB Concentrations
Untreated Soil
TCLP Leachate
ug/L mg/kg
B-6
B-7
B-8
B-ll
B-12
B-13
U J, W
B-16
B-17
B-21
B-22
C-l
C-3
C-7
C-10
650
460
220
950
140
250
d W W
300
495
___
—
98
94
150
86
12.
400.
<1.
7.
1.
xi
s!
3.
--
--
<1.
<1.
<1.
<1.
0
0
0
2
1
n
\j
7
0
-
-
0
0
0
0
(15.0)**
(250)**
(0.33)**
(0.50)**
(1.0)**
49 <]
82 <]
9.6
170 <]
16 <]
100 <]
100 <]
60 <]
114 <]
20 <]
57 <]
22 <]
80 <]
L.O
[.0
<1 ,
1.0
1.0
L.O
..0
..0
..0
..0
..0
..0
..0
(0
(0
,0
(<
(<'
(0
(
-------
leach tests that PCBs were probably
immobilized. In addition, the low
permeability of the solidified
mass, minimizing water contact with
the PCBs, reduced PCB mobility.
Nevertheless, PCB immobilization as
a result of the IWT/Geo-Con process
was not confirmed by the SITE
demonstration.
Durability of the Solidified Mass
The ability of the solidified
mass to maintain its integrity over
a long period of time cannot be
determined quantitatively. How-
ever, tests were performed that
indicated the potential for long-
term durability. Unconfined compres-
sive strength provides an indirect
measure of structural integrity.
The results obtained from the demon-
stration test ranged from 75 psi to
866 psi and averaged about 410
psi. These values easily exceed
the EPA guideline minimum of 50 psi
(6). Most other stabilization/solid-
ification processes typically give
UCS values in the range of 15 psi
to 150 psi (7). Therefore, these
results were quite satisfactory.
The effects of weathering can
break down the internal structure
of the solidified soil potentially
producing paths for water flow,
which would increase permeability
and the potential for contaminant
leaching. Wet/dry and freeze/thaw
weathering tests were performed.
These tests, which involved condi-
tions more severe than exposed
solidified material would see in
the field, provided an indication
of short-term treated soil integ-
rity under natural weathering
conditions.
The results for the wet/dry tests
showed very small cumulative rela-
tive weight losses,averaging approx-
imately 0.1% difference between
test specimens and controls. How-
ever, the results for the freeze/
thaw tests were unsatisfactory,
with the cumulative relative weight
losses ranging from 0.5% to 30%,
and averaging about 6.3%. On a few
of the weathered samples, UCS tests
were performed. For the wet/dry
test specimens, the values obtained
were equivalent to the unweathered
samples. However, for the freeze/
thaw tests, samples with cumulative
relative weight losses greater than
3.0% showed reduced UCS values,
approaching zero at 10% weight
loss. Permeability tests were per-
formed on a few freeze/thaw speci-
mens with weight losses up to 6.0%,
with results the same as for the
unweathered samples.
Microstructural studies were per-
formed on untreated and treated
soil samples. Each sample was
studied by scanning electron micro-
scopy, optical microscopy,and x-ray
diffraction. Treatment of the con-
taminated soil produced a dense,
homogeneous mass with low porosity.
There were no variations in quartz
and calcite quantity in the verti-
cal and horizontal directions, even
though the original soil was a
layered structure. These two sets
of observations indicated that the
Geo-Con auger provided good mixing
for the injection of HWT-20 addi-
tive into the soil. These results
indicated that the solidified mass
has a potential to be highly
durable.
Other physical properties of the
untreated and treated soils were
measured, such as bulk density,
moisture content, total organic
carbon, and particle size distribu-
tion. The bulk density of the soil
upon treatment increased 21% as a
result of addition of additive plus
157
-------
water of 32 wt%. This resulted in
a volume increase of 8.5%. While
the volume increase was modest, for
a remediation to a depth of 18 ft
the ground rise was about 18 in.,
which could provide land contouring
difficulties in many locations. The
organics content of the soil was
quite low, usually below 0.1 wt%,
and would not provide any interfer-
ences in the cement hydration reac-
tions. The average of the results
for untreated and treated soils are
provided in Table 2.
Operations
Equipment performance during the
six-day demonstration was smooth,
and there were a minimum of opera-
tional problems. Each soil column
treatment took 30 min. A few opera-
ting difficulties were encountered
and can be eliminated with some
simple engineering design changes.
The difficulties were:
o Control of the various flow
streams could not be maintained
automatically, and manual flow
control was needed. This resulted
in some uneven additive addi-
tions, which with sufficient mix-
ing by the injection auger, did
not lead to discernible physical
property variations.
o The auger positioning deviated
from the targeted point in some
This produced areas
column overlap, which
in areas low in
locations.
of poor
resulted
additive.
o Supplemental water addition was
lost late in the program as a
result of a major leak in the
auger header. To save time,
operations continued without the
supplemental water.
Costs
The IWT/Geo-Con in situ stabiliza-
tion/solidification system is econo-
mical. Remediation costs using the
1-auger machine, used for the demon-
stration,are $194/ton ($150/yd3).
For larger applications, Geo-Con
would use its 4-auger machine and
costs would be lower.
CONCLUSIONS
The following conclusions can be
drawn from the results of the SITE
project demonstration:
o PCB do appear to be immobilized
by the process. However, due to
the very low PCB concentrations
in the soil and leachates, it
could not be confirmed.
o The physical properties of the
treated soil were satisfactory,
which would indicate a potential
for long-term durability. For
each of the test samples, high
unconfined compressive strength,
low permeability, low porosity,
and poor freeze/thaw weathering
test results were obtained.
o Operations were well organized
and ran smoothly; the difficul-
ties experienced should be
readily correctable.
ACKNOWLEDGEMENTS
The analytical services for the
physical and chemical tests were
performed by NUS Corporation of
Pittsburgh, Pa. The microstruc-
tural analyses were performed by
Louisiana State University (LSU)
under the supervision of Scientific
Waste Strategies, a consulting
organization of professors at LSU.
158
-------
Table 2. Average soil properties.
Sector B
Untreated Treated
Moisture Content, wt% 11.8 19.0
Bulk Density, g/mL 1.51 1.85
Permeability, cm/s 1.46xl(T2 5.5xlO'7
Unconfined Compressive
Strength, psi -- 290
Weathering Tests
Wet/Dry, wt% lost -- 0.39*
Freeze/Thaw, wt% lost -- 7.2*
pH 8.1
Oil and Grease, wt% 0.3
TOC, mg/kg 4,380
Sector C
Untreated Treated
13.2 17.3
1.56 1.94
3.5xlO"2 2.7xlO'7
536
0.34*
6.0*
8.5
0.1
2,300
* These values represent the weight loss of the test specimens. The
wet/dry weight losses of the controls were approximately 0.1% less.
For the freeze/thaw controls, the absolute weight losses were in the
range of 0.3% to 0.4%.
159
-------
The authors wish to express their
appreciation for the services
provided.
REFERENCES
1. Superfund Innovative Technology
Evaluation (SITE) Strategy and
Program Plan, EPA/540/G-86/001,
1986.
2. International Waste Technologies
In Situ Stabilization/Solidifica-
tion Hialeah, Fla., EPA Technol-
ogy Evaluation Report, May 1989.
3. Tittlebaum, M.E., et al. State-
of-the-Art on Stabilization of
Hazardous Organic Liquid Wastes
and Sludges,CRC Critical Reviews
in Environmental Control, Vol.
15, Issue 2, 1985.
4. Handbook--Remedial Action at
Waste Disposal Sites, EPA/625/
6-85/006, 1985.
5. Sheriff, T.S., et al. Modified
Clays for Organic Waste Dis-
posal, Environmental Technology
Letters, Vol. 8, 1987.
6. Prohibition on the Placement of
Bulk Liquid Hazardous Waste Land-
fills—Statutory Interpretative
Guidance, EPA 530/SW-86/016,
1986.
7. Guide to the Disposal of Chemi-
cally Stabilized and Solidified
Waste, SW-872 Revised, 1982.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
160
-------
APPLICATIONS OF GEOPOLYMER TECHNOLOGY
TO WASTE STABILIZATION
Douglas C. Comrie
John EL Paterson
Douglas J. Ritcey
D. Comrie Consulting Ltd.
120 Traders Boulevard East
Suite 209
Mississauga, Ontario
L4Z 2H7
ABSTRACT
Hazardous wastes can be rendered innocuous
through chemical (waste stabilization) or
physical (waste encapsulation) methods. A
research program has been conducted at D.
Comrie Consulting Ltd. in order to evaluate
geopolymer as an agent for both waste
stabilization and encapsulation.
Physical properties of solidified waste and sand
mortar mixes have been examined on the basis
of compressive strength testing. Abilities of
geopolymer to render waste chemically
innocuous have been assessed on the basis of
leachate testing carried out in accordance with
Ontario's Ministry of Environment Regulation
309 guidelines.
Preliminary results show this inorganic binder
as extremely effective in reducing metal
leachability in a wide range of wastes, and that
the physical properties of solidified products
make it an ideal candidate for waste
stabilization or encapsulation.
INTRODUCTION
Waste treatment can be effected through both
physical and chemical processes. Waste
treatment and stabilization may be carried out
either independent of, or in conjunction with
physical encapsulation. While chemical
treatment may render the waste itself essentially
chemically inert and no longer susceptible to
releasing toxins in response to leaching, physical
encapsulation concentrates on isolating the
waste from interaction with surrounding surface
and groundwater.
There are many stabilization agents available
for the treatment of toxic waste. Testing of
these materials has met with varying degrees of
success at reducing the leachability of
contaminants in the waste. The purpose of this
research program was to evaluate the
performance of one such stabilization agent - an
inorganic binder known as geopolymer.
161
-------
GEOPOLYMERS
Geopolymers are inorganic binders consisting
of two components: a very fine and dry
powder, and a syropy, highly alkaline liquid.
Liquid and powder portions are combined to
produce a mixture of molasses-like consistency
which is then reacted with the desired waste or
aggregate. Blending of the two binder
components can be readily carried out with the
aid of conventional cement mixing technology.
The resulting waste-binder mixture can be
poured into molds (for example, fibre drums
such as might be obtained from a concrete
products supplier); drier mixes have potential
for extrusion.
The geopolymeric reaction occurs as a result of
reacting alumino-silicate oxides with alkali
(NaOH, KOH) and soluble alkali polysilicates.
Resulting from this reaction is the formation of
SiO4 and A1O4 tetrahedra linked by shared
oxygens. A mildly exothermic reaction in the
alkali activated mixture is accompanied by
hardening and polycondensation.
The basis of the sialate (or, silicon-oxo-
aluminate) network of SiO4 and A1O4
Figure 1: AlO4 and SiO4 tetrahedra - basis
of the sialate network.
tetrahedra is illustrated in Figure 1. Positive
ions, such as sodium, potassium, lithium,
calcium or barium, must be present in the
framework cavities in order to achieve charge
balance.
PHYSICAL CHARACTERISTICS
Geopolymers are characterized by a number of
interesting physical characteristics, including
thermal stability, high surface smoothness,
precise moldability and hard surfaces (4-7 on
Moh's scale) (Davidovits and Davidovits, 1987).
These properties are commonly imparted to
products created through a combination of
binder with waste materials. In terms of waste
stabilization or encapsulation, the most
significant physical property imparted by
geopolymers is their ability to transform soft,
disaggregated or sludge-like wastes into hard,
cohesive solids in remarkably short time frames.
Physical testing has been carried out on
unconfined cubes created from mortar mixes of
sand and geopolymer. Compressive strengths of
40 MPa have been achieved over a 28 day cure
period; strengths of 30 MPa (75% of final
strength) are acquired in the initial two days of
curing (Figure 2). Strengths are acquired with
considerable speed in comparison to mortars
developed from regular Portland cement (Figure
0
\
zSafl^A
/ — —
/ ^^-rfTTTT
«•
Tfflfll
/
1
If
A W
Ul/
/2 [
>8Day T
Jau/tesU
m
^S^oOTcsaiBb**
5T
Compressive Strength CMPa3
Z 4 B 8 10 K M 10 18 20 & » ffl 28 30
Geopolymsr Content (Percent)
Figure 2: Comparison of 2 and 28 day
strengths in inorganic binder
based mortars.
162
-------
4 8 8 TO 12
Setting Tine (Hours)
Figure 3: Comparison of Normal
Portland and Inorganic Binder
Based Concrete Initial
Strengths.
In fact, some inorganic binders have been
developed to form solid, cohesive masses within
one to two minutes of mixing with aggregate or
waste forms. Of course, compressive strengths
achieved depend on the nature of the aggregate
or waste form employed in conjunction with the
binder, and the amount of binder used.
CHEMICAL CHARACTERISTICS
In addition to their physical properties,
geopolymers are characterized by an ability to
resist chemical attack. This is attributed to the
fact that, unlike cements, lime does not play a
part in the lattice structure which provides the
structural strength to the product. The
resistance provided against chemical attack is
best illustrated by the ability of geopolymer to
retain toxins during acidic leaching.
In order to examine this property, a variety of
toxic waste samples have been collected from
different waste generators. Each sample was
split into three and the first portion was
assayed for metal concentrations. The second
portion was subjected to leachate testing in
accordance with Ontario's Ministry of the
Environment Regulation 309, Schedule 4
guidelines, calling for pulverization of the
sample followed by tumbling in acetic acid. The
third portion of each waste was treated with
geopolymer and, after curing for 7 days, leach
tested. Results were compared with the
maximum permissible leachate concentrations
allowed by the Ministry of Environment.
Tests conducted on waste at a southern Ontario
scrap yard have proved most successful.
Bulldozer and soil movement operations at the
scrap yard, which has been used for crushing
and compacting both automotive and industrial
scrap, have resulted in contamination which
extends approximately 2 meters down into the
soil. Of prime concern is lead contamination,
resulting from scrapped car batteries. Analysis
of soil from the yard indicates lead
concentrations in the order of 6,000 ppm, with
25.3 mg/1 leaching out during Regulation 309
style tests. In contrast, samples treated with a
geopolymeric inorganic binder leached out only
2.1 mg/1, well below the Ministry of
Environment limit for safe emissions of 5 mg/1.
Figure 4: Comparison of Leachate Quality
in Treated and Untreated Scrap
Yard Waste.
Similar tests were conducted on contaminated
soils surrounding a southern Ontario refining
and smelting operation. A series of leaks and
spills during the plant's forty year operational
life have resulted in lead contamination to a
-------
depth of three meters. Concentrations of lead
in the soil average about 285,000 ppm, with 204
mg/1 released to leachate during Regulation
309 style leaching. Geopolymerization brings
this level down by an order of magnitude, to
18.3 mg/1.
Figure 5 illustrates the effect on leachate
quality as a result of increasing geopolyiner
content. Ore processing waste (tailings) from
an Ontario base metal mining operation was
solidified with 10, 15 and 25 weight percent
geopolymer.
Figure 5: Comparison of % Loss During
Leaching - Mine Tailings with
10%, 15% and 15% Binder.
WASTE TREATMENT APPLICATIONS
The physical and chemical properties of
geopolymers and of products created with
geopolymers makes them attractive candidates
for a variety of waste stabilization and
encapsulation situations (Comrie, 1988;
Comrie and Davidovits, 1988; Davidovits and
Comrie, 1988). The automotive scrap yard and
the refining and smelting operation noted
above represent good examples of possible
applications.
Work has already been initiated at the
automotive scrap yard on a pilot plant (20
tonne) scale, and contaminated soil has been
screened to remove inordinately coarse material
which might interfere with the blending
equipment. Mixing will be carried out using
conventional concrete mixing technology, and
the resulting inorganically bound mixture will be
poured into fibre barrels. After a short curing
time, this solidified and stabilized waste can
then be transported to a non-hazardous landfill
site, at considerable cost savings over disposal at
a hazardous waste facility. This alternative is
made possible by the fact that solidification with
geopolymer reduces leachate levels to
concentrations significantly below those.
considered hazardous by the Ministry of
Environment in Ontario.
Plans for remediation of the refining and
smelting site noted above call for a combination
of chemical waste stabilization and waste
encapsulation. The most contaminated area on
the site will be excavated to create an empty
"vault", with the removed soil stored nearby.
While the vault is lined with low permeability,
high strength geopolymer concrete, the soil set
aside will be treated with geopolymer to create
a solid of minimal leachate hazard. The treated
soil will then be returned to the vault and
capped with 100 mm of geopolymer concrete,
on top of which will be placed an additional 600
mm of top soil. The entire perimeter of the
vault will be surrounded with peripheral surface
water ditches, resulting in a waste management
unit which contains a waste of minimal
reactivity and which is isolated from interaction
' Topaotl
Low permeability cop
• Low permeability liner
Figure 6: Cross-section of Waste
Management Unit for Treatment
of Smelter Yard Waste.
-------
with ground and surface water (Figure 6).
The unique properties of geopolymer and
inorganic binders, including rapid achievement
of high strengths and the ability to immobilize
chemical toxins even under conditions of acidic
leaching, should result in some interesting
future applications.
REFERENCES
COMRIE, B.C. (1988):
New Hope for Toxic Waste, IN: The World & I,
August, 1988, Publ. Washington Times, pp.171-177.
COMRIE, B.C. and BAVIBOVITS, J.(1988):
Waste Containment Technology for Management of
Uranium Mill Tailings. Presented at the 117th
Annual Meeting of the AIME/SME, January, 1988,
Phoenix, Arizona.
REFERENCES (continued)
BAVIBOVITS, J. and COMRIE, B.C.(1988):
Archaeological Long-Term Burability, of Hazardous
Waste Bisposal - Preliminary Results with Geopolymer
Technologies. Biv. Env. Chemistry, American Chemical
Society, Toronto, June, 1988. Preprint.
BAVIBOVITS, J. and COMRIE, B.C. (1989):
Applications of Geopolymeric Grouts in the Prevention
of Environmental Contamination. American Chemical
Institute National Convention, Atlanta, February, 1989.
Preprint.
BAVIBOVITS, J. and BAVIBOVITS, M.(1987):
Geopolymer Poly(sialate)/Poly(sialate-siloxo) Mineral
Matrices for Composite Materials. Proc. Vlth
International Conference on Composite Materials,
Imperial College, London, UK, July 20-24, 1987.
Preprint.
Synthesis of new high-temperature Geopolymers for
reinforced plastics and composites, SPE PACTEC '79,
Costa Mesa, California, Society of Plastics Engineers,
USA, 1979, pp. 151-154.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and ho official endorse-
ment should be inferred.
-------
INVESTIGATION OF STABILIZING
ARSENIC-BEARING SOILS AND WASTES USING CEMENT CASTING
AND CLAY PELLETIZING/SINTERING TECHNOLOGIES
John J. Trepanowski, David D. Brayack
NUS Corporation
Devon, PA 19087
and
Jeffrey A. Pike
U.S. Environmental Protection Agency
Philadelphia, PA 19107
ABSTRACT
Cement casting and clay pelletizing and sintering are two treatment
techniques for hazardous wastes containing toxic metals. Research in the
past has indicated these techniques may be effective in stabilizing arsenic
wastes. Treatability studies using these techniques were conducted on
arsenic-bearing soils and wastes present at the Whitmoyer Laboratories
Superfund Site.
Treatability study results are presented. The experimental results
revealed that cement casting was somewhat effective in stabilizing a
calcium-arsenic sludge. Stabilization was enhanced by incorporation of a
pre-roasting step prior to hydration or by thiourea addition.
Cement casting was not effective in fixing the arsenic in an iron-
arsenic sludge and a sludge/soil mixture. However, pre-roasting the cement-
waste combination prior to hydration was effective in promoting waste
stabilization.
Clay pelletizing and sintering was partially successful in stabilizing
the calcium-arsenic sludge. This technology was not effective on the iron-
arsenic wastes, however. In all cases, significant quantities of arsenic
volatilized during treatment.
INTRODUCTION
The Whitmoyer Laboratories
Superfund Site is located in Jackson
Township, Lebanon County,
Pennsylvania. Veterinary
pharmaceutical products, including
organic arsenicals, were manufactured
at the site. Prior to 1964, the site
operators treated arsenic-bearing
wastewater by adding lime to effect
arsenic precipitation in an unlined
lagoon. When environmental
consequences from this practice were
noted, the sludge was excavated and
placed into a concrete vault, where
it currently resides (2). An
estimated 2,000 tons of sludge
("vault sludge") containing 370 tons
(740,000 pounds) of arsenic were
placed into the vault. Other site
wastes, including organic chemicals,
were also placed in the vault. These
other wastes have partially mixed
with the sludge. There are concerns
about the
integrity.
vault's structural
The site operator conducted a
groundwater pump-and-treat program at
-------
the site from 1964 to 1971. During
this period arsenic-bearing ground-
water was extracted and treated with
ferric sulfate and lime to
precipitate an iron-arsenic sludge in
two sets of lagoons.
In 1976-1977, the site operator
excavated the iron-arsenic sludge
from one set of lagoons and placed
this sludge on top of the sludge
present in the other lagoons. To
improve the sludge's bearing
capacity, soils were added to the
sludge. Thus, there are two distinct
sludge materials present in these
"consolidated lagoons": a relatively
pure sludge ("lagoon sludge"),
overlain by a mixture of sludge and
soil ("sludge/soil mixture"). An
estimated 4,000 cubic yards of lagoon
sludge and 10,000 cubic yards of
sludge/soil mixture are present in
the consolidated lagoons.
At the time the vault and lagoon
sludges were generated, they were
believed to be environmentally
stable. Recent research (9,10) has
shown, however, that these pre-
cipitates leach an environmentally-
significant amount of arsenic. There
is a concern that the sludges are
contributing to soil, ground- and
surface-water contamination at the
Whitmoyer site.
PURPOSE
The Superfund Amendments and
Reauthorization Act of 1986 (SARA)
requires the U.S. Environmental
Protection Agency (USEPA) to use
treatment technology to the maximum
extent practicable for remediation of
Superfund wastes. To achieve this
goal on the Whitmoyer sludge wastes,
a technical memorandum evaluating the
applicability of treatment tech-
nologies was prepared (1). Based on
cost, potential effectiveness, and
implementability, two technologies,
cement casting (with and without a
pre-roasting step) and clay
pelletizing and sintering, were
recommended for further consideration
as treatment technologies. Since
these technologies were unproven on
the Whitmoyer wastes, bench-scale
treatability studies were initiated.
APPROACH
Background-Cement Casting
Research by Mehta (7) has
indicated that pure calcium arsenate,
ferric arsenate, and arsenic trioxide
can be stabilized (as defined by a
distilled water leaching procedure)
by binding the arsenic in a cement
matrix. The efficacy of the cement
casting technology on other forms of
arsenic (iron and calcium arsenites
and organically-bound arsenic) was
not investigated by Mehta. Johnson
and Lancione (3) tested fourteen
proprietary and nine generic fixation
processes on residues from the
production of arsenical herbicides
(organic arsenicals). They reported
that all of the samples of treated
herbicides released between 28% and
100% of their arsenic when subjected
to a distilled water leach (after
crushing). Lopat Enterprises,
Inc. (6) chemically fixed three soil
samples from the Vine!and Chemical
Company Superfund Site (herbicide
manufacturing plant) using mixtures
of activated carbon, cement, fly ash,
lime and their proprietary ingredient
"K-20 LS." All of the samples
leached less than 0.36 mg/1 arsenic
when analyzed using the USEPA RCRA
Extraction Procedure (EP) Toxicity
test. (USEPA defines a RCRA
characteristic hazardous waste for
arsenic as one with an EP extract
above 5.0 mg/1.) Tetsuro and
Matsunaga (12), as cited in Kawashima
et al. (4), reported that arsenic-
167
-------
containing sludges can be treated
with a 5% aqueous solution of
thiourea, sand and Portland cement to
create a concrete with arsenic
leachate levels "far below regulation
levels."
Mehta (7) demonstrated that the
arsenic solubility of the arsenic-
cement mixtures could be lowered even
further by pre-roasting the mixtures
at 600°C prior to casting and curing.
This reduction in solubility could be
due to the dehydration-crystalliza-
tion effects discussed below.
Nishimura and Tozawa (8)
reported that a crystalline,
anhydrous calcium arsenate with
limited arsenic solubility could be
created by calcining pure calcium
arsenate and calcium arsenite above
700°C. Nearly all of the arsenite
was oxidized to arsenate at roasting
temperatures above 600°C. Stefanakis
and Kontopoulos (11) confirmed
Nishimura and Tozawa's work.
Wenshao (14) reported that this
calcination process was effective on
an industrial calcium arsenate
sludge.
Tozawa et al. (13) demonstrated
that calcining pure ferric arsenate
above 600°C reduced the arsenate's
solubility in the pH range of 2-7 by
forming crystalline, anhydrous ferric
arsenate. However, above 800°C, the
ferric arsenate thermally decomposed
to a more soluble compound. Also, at
this temperature, a significant
amount of arsenic was volatilized.
Background Clav Palletizing
and Sintering
The second recommended
technology, clay pelletizing and
sintering (roasting to form a
coherent mass), was shown by Mehta to
be effective in stabilizing pure
calcium arsenate, iron arsenate,
arsenic trioxide, and arsenic
pentoxide (7). The treated arsenic
trioxide and pentoxide typically
leached arsenic in distilled water in
concentrations above 5 mg/1; however,
this level is much lower than their
untreated arsenic solubilities in
water. The iron and calcium arsenate
leachate concentrations were all
below 0.3 mg/1.
The effects of impurities,
specifically organics and soil, and
other arsenic forms on the clay
mixtures have not been investigated
in the past.
Testing Program and Methodology
To evaluate the efficacy of the
cement casting technology, 42 experi-
ments were conducted on the three
arsenic wastes. Each test's
conditions (as well as the test's
results) are presented in Tables 2
to 4. Key testing variables included
cement/waste ratio, lime/waste ratio,
incorporation of a pre-roasting step
prior to hydration, roast
temperature, and cure time. All
wastes were air-dried and milled (to
break up chunks) prior to treatment
initiation. Wastes and additives
were mixed in a rolling mill. When
conducted, roasting generally
occurred in a muffle furnace;
roasting for tests 5-14 for the vault
waste and 5-8 for the lagoon sludge
took place in a tube furnace.
Mixtures were roasted for 1 hour at
the designated temperature. Between
28% and 48% water was added to the
mixtures during hydration. Cast
samples were cured at 22°C (70°F) and
100% humidity.
Clay pelletizing and sintering
was tested at two ratios,
3:1 clay/waste and 1:3 clay/waste.
Wastes were air dried and milled to
break up chunks; bentonite clay was
added, and the mixtures blended in a
168
-------
rolling mill. The mixtures were
pelletized in a disk pelletizer and
then sintered at 1,000°C. Test
conditions (and results) are
presented in Table 5.
To measure the success of the
treatment tests, three leachate
procedures were used: the USEPA
Toxicity Characteristic Leaching
Procedure (TCLP), which uses acetic
acid as an extractant; a modified
ASTM leachate procedure (Method
D3987-85) where the same treated
sample is extracted for 2 days in
fresh distilled water three times,
with each extract being analyzed
separately; and a second modification
of the ASTM leachate procedure with
identical conditions to the
previously-described test, except
3 g/1 NaHCOs solution is used instead
of distilled water. All three test
procedures require a 20:1 liquid/
solid ratio.
The TCLP procedure is used by
the USEPA to determine if a hazardous
waste needs treatment (or additional
treatment) prior to land disposal.
Krause and Ettel (5) demon-
strated that iron arsenate is more
soluble at neutral and alkaline pHs
than at pHs less than 5. The
modified ASTM distilled water
leaching procedure was selected to
test treated product Teachability
with neutral extractant.
Work by Robins and Tozawa (10)
and Nishimura et al. (9) indicated
that, in the long term, calcium
arsenate and arsenite are decomposed
by carbon dioxide to calcium
carbonate, releasing arsenic acid.
There is a concern that cement-waste
matrices may be attacked by
atmospheric carbon dioxide (C02), C02
dissolved in rain water, and
carbonate derived from mineral
leaching, resulting in decomposition
and arsenic release. To evaluate
this possibility, the bicarbonate
leaching procedure was developed.
This process also serves to test the
Teachability of the treated products
under alkaline conditions.
The project team set a primary
treatment objective of 1.0 mg/1
leachable arsenic for each of the
three leachate procedures. A
secondary objective of a 90%
reduction in total organic carbon
Teachability was set, using the two
non-acetic acid based leachate
procedures. This secondary objective
was only applicable if significant
organics were detected in the wastes.
RESULTS
Haste Characterization
Initial analyses of the
untreated waste samples are presented
in Table 1. As can be seen, the
vault sludge has insufficient calcium
to effectively bind the arsenic as
caTcium arsenate or arsenite. The
Ca/As moTar ratio is 0.92:1, which is
significantTy Tess than the stoichio-
metric ratios of 1:1 and 2:1 required
to form caTcium arsenite* and caTcium
arsenate»calcium hydroxide, re-
spectiveTy (11). The vauTt sTudge's
iron and magnesium content are
reTativeTy smaTl. Thus, a
significant portion of the vault
sludge's arsenic is either present as
a sodium-arsenic compound, organi-
cally-bound arsenic, or in some other
soluble form. The untreated vault
sludge's distilled water leachate
contained 1,650 mg/1 arsenic,
851 mg/1 arsenic, and 512 mg/1
arsenic for the three extractions
(see Table 2). The quantity of
arsenic leached from the sample
during the three extractions was 38%
of the arsenic present. This
supports the contention that much of
169
-------
the arsenic in the vault sludge is
present in a readily soluble form.
The initial analyses for the
lagoon sludge and sludge/soil mixture
are also shown in Table 1. The Fe/As
ratios for both the lagoon sludge and
the sludge/soil mixture were both
greater than 4:1. Thus, there
probably is sufficient iron in the
samples for the arsenic to be bound
in iron-arsenic compounds. This is
supported by the samples' limited
arsenic Teachability as shown in
Tables 3 and 4. Calcium is also
abundant, although it may be bound to
sulfate ions.
Cement Fixation
Cement casting (without a pre-
roasting step) reduced the vault
sludge's TCLP leachate concentration
by a minimum of 91% (99.8% for the
3:1 cement sample mixture -28 day
cure). Significant reductions in the
sludge's distilled water leachate
concentration were also noted. The
TCLP leachate concentration of the
3:1 cement mixture was an order of
magnitude lower than the 1:1 cement
mixture's concentration.
Cement casting without pre-
roasting increased the lagoon sludge
and sludge/soil mixture's leachate
arsenic concentrations in all cases.
This phenomenon may be due to the
increased solubility of iron-arsenic
compounds at pHs above 5 (5). The
cement addition resulted in solution
pHs above 9. Arsenic concentrations
in the leachate decreased with
increasing cement/waste ratios.
Pre-Roastinq
Pre-roasting reduced the
leachate arsenic TCLP and distilled
water concentration of the vault-
cement mixtures by an additional one
to two orders of magnitude beyond
that of cement fixation only;
leachate concentrations were reduced
by approximately two orders of
magnitude for the cement-lime-waste
mixtures. Greater reductions in
leachate concentrations were
generally achieved at higher Ca/As
ratios. TCLP and distilled water
leachate concentrations as low as
0.99 mg/1 and 0.15 mg/1 arsenic were
achieved for the vault waste.
Between 0% and 7% of the arsenic was
volatilized during the roasting step.
Cement casting with pre-roasting
for the lagoon sludge and sludge/soil
mixture similarly was successful in
reducing the TCLP and distilled water
leachate arsenic concentrations
(except for the 1:5 cement/waste
mixture).
To differentiate the importance
of the roasting step from the cement
casting step, a series of tests
(vault waste tests 5-14 and lagoon
sludge tests 5-8; see Tables 2 and 3)
were conducted where the wastes and
waste fixative mixtures were roasted
but not cast (hydrated). For the
vault wastes, TCLP leachate arsenic
concentrations approached levels
reached for the roasted, cast, and
cured samples only when the mixtures
were roasted and excess calcium was
added. This indicated that the
anhydrous calcium arsenate with
limited arsenic solubility described
by Nishimura and Tozawa (8) may have
been formed during roasting.
Roasting temperature increases above
600°C had little effect on the TCLP
leachate arsenic concentration.
Since the lagoon sludge had Fe/As
molar ratios greater than 4:1,
roasting without additives was
conducted at temperatures between
600°C and 1,000°C to see if the low
arsenic solubility, anhydrous ferric
arsenate described by Tozawa
et al. (13) could be created. These
tests were not successful, as up to
170
-------
12% of the arsenic volatilized at the
higher temperatures and the products'
TCLP leachate arsenic concentration
were greater than the concentration
for the untreated sludge.
The mechanism for the increased
stabilization achieved for the iron-
arsenic sludges from roasting is not
readily apparent.
Lime Addition
Lime addition to the cement-
vault waste mixture without pre-
roasting did not discernably enhance
arsenic stabilization. However lime
addition with pre-roasting
significantly reduced the TCLP and
distilled water extracts' arsenic
concentration. This phenomenon
occurred even when cement was not
added to the waste mixture and the
mixture was not hydrated (see
Table 2, experiment No. 14). As
mentioned above, the arsenic
stabilization may be due to the
formation of calcium arsenate.
Thiourea Addition
When 5% thiourea (by weight) was
added to the 1:1 cement-vault sludge
mixture a tenfold reduction in the
arsenic TCLP leachate concentration
was noted. However, when 1% thiourea
(by weight) was added to the
1:1 cement-sludge mixtures, no
appreciable reduction was identified.
Cure Time
TCLP leachate arsenic
concentrations were measured after
both 5 and 28 days of curing. When
the 5-day TCLP leachate arsenic
concentration was greater than
10 mg/1, a significant reduction in
the arsenic concentration was noted
for the 28-day samples. However,
when the 5-day concentration was less
than 10 mg/1, no further reduction
was achieved.
Effect of Carbonate and Cement Casts,
Samples
The cement-waste matrices were
attacked by the aggressive
bicarbonate leaching solution. In
all cases, the NaHCOs extracts
contained more arsenic than either
the TCLP or distilled water extracts.
The NaHCOs effect was most pronounced
on the roasted samples. By the third
extraction, the leachate arsenic
concentration approached the
concentration for the unroasted
samples, i.e., the NaHCOs solution
negated the additional stabilization
provided by roasting.
Clav Pelletizinq and Sintering
The clay pelletizing and
sintering technology was moderately
successful with the vault sludge.
The arsenic leachate concentration
(TCLP) was reduced by a minimum of
95%. However, between 11% and 33% of
the arsenic was volatilized during
sintering.
The clay pelletizing and
sintering technology was not
successful on the lagoon sludge and
sludge/soil mixture samples. The
arsenic leachate concentrations
increased by an order of magnitude
after treatment, and between 20% and
80% of the arsenic volatilized during
sintering.
The clay pelletizing and
sintering technology failed to meet
the project treatment objectives.
Two possible reasons for this failure
are that the organic content of the
samples caused reducing conditions
during sintering, resulting in
arsenic trioxide formation and
volatilization, and that excess iron
in the samples reacted with iron-
171
-------
arsenic compounds to form
Fe203-2FeAs04, with accompanying
arsenic volatilization. This iron-
arsenic compound was noted by Tozawa
et al. (13) to form above 800°C, and
1s more soluble than basic ferric
arsenate.
Conclusions
Cement casting was somewhat
effective in stabilizing the calcium-
arsenic (vault) sludge. Additional
stabilization was provided by
Incorporating a pre-roasting step
prior to hydration. Thiourea
addition to the cement-waste mixture
appeared to enhance stabilization of
the vault waste.
Cement casting was not effective
In fixing the iron-arsenic lagoon
sludge and sludge/soil mixture.
However, cement casting combined with
a pre-roasting step effectively
stabilized these wastes.
Bicarbonate leach test results
indicate that carbonate derived from
atmospheric carbon dioxide,
Infiltrating rainwater and/or
limestone immediately underlying the
site may attack cement-stabilized
wastes over time. If this technology
1s selected for remediation, the
treated waste landfill must be
designed to counter this effect.
Clay pelletizing and sintering
does not appear to be well suited to
the Whitmoyer wastes. This
technology did not meet the treatment
objectives for stabilizing the
wastes.
REFERENCES
Ebasco Services, Inc., 1988,
Technical Memorandum --
Evaluation of Treatment Tech-
nologies, Vault and Lagoon
Arsenate Wastes, Whitmoyer
Laboratories, 25 p.
Environmental Protection Agency,
1981, Remedial Actions at
Hazardous Waste Sites: Surveys
and Case Studies, EPA 430/9-
81-05, U.S. EPA Solid and
Hazardous Waste Research
Division, Cincinnati, OH.
Johnson, J . C., and
R. L. Lancione, 1982, Stabili-
zation, Testing, and Disposal of
Arsenic Containing Wastes, EPA-
600/D-81-104, U.S. EPA Municipal
Environmental Research
Laboratory, Cincinnati, OH.
Kawashima, H., D. M. Misic, and
M. Suzuki, 1985, Review of
Current Practices for Removal
and Disposal of Arsenic and Its
Compounds in Japan, In:
Proceedings, International
Conference on New Frontiers for
Hazardous Waste Management, EPA-
600/9-85/025, U.S. EPA Hazardous
Waste Engineering Research
Laboratory, Cincinnati, OH.
Krause, E. and V. A. Ettel,
1985, Ferric Arsenate Compounds:
Are They Environmentally Safe?
Solubilities of Basic Ferric
Arsenates, In: Proceedings,
Impurity Control and Disposal,
24th Annual Conference of
Metallurgists, August,
Vancouver, Canada.
4.
5.
172
-------
10.
11.
Lopat Enterprises, Inc., 1987,
Treatability Studies for
Chemical Fixation of Arsenic and
Solidification of Soil, Vine!and
Chemical Company Site Project,
Vine!and, New Jersey, 10 p.
Mehta, A. K., 1981, Inves-
tigation of New Techniques for
Control of Smelter Arsenic
Bearing Wastes, Vol. I and II,
EPA-600/2-81-0496 U.S. EPA
Industrial Engineering Research
Laboratory, Cincinnati, OH.
Nishimura T., and K. Tozawa,
1985, Removal of Arsenic from
Waste Water by Addition of
Calcium Hydroxide and
Stabilization of Arsenic-Bearing
Precipitates by Calcination, In:
Proceedings, Impurity Control
and Disposal, 24th Annual
Conference of Metallurgists,
August, Vancouver, Canada.
Nishimura, T., C. T. Ito,
K. Tozawa, and R. 6. Robins,
1985, The Calcium-Arsenic-Water-
Air System, In: Proceedings,
Impurity Control and Disposal,
24th Annual Conference of
Metallurgists, August,
Vancouver, Canada.
Robins, R. 6., and K. Tozawa,
1982, Arsenic Removal from Gold
Processing Wastewaters: The
Potential Ineffectiveness of
Lime, CIM Bulletin, Vol. 75,
pp. 171-174.
Stefanakis, M., and
A. Kontopoulos, 1988, Produc-
tion of Environmentally
Acceptable Arsenites-Arsenates
from Solid Arsenic Trioxide, In:
Proceedings, Arsenic Metallurgy
Fundamentals and Applications,
1988 TMS Annual Meeting,
January, Phoenix, AZ.
12. Tetsuro, Y., and S. Matsunaga,
1977, PPM Journal, 8, pp. 8-21.
13. Tozawa, K., T. Nishimura, and
Y, Umetsu, 1977, Removal of
Arsenic from Aqueous Solutions,
Presented at: 16th CIM
Conference of Metallurgists,
August, Vancouver, Canada.
14. Wenshao, W., Fixation of Arsenic
in Industrial Calcium Arsenate
Sludge at Moderate Tempera-
tures - Shenyang Smeltery,
p. 10.
DISCLAIMER
This material has been funded wholly
or in part by the Environmental
Protection Agency under Contract
Number 68-01-7250 to Ebasco Services,
Inc. It has been subject to the
Agency's review, and it has been
approved for publication. Mention of
trade names or commercial products
does not constitute endorsement or
recommendation for use.
UNTREATED WAST! CONTAMINANT CONCENTRATIONS
(mg/kg unless otherwise indicated)
Aluminum
Antimony
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Menoanese
Mercury
Nickel
•otatttuin
Selenium
Silicon
5ilv«r
tedium
Thallium
Titanium
vanadium
Zinc
Total Organic Carbon (tt)
Sulfate
Anil in*
Vault
Sludge
10.700
ISO
157.000
-------
ANALYTICAL RESULTS
VAULT WASTE TREATED IV CEMENT CASTING
«"*
N*
t
t
}
i
I
)
4
i
t
~8~
t*
I9»
II*
IJ'
ir
14'
Ts~
It
if
it
it
M
CemeM
VludUe
Ratio
j
i
I
1
1
I
1
J
)
1
]
t
1
0?
07
Lmtt
Vatgt
RallO
™~i
i
i
i
OS
06
KTh.0
wea
(Total
Weight)
S
I
—
Roast
Temp
TO
(60 mm)
600
COO
$00
600
100
700
looo
700
COO
700
930
700
•-
700
Cure
ttm«
Ways)
S~
28
28
28~
3~
28
28
28
•-
...
Ca/A
Ratio
57
22
22
22
OS6
086
086
TT
93
S7
57
57
S7
22
22
It
70
70
HAl
Vota.
tilized
TT"
J7"
2~T
rr
"si
37
~~86
~
5T
5T
U
Cement
ilutf«t
Ratio
—
1
1
1
J
1
1
1
1
-
~
-
05
05
02
02
1
HTr»o-
urea
(Total
W«ijhl)
—
—
—
•-
—
—
~
_
~*
-
•"
-
-
—
—
—
1
Riilt
Temp
ro
CO mm
—
—
—
500
COO
'"
—
600
COO
COO
700
800
1000
_
700
...
700
~
Cure
Time
(d.yl)
—
S
28
S
28
S
28
S
28
-
—
-
28
28
28
29
28
SA1
Vola-
tilized
—
—
_.
—
_.
"•
—
-
-------
ANALYTICAL RESULTS
SLUDGE/SOIL MIXTURE TREATED BY CEMENT CASTING
E«p
No.
0
1
1
2
2
3
3
4
4
5
6
7
a
9
Cement.
Sludge
Ratio
3
3
3
3
1
1
1
1
OS
05
0.2
0.2
1
H Thio-
urea
(Total
Wetgm)
-
...
...
...
...
1
Roast
Temp
CC)
160 mm)
600
600
...
600
EOO
...
700
...
700
...
Cure
Tim*
S
28
5
28
S
28
5
28
28
28
28
28
28
HAS
Vola-
tilized
7.7
7.7
...
...
62
62
...
5.9
32.5
TCLP
Leachate
Results
Al
mg/L
236
791
881
005
0.32
37.9
30.1
0,22
028
51 9
1.0
74.2
348
39.2
pH
62
It S
11 a
11 6
11,8
11 4
11 5
11.4
11.5
11 2
II 2
89
9.1
11.7
Distilled HjO Leach
1st Extract
As
mg/L
S 18
924
...
003
...
407
...
027
824
020
101
0.51
494
TOC
mg/L
It
13 1
1.51
...
40.1
...
187
765
326
988
2,66
122
pM
79
124
12.4
...
12.3
...
123
11.9
11.9
116
11.5
12 1
Distilled H;0 Leach
2nd Extract
As
mg/L
447
60,7
003
...
155
-
0.16
23.9
004
15.7
0.41
12.4
TOC
mg/L
128
11,6
7.36
—
19.7
—
4.52
312
6 14
259
3.15
31 7
pH
7.7
124
12.4
-
12.3
...
123
122
12 I
119
11.7
12.3
Distilled H;O Leach
3rd Extract
AS
mg/L
622
5.38
0.04
-
170
....
020
15 1
007
7.18
051
917
TOC
mg/L
98
294
26.7
...
297
—
998
210
123
'31
346
188
pH
86
12.4
12.6
—
11.8
—
12.3
12.1
12.1
11.8
118
124
NaHCOj Leach
1st Extract
At
mg/L
267
126
0.44
—
359
...
102
106
0.66
127
1.33
709
TOC
mg/L
366
25.0
6.4
47.3
8.16
935
362
122
1 16
124
pH
84
129
12.9
128
-
128
122
122
116
11.7
12.2
NaHCOj Leach
2nd Extract
As
mg/L
105
89 1
48.4
24.6
...
159
499
6.04
130
65.4
14.7
TOC
mg/L
119
128
5,99
234
...
5 13
358
243
528
165
31 7
pH
107
125
12.5
12.4
...
125
12 1
12.1
11.8
11 5
12.2
NaHCOj leach
3rd Extract
As
mg/L
115
11 7
13.2
402
...
30.2
706
387
836
82.4
247
TOC
mg/L
182
„.
23.9
17 1
26 1
...
127
263
479
291
3.92
17.6
pM
109
...
12.4
12.4
123
...
12.3
11 9
120
100
10.5
12.1
Uncon-
fined
Compres-
sive
Strength
(psi)
...
...
5.220
5.350
5.000
...
5.570
3.670
2.890
...
•-
3.730
. Uncast c*mtm samples
- ; Not analyzed or reagent not added
ANALYTICAL RESULTS
CLAY f»€UET!ZED AND 51 NT EH ED SAMPLES
Clay:
Sample
Ratio
Sinter
Temp.
(VC)
(IS mm)
HAS
Volatilized
TCLP Leachate
Results
At
mg/L
pH
Diitilled HjO Leach
1st Extract
At
mg/L
TOC
mg/L
pH
OiitilledH/OLeacti
2nd Extract
As
mg/L
TOC
mg/L
pH
Distilled HjO Leach
3rd Extract
As
mg/L
TOC
mg/L
pH
NaHCOj Leach
1st Extract
As
mg/L
TOC
mg/L
PH
NaHCOj LeacH
2nd Extract
At
mg/L
TOC
mg/L
PH
Na HCOj Leach
3rd Extract
As
mg/L
TOC
mg/L
pH
Unconfined
Compfeistve
Strength
(P*0
3
0.33
1,000
1.000
...
33
11
2.260
216
782
70
44
SO
1.650
174
159
1.019
1 19
1 17
86
98
11 1
851
78
12.0
1 28
1 27
85
106
11 5
512
74
9.9
166
143
11 7
9.0
9.7
It 2
2.060
182
364
147
8.7
8.8
9,6
2.710
5-5
659
9.75
60
104
8.7
92
1.900
62
226
5.59
796
10.7
87
9 1
LAGOON S
3
033
LUDGE
1.000
1.000
77 J
25
462
22.9
741
61
47
53
503
0.53
10
245
743
76
11 6
119
488
0.73
073
85
1 82
264
76
116
11 8
186
1 4
1 14
66
124
495
8.4
It 1
114
105
83
2 !
175
6.80
134
8,9
99
10.7
126
232
7.27
8.56
93
98
101
57
149
208
5.39
630
91
9.6
SHIDr.FjSOIL MIXTURE
3
033
1.000
23.1
2.36
37.2
62
56
96
5 18
007
0 16
H
1.49
101
7.9
11 9
122
447
031
0.14
123
1 48
186
77
119
12.1
622
058
026
98
11 5
138
8.6
113
11 7
267
89
067
36.6
14.8
738
8.4
10.6
12.5
105
14,0
30
119
4.37
5.17
10.7
9.6
104
IIS
9.7
58
182
617
5.23
10.9
92
99
Uncast cement samples
Not analyzed or reagent not added
175
-------
EVALUATION OF THE SOLIOITECH
SOLIDIFICATION/STABILIZATION TECHNOLOGY
Walter E. Grube, Jr.
Risk Reduction Engineering Laboratory
I). S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
The Soliditech technology demonstration was conducted at the Imperial
Oil Company/Champion Chemicals Superfund Site in Monmouth County, New Jersey.
Contamination at this site includes PCB's, lead (with various other metals),
and oil and grease.
The Soliditech process mixes the waste material with proprietary
additives, pozzolanic materials, and water, in a batch mixer. Technical
criteria used to evaluate its effectiveness include (1) short-term
extraction and engineering tests, (2) long-term extraction and leaching
tests, (3) petrographic examination, and (4) structural integrity
observations.
Three different waste types—contaminated soil, waste filter cake
material, and oily sludge—and a sand blank were treated. Fourteen cubic
yards of treated waste monoliths, and nearly 300 cast cylindrical mold samples
were produced.
Neither PCB's nor volatile organic compounds were detected in the TCLP
extracts of treated wastes. Significantly reduced amounts of metals were
detected in the TCLP, EP, BET, and ANS 16.1 extracts of treated wastes
compared to untreated. Low concentrations of phenols and cresols were
detected in post-treatment TCLP extracts. The pH of treated waste was near 12.
Unconfined compressive strength of treated wastes was high; permeability was
very low. Weight loss after wet/dry and freeze/thaw cycles was very low.
Portland cement contributed several metals to the treated waste.
Physical stability of treated wastes was high. Data from all extraction and
leaching tests showed negligible release of contaminants. Phenols and cresols
appeared to be formed during the stabilization reactions.
176
-------
INTRODUCTION
The EPA's Office of Research and
Development has been carrying out the
Agency's forma"! program to accelerate
the development, demonstration, and
use of new or innovative technologies
which can provide permanent cleanup
solutions for hazardous waste sites.
The Soliditech, Inc.'s waste
solidification/stabilization process
was the seventh technology to be
demonstrated within this Supsrfund
Innovative Technology Evaluation
(SITE) program.
In cooperation with EPA's Office
of Solid Waste and Emergency Response
(OSWER), the Imperial Oil/Champion
Chemical Superfund site in New Jersey
was selected as the location to
demonstrate the Soliditech SITE
technology. This site is currently
partially occupied by an active
private company involved in blending
and packaging oil products. Techni-
cal staff of the New Jersey Depart-
ment of Environmental Protection
(NJDEP) provided data describing the
characteristics and extent of con-
tamination at this site and assisted
EPA in public relations aspects of
the demonstration.
This technology demonstration was
conducted in early December, 1988. A
batch-mixer, a supply of portland
cement, L)rrichem(tm) reagent, other
additives for their formulation, and
accessory equipment were provided by
Soliditech, Inc. The EPA's support
contractor provided a sampling team.
The demonstration was completed over
a five-day period, resulting in
nearly 14 cubic yards of solidified
material , and over 300 individual
samples for analyses of the numerous
parameters applied to evaluate this
technology.
PURPOSE
The primary goal of the SITE
program is to evaluate the effective-
ness of a technology by conducting a,
field-scale demonstration of each
technology, collecting samples of
treated waste materials and
performing laboratory tests.
TEST METHODS
The Soliditech SITE technology
evaluation was based on the results
of laboratory tests on samples of
waste material before and after
treatment. Physical tests included
ASTM procedures for particle size
analysis, water content, and
unconfined compressive strength (2);
and TMSWC tests for water content,
bulk density of treated waste,
permeability of treated waste, and
wet/dry and freeze/thaw tests on
treated waste (3). Extraction tests
included TCLP extraction, EP
Toxicity, Batch Extraction Test,
American Nuclear Society 16.1, and
Waste Interface Leaching Test (1,4).
EPA SW-846 methods were applied for
pH, Eh, total dissolved solids, total
organic carbon, oil and grease,
volatile organic compounds, semi-
volatile organic compounds, polychlo-
rinated biphenyls, and metals (6).
The Soliditech SITE Technology
Demonstration was described in detail
within the Demonstration Plan, Which
was written and peer-reviewed prior
to initiation of field activities
(5). This demonstration plan also
included an approved Quality Assur-
ance Project Plan, which described
all planned sample acquisitions and
analytical methods.
177
-------
APPROACH
Contaminated soil was excavated
from a pit approximately 5 feet wide,
3 feet deep, and 8 feet long in the
designated "Off-Site Area One" of
this Superfund site. Filter cake was
collected from the open face of a
waste pile (Figure 1). Oily sludge
was scooped from an abandoned storage
tank with a bucket, and stored in
drums until the waste batch was
processed. Approximately equal
parts of oily sludge and filter cake
were mixed to form the third waste
type processed. All waste feedstock
was screened to prevent large objects
such as rocks, roots, bricks, or
other debris from being incorporated
into the treated waste. Although
this debris would not have interfered
with the Soliditech process, it was
removed to prevent inclusion within
samples taken for analytical testing.
Water was added to the waste
within the mixer to provide the
proper mixing consistency. Portland
cement, other specific additives
formulated by Soliditech staff, and
Urrichem were then added and mixing
was completed. The mixture was
discharged from the mixer into one-
cubic-yard plywood forms (Figure 1).
Aliquots of the unsolidified mixture
were taken from the forms and poured
into 267 waxed cardboard and PVC
cylindrical forms, of several dif-
ferent sizes, to provide material for
the evaluation analyses (Table 1).
All of these materials were
allowed to set for 28 days inside a
heated warehouse; the cylindrical
samples were then transported to the
storage area of the analytical
laboratory. The nearly 14 cubic
yards of treated waste contained
within the plywood forms to form the
treated waste monoliths (TWM) were
placed in a two-tiered stack and
covered with a plastic sheet for
subsequent long-term examination
(Table 2).
Figure 2 shows the approaches
that were used to evaluate the
effectiveness of Soliditech's pro-
cess. The Quality Assurance Project
Plan, within the project's Demonstra-
tion Plan (5), specified the details
of sample collection and preserva-
tion, analytical protocols, matrix
and surrogate spike procedures,
blanks, replicate analyses, and
statistical procedures to be applied
to data evaluation. Triplicate
samples were provided for all analyt-
ical parameters in the treated mate-
rial. The Demonstration Report (7)
presents the complete data set
resulting from this technology
evaluation.
RESULTS
Table 3 shows the compositions of
the waste treatment mixtures. The
"Reagent Mixture" includes clean sand
as a substitute for waste; the "Co-
mixture" consists of approximately
equal parts of filter-cake and oily
sludge, because Soliditech preferred
not to treat the oily sludge in its
pure state.
Table 4 shows that, both before
and after treatment by Soliditech,
the density increased and water
content decreased in all cases. The
permeabilities of treated waste were
very low; at values of 1 x 10-3
cm/sec and lower. The unconfined
compressive strengths of the
solidified wastes greatly exceeded
those of the friable, and liquid in
the case of the oily sludge,
untreated materials.
Total chemical analyses of
untreated and treated wastes showed a
varying effect of dilution, depending
178
-------
upon the particular compounds of
interest. PCB's varied from no
observable change to over one-third
less in the treated waste. Analyses
of pure sand solidified using the
Soliditech process showed that
arsenic was present at 59 mg/Kg.
Chromium, copper, lead, nickel , and
zinc were noted to the extent of a
few tens mg/Kg in this sand plus
reagent mixture. A few mg/Kg phenols
and cresols were detected in analyses
of the treated wastes for semi-
volatile organic compounds. The
origin of these phenolics is uncer-
tain, but laboratory contamination
and contribution by the Soliditech
additives and reagents have been
ruled out. Volatile organic com-
pounds were detected in levels up to
nearly ten mg/Kg in the Off-site Area
One soil and filter cake/Oily sludge
mixture; except that 32 mg/Kg xylene
was found in the latter. No vola-
tiles were detected in analyses of
the treated wastes; neither were any
detected by monitoring the environ-
ment above the mixer as waste batches
were being processed.
Table 5 shows the analytes found
in TCLP extracts. Extracts of both
untreated and treated wastes showed
undetectable quantities of PCB's.
Arsenic in the extract from treated
Offsite Area One, at 0.020, and lead
in the extract from treated Filter
Cake, at 0.0020 mg/Kg were the
highest levels of metals of concern
detected. Chromium was found at
0.060 mg/L in extracts from both
treated Filter Cake and treated sand
reagent mix.
Analyses of EP extracts showed no
detectable PCB's from either
untreated or treated wastes. Table 6
shows the reductions in extractable
contaminants after treatment.
The Batch Extraction Test (BET)
includes crushing the sample to pass
an ASTM No. 100 sieve (150 urn),
followed by 7-day extraction with
distilled water of samples at the
three sol id-to-liquid ratios of 1:4,
1:20, and 1:100 (Cote, 1988). Data
from this procedure provide an
indication of the capacity of the
sample (as a reservoir) to provide a
source of Teachable solutes. No
PCB's were detected in any of these
extracts. Aluminum, barium, calcium,
and sodium were contributed by the
Portland cement added to the mixture;
these were the major inorganics
detected in the BET extracts. Lead
was less than 0.05 mg/L (detection
limit) in extracts at all three
solid/liquid ratios; this
immobilization occurred even where
the untreated waste (co-mixture) ;
released 1.7 mg/L into the 1:4
extract. Arsenic was present only at
hundredth mg/L levels in all extracts
of treated wastes, and decreased with
decreasing solid/liquid ratio.
Data from ANS 16.1 leaching over
28 days show no detectable levels of
PCB's, chromium, copper, lead,
nickel , or zinc removed from any of
the three treated wastes. Arsenic
was present at 0.005 - 0.008 mg/L in
all extracts from the treated Off-
Site Area One waste. This repre-
sented the only presence of a contam-
inant of potential concern, and its
concentration was quite low. A
nearly constant quantity of Oil &
Grease, between 1-3 mg/L, was
removed during the 28-days of leach-
ing. Thus, no contaminants of con-
cern were removed in amounts
sufficient to allow calculation of
"Leachability Index," as prescribed
by the ANSI/ANS-16.1-1986 procedure.
179
\
-------
The Waste Interface Leaching Test
(WILT) includes submersion of a mono-
lithic material in distilled water,
with drainage and analysis of solutes
at two-week intervals. No RGB's
could be detected in the WILT leach-
ate fro;n any of the treated wastes.
Arsenic decreased by factors ranging
from 20 to 100, to as low as 0.001
mg/L, among the three wastes treated,
fro:n the first to the sixth leaching
increment. Lead was not detectable
(<0.05 mg/L) in any of the leachates
from treated wastes. Total dissolved
solids decreased by about a factor of
three from the first to the sixth
leaching. Calcium, a good indicator
solute derived from the port!and
cement, also decreased by about a
factor of three from the first to the
sixth leaching.
Petrographic examination of the
solidified, treated wastes was
planned in order to characterize the
homogeneity of mixing, extent of
curing of the concrete-like matrix,
raineralogic composition of the
solidified mass, voids within the
solid matrix, and potential long-term
weathering effects. In addition,
morphologic examination of the
treated waste monoliths (TWM) was
planned to provide long-term data
which describe how well these large
blocks withstand environmental
exposure. Preliminary observations
show that the oil and grease appear
widely dispersed in globules
throughout both the cast cylinders
prepared for laboratory study and the
TWM's. Detailed characterization
data will appear in a later report.
Morphologic examination of the TWM's
after 28-day initial curing showed a
few large masses of oil and grease,
suggesting that the first batch of
waste processed in this technology
demonstration may not have been
thoroughly mixed. A few stress-
relief cracks were noted along
corners of a few of the TWM blocks.
CONCLUSIONS
The high unconfined compressive
strength, very low permeability, and
high resistance to wet/dry and
freeze/thaw deterioration demonstrate
a high degree of physical stability
of the three treated wastes. Calcu-
lation of contaminant release from
these solidified wastes by use of
known diffusion coefficients can be
expected to be accurate, since
advective flow is so low.
Since the concentrations of all
contaminants found in the EP and TCLP
extracts of treated samples are so
low, the Soliditech process has
stabilized the contaminants of con-
cern at the site of this demonstra-
tion. It is significant that as
measured by TCLP, EP, BET, ANS 16.1,
and WILT procedures, lead is barely
detectable in extracts of treated
wastes. This indicates a high degree
of stability in a poorly Teachable
form within the treated wastes. The
BET data confirm the stability of the
treated wastes against leaching loss
of lead and arsenic. The extremely
low amounts of contaminant solutes
found in the WILT leachates confirm
the parallel findings in the shorter-
term extraction tests--TCLP, EP, BET,
and ANS 16.1.
Measurable amounts of arsenic,
barium, chromium, copper, lead,
nickel, and zinc appeared in the
treated wastes. The source of these
elements is believed to be portland
cement. Decreases in loss-on-
ignition are most likely due to
dilution by the added cement.
The absence of any mechanical
equipment problems during the demon-
stration illustrated the reliability
of the Soliditech system. After the
equipment operator gained familiarity
with waste materials at this site,
the process mixed all components into
a homogeneous solidified product.
180
-------
ACKNOWLEDGEMENTS
The EPA was assisted by PRC
Environmental Management, Inc. in
conducting this SITE technology
evaluation; Dr. Kenneth G.
Partymiller is the PRC project
manager. The sampling and analyses
of untreated and treated waste
materials was supervised by
Dr. Danny R. Jackson, Radian
Corporation. Bob Soboleski, Site
Manager in the New Jersey Department
of Environmental Protection, provided
valuable support in demonstration
site selection and public information
in New Jersey. Mr. George C. Kulick,
Jr., Vice-President of Imperial Oil
Company, provided access to the
property on which the demonstration
was conducted.
REFERENCES
(1) American Nuclear Society. 1986.
ANS 16.1 Laboratory Test Procedure.
American Nuclear Society, LaGrange
Park, IL.
(2) ASTM. 1987. annual BOOK or
ASTM Standards, Vol. 4.08. Ameri
Society for Testing and Materials
Philadelphia, PA.
Annual Book of
4.08. American
(3) Cote, P. 1988. (Draft)
Investigation of Test Methods for
Solidified Waste Characterization
(TMSWC). Wastewater Technology
Centre, Burlington, Ontario.
Prepared for RREL, USEPA, Cincinnati,
OH 45268.
(4) Jackson, D. R. 1988.
Comparison of Laboratory Batch
Methods and Large Columns for
Evaluating Leachate from Solid
Wastes. Prepared for RREL, USEPA,
Cincinnati, OH 45268.
(5) PRC Environmental Management,
Inc. 1988. Demonstration Plan for
the Soliditech, Inc., Solidification
Process. WA 0-5, Contract No. 58-03-
3484, USEPA, Cincinnati, OH 45268.
(6) USEPA. 1986. Test Methods for
Evaluating Solid Waste (SW-846),
vol's. IA, IB, 1C and II, Third
Edition. USEPA Doc. Control No.
955-001-00000-1.
(7) USEPA, 1989 (in press).
Technology Evaluation Report SITE
Program Demonstration Test,
Soliditech, Inc., Solidification
Process, EPA/540/x-89/xxx. RREL,
USEPA, Cincinnati, OH 45268.
Disclaimer
This material has been funded wholly or in part by the United States
Environmental Protection Agency under Contract No. 68-03-3484 to PRC
Environmental Management, Inc. It has been subject to the Agency's review
and it has been approved for publication as an EPA document. Mention of
trade names or commercial products does not constitute endorsement or
recommendation for use.
181
-------
QQQ
A - Proprietary additives M -
B - Portland cement supply s -
D - Drums containing oily sludge U -
F - Forms for treated waste monoliths W -
Mixer
Sample preparation
Urrichem supply
Filter cake waste pile
Figure 1. Soliditech Technology Demonstration Operations
182
-------
o
3
o
X
*, O
I H
2 X
^o
UJ
5
o
CO
£
11
Q o
UJ y
OC
5 -c .9
2 ® «
2 f 5
JJ|8
%*«
£ §6
M^
a
O CL
P UJ
O
CO
CO
to
O ^
"O —J CO k—
co O a. 2 UJ
^ h- LU < oa
CO
0) CO
(O CD
.>» CO
CO ^
CD
6
CO
I
3
'co
is:
-------
IX)
_1
CO
UJ
DCXZ
U-Og
#OH
O
Ul
CO
LLI
CO
O
UJ
9
UJ
UJ
UJ
HI
CO
CO
UJ
CO
w
Q>
$ "S
CB
I
O
0)
O)
O
(0
2 3.0 8
0)
T3 O
(0
.^ CO
±2 (D
S i-
o
o
UJ
•= o
« c=
S
•= .2 to
2 OS
G)
Sfi 2
O
£
O
II I II I
-------
UJ
Pie
^o^
QC(0
UJ
D
o
o
W
Q)
S^
2
I
O
CO
(0
a-S
li
*8
I-
O re
O
uu
Q
Z
UJ
O
DC V)
UJ ^
CQIL.Q
200
z m
w
.Ȥ
A O
3 Q.
°0
in co
^ co
o T^
Q) CD
^ re
re o
o *•
o u
*•• g>
5S
!5.°
N>Mt CO
W 3
O W'K
slS
3 w re
.^2 _ ^ ro TO
2oi £2
u
S
o
(/>
W
re.ts
o-o
0) re
•=•0
"- c
j> re
•5.2
3S
U)
re re
o
u
0>
o
00
0)
15. w
o
o »
O "O
±! (0
II
4= O
— D)
"o S
(0
•D
0)
w
0 3
JQ O
00 T-
O T-
co oT
3
JQ Q.
IO T-
CO
-------
co
CO
CD
CU
CO
CO
LO CM
co
O
CO
CO
4->
10
CD
co
CD
u.
UJ
O
UJ
CU
a>
^5
CO X
CJ T3
a. -r- co
CO
co
co
to
co
co
o
CO
CO
O)
CU
CO
T-I CD
CO
CO
(O
CO
eu
O •!-
CJ in
-------
TABLE 5
MAJOR CONSTITUENTS DETECTED IN ANALYSES OF TCLP AND EP EXTRACTS
Volatile Organic Compounds
Acetone
Benzene
2-Butanone
Ethyl benzene
4-Methyl-2-pentanone
Methylene chloride
Tetrachloroethene
Toluene
1,1,1-Trichloroethane
, .Trichloroethene
Xylenes
Semi volatile Organic Compounds
Benzyl alcohol
Butyl benzyl phthalate
o-Cresol
p-Cresol
2,4-Dimethylphenol
bis (2-Ethylhexyl) phthalate
2-Methylnaphthalene
Naphthalene
Phenol
PCBs
Aroclor-1242
Aroclor-1260
Metals (AA)
Arsenic
Lead
Mercury
Selenium
Thai!ium
Metals (ICPES)
Al uminum
Barium
Beryllium
Cadmium
Calcium
Chromium
Copper
Lead
Nickel
Sodium
Zinc
Other Chemical & Physical Tests
Eh
Filterable Residue (TDS)
Oil & grease, infa.red
PH
187
-------
o o
ss
d d
v v
o o
si
°-8
* o ^i
O O Z; O
ocoooor-ooo
a^sasssas
0 S5~dddd
00 v v v
o
V
O O O I
oo" v '
Treated
(N
d
v
•o
1
c
d
v
0)
d
v
o
o\
d
v
o o SB
o o g
d
v
OO <"*! O
dog
< 0.0050
|
°
< 0.0010
0
g
d
v
o
CN
O
0
*r
d
d d
V V
oo o r*» o
o' d
V
o
1
o o
V
o
8
o
v
0 0
ss
o o
V V
o
d
v
o r- o «— ' ^o
8 g 8 <-i 2
d
v
O 0
V
o o f< r-
\o oo
LU
_J
CQ
(3'
E ^
ss
d d
V V
o o
V V
o o o
=. d d
° V V
r-t *O O O CO
S 3 r-f d o" d d
° v v v v
o o •-• oo
°?2 m rJ
o
V
.
.00
Rd d
°v
O r-» O Ol I-H
CN • r-i oo t-» . .
° ss Q
V 00
V
M O 00 <
»O ol vj f
dQ ,
d d ° cj
v v v
O
r-l f
M
I O 00
v
o m o o o
d d
V V
o o
v v
8
o ^r o <
r>l ^ f^l i
O O •-! O O
o v v
o
V
- ^ of d d d d
O O V V
v v
O T-* O O C
o o g § '
v do
V V
o o
V V
O
CJ
O O O Cv
— Oi
11
rii
J2 <.
U fl S = S
t. a tt. o a.
188
-------
The Development of Screening Protocols
to Test
the Efficacy of Bioremediation Technologies
John A. Glaser
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 Martin Luther King Drive
. Cincinnati, Ohio 45268
ABSTRACT
The selection of technologies for the cleanup at National
Priority List Sites often is made difficult with the diversity of
proposed technologies and their relative abilities to treat the
waste materials. Claims of treatment capability based on
treatment of non-site materials may have no bearing on the
treatment efficacy at a specified site. To profitably use the
time period allotted to the evaluation of technology a more
systematized approach is necessary. This can be accomplished is
through the use of a, defined bench scale evaluation of the
candidate technology. Such evaluations should specify control
conditions so that the results can provide an objective means of
technology comparison and utility for the specified waste. Of the
various technologies available for application to remediate
hazardous waste sites, biological treatment is only now beginning
to receive its deserved attention. Clearly, bioremedation is less
developed, for this application, than some of the other
technology options. In keeping with the general objectives of
promoting the use of biological treatment and accomplishing
quality cleanup objectives, this protocol development
provide a means to distinguish, at a small scale, promising
biological treatment specifically targeted to the aerobic
treatment of soil at a specific site. The protocol design permits
the testing of a candidate biological treatment by a third party
to provide an unbiased assessment of efficacy.
strives to
189
-------
INTRODUCTION
The use of treatment
technologies to achieve
quality site cleanups has been
emphasized as the only means
to accomplish the intended
goals of Superfund activities
[1J. Biological degradation
offers the potential for
effective, safe, and
inexpensive cleanup of
hazardous waste contaminants
at many Superfund sites.
Biological treatment utilizes
the various material cycles ie
carbon, nitrogen and etc. to
transform the toxic waste into
benign products within the
biosphere. Current biological
technologies are based on the
use of microorganisms native
to a site or natural
populations selected from
sources apart from the site.
The selection of native
populations or organisms from
other natural sources is made
on the basis of compatibility
with site conditions and/or
the ability to degrade the
organic pollutants at the
site. Native organisms can be
cultured in place through the
addition of nutrients and
growth supporting substrates.
Application of biological
treatment to soils is in a
development stage in terms of
both biology and engineering.
Simple effective inoculation
with non native organisms to
contaminated sites can be
daunting. Novel bioreactor
configurations (such as soil
slurry reactors) are under
study with very little
performance data to support
operation. In many cases
optimal operation conditions
are not known nor have these
type of reactors received
adequate attention even for
proper performance estimates.
In fact, current operations
may be significantly distant
from optimum conditions. The
distinct possibility that
abiotic processes(such as
volatilization or sorption)
significantly contribute to
the overall "apparent"
treatability of a biological
process looms as a general
unknown on the horizon[2,3,4].
Any evaluation of new and
proposed treatment technology
must be cognizant of the
potential of abiotic processes
to contribute significantly to
the dispersal of pollutants.
Abiotic transformations may
not contribute to the intended
conversion into non toxic
products. The mere conversion
of the contamination from one
polluted phase to another,
e.g. soil contamination
volatilized to form air
pollution, does not constitute
a quality cleanup. From an
economic viewpoint, it is
important to design biological
processes that treat the waste
directly without relying on
abiotic means for contaminant
loss or dispersal.
Treatability claims based
on treatment of waste
materials other than those
derived from the site slated
for cleanup may give rise to
treatment expectations that
can never be realized at field
scale. Other microflora, heavy
metals and other environmental
factors may negatively affect
the ability of a biological
treatment technology to meet
its treatment goals. When
190
-------
non-native organisms are added
to the site as part of the
cleanup technology, there is
the distinct possibility that
predation by native organisms
may occur. In one instance,
physical separation and
apparent loss of a complex
pollutant from contaminated
soil could have been
misinterpreted as successful
biological treatment[5]. The
absence of a hazardous waste
treatability data base for
biological processes and the
corresponding knowledge of
operational conditions for
these processes has prompted
the development of this
protocol [6].
PURPOSE
A performance protocol
was devised for the testing of
candidate biological soil
treatment(s) proposed for
hazardous waste site cleanup
using site specific materials
to determine on a small scale
the efficacy of treatment
based on the preceding
considerations. This protocol
is viewed as a means to
distinguish between the
relative merits of
technologies competing as
candidate technologies to
accomplish site cleanup. The
selection of optimal
technology will assist the
achievement of quality
cleanups. This protocol is
designed to narrow the field
of competing cleanup
biological technologies at the
remedial investigation/
feasibility stage of site
cleanup.
APPROACH
This paper presents the
first of a series of protocols
devoted to the formulation of
an objective criterion to
measure treatment. It is
devoted to screening aerobic
soil treatment technologies.
The protocol is scheduled for
use during the remedial
investigation/ feasibility
study(RI/FS) portion of the
remedial process [7]. The data
derived from the remedial
investigation permits the
targeting of technology to
different portions of the site
based on identity, quantity,
and disposition of site
contaminants. With selected
biological technologies, it is
possible screen for the most
promising treatment using
contaminated site materials.
The protocol provides the
detailed, and tested set of
instructions necessary for the
evaluation of a biological
treatment technology to
determine the efficacy and
suitability of a proposed
treatment for a specific site
using actual site materials.
Essentially, the protocol is
designed to assess the rate of
transformation or treatment of
contaminants through the
analysis of contaminant
concentrations and their
changes with time. Short time
periods are required for this
analysis. The entire contents
of the reactor flask are
analyzed for each specified
time frame in replicate
numbers. The apparatus for
this procedure is depicted in
Figure 1. For each analytical
time period specified in the
rate analysis, a separate
apparatus is required. For
instance a rate study using an
experimental plan of 3 periods
in triplicate would require
nine sets of experimental
191
-------
apparatus plus the control
assemblies. The apparatus is
designed to control and assess
the effects of abiotic
loss(eg. volatilization) and
biotransformation.
All the appropriate best
laboratory practices, sample
handling instructions,
appropriate experimental
designs, and suitable quality
assurance/quality control
measures to provide a clear
unequivocal estimate of
treatment efficacy for a
specific site are incorporated
in the protocol. The new
protocol uses information
developed from field testing
studies [8] and other
protocols available in the
literature [9,10,11].
Instructions leading to
selection of samples and
appropriate sampling
techniques are presented.
Directions for use of
appropriate and necessary
analytical chemistry
measurements and their control
are included. Reliance on
existing and tested EPA
analytical methodology was
used to simplify the multitude
of analytical options [12].
Design of the equipment
assembly called for by the
protocol is a reflection of
the concern over control of
avenues of abiotic loss
attributable to
volatilization, sorption, and
photochemical conversions
[13].
The protocol is designed
to allow for general
modifications recommended by
the technology developer to
permit the bench scale test to
closely mimic large scale
operating conditions. A
reaction flask arrangement has
been designated as the normal
vessel containing the
treatment Figure 1. A balanced
gas trapping system is
attached that permits the
analysis for volatile losses
completes the general
assembly. The testing assemble
is designed for ease of
dismantling and the ability to
wash down all surfaces that
would potentially support
deposition of waste
substrates. Modifications of
the simple flask configuration
are permitted as long as the
general objectives of
manipulation and control of
contaminant loss are
maintained. Multiple
assemblies are necessary for
the conduction of the protocol
since the flask contents are
sacrificed to measure the
extent of treatment. Replicate
assemblies are necessary for
each time period with the
background control.
LIMITATIONS
This protocol has been
designed to screen candidate
biotechnologies designated for
potential use for cleanup of a
specific site. The data
collected from this
treatability testing provides
a initial evaluation of rate
and extent of conversion of
site pollutants to detoxified
products. The protocol has
been designed to be
economical. It therefore
doesn't strive to completely
define the rate of conversion
since the scale of testing may
not be totally comparable with
full field scale operations.
The testing assembly has been
designed to insure control
over all environmental loss
pathways that may contribute
192
-------
to reduction of pollution
concentrations. The use of
mass balance is involved to
ascertain that the loss of the
hazardous waste chemical is
the result of biodegradation
and not some other abiotic
process, such as chemical
decay, volatilization,
stripping or sorption, that
may be encountered with full
scale operations [13].
The data derived from
these screening studies is not
intended to predict the rate
or extent of biodegradation at
field scale [14]. The use of
this data to predict costs of
technology operational full
scale or the time necessary
for site closure is not
supported by the design or
intent of the designers of the
protocol.
A new feature of this
protocol is a section
specifying the genotoxic
testing of effluents from the
treatment processes. Since the
avowed direction of the
technology is to detoxify
contaminated materials, the
protocol design team agreed
that this is a desirable
addition and may with time
become a stand- along
protocol. Current requirement
for toxicity testing are those
related to ecological testing
and other tests such as the
Ames test. It is quite
possible that such toxicity
testing may replace the
current analytical
requirements to assay the
extent of treatment. Such a
change is estimated to be less
costly than the current
analytical chemistry
methodologies such as GC/MS.
Until sufficient confidence
can be placed in these new
toxicity measurements, their
effectiveness will be
monitored by correlation with
the results of GC or GC/MS
studies.
DISCUSSION
Due to the diversity of
treatment technology applied
to a variety of contaminated
environmental phases, a
protocol for the evaluation of
aerobic soil treatment
technology was deemed to be
the most desired by the site
managers. A draft protocol has
been peer reviewed and
readjusted to correct points
of deficiency. A testing
phase is currently underway to
determine the suitability of
instructions given within the
protocol. The results of this
testing will be incorporated
in the final version of the
protocol. Protocols for the
aerobic and anaerobic
treatment of contaminated
liquids are scheduled for
future development. A protocol
for the evaluation of
anaerobic soil treatment
technology will be developed
in subsequent years.
ACKNOWLEDGMENTS
This screening of
biological technology protocol
development has been shared by
the U.S. EPA Biosystems
Technology Development
Program, Scientific Steering
Committee and The Soil
Treatment Processes Committee
of the R.S. Kerr Environmental
Research Laboratory.
REFERENCES
1. U.S. Congress, Office of
Technology Assessment. 1985.
193
-------
Superfund Strategy OTA-ITE-252
p 223-253.
2. Hamaker, J.W. 1972.
Diffusion and Volatilization.
In Organic Chemicals in the
Soil Environment vol 1 (eds)
C.A.I. Goring and J.W. Hamaker
pp. 342-397.
3. Plimmer, J.R. 1976.
Volatility. In Herbicides -
Chemistry, Degradation, and
Mode of Action. Vol 2. 2nd Ed.
(eds) P.C. Kearney, and D.D.
Kaufman, pp. 892-934.
4. Dragun, J. 1988. The Soil
Chemistry of Hazardous
Materials. Hazardous Materials
Control Research Institute.
Silver Spring, MD.
5. Unterman, R., M.J. Brennan,
R.E. Brooks, and C. Johnson.
1987. Biological Degradation
of Polychlorinated Biphenyls.
U.S. Army, CERL, Champaign IL.
N-87/12 pp 379-389.
6. U.S. EPA. 1988. Interim
Protocol for Determining the
Aerobic Degradation Potential
of Hazardous Waste
Constituents in Soil. (Draft
Document for Peer Review).
7. U.S. EPA. 1987. The RPM
Primer. EPA 540/G-87/005.
8. Sims, R.C. 1986. Waste/Soil
Treatability Studies for
Hazardous Wastes:
Methodologies and Results.
Vols 1 and 2. EPA/600/6-
86/003a and b.
9. Laskowski, D.A. et.al.
1981. Standardized Soil
Degradation Studies. Test
Protocols for Environmental
Fate and Movement of
Toxicants. In Proceedings of
94th Annual Meeting of the
Association of Official
Analytical Chemists,
Washington, D.C. October
21-22, 1980.
10. Laskowski,
R.L.Swann, P.J
H.D. Bidlack
Degradation Studies
85, 139.
D.A. ,
McCall, and
1983 Soil
Res. Rev.
11. Howard, P.H., J.Saxena,
P.R. Durkin, and L.-T. Ou.
1975. Review and Evaluation of
Available Techniques for
Determining the Persistence
and Routes of Degradation of
Chemical Substances in the
Environment. EPA 560/5-75/006.
pp.223-282.
12. U.S. EPA 1986. Test
Methods for Evaluating Solid
Waste. EPA SW-846.
13. Thibodeaux, L.J. 1979.
Chemodynamics. John Wiley &
Sons. New York.
14. Hamaker, J.W. 1972.
Decomposition: Quantitative
Aspects. In Organic Chemicals
in the Soil Environment vol 1
(eds) C.A.I. Goring and J.W.
Hamaker pp. 255-340.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
194
-------
Table 1. Properties Assessed By Protocol
Biodegradability of Contaminants
Effectiveness of Nutritional Ammendments
Inorganic Supplements (N,P,S)
Electron Acceptors
- Organic Growth Supplements
Effectiveness of Innocula
Cultures of Native Organisms
Specific Degraders
Abiotic Losses
Volatilization
- Sorption
Leaching
Genotoxicity of the Waste
195
-------
FUNGAL BIOTRAP FOR RETRIEVAL OF HEAVY METALS
FROM INDUSTRIAL WASTEWATERS
Theodore C. Crusberg
Pamela Weathers
Ellen Baker
Department of Biology and Biotechnology
Worcester Polytechnic Institute
Worcester, MA 01609
ABSTRACT
Biotraps are living cells or specific cell components capable of
removing or stabilizing toxic substances from waste streams. The fungus
Penicillium ochro-chloron was discovered growing in an electroplating
wastewater stream in Japan. It is not only tolerant to very high
concentrations of divalent metal ions, but it can effectively remove heavy
metals from almost any aqueous waste stream. P. ochro-chloron biotrap was
prepared by growing spores in a glucose-minimal salts medium supplemented
with 0.5 percent Tween 80 for 5 days with constant gentle agitation. The
white raycelia beads 4-6 mm dia. were treated in a Buchner funnel with 80%
ethanol to kill the cells, 15 percent sodium carbonate/bicarbonate pH 9.5,
and then resuspended in an aqueous slurry at pH 4.0. The mycelia beads were
used as an adsorbent in a batch experiment to determine copper-to-mycelia
binding. The distribution coefficient K (concentration of copper ion in a
solution/amount of copper bound to the mycelia) of 0.02 indicates this fungus
should provide for efficient separations of copper by the mycelia in a
continuous batch treatment system. Contercurrent distribution theory was
employed to predict the behavior of a hypothetical copper waste stream
passing through a continuous batch system employing the fungal biotrap as the
immobile phase. This system should be capable of heavy metal uptake and
recovery from both electroplating wastewaters and contaminated aqueous
environments. The use of this fungus biotrap will rival synthetic cation
exchange resins because of lower cost, lower weight per unit of exchange
capacity and ease of application.
INTRODUCTION
Industries and government
facilities which generate heavy metal
wastes are now subject to a great
deal of scrutiny by regulatory
agencies, and they are also deluged
with record keeping nightmares.
Effluent streams from electroplating
processes usually require chemical
precipitation to remove contaminants
before the water can be safely
discharged into the environment.
This in turn generates another form
196
-------
of hazardous waste, either a heavy
metal sulfide or a heavy metal
hydroxide sludge, which also must be
disposed of according to strict
regulations and these disposal
practices are becoming more and more
expensive (9). New processes
developed through applications of
biotechnology may allow industries to
retrieve metals and to minimize
disposal costs with corresponding
economic and political benefits.
The resistance of many organisms
to dissolved heavy metals (copper,
cadmium, nickle, mercury, uranium,
gold, silver for example) has been
noted in bacteria, algae, water
plants and some fungi (3-7). Most of
the heavy metal resistant organisms
have been derived from metal
contaminated sites and by "enrichment
selection".
PURPOSE
The purpose of this study was to
evaluate one particular organism
which was previously shown to survive
and even reproduce in high concen-
trations of copper sulfate (4,5,8,10)
as a potential biotrap for heavy
metal retrieval and recovery from
industrial wastewaters. The choice
of organism for accomplishing the
goal was Penicillium ochro-chloron
(ATCC 42177) which was reported to be
able to tolerate copper, zinc and
manganese levels up to 100 g/L and
cadmium to 20 g/L. Copper was also
found to be accumulated by the
organism up to 1 percent dry weight
(3).
APPROACH
Fungus Cultures
p. ochro-chloron spores were
harvested by suspension in deionized
water from cultures growing in 15 cm
Petrie dishes on corn meal agar
(Scott Laboratories, Fiskeville, RI)
after 4 days at 30°C. Spores are
stored at 4° in deionized water. P.
ochro-chloron mycelia grew in the
form of dense spheres 3-6 mm diameter
when spores were inoculated into 150
mL glucose minimal salts (CMS) medium
(8) with 0.5% Tween 80 at 30°C, and
swirled at 150 rpm in a rotatory
shaker. These spheres or "beads"
were then washed and treated with 80
percent ethanol for 30 min. Uptake
of copper was enhanced if the
ethanol-treated beads were then
washed with a solution of 90 g.
sodium carbonate and 60 g. sodium
bicarbonate per L for 30 min. (pH
9.5) (7), and then exhaustively
washed with distilled water adjusted
to pH 4.0 with IN HC1 prior to use.
Cu-to-Mycelia Binding
A copper-to-mycelia binding
experiment was then carried out. Two
hundred mg of wet mycelia (approx 40
mg dry wt.) was placed into a 50 mL
flask. Ten mL of a solution contain-
ing an appropriate concentration of
copper sulfate at pH 4.0 was added
and a 15 min. incubation was carried
out swirling at 100 rpm at 25°. An
aliquot of the solution was removed
and assayed for copper by either
atomic absorption or by colorimetric
analysis using the bathocuproine
method adjusted to a maximum sample
size of 5 mL (1). Bound copper was
determined from the difference
between the initial amount of copper
present and that found by assay.
Copper uptake was optimal at pH
values below 4.0 (10). Copper could
be removed from mycelia with a wash
in 1 percent HC1. The dry wt. of
mycelia was then determined.
197
-------
Electron Microprobe Analysis
In situ elemental analysis of
fungal mycelia was carried out by
electron microprobe analysis.
Mycelia beads were dried, coated with
a 100 A layer of carbon in a vacuum
evaporator and subjected to energy
dispersive X-ray microanalysis in a
JEOL 300C Scanning-Transmission
(STEM) electron microscope with a
Kevex X-ray system.
PROBLEMS ENCOUNTERED
The greatest problem encountered
was loss of heavy metal binding
ability by the fungus when it was
maintained on non-selective medium.
It is recommended that cultures be
maintained on 2000 mg/L Cu . Some
brands of corn meal agar do not
promote spore formation readily, but
use of the Scott Laboratories product
resulted in spore formation in only 5
days of incubation. Spores also lost
viability if stored in phosphate
buffered saline but not if stored in
deionized water at 4°C.
RESULTS
Mycelia which were extensively
washed with deionized water accumu-
late and retain copper as shown in
the X-ray elemental microanalysis in
Fig. 1. Lighter elements are also
present in amounts which can be
detected. In this analysis copper is
present to about one percent dry wt
of mycelia.
A titration experiment to
determine ^binding parameters of
copper (Cu ) to P. ochro-chloron
mycelia is shown in Fig. 2a. The
ordinate represents the amount of
copper bound (in micrograms) per mg
dry wt. of mycelia. The abscissa
represents the amount of free (un-
bound) copper in solution at equil-
ibrium. The data are means of three
independent determinations. The
solid line represents the curve
calculated from linear regression
parameters shown in Fig. 2b. Here
the data are represented as double
reciprocal plots of values from Fig.
2a. The linear regression equation
is given, with a correlation coeffi-
cient of 0.99 and p<.001. From the
intercept with the ordinate the maxi-
mum binding of copper to mycelia was
calculate to be 3.73 micrograms/mg.
The distribution coefficient K
is defined as the amount of copper in
solution (mg/L) in equilibrium with
copper bound to mycelia (mg/Kg).
This is analogous to a two phase
system consisting of water and an
immiscible solvent into which the
solute can dissolve. In this
situation the immiscible "solvent" is
the mycelia. The data in Fig. 2
allow Kp to be determined at low
values of copper in the linear
portion of the curve. The value of
Kp was found from this analysis to be
0.02.
Craig and Craig (2) showed that
in a series of tubes each containing
the same volume of two immiscible
solvents a solute would partition
according to K when the solute was
first introduced into the first tube
(tube 0). After equilibrium is
achieved the upper solvent from tube
0 is transferred into the next tube
(tube 1), new solvent is added to
tube 0, and the tubes are shaken. At
equilibrium the same type of transfer
is again carried out with solvent
being transferred from tube 1 to tube
2, from tube 0 to tube 1, and new
solvent is added to tube 0. An
expression was derived for the
fraction of solute (Fn,r) in any tube
after any number of transfers. K is
the partition coefficient, n
198
-------
represents the number of the tube and
r the number of the transfer:
n.r
X
r+1
/U+K)n]
Although these relationships
were derived for countercurrent
distribution separation processes
used by the pharmaceutical industry,
they apply to the system described
here as well. This is because fungal
mycelia may be treated analogously to
the lower immiscible solvent layer.
A simple computer program was
written to enable the calculations of
the fraction of solute in each tube
in a series of 20 tubes after 19
transfers (Fig. 3) for various values
of K. For values of K<0.1 the solute
is retained in the lower solvent or
in this case the mycelia of the first
few tubes. In the case of P. ochro-
chloron beads serving as adsorbent
for copper ion where K = 0.02 similar
retention in a continuous batch
process would be expected. The value
of K=10 is also shown because this
represents a probable value of K. in
the presence of 1 percent HC1. In
this case the solute originally
adsorbed to mycelia in the first few
tubes would be expected to rapidly
elute from the mycelia and could be
collected in a small volume.
Computer analysis for smaller numbers
of tubes (5-11) using K = .02 shows
similar results with the majority of
the water soluble solute appearing in
the upper solvent of the first tube.
A pilot plant or industrial
continuous batch process might be
expected to contain a more convenient
number of tanks (represented by tubes
in this analysis), perhaps less than
10. In addition, the ratio of liquid
(wastewater) to mycelia in an actual
process would not be 1:1 (an
assumption made to derive the Craig
equation) but more likely 10:1.
CONCLUSION
These studies are important in
developing new methods for industrial
wastewater treatment. The use of
ehtanol-killed mycelia beads elimi-
nates the need to immobilize cells or
use heavy mats of fungal mycelia.
The beads are easy to handle and
process in column or batch systems.
Heavy metal uptake occurs at sites
within the cell wall matrix (5) and
not within the cytoplasm. Uptake and
release are achieved under conditions
which are easy to manipulate. The
fungus is a renewable product which
should provide a cost-effective
method for dealing with retrieval of
heavy metals from contaminated waste-
waters by continuous batch or column
technologies.
REFERENCES
1. Am. Public Health Assoc., 1981,
Standard Methods for
the Examination of Waters and
Wastewaters, 15th Ed., Procedure
313C.
2. Craig, L.C., and Craig, D., 1950,
Technique in Organic Chemistry,
Weissberger, A., Ed., Vol. Ill,
Wiley Interscience, N.Y.
3. Filip, D.S., et al., 1979,
Residual Heavy Metal Removal by
an Algae-Intermittent Sand
Filtration System, Water Res.,
Vol. 13, pp. 305-313.
4. Fukami, et al., 1977, Uptake of
Heavy Metals by a copper-tolerant
Fungus, Penicillium ochro-
chloron, Agricultural Biochem.,
Vol. 41, pp. 17-22.
5. Fukami, M., et al., 1983,
199
-------
8.
9.
6.
Distribution of Copper in the
Cells of the Heavy Metal Fungus
Penicillium ochro-chloron
Cultured in Concentrated Copper
Medium, ^Agricultural Biochem.
Vol. 47, pp. 1367-1369. "
Galun, M. , et al., 1983, Removal
of Uranium (IV) from Solution by
Fungal Biomass and Fungal Wall-
related Biopolymers, Sci. , Vol.
219, pp. 285-286.
7. Macaskie, L.C., and Dean, A.C.R.,
1985, Uranium Accumulation by
Immobilized Cells of a Citro-
bacter sp., Biotech Lett.. Vol.
7, pp. 457-462.
Stokes, P.M., and Lindsay, S
1979, Copper Tolerance
Accumulation
ochro-chloron
Copper-plating
Mycologia. Vol
E.,
and
in Penicillium
Isolated from
Solution,
71, pp. 788-806.
U.S. Environmental Protection
Agency, 1987, Meeting Hazardous
Waste Requirements for Metal
Finishers, EPA/625/4-87/018.
10. Crusberg, T.C., et al., 1989,
Bio trap for Removal of Heavy
Metals, in Adv. in ion
Chromatography, Vol. 1, Jandik,
P., and Cassidy, R.M., Eds.,
Century International, Inc.
pp. 247-260.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
200
-------
CO
3F P. OCHRO-CHLORON TREATED WITH
ERGY DISPERSIVE X-RAY MICROANALYSIS (
w
o O :::::::«!,
Hi 0.1 ::::::%
CD CD : : ::: jr-
133 133 ^3 : : : TjJ-
! lljiil-
~~\ "*m
n ::::::: :::::::::::::::::::::: I;-:::::::::::::::::::::: f
OJ CD [[[ 4
0.1 rci ::::::: [[[/
Q. I::::::::::::::::::::::::::::::::::::: ::::::: ;5«
.Hi w :::::::::::::::::::"""":::!::iii;;iii;i!ii!i!i;!i;<
^ -p [[[-"
C. :::::::::::::::::::::::::::::::::::::::::cn :::::t::.-=.
CD 3 ::::::::::::::: :1 1:::::::::::::::::::::::::::::::: :•!" *
Cxj o ::::::::::::::::::::::::::::::::::::::::::::: i :: 1 : 1 i i : i :::::: l"^1
CO O :: :::•.•.:::::::::::::::::: :£*
'•• "* i_u.iJ ig
£_ ¥ ' *
0 II .-. O ".
>: i • -f-> :::::;;:::::::;:::;;:; "::::::::;:;::;::::"i";
-------
2.0
I
1.0
U
CD
0
o
i t
10 20 30 40 50 60 70 80 90
Jfree mg/L
Fjg.2a. Titration of fungal mycelia with Cu+2.
1
*MBMM
"V
2.5
2.0
1.5
1.0
.5
Y=.27+15.8X
r=0.99
P<.001
J I
8 10 12 14
X 102
0 2
Fig. 2b. Double reciprocal plot of data from Fig. 2a.
202
-------
O
cs
CO
— o
CN
f
o
4J
rH
O
to
OJ
1
1 **
• fm~m
1
1
1
CN
O
*~~
CO
o
^^
CD
_n
E
D
C
0)
••—
,
M-l
O
C
O
•rl
4J
•H
4-1
CO
•H
4-1
Pi
n
3
0
}_i
(U
4J
C
3
O
0
»
w
^-1
0
to
0)
1
to
§
•H
>
}_,
0
«w
to
Q)
<^
^
CO
6)
D O
>o
CN
O
CN
10
O
203
-------
r
COMPOSTING OF EXPLOSIVES AND PROPELLANT
CONTAMINATED SEDIMENTS
Richard T. Williams* and P. Scott Ziegenfuss,
Roy F. Weston, Inc., West Chester, Pennsylvania,
and Gregory B. Mohrman and Wayne E. Sisk, U.S. Army
Toxic and Hazardous Materials Agency
Two field-scale demonstrations were conducted to
investigate composting as a technology for remediating
explosives and propellant contaminated sediments. Test
sediments at the Louisiana Army Ammunition Plant contained
approximately 76,000 ppm of total explosives, including TNT
(66% of total explosive), RDX (25%), HMX (9%), and tetryl
(0.3%). The mixture that was composted consisted of
straw/horse manure, alfalfa, horse feed, and sediment. Two 12
cubic yard piles were constructed, one was maintained at
approximately 35°C and the second at approximately 55°C.
After 22 weeks, total explosives were reduced by 99% (from
17,872 to 74 ppm) in the thermophilic (55°C) pile.
Transformation products peaked in concentration at
approximately 20 days and subsequently fell to near detection
limits. At the Badger Army Ammunition Plant, test sediments
contained approximately 18,000 .ppm of nitrocellulose (NC). NC
was reduced from 13,086 ppm to 16 ppm after 101 days in a
thermophilic pile.
204
-------
Introduction
The manufacture and handling of explosives and
propellants has resulted in soil and sediment contamination at
U.S. Army munitions facilities, often as a result of
previously acceptable waste disposal practices. The United
States Army is currently investigating several technologies
for decontaminating propellant- and explosives-contaminated
soils. Among these candidate technologies is composting.
Composting is a process during which organic materials
are biodegraded, resulting in the production of organic and/or
inorganic transformation products and energy in the form of
heat. This heat is trapped within the compost matrix, leading
to the self-heating that is characteristic of composting.
Composting for the purpose of hazardous substance destruction
is initiated by mixing a matrix contaminated with
biodegradable compounds (explosives and propellants in the
present investigations) with organic carbon sources and
bulking agents, which are added to enhance the porosity of the
mixture to be composted.
Contaminants of concern at the Louisiana Army Ammunition
Plant (LAAP) include 2,4,6-trinitrotoluene (TNT),
hexahydro-l,3,5-trinitro- 1,3,5-triazine (RDX),
octahydro-l,3,5,7-tetranitro-l,3,5,7-tetraazocine (HMX) and
N-methyl-N,2,4,6-tetranitroaniline (tetryl). Structural
formulas for these compounds are presented in Figure 1. These
explosives are found as contaminants in lagoon sediments as a
result of the disposal of "pink water", which is generated
from wash-down operations during munitions packing and
loading. The name originates from the pink coloration the
explosives produce in solution.
Previous research has indicated that TNT is microbially
transformed, but is not completely mineralized to inorganic
products [1,2,3]. Conditions that have been found to enhance
the biotransformation of TNT include high organic carbon
concentration and aerobic conditions [4]. Microbes generally
catalyze nitro-group reduction of the TNT molecule [5]. A
number of TNT biotransformation products are known. Evidence
exists that these metabolic transformation products adsorb
strongly to organic materials [6].
Anaerobic conditions and high organic carbon content have
been found to enhance the biotransformation of RDX and HMX
[4,7], Laboratory-scale composting studies with 14C-RDX have
demonstrated high levels of *4CC>2 production (37 to 46 percent
of initial 14C activity added), suggesting ring cleavage and
complete mineralization [8].
The contaminant of concern at the Badger Army Ammunition
Plant (BAAP) was nitrocellulose (NC) . NC may contain from
11.1 percent nitrogen (cellulose dinitrate) to 14.5 percent
nitrogen (cellulose trinitrate). NC has been found to be
susceptible to microbial attack in lab and pilot-scale
composting systems [8].
205
-------
The primary objective of these studies [9, BAAP report in
preparation] was to evaluate the utility of aerated static
pile composting as a technology for remediating soils and
sediments contaminated with the explosives TNT, HMX, RDX, and
tetryl and the propellant nitrocellulose. Secondary
objectives included evaluating different materials handling
and process control strategies and determining transformation
products when Standard Analytical Reference Materials (SARMs)
were available.
Materials and Methods
Two 8-inch-thick concrete test pads were constructed 20
feet apart adjacent to the pink water lagoons -at LAAP.
Drainage channels in the pads were connected to a sump located
below grade. Water from the sump was reapplied to the compost
piles.
The mixture to be composted at LAAP was prepared using
horse manure and soiled bedding (straw), alfalfa, horse feed
(Purina Balanced Blend 14) , and contaminated sediment.
Sawdust, wood chips, and baled straw were used to construct
bases and insulating covers (see Figure 2).
A mechanical feed system, developed initially to meter
explosives-contaminated soil into an incinerator, was used to
homogenize sediment and to mix the material to be composted.
Dual-channel strip-chart recorders (Omega Engineering)
were used to continuously record compost temperatures. Two
landfill thermocouple probes (Atkins Technical) were placed in
each pile. One probe was placed 2 feet inside each pile, 3
feet above pad level, and 2 feet from the blower end of the
pile. The other probes were inserted into the center of each
pile. A landfill probe equipped with a hand held digital
thermometer was used to monitor temperatures daily at 6
locations in each pile.
Each compost pile contained a system of perforated and
nonperforated polyethylene drainage tubing (ADS, 4-inch
diameter) placed on top of a wood chip base and connected to
an explosion-proof radial-blade blower. The blowers were used
to pull air through the compost piles. Blower cycling was
controlled by both timer and temperature feedback systems.
The temperature feedback system consisted of soil thermistors
that measured compost temperature and panel-mounted Fenwal
series 551 thermistor sensing temperature controllers.
The mixture to be composted was prepared as follows.
Sediment (excavated and homogenized 2 weeks previously and
analyzed for explosives concentration) was reprocessed through
the feed system once. Horse feed was mixed into the sediment
on the mixing pad using an excavator bucket. Straw/manure,
alfalfa, and 35 pounds of fertilizer (13-13-13) containing
nitrogen (13 percent w/w), phosphorous (5.7 percent w/w), and
potassium (10.87 percent w/w) were added to the mixture. The
materials were mixed using an excavator and front-end loaders
206
-------
for 30 minutes. The mixture was processed through the feed
system once. Each load of compost was moistened with water as
it was delivered to the pads. Approximately 400 gallons of
water were applied to each compost pile. Table 1 shows the
amount of each material in the mixture.
Pile construction was completed and the temperature
control systems and recorders started on 25 February 1988.
Samples were analyzed for contaminant concentration nine times
over a 153-day test period. The compost piles were
individually dismantled, remixed, and remoistened at day 33,
60, and 111.
Analyses for lead, selenium, and arsenic were -conducted
by procedures in Standard Methods for Chemical Analysis of
Water and Wastes (U.S. EPA 600/4-79-020, 1979). Compost
samples were analyzed for TNT, RDX, HMX, tetryl, and
transformation products by USATHAMA Method LW02, modified for
the extraction and analysis of compost. Compost samples were
analyzed for NC by USATHAMA Method LY02.
The BAAP field demonstration was similar to the LAAP
demonstration except that cow manure was used instead of horse
manure/straw, an agricultural mixer (Knight Manufacturing) was
used to prepare the mixture, and 6 thermocouple probes were
used in each pile. Table 2 shows the amount of each material
used in one of the four piles constructed at BAAP.
Results
The LAAP sediment contained TNT (56,800 mg/kg), RDX
(17,900 mg/kg), HMX (2,390 mg/kg), and tetryl (65O mg/kg).
The sediment was combined with the other components of the
mixture to be composted according to the materials balance
presented in Table 1.
The BAAP sediment for piles 1 and 2 contained 18,800 ppm
of NC, and for piles 3 and 4, 17,027 ppm of NC. The sediment
was combined with the other components of the mixture to be
composted in pile 3 according to the materials balance
presented in Table 2. Piles 1 and 2 contained 19 percent soil
by weight. Pile 3 contained 22 weight percent soil and pile 4
contained 32.5 weight percent.
Total explosives (summation of TNT, RDX, HMX, and tetryl)
concentrations in piles 3 and 4 at the beginning of the study
were 16,460 and 17,870 mg/kg, respectively! After 153 days,
the concentration of total solvent extractable explosives in
pile 4 was 74 mg/kg. A linear plot of total explosives
concentration versus time for the thermophilic pile , is
presented in Figure 3. The mean and standard deviation at
each time point represents at least three and as many as nine
replicate samples.
Samples of water from the sump were analyzed for
explosives and transformation products on days 0, 16, 22, and
153. The analytes were below detection limits, which varied
207
-------
depending upon the age of the sump water, in all sump water
samples tested.
Day zero compost samples were analyzed for TNT
transformation products. Concentrations of transformation
products peaked in concentration at approximately 20 days and
subsequently fell to near detection limits.
The physical appearance of the compost changed
considerably over the 153-day test period. When the compost
was initially mixed, it had a highly fibrous appearance, a
rough texture, and it smelled conspicuously of the manure,
urine, and feed used to prepare it. After approximately 100
days, the compost had become more soil-like and less fibrous
in appearance. At the end of the test period, the compost had
both the appearance and smell of loamy soil.
NC in pile 2 (thermophilic) at BAAP was reduced from
3,039 ppm at day zero to 59 ppm at day 70. A linear plot of
NC concentration versus time for pile 2 is presented in Figure
4. The NC concentration did not change significantly over the
remainder of the 153 day test period.
The variation in NC concentration for any given sampling
time (representing 5 individual samples) decreased with time.
Duplicate analyses for any one sample showed good
reproducibility. However, at earlier stages of the test
period, significant differences were observed in NC content
between different sampling locations within all four piles
studied at BAAP.
Discussion
The concentration of solvent-extractable total explosives
was significantly reduced during the 153-day LAAP test period.
Fate mechanisms that may have been responsible for contaminant
reduction include sorption of explosives and transformation
products to the compost matrix,, incorporation of explosives
and transformation products into environmentally stable
molecules, and mineralization of explosives to carbon dioxide,
water, and other inorganics.
Previously published literature indicates that
biotransformation of TNT, RDX, and HMX does occur. However,
with the exception of RDX [8], significant mineralization of
these compounds has not been demonstrated. Work by Kaplan and
Kaplan [6] indicated that incorporation of chemically reactive
transformation products into compost matrices increased with
increasing compost age and was the primary fate process
involved in explosives composting. These studies were,
however, conducted at laboratory scale in externally heated
flasks. Consequently, microbial community development and the
corresponding metabolic activity may have been less than that
observed in a large scale compost. The significance of
mineralization in the present study cannot be determined from
the data generated.
208
-------
Previous research on explosives composting with
14C-labeled test materials has demonstrated extensive (and
apparently nonreversible) binding of 14C activity to a compost
matrix over time [3,8]. In one study [8], less than 1 percent
of the initial 14C activity (spiked as J-4C-TNT into compost)
was recovered as 14CC>2 after 6 weeks of composting, while 66
percent of the 14C activity became bound to the compost matrix
and was unextractable. The production of organic transforma-
tion products from TNT, RDX, and HMX is well documented
[10,11]. In the present study, TNT transformation products
were detected in initial compost samples, increased in
concentration over the first several weeks of the test period,
and decreased to low mg/kg levels thereafter. These data do
not support one environmental fate process over another.
The concentration of solvent-extractable NC was
significantly reduced during the 151-day test period for piles
1 and 2 at BAAP. The same fate processes potentially
responsible for the loss of explosives at LAAP also could be
responsible for NC loss at BAAP. Previously published
literature [8] indicates that mineralization of NC does occur.
Therefore, it is probable that NC was mineralized to some
extent during the BAAP field demonstration. However, the
exact contribution of mineralization by microbial activity
cannot be determined from the data collected.
The results of these field demonstrations indicate that
composting is a feasible technology for reducing the
extractable concentrations of NC and explosives/transformation
products in contaminated soils and sediments. Consequently,
composting may be a suitable technology for remediating
propellant- and explosives-contaminated soil and sediment.
The compost residue, however, must be acceptable for disposal
in a manner which makes composting cost effective. Additional
chemical characterization of the residue as well as residue
toxicity studies are recommended.
Acknowledgement
The authors thank W. Sniffen, L. Morse, E. Schaefer, G.
Perry, and E. McGovern for technical support. This work was
sponsored by the U.S. Army (contract purchase order No.
DAAK11-85-D-0007).
The
views, opinions, and findings
contained in this publication are those of the authors and
should not be construed as an -official Department of the Army
position, policy, or decision unless so designated by other
documentation.
References
1. McCormick, N.G., F.E. Feeherby, and H.S. Levinson.
"Microbial Transformation of 2,4,6-Trinitrotoluene and
Other Nitroaromatic Compounds." Appl. Environ.
Microbiol. 31: 949-958 (1975).
2. Kaplan, D.L. and A.M. Kaplan. "Composting Industrial
Wastes - Biochemical Considerations." Biocycle 23: 42-4
(1982) .
209
-------
3. Isbister, J.D., R.C. Doyle, and J.K. Kitchens,
"Composting of Explosives." U.S. Army Report DRXTH-TE
(1982) .
4. Spanggord, R.J., T. Mill, C. Tsong-Wan, W.R. Mabey, J.H.
Smith, and S. Lee. "Environmental Fate Studies on Certain
Munition Wastewater Constituents." Final Report No. AD
A082 372, Phase 1 - Literature Review (1980).
5. Carpenter, D.F., N.G. McCormick, J.H. Cornell, and A.M.
Kaplan. "Microbial Transformation of 14C-Labeled
2,4,6-Trinitrotoluene in an Activated Sludge System."
Appl. Environ. Microbiol. 35: 949-954 (1978).
6. Kaplan, D.L. and A.M. Kaplan. "Reactivity of TNT and
TNT-Microbial Reduction Products with Soil Components."
United States Army Technical Report, Natick/TR-83/041
(1983) .
7. McCormick, N.G., J.H. Cornell, and A.M. Kaplan. "The
Anaerobic Biotransformation of RDX, HMX, and Their
Acetylated Derivatives." U.S. Army Technical Report,
Natick/TR/85/007 (1984).
8. Doyle, R.C., J.D. Isbister, G.L. Anspach, and J.F.
Kitchens. "Composting Explosives/Organics- Contaminated
Soils." U.S. Army Report AMXTH-TE- CR-86077 (1986).
9. Williams, R.T., P.S. Ziegenfuss, and P.J. Marks. "Field
Demonstration - Composting of Explosives Contaminated
Sediments at the Louisiana Army Ammunition Plant (LAAP).
U.S. Army Report AMXTH-IR-TE-88242 (1988).
10. Kaplan, D.L. and A.M. Kaplan. "Thermophilic
Biotransformations of 2,4,6-Trinitrotoluene Under
Simulated Composting Conditions." Appl. Environ.
Microbiol. 44: 757-760 (1982).
11.
Won, W.D. and R.J.
Trinitrotoluene." U.S.
(1974).
Heckly. "Biodegradation of
Army Report No. AD 921 232L
Disclaimer
The work described in this paper was not funded by .the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
210
-------
Table 1.
Materials balance of mixture to be composted at LAAP.
Material
Volume
(cu yd)
Mass
(Ib)
Percent
Volume
Mass
Sediment
Alfalfa
Straw/manure
Horse Feed
1
13
16
4
2,300
940
2,480
4,000
3
38
47
12
24
10
25
41
Total
34
9,720
100
100
211
-------
Table 2.
Materials balance of mixture to Ibe composted in Pile 3 at BAAP.
Material
Soil
Alfalfa
Feed
Wood chips
Manure
Volume
(yd1)
1.50
5.2
1.6
0.7
3.7
Mass
(Ib)
2,550
904
1,776
440
5,800
Percent
Volume
11.8
40.9
12.6
5.5
29.1
Mass
22.2
7.9
15.5
3.8
50.6
Total
12.7
11,470
100.0
100.0
212
-------
NO2
2,4,6-Trinitrotoluene
TNT
N
i
NO2
Hexahydro-1,3,5-Trinitro-
1,3,5-Triazine
BOX
o2N
02N
NO2
Octahydro-1,3,5,7-Tetranitro-
1,3,5,7-Tetraazocine
HMX
NO2
N-Methyl-N,2,4,6-ietra-
nitroaniline
Tetryl
Figure 1. Structures of explosives of concern at LAAP.
213
-------
Roof
Wood chip
cover and
base
/. jr. • •. -.-_••. . -•-_*-•";
Concrete pad (18'X30'X8" thick)
Note: Schematic only, not to scale.
Figure 2. Cross-section schematic of compost pile
with roof, Louisiana Army Ammunition Plant.
214
-------
T3
•B
CO
•a
In
+\
Q)
<3
g
I
03
JD
I
O
TO
s
O
• CM
O
co
o
CO
V)
>.
(0
TJ,
0)
E
.Ea
«5
o> <
>-j
-------
O.I
l
II
CD "2
in
D)
O)
T I
CO CM
(spuesnoiit)
UOUBJIUOOUOQ
ON
O
— o
i
- Q 2.
O.
£3
O GO
co *;
O «
co
o>
_ o
co
-
i: Q.
c o
O E
o £
C 0)
O £
OF
I
O)
_ o
CM
216
-------
A NEW BIOTECHNOLOGY FOR RECOVERING HEAVY METAL IONS FROM WASTEWATER
Dennis W. Darnall
Alice Gabel
Bio-recovery Systems, Inc.
P.O. Box 3982
4200 S. Research Drive
Las Cruces, New Mexico 88003
ABSTRACT
Bio-recovery Systems has developed a new sorption process for removing
toxic metal ions from water. This process is based upon the natural, very
strong affinity for biological materials, such as the cell walls of plants and
microorganisms, for heavy metal ions. Biological materials, primarily algae,
have been immobilized in a polymer to produce a "biological" ion exchange
resin, AlgaSORB®. The material has a remarkable affinity for heavy metal
ions and is capable of concentrating these ions by a factor of many thousand-
fold. Additionally, the bound metals can be stripped and recovered from the
algal material in a manner similar to conventional resins.
INTRODUCTION
Bio-recovery Systems has
developed a new sorption process for
removing toxic metal ions from
water. This process is based upon the
natural, very strong affinity of
biological materials, such as the cell
walls of plants and microorganisms,
for heavy metal ions. Biological
materials, primarily algae, have been
immobilized in a polymer to produce a
"biological" ion exchange resin,
AlgaSORB®. The material has a
remarkable affinity for heavy metal
ions and is capable of concentrating
these ions by a factor of many
thousand-fold. Additionally, the
bound metals can be stripped and
recovered from the algal material in
a manner similar to conventional
resins.
PURPOSE
This new technology has been
demonstrated to be an extremely
effective method for removing toxic
metals from groundwaters. Metal
concentrations can be produced to
very low parts per billion (ppb)
levels. An important characteristic
of the binding material is that high
concentrations of very common ions
such as calcium, magnesium, sodium,
potassium, chloride and sulfate do
not interfere with the binding of
heavy metals. Waters containing a
total dissolved solids (TDS) content
of several thousand and a hardness of
217
-------
several hundred parts per million
(ppm) can be successfully treated to
remove and recover heavy metals.
The process is particularly effective
in removing mercury, lead and
cadmium, but also works well for
other metals such as chromium,
copper, nickel, zinc, uranium, cobalt
and manganese.
Past waste disposal practices
have caused serious heavy metal
contamination to many groundwater
supplies. Both acute and chronic
illnesses in humans, animals and
plants can be caused by even very low
concentrations of heavy metals. A
major source of these contaminants
is leachates from legal and illegal
landfills and drainages from old
mines. This problem is increasing,
and attention is being focused on the
development of new methods to
remove heavy metals from
groundwater.
Scientists at Bio-recovery
Systems have developed a proprietary
sorption technique for removing
heavy metals from contaminated
wastewaters. This process is based
upon a very strong affinity of heavy
metal ions for functional groups on
algae cells. The algae cells are non-
living and have been immobilized in a
silica polymer to produce a
"biological" ion exchange resin called
AlgaSORB®. Figure 1 shows the
reaction of divalent and trivalent
metal ions with an algal cell. The
metal ions interact with various
chemical groups on the cell surface
to produce a very strong complex.1
AlgaSORB® functions as a
biological ion-exchange resin. Like
ion exchange resins AlgaSORB® can
be recycled. Metal ions have been
sorbed and stripped for as many as 75
cycles over a two-year period with no
noticeable loss in efficiency.
AlgaSORB® is superior to synthetic
ion exchange resins when the
wastewater being treated is high in
total dissolved solids or organic
materials. Under these conditions
synthetic resins would be either
inefficient or would be unoperational.
Thus, AlgaSORB® is particularly
efficient in removing heavy metals
from groundwaters which naturally
contain high TDS.
Over the past two years Bio-
recovery Systems has been awarded
two Small Business Innovative
Research (SBIR) contracts and a
contract as part of the Emerging
Technologies program under the
auspices of the Superfund Innovative
Technologies Evaluation (SITE)
program. All three of these contracts
were from the United States
Environmental Protection Agency to
research and develop the AlgaSORB®
technology for commercial
applications. Results from these
contracts, some of which are
summarized below, clearly show the
efficiency of AlgaSORB® for
removing heavy metals from a variety
of sources and have led to
commercialization of the technology.
APPROACH
The testing of metal ion
binding by algae was performed in a
column configuration. These
experiments were performed by
passing the metal-containing
solutions through a column that was
218
-------
packed with AlgaSORB®. Metal ion
analyses were preformed on both the
influent and effluents from the
column, and the metal binding was
determined from the difference in the
metal concentration in influent and
effluent fractions. Effluents were
often collected from these columns
untii "breakthrough" of metal
occurred, and often sufficient
volumes of metal-containing
solutions were passed through the
columns until the effluent
concentrations of metal ion was the
same as influent concentration. At
that point the column was washed
with deionized water and a stripping
solution was passed over the column.
Effluents containing stripped metal
ions were then analyzed for metal
content.
All experiments were
performed in duplicate or triplicate
to insure that results were
reproducible. Careful attention was
given to the control of pH, and cell-
free control experiments were
performed to insure that no
precipitation of metal ions (such as
metal hydroxides) occurred as a
result of pH adjustments or the
addition of other ions.
The techniques used for metal
ion analyses included atomic
adsorption spectrophotometry,
graphite furnace analysis, direct
current argon plasma
spectrophotometry, and/or
ultraviolet-visible
spectrophotometry. Solutions and
standards were matrix matched, or
the method of standard additions was
used wherever necessary. Complete
details of metal ion analyses are
found in recent publications. 1-6
RESULTS
Removal of Heavy Metals from
Ground water
A= Removal of Copper from
Contaminated Groundwaters
Containing Halogenated
Hydrocarbons. Bio-recovery Systems
obtained groundwaters which had
been contaminated with copper,
tetrachl oroethy I ene and
dichloroethylene by a printed circuit
board manufacturer. These waters
contained a total dissolved solid
content (TDS) of nearly 2000 ppm and
had a total calcium and magnesium
content of approximately 300 ppm.
Past experience had shown that ion
exchange resins were not effective in
treating these waters for copper
removal because of i) the high
mineral content and ii) the propensity
of the resins to become clogged with
the organics in these waters.
However, experiments showed that
400-bed volumes of the copper
containing waters could be passed
through a column (0.7 cm ID x 13 cm
high) containing AlgaSORB® without
effluents from the column containing
more than 0.01 ppm of copper. The
experiments were stopped at 400-bed
volumes, so undoubtedly larger
volumes of waters could have been
treated before unacceptable levels of
copper appeared in the effluents.
After 400-bed volumes had
been passed through the AlgaSORB®
column, the bound copper was, within
experimental error, completely
stripped from the column by the
passage of 0.5M H2SO4 through the
219
-------
column. Again, as with the cadmium
stripping, the copper was almost
completely stripped within the first
few bed volumes.
fi, Removal of Cadmium
from Waters at a Superfund Site.
Officials from U.S. EPA Region II
arranged to supply samples from a
well at a Superfund site in New
Jersey, the Waldick Aerospace
Devices site. These waters
contained, among other things,
cadmium at a level of 0.13 mg/L. The
waters, at a pH of 6.0-7.1, also
contained 0.66 mg/L of a halogenated
hydrocarbon, tetrachloroethylene as
well as other organics. Organics, of
course, are well known to interfere
with the function of traditional ion
exchange resins.
A column containing
AlgaSORB® (0.7 cm ID x 13 cm high)
was prepared, and the Waldick
Aerospace waters were passed
through the column. Five rnilliliter
fractions of water exiting the column
were collected until 500 ml (100-
bed volumes) of Waldick waters were
passed through the column at a flow
rate of one-sixth of a bed volume per
minute (total bed volume was 5.0
mL). Each fraction of effluent was
analyzed for cadmium using graphite
furnace atomic absorption
spectrometry. All effluent fractions
showed that cadmium concentration
was near or below 0.001 mg/L
through the passage of the 100-bed
volumes of the cadmium-containing
solution. Because the experiment
was stopped after the passage of
100-bed volumes through the column,
it is not possible to state explicitly
what volume of solution could be
treated before cadmium breakthrough
occurred. However, experience has
shown that if a test material is
capable of treating at least 100-bed
volumes, it will be economically
feasible to use the material. The
essential point is that the
AlgaSORB® removed cadmium well
below those levels which are allowed
in drinking water. The current
drinking water levels for cadmium
stand at 0.005 mg/L.
Once 100-bed volumes of the
cadmium-containing solution had
passed through the AlgaSORB®-
containing column, cadmium was
stripped from the column by passing
0.15M H2SO4 through the column.
Analysis of the column effluents
showed that nearly 90 percent of the
cadmium was stripped from the
column with the passage of two-bed
volumes of sulfuric acid through the
column. Most of the remainder of the
cadmium appeared in the next two-
bed volumes. Mass balance
calculations showed that, within
experimental error, all of the bound-
cadmium was stripped from the
column. Once the cadmium was
stripped from the column, the column
was ready for reuse after rinsing
with distilled water. Subsequent
passage of cadmium-containing water
through the column showed similar
cadmium binding properties.
C. Removal of Nickel and
Chromium from Contaminated
Groundwater. A sample of
groundwater that had been
contaminated with nickel and
chromium was obtained directly from
the electroplating business which
was responsible for the
220
-------
contamination. Presently the
electroplater is pumping and treating
these groundwaters with a
conventional (precipitation)
wastewater treatment system. The
initial pH of the groundwaters was
near pH 7. The chromium content
(essentially all hexavalent chromium)
was near 0.9 mg/L and nickel content
was at 2.7 mg/L. The customer's
discharge levels are 0.5 mg/L and
0.25 mg/L for nickel and chromium,
respectively.
Passage of these waters
through a column (6 mL total volume)
containing AlgaSORB® resulted in
effluents that were below 0.5 ppm in
nickel after elution of 175-bed
volumes. The nickel could be easily
stripped by passage of acid through
the column. However, chromium
broke through rather rapidly after the
passage of only five-bed volumes ©f
the metal-contaminated groundwater
through the column. This was
actually what had been anticipated
since other work has shown that
chromium(VI) is most strongly bound
to AlgaSORB® at pH 3.5 and is not
bound at pH values near 7. Thus, after
adjustment of the pH of another
portion of these waters to pH 3.0 and
passage through another AlgaSORB®
column, results showed that after
elution of 225-bed volumes of the
chromium-bearing waters through the
column, chromium content in the
effluents was near or below 0.3 mg/L.
Thus, these waters can be
successfully treated using two
AlgaSORB® columns, if the pH of the
effluent from the first column is
adjusted to pH 3 before passage
through the second column.
jQ, Removal of Mercury from
Contaminated Groundwaters. Bio-
recovery was provided with water
samples from a mercury-
contaminated groundwater site. The
site had been contaminated with
mercury years ago through the
process used to manufacture chlorine
from seawater. The groundwaters
contained 2-3 ppm of mercury (both
inorganic and organic mercury), had a
total dissolved solid content of 7,200
mg/L and contained over 900 mg/L of
calcium and magnesium. Passage of
these mercury-containing waters
through an AlgaSORB® column (0.7 cm
ID x 13 cm high) resulted in
effluents which ranged in mercury
content below 0.006 mg/L as
determined by analysis using cold
vapor generation and atomic
absorption spectrometry. The
customer requires effluents of below
0.01 mg/l for discharge.
These experiments show, as
had earlier experiments, that
AlgaSORB® is effective in removing
both inorganic and organic mercury
from aqueous solutions even in the
presence of very high calcium,
magnesium and salt concentrations.
CONCLUSION
AlgaSORB® is a biological ion
exchange resin which has proved to
be particularly effective in removing
heavy metals from various types of
wastewaters.
In the presence of high total
dissolved solids AlgaSORB® is
particularly effective in reducing
heavy metal ion concentrations down
to the parts per billion level.
221
-------
Underground aquifers, leachates, mine
drainages and often industrial
wastewaters contain high
concentrations of total dissolved
solids. These total dissolved solids
make it difficult and costly if not
impossible to recover heavy metals
by any method other than the
AlgaSORB® technique.
Darnall, D.W., B. Greene, J.M.
Hosea, R.A. McPherson, M.T.
Henzl and M.D. Alexander,
"Recovery of heavy metals by
immobilized algae", in Trace
Metal Removal from Aqueous
So I uti o n . Edited by R.
Thompson, The Royai Society of
Chemistry, London, 1986, pp. 1-
24.
Darnall, D.W., Greene, B., Henzl,
M.T., Hosea, J.M., McPhsrson,
R.A., Sneddon, J., and Alexander,
M.D., "Selective Recovery of
Gold and Other Metal Ions from
an Algal Biomass", Environ. Sci.
Technol.. 20. 206 (1986).
Greene, B., Darnall, D.W.,
Alexander, M.D., Henzl, M.T.,
Hosea, J.M., and McPherson,
R.A., "The Interaction of Gold(l)
and Gold(lll) Complexes with
Algal Biomass", Environ. Sci.
Technol.. 20. 627 (1986).
Greene, Benjamin, Robert
McPherson and Dennis W.
Darnall, "Algal sorbents for
selective metal ion recovery,"
Metal Speciation. Separation
and Recovery. Lewis
Publishers, Chelsea, Ml (1987)
pp 315-332.
5. Hosea, J.M., Greene, B.,
McPherson, R.A., Henzl, M.T.,
Alexander, M.D., and Darnall,
D.W., "Accumulation of
elemental gold on the alga
Chlorella vulgaris", Inorganica
Chimica Acta. Bioinorganic
Chemistry. 123. 161 (1986).
6. Watkins, J.W., Elder, R.C.,
Greene, B. and Darnall, D.W.,
"Determination of gold binding
in an algal biomass using
EXAFS and XANES
spectroscopies", Inorganic
Chemistry. 26. 1147 (1987)
Figure Legends
Figure 1. The Reaction of Divalent
and Trivalent Metal Ions with
an Algal Cell. The metal ions
in an aqueous solution are
rapidly complexed by the
biopolymers in an algal cell
wall.
Figure 2. Removal and Recovery of
Copper from an Ammoniacal
Etch Waste Stream in a Printed
Circuit Board Manufacturing
Facility. The waste solution
containing copper as the
ammonia complex was passed
through a column containing 0.5
g of AlgaSORB®. After the
column became saturated with
copper as evidenced by the
presence of copper in the
effluent (near 75 mL) sulfuric
acid was passed through the
column to elute the bound
copper.
222
-------
223
-------
o
0
0
5800
5700
5600
5500
VN
E
Q.
Q.
200
100
ppm Cu
incoming waste
elute with
0.50M H2SO4
ol ~ ~ ~ _
25
50 75
mis eluted
100
125
224
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
225
-------
HAZARDOUS WASTE MANAGEMENT IN RESEARCH LABORATORIES
George Sundstrom
Safety and Health Policy Staff
Agricultural Research Service
Hyattsville, MD 20782
ABSTRACT
Hazardous waste management in research laboratories benefits from a
fundamentally different approach to the hazardous waste determination from
industry's. This paper introduces new, statute-based criteria for
identifying hazardous wastes and links them to a forward-looking compliance
system. Geared toward the unique structural and operational characteristics
of laboratories, the overall system integrates hazardous waste management
activities with other environmental and hazard communication initiatives. It
is generalizable to other waste generators, including industry.
After the brief, labor-intensive phase necessary to implement this system
on a site-specific basis, conservative hazardous waste management decisions
become routine, and new liabilities for improper disposal practices are
controllable. Although only the waste identification and classification
aspects of the system are outlined in detail here, four other components are
defined or supported, namely:
-routine and contingency practices
-waste treatment/disposal option definition and selection
-waste minimization, recycling, reuse, and substitution opportunities
-key interfaces with other systems, including pollution prevention
All the components superficially resemble those already in place for
industry, but use of the statutory definition of hazardous waste and toxicity
information from material safety data sheets (MSDSs) as the basis for
classifying waste hazard has far-reaching implications.
INTRODUCTION
A cornerstone of hazardous waste
management in the United States is
the "cradle-to-grave" concept. Once
a material is identified as a
hazardous waste (i.e., enters the
cradle), it must be managed and
tracked for its entire life, to its
final disposition (i.e., to its
grave). The hazardous waste
determination (1) is the process for
identifying hazardous waste, and the
manifest (2) is the mechanism for
offsite tracking. Facilities are
authorized to treat, store, or
dispose specified hazardous wastes,
provided they meet certain
regulatory and/or design criteria.
Since enactment of the Resource
Conservation and Recovery Act (3)
226
-------
(RCRA) in 1976 and the RCRA
regulations (4) in 1980, the
regulatory programs for hazardous
waste have changed in significant
ways. They were extended to
generators of smaller quantities of
hazardous waste. Uniform manifest
forms were introduced and modified.
Facility standards were made more
stringent. Consistent with RCRA's
liability provisions and reflected
in the landfill ing bans, even the
operational definition of the
regulatory endpoint has changed:
the period of care is now generally
recognized to extend until the
hazardous waste is no longer
hazardous (i.e., is treated by
incineration, neutralization, etc.)
and/or is no longer a waste (i.e.,
is recycled/reused).
While evolution of the RCRA
regulatory program has been dramatic
at times, its maturation in the area
of identification of hazardous waste
has lagged noticeably. Some changes
in the hazardous waste lists and
characteristics have occurred, but
RCRA's regulatory (5) and statutory
(6) definitions still differ
greatly. This discrepancy might be
less important if only RCRA's listed
and characteristic items were in
fact hazardous as wastes, if the
Superfund (7) list of hazardous
substances (8) were not nearly twice
as long as the RCRA lists, if
liability and costs for cleaning up
contamination were not so high, and
if negative public reaction to even
the perception of improper disposal
were not so prevalent.
It is predictable that the
hazardous waste regulations will
eventually be changed to better
reflect the statutory definition.
Some movement to expand the "cradle"
is already evident in areas such as
infectious wastes, radiological
mixed wastes, and waste oil.
Research laboratories are repre-
sentative of the type of organiza-
tion that will be most affected by
changes in the way in which hazard-
ous wastes are defined. Their
variable and sporadic wastestreams
do not easily fit the model of the
hazardous waste regulations. On the
other hand, they are particularly
fertile ground for application of a
hazardous waste determination system
that focuses first on hazardousness,
then on wastes. Such a system is
described herein.
PURPOSE
The principal purpose of this
paper is to introduce, describe, and
illustrate an advanced hazardous
waste determination methodology
applicable to the research labora-
tory but generalizable to other
types of hazardous waste genera-
tors. The methodology is based on
the statutory definition of hazard-
ous waste and can be meshed with
other management, compliance, and
regulatory initiatives, including
hazard communication, community
right-to-know, waste minimization,
and pollution prevention.
In re-defining the parameters of
the hazardous waste determination as
shown in Figure I, a decision on a
material's "hazardousness" precedes
(and drives) its waste classifica-
tion, which in turn determines how
it is to be evaluated then managed
as a waste. Applying this method-
ology, the prudent manager is able
to make and document science-based
operational and waste-disposal
decisions that will reduce liability
exposure, now and in the future.'
APPROACH
A four-step approach is utilized
to develop and then implement this
model hazardous waste
227
-------
FIGURE I
DECISION TREE FOR THE MODIFIED HAZARDOUS WASTE DETERMINATION
MATERIAL RECEIVED OR IN STOCK
IS IT
HAZARDOUS
BY HCS?
DOES ITS
USE PRODUCE
WASTE?
COMPLY WITH
USE & PRODUCT
REGULATIONS
DOES ITS
USE PRODUCE
WASTE?
ASSUME WASTE HAZARDOUS
IS IT
HAZARDOUS BY
CRA/STATE
MANAGE BY RCRA/
STATE STANDARDS
IS IT
HAZARDOUS
BY HCS?
USE PROPER DISPOSAL,
TREATMENT, RECYCLING
OPTIONS/METHODS
MANAGE CONSISTENT WITH
RCRA/STATE STANDARDS
FOR HAZARDOUS WASTE
Identification/management system.
First, hazard classes are defined,
and objective criteria are used to
classify the various substances
found in the laboratory. Second,
the boundary between substance and
waste is rigorously established, and
the same criteria are applied to the
wastes resulting from the use of
these substances. Third, criteria
228
-------
and options for treatment and
disposal of wastes are identified.
Finally, site-specific procedures
are implemented, monitored, and
evaluated.
Assigning hazard classes. This
step is the inventory and hazard
assessment portions of the
Occupational Safety and Health
Administration Hazard Communication
Standard (9) (HCS). The laboratory
is inventoried for all chemical and
biological agents. If possible, the
inventory is computerized, and the
database is designed to accept
information on acquisition and
disposal. Searchable fields in the
database are used to designate each
substance as hazardous or not, its
hazard type, management practices
required, disposal options,
quantities used, and so forth.
The principal sources of data
for these assessments are MSDSs and
the references listed in Appendix C
of the HCS. Appendices A and B of
the HCS, summarized in Table I, are
the criteria for the classification.
identifying and classifying
wastes. This step entails the
process presented in Figure I. The
underlying principle in this model
is conservative and departs sig-
nificantly from the RCRA approach
and regulations: until specifically
found to be nonhazardous, a waste
that is - or contains - a hazardous
chemical or biological agent (as
these terms are defined by the HCS)
is assumed to be a solid waste and
is a hazardous waste.
A substance/material is not a
waste - and so cannot be either a
hazardous or nonhazardous waste -
until it is no longer needed (e.g.,
will no longer be used), is to be
discarded or disposed (e.g., to be
gotten rid of, in fact or in
effect), is no longer useful for its
intended purpose (e.g., doesn't meet
specifications), is spent (e.g., a
contaminated degreaser or extraction
solvent), is inherently waste-like
(e.g., dirty vacuum pump oil),
and/or is to be abandoned (e.g.,
removed from active, managed
inventory), recycled (e.g.,
re-distilled or sent to a
recycler/reclaimer), or used in a
manner constituting disposal (e.g.,
used motor oil used for dust
suppression). When one or more of
these conditions is met in fact (not
just by generator declaration), a
waste has been generated.
The RCRA statute provides the
qualitative starting point for
re-defining the hazardous waste
determination: a hazardous waste is
a solid waste or combination of
solid wastes, which because of
quantity, concentration, or
physical, chemical, or infectious
characteristics, may cause or
contribute to increased mortality,
serious chronic illness,
incapacitating acute illness, or a
substantial present or potential
hazard to human health or the
environment if improperly treated,
stored, transported, disposed, or
otherwise managed. Many more
substances than appear on the RCRA
lists are hazardous by this
definition.
The best available quantitative
measure of RCRA's qualitative
criteria (at least as far as human
health effects are concerned) are
the same Appendices A and B of the
HCS summarized in Table I.
Solutions, mixtures, and other forms
in which a hazardous agent are used
are also considered hazardous until
explicitly found otherwise.
If the waste is derived from, or
contains, a hazardous substance or
229
-------
TABLE I
CRITERIA FOR IDENTIFYING PHYSICAL AND HEALTH HAZARDS
Criterion
Carcinogen
Corrosive (bio)
Highly Toxic
Irritant
Sensitlzer
Toxic
Target Organ
Effects
Physical hazard
Representative Range
Proven or potential carcino-
gen, mutagen, or teratogen
Destroys or irreversibly
changes living tissue
LD50 < 50 mg/kg,
LDso < 200 mg/kg,
LCso < 200 ppm,
LCso < 2 mg/1
Reversible inflammatory effect
on skin or eyes
Allergic reaction possible in
exposed individuals
50 < LDso < 500 mg/kg,
200 < LDso < 1000 mg/l
-------
appropriate, the waste must be added
Into the lab's hazardous waste
generation figures and be managed
and tracked as a hazardous waste,
whether it is treated or disposed
onsite, shipped offsite, or
otherwise managed. If the state of
nonhazardousness was reached after
the substance became a waste, the
process of rendering it nonhazardous
is considered treatment. Depending
upon the quantity and type of
hazardous waste generated at the
facility and other factors, a RCRA
or other permit for the treatment of
hazardous waste may be required.
Identification/selection of
waste treatment/disposal
alternatives. Feasible and
site-specific preferred alternatives
are identified, as are regulatory
constraints/requirements and the
level of risk/liability exposure
management is willing to assume in
the short- and long-term. Chemical
supplier catalogs and Prudent
Practices (10) can be valuable
resources for identifying potential
treatment/disposal options that may
be used in many cases. However, it
must be kept in mind that State
regulations, local sewer-use rules,
and other constraints may foreclose
some of these possibilities.
Management should issue specific
guidance on identifying and select-
ing waste treatment/disposal options
that are sensitive to the natural
and sociopolitical environment.
Program implementation. Once
the hazardous waste identification
system is developed, it must next be
implemented, integrated with the
rest of the lab's hazardous waste
management and other systems, and
periodically evaluated. This proc-
ess is made easier if management and
individual(s) responsible for the
day-to-day execution of the neces-
sary work realize that this program
has fundamentally different goals
from the facility's safety and in-
dustrial hygiene initiatives: it is
aimed at protecting the public and
environment outside the workplace.
PROBLEMS ENCOUNTERED
The impediments to implementing
this system lie in 5 distinct areas:
-the nature of research
operations,
-the availability of resources,
-existing stocks of materials,
-letting wastes be wastes, and
-application of hazard
information.
Nature of research operations.
The work carried out in a research
laboratory typically involves a
number of relatively autonomous
units, possibly with differing
institutional affiliations, working
in relative isolation without much
in the way of central administrative
controls. Composition of the limited
volumes of waste, which are general-
ly produced only sporadically, can
vary widely, depending on research
objectives and direction. Most
chemicals are used predominantly in
relatively dilute solutions.
Availability of resources. In
recent years, negative growth in
fiscal and manpower resources has
been the norm. Administrative and
other support services have been
squeezed, and the laboratory manager
and researcher are both under
pressure to minimize non-research
costs and administrative workload.
Typically, full hazardous waste man-
agement responsibility is given to
the lab's marginally-trained col-
lateral-duty safety officer. Under
these circumstances, even current
RCRA hazardous waste determination
requirements are not always properly
met. Also, little help is available
231
-------
from Immature and understaffed
regulatory agencies still trying to
catch up with hazardous waste
generators who are not in the "gray
areas" of RCRA.
Existing stocks of materials. A
decentralized inventory of thousands
of commercial and experimental com-
pounds is commonly present, many in
poor condition and in dead storage
under no one's control. Labels may
be missing and records of contents
lost or dispersed as a result of
retirements or organizational
changes. Where inventory records do
exist, they are typically on file
cards and have not been incorporated
into the laboratories' hazard
communication program, so MSDSs and
hazard labels are not present.
Unfortunately, there is no good
substitute for the labor-intensive
physical inventory of the lab's
chemical and biological stocks.
Letting wastes be wastes.
Budget uncertainties, plan changes,
personality traits, and other
factors make discarding anything
difficult.
Application of hazard data.
Even if hazard data are available,
most researchers are not trained in
chemical and biological hazard
identification and are cognitively
unaware of any waste problems.
Taken together, these 5 problem
areas tempt management into one of
two extreme positions. The first can
be characterized as inertia and
neglect, the second as overkill. In
the former, little is done before
(or even after) a crisis arrives in
the form of an inspector from a
regulatory agency or the discovery
of environmental contamination by
others. In the second approach, the
"quick fix" is grabbed, whatever the
cost and whether or not it fits the
lab's needs or situation.
RESULTS
This hazardous waste identifica-
tion system takes its place beside
other management systems (Table II)
to ensure that conservative,
informed decisions are made with
respect to laboratory chemicals and
the wastes resulting from their
use. Application of the same hazard
criteria to both materials and
wastes yields a defensible
consistency, a smooth transition at
an important regulatory boundary,
and a way to avoid legally but
inadvertently producing new
Superfund sites.
The system simultaneously
advances the developing "pollution
prevention" ethic, which is rooted
in the economics of cleaning up
environmental contamination. It
provides real incentives for meeting
the waste minimization objectives of
RCRA. It mandates a conscious
process at the transition point
between materials and wastes.
By implementing this last link
in the chain of materials and waste
management and integrating it into
other administrative systems,
facility managers are able to assure
themselves and the public that
responsible and scientifically-
defensible decisions are being made
and documented.
REFERENCES
1. 40 Code of Federal Regulations,
Part 262.11 (40 CFR 262.11).
2. Appendix, 40 CFR 262
3. Resource Conservation and
Recovery Act of 1976, as
amended. 42 United States
Code, Sections 6901 et. seq.
(42 USC 6901 et. seq.)
4. 40 CFR 260-280
23:
-------
MATERIALS
System
Hazard communication
Hazardous materials
management
Inventory management
Emergency planning/
Community right-to-know
Experimental design
Waste identification
Hazardous waste
identification
Waste minimization
Hazardous waste
management
Recordkeeping
Emergency/conti ngency
planning
Pollution prevention
TABLE II
AND WASTE MANAGEMENT SYSTEMS
Objectives
Inform employees of chemical hazards in the
workplace
Minimize personal injury and property damage
from incidents involving hazardous materials
Cost and overhead control
Inform public and emergency responders of
hazards present
Logistical planning from concept to cleanup
Housekeeping and space utilization
Ensure proper handling, treatment, and
disposal of hazardous wastes
Cost and overhead control
Ensure proper handling, treatment, and
disposal of hazardous wastes
Documentation of activities
Ensure employees know what to do in an
emergency
Source reduction and recycling
5. 40 CFR 261.3
6. 42 USC 6903
7. Comprehensive Environmental
Response, Compensation, and
Liability Act of 1980, as
amended (CERCLA). 42 USC 9601
et. seq.
8. 40 CFR 302
9. 29 CFR 1910.1200, Hazard
Communication
10. Prudent Practices for Disposal
of Chemicals from Laboratories,
National Research Council,
National Academy Press, 1983
Disclaimer
The work described in this paper was
not funded or supported by the
Agricultural Research Service or the
United States Department of
Agriculture. The contents do not
necessarily reflect the views or
policy of ARS or USDA, and no
official endorsement should be
inferred.
233
-------
WASTE REDUCTION — WHAT IS IT? HOW TO DO IT?
Roger L. Price, P.E.
Center For Hazardous Materials Research
University of Pittsburgh Applied Research Center
Pittsburgh, PA 15238
ABSTRACT
Numerous case studies indicate that the sound management of resources
results in simultaneous economic and ecological benefits. If less waste
is produced, there is less potential for damage to the environment. Con-
sequently, waste reduction is sound economically as well as ecologically.
It's simply good business. However, beyond simply understanding these
technical approaches to waste reduction, it is widely recognized by the
experts that waste reduction is first and foremost an attitude.
The old attitude that environmental compliance costs money (and will
be considered only when absolutely necessary) must give way to a new atti-
.tude that waste reduction is a sound investment which makes good business
sense. Businesses need to see that they can benefit economically by com-
plying with environmental regulations through waste reduction. Environ-
mental protection must become an integral part of the day-to-day business
decision-making process.
Since its inception, the Center for Hazardous Materials Research
(CHMR), through its technical assistance program, has been assisting busi-
nesses to identify opportunities for waste reduction and to implement
waste reduction programs. Through this technical assistance, essential
elements for successful waste reduction were identified, which include:
management initiatives; waste audits; improved housekeeping; substitute
materials; redesigning equipment; recycling and reuse; and waste exchanges.
This paper goes beyond simply explaining each of the elements neces-
sary to achieve waste reduction. Examples of waste reduction successes
are^provided and personal experiences related which emphasize the role of
attitude changes in implementing any successful waste reduction program.
INTRODUCTION
Approaches to waste minimiza-
tion are primarily low-cost, low-
risk alternatives to hazardous
waste disposal. Most of the ap-
proaches do not require a great-
deal of sophisticated technology
234
-------
and can be relatively inexpensive.
In short, waste minimization
approaches are: technically fea-
sible, economically viable, and
ecologically beneficial.
In general, any waste minimi-
zation program should include or
consider: (1) management initia-
tives, (2) waste audits, (3) im-
proved housekeeping, (4) substi-
tute materials, (5) redesigned
equipment, (6) recycling and re-
use, and (7) waste exchange.
This paper will introduce
these various elements of waste
minimization. Case studies are
selected which emphasize the role
of attitude changes in implement-
ing any successful waste reduction
program.
However, many waste genera-
tors are still unaware of waste
reduction opportunities or don't
know how to go about developing
their own waste reduction program.
As a result, they have done little
or nothing to reduce their wastes.
Much work is still needed in
order to help individuals in busi-
ness understand the technical ap-
proaches to waste reduction and
identify waste reduction opportun-
ities in their operations. How-
ever, beyond simply understanding
the technical approaches, waste
reduction is first and foremost a
problem of attitude. Every suc-
cessful waste reduction program
begins with an attitude change
regarding how we view the genera-
tion and management of waste.
PURPOSE
Over the past two decades,
there has developed an increased
awareness of the harmful effects
to human health and the environ-
ment from uncontrolled releases of
pollutants and hazardous substanc-
es. Initially, this led to a na-
tional waste management strategy
which emphasized the control and
cleanup of pollution by hazardous
substances after they are gener-
ated and no longer serve a produc-
tive function. Now the nation is
turning its attention to prevent-
ing hazardous waste problems by
cutting down on the generation of
hazardous waste at its source.
Many individuals and facili-
ties have already identified and
implemented waste reduction oppor-
tunities, although data suggest
that current waste generation can
be further economically reduced
nationwide with existing technol-
ogy by about 50%.'
APPROACH
Research was conducted to
identify the key elements of a
successful waste reduction pro-
gram. These key elements were
reviewed and further developed in
the conduct of a statewide waste
reduction technical assistance
program which included on-site
audits, training seminars, and
other interactions with individu-
als, in industry and small busi-
ness, who were concerned about
reducing the wastes they generate.
Numerous case studies on suc-
cessful waste reduction efforts
were collected and reviewed. Sev-
eral case studies were selected
and summarized which emphasize the
role of attitude changes.
PROBLEMS ENCOUNTERED
Centralized sources of infor-
mation on waste reduction were
235
-------
lacking, and problems were encoun-
tered in collecting information on
waste reduction technologies.
There is insufficient dissemina-
tion of information regarding re-
duction methods or successful ap-
proaches to developing and imple-
menting facility specific waste
reduction programs.
Because of the need to sort
through the vast amount of old and
new information becoming availa-
ble, technical assistance is need-
ed by all generators to fully un-
derstand options, costs, and bene-
fits of various waste reduction
strategies. Technical assistance
programs are needed which can also
serve as clearinghouses for tech-
nical and relevant regulatory in-
formati on.
RESULTS
Numerous case studies 2-7
exist which indicate that the
sound management of resources
results in simultaneous economic
and ecological benefits regardless
of the size of an organization.
These case studies show that: (1)
waste reductions can range from
20% to 98%; (2) payback periods
for waste minimization investments
typically range from immediate to
5 years; and (3) firms which
handle fewer hazardous materials
reduce hazards to their workers
and the environment—and experi-
ence fewer longterm liability and
victim compensation claims.
The Role of Attitudes
However, review of these case
studies, combined with experiences
in performing waste reduction au-
dits, training programs, and other
interactions with individuals in
the business community, clearly in-
dicates that attitudinal change is
at the heart of expanding voluntary
waste reduction practices.
The old attitude that envi-
ronmental compliance costs money
must give way to a new attitude
that waste reduction can be a
sound investment and makes good
business sense. Businesses must
see that they can benefit econom-
ically by complying with environ-
mental regulations through waste
reduction.
The old attitude that environ-
mental protection is best achieved
through "end of pipe treatment"
must give way to a new a.ttitude
that environmental protection is
best achieved by reducing waste
generation at the source.
The old attitude that waste
is inevitable must give way to a
new attitude that waste reduction
is a dynamic opportunity contin-
gent on a host of changing techni-
cal, economic, and institutional
factors. Substantially more waste
reduction is currently feasible
and more will become feasible in
the future.
Basic attitude changes take
time, and waste reduction cannot
be achieved overnight. However,
the process of changing attitudes
must begin immediately, and envi-
ronmental protection through waste
reduction must become an on-going,
integral part of the day-to-day
business decision-making process.
The remainder of this paper
sets forth essential elements of a
waste reduction program, which
when combined with basic attitu-
dinal changes, will result in suc-
cessful reduction of waste genera-
tion.
236
-------
Developing Management Initiatives
The commitment to waste mini-
mization must come from the top—
the management of a business or
organization. Management initia-
tives are vital to the success of
any waste minimization efforts.
The following two management
actions are crucial to a success-
ful waste minimization program.
(1) Communication—Management must
make all employees aware of the
waste minimization effort. (2)
Incentives—Management should pro-
vide incentives for the develop-
ment of useful waste minimization
ideas just as incentives are used
to boost employee productivity.
Although a waste minimization
commitment should begin with man-
agement, the employees are often
able to suggest improvements in
the day-to-day operations of the
business. The new management ini-
tiatives should foster the follow-
ing elements of waste minimization
success: increased awareness and
attention to hazardous chemicals;
motivation to change old work pat-
terns; knowledge of options for
change; and willingness to inno-
vate and change. Another impor-
tant management tool is employee
training.
Haste Audits
The waste audit is the most
basic of all of the approaches to
waste minimization. The waste
audit tracks hazardous waste by
monitoring all of the waste which
is produced at a place of business
to learn where it was generated.
One can determine where hazardous
materials are used and where mate-
rials are being wasted. As a re-
sult, areas of a business that
produce waste may be discovered
which had not been recognized be-
fore the audit.
The waste audit can be divid-
ed into the following six steps:
(1) identifying hazardous substan-
ces in waste or emissions; (2)
identifying the sources of these
substances; (3) setting priorities
for various waste reduction ac-
tions to be taken; (4) analyzing
some technically and economically
feasible approaches to waste mini-
mization; (5) making an economic
comparison of waste minimization
and waste management options; and
(6) evaluating the results.
The waste reduction audit is
a systematic and periodic survey
of a company's operations and is
designed to identify areas of po-
tential waste reduction.
Improving Housekeeping
Improved housekeeping, or
"good operating practice," is the
simplest waste minimization prac-
tice. Improved housekeeping re-
lies on using common sense and is
often the most effective first
step toward waste reduction.
Good housekeeping practices
involve the procedural or organi-
zational aspects of a manufactur-
ing process and include elements
such as: inventory control, waste
stream segregation, material han-
dling improvements, spill and leak
prevention, improved scheduling,
and preventive maintenance.
One relatively simple house-
keeping method is waste segrega-
tion. In many cases, segregation
of wastes allows for . certain
wastes to be recycled or reused.
For example:
237
-------
o In a business using both chlor-
inated and nonchlorinated sol-
vents, these waste types should
be kept separate.
o At a printing company, waste
toluene from printing press
cleanup can be eliminated by
segregating this solvent ac-
cording to the color and type
of ink cleaned from the press.
Each segregated batch of tolu-
ene can be reused for thinning
the same color ink.
Improved labeling allows em-
ployees to know precisely what a
container or pipeline holds and
guards against accidental spills
and unnecessary use—both a waste
of materials. All substances used
in the workplace should be proper-
ly labeled. In additional, all
wastes, once segregated, should be
labeled as well. This procedure
helps to ensure safe handling of
wastes, and can point out contain-
ers of waste which have the poten-
tial for recycling, reuse, or even
resale.
Substituting Materials
A waste audit may identify
specific materials within a busi-
ness which are producing hazardous
waste. If this is the case, it
may be possible to find a substi-
tute material which is less haz-
ardous. Although material substi-
tution is only applicable in cer-
tain situations, it can prove to
be an efficient waste minimiza-
tion approach. For example:
o A painting business uses a hy-
drocarbon solvent (toluene) for
cleanup of hydrocarbon-based
paint. By switching to water-
based paint, water can be used
for toluene during cleanup.
o Hater-soluble cleaning agents
can often replace organic sol-
vents or degreasers. One com-
pany did this and successfully
reduced its 1,1,1-trichloro-
ethane use by 30X, resulting in
a $12,000 annual savings.
o ITT Telecom reduced the quan-
tity of waste solvents they
generate by merely replacing a
solvent-based, photo-resist
system with an aqueous-based
system. The new system reduces
hazardous waste generation and
also improves product quality
while reducing production time.
Technology Modifications
In many instances, technolog-
ical modifications or material
substitutions are also very effec-
tive in minimizing wastes. Some
products can be manufactured by
two or more distinct processes,
and one process may produce less
hazardous waste than the other.
Modifying equipment within a given
process is another way to reduce
waste generation.
Technological modifications
can be generally categorized as:
process modifications, equipment
modifications, process automation,
changes in operation settings,
water conservation, and energy
conservation.
Production processes may be
responsible for the production of
hazardous waste. Old or ineffici-
ent processes could be sources of
hazardous waste. By changing to a
newer, more efficient process, a
company could decrease the amount
of waste it generates. In
addition, many companies can ex-
perience improved production ca-
pacity and product quality and
realize savings in expenditures
238
-------
for utilities and raw materials.
For example:
o In printed circuit board manu-
facture, the use of screen
printing for image transfer
i nstead of photo1i thography
eliminates the use of develop-
ers.
o By replacing a solvent-based
painting system with a water-
based electrostatic immersion
painting system, the Emerson
Electric Company has reduced
waste solvent and paint solids
generation by over 951.
o A chemical company reduced phe-
nol resin waste by 95%. Old
tank cleaning procedure (fill-
ing with water which was then
discharged for treatment) was
replaced with a two-step pro-
cess where an initial small
volume rinse produces a concen-
trate which could be recycled
as raw material. Second rinse
produces waste stream with re-
duced phenol resin content.
Process modifications often
entail subsequent equipment modi-
fications. Equipment modifications
accomplish waste reduction by re-
ducing or eliminating equipment-
related inefficiency. An equip-
ment modification leaves the pro-
duction process intact and also
unchanged because it modifies only
the equipment which comprises the
process. For example:
o A simple dragout recovery sys-
tem was installed on a nickel
plating machine. Less than
$1,000 was invested for a
storage tank, which saved the
firm $4,200 worth of nickel per
year and reduced nickel sludge
generation by 9,500 pounds per
year.
Process automation involves
the use of automatic devices to
assist or replace human employees.
Automation can include monitoring
and subsequently adjusting process
parameters by computer or mechani-
cally handling hazardous substan-
ces. Waste minimization is accom-
plished by reducing the probabil-
ity of employee error (which can
lead to spills or off-spec prod-
ucts) and by increasing product
yields through the optimum use of
raw materials.
Often the generation of haz-
ardous waste may not be the fault
of the equipment. Instead, the
fault may lie in the way in which
equipment is set to operate. These
are often the most easy and inex-
pensive of equipment changes. For
example:
o Many spraying processes operat-
ing at decreased pressures have
less overspray and subsequently
less waste.
o In formulating their cyanide
copper plating baths, the Stan-
adyne Company determined that
lower chemical concentrations
can be used. By running the
potassium cyanide concentration
at 2.5 ounces per gallon, in-
stead of 3.5 ounces, the cyan-
ide dragout concentration was
reduced by 28% without any ad-
verse effect on plating qual-
ity.
Most equipment has optimum set-
tings at which it operates most
efficiently. By determining the
optimum settings for certain pa-
rameters (such as optimum tempera-
ture and pressure), less waste is
generated as a by-product.
Although not as significant as
other approaches, water conserva-
239
-------
tion can have an effect on mini-
mizing hazardous waste generation.
For example:
o By reducing the amount of water
used for washing some organic
cheralcal products, compani es
can lower the amount of waste
water which must be pretreated
before disposal.
Energy conservation minimizes the
waste associated with the treat-
ment of raw water, cooling water
blowdown, and boiler blowdown. In
addition, lower energy usage means
a reduction in the generation of
ash and other wastes associated
with combustion. Energy conserva-
tion can be accomplished through a
series of heat exchanges within
the production process.
Recycling and Reuse
Recycling and reuse of hazardous
wastes can be a very economical
undertaking. Many companies have
discovered that the cost of in-
stalling on-site recycling equip-
ment can be quickly recovered, and
future profits gained, by savings
in waste management and raw mate-
rial costs. For example:
o A pesticide manufacturer gener-
ated pesticide dust from two
major production systems. The
firm replaced the single bag-
house with two separate vacuum-
air-baghouse systems specific
to the two production lines for
$9,600. The collected materi-
al was recycled to the process
where it was generated. The
firm has eliminated over $9,000
annually in disposal cost, and
they estimate that the recover-
ed material is worth more than
$2,000 per year.
o The Rexham Corporation facility
in Greensboro, North Carolina,
installed a distillation unit
to reclaim n-propyl alcohol
from waste solvent for a total
installed cost of $16,000. The
distillation unit recovers B5Z
of the solvent in the waste
stream, resulting in a savings
of $15,000 per year in virgin
solvent costs, and in a $22,800
savings in hazardous waste dis-
posal costs.
In addition, there are many off-
site recyclers who will take a
company's waste, recycle it, and
sell the refined product back to
the company at a price signifi-
cantly less than the cost of vir-
gin material. Additionally, that
company will not have to entail
waste disposal costs.
o The Hamilton Beach Division of
Scovill, Inc., operation re-
quires the solvent 1,1,1-tri-
chloroethane to degrease metal
stampings. Ashland Chemical
Company was contracted to recy-
cle the waste by distilling
1,1,1-trichloroethane. Substi-
tuting the recycled solvent for
the virgin product has reduced
Hamilton Beach's overall raw
material costs by $5,320 per
year. Scovill also eliminated
all of their previous waste
disposal costs, estimated to be
about $3,000 per year.
Participating in Haste Exchanges
Waste exchanges are networks of
businesses which attempt to find
markets for the wastes they gener-
ate. Remember that hazardous waste
to one business can be a valuable
resource to another. The exchange
attempts to match one business
waste with another business raw
material requirements. Small busi-
240
-------
nesses can also find excellent
recycling opportunities through
such organizations. Often a "buy-
er" company is able to purchase,
recycle, and subsequently reuse
another's waste. In this way, the
buyer is able to save on raw mate-
rial costs, and the hazardous
waste generator is able to market
a new product as opposed to dis-
posing a hazardous by-product.
ACKNOWLEDGMENTS
This research was performed to
prepare a manual published by CHMR
entitled Hazardous Waste Minimiza-
tion Manual For Small Quantity
Generators in Pennsylvania, April
1987. The manual was prepared
under a grant from the U.S. Envi-
ronmental Protection Agency, Re-
gion III, RCRA Support Section of
the Waste Management Branch.
We also extend our appreciation to
those leaders of industry who pro-
vided information on their indus-
trial processes and on the ways
they have minimized their wastes.
REFERENCES
1. Joel S. Hirschhorn, Ph.D.,
Office of Technology Assess-
ment, U.S. Congress, paper
presented at Pennsylvania
Conference on Hazardous Waste
Minimization and Source
Reduction, Pittsburgh, PA, Nov.
16, 1987.
2. Center for Hazardous Materials
Research, April 1987, Hazardous
Waste Minimization Manual For
Small Quantity Generators in
Pennsylvania.
3. U.S. EPA, 1986a, Report to
Congress: Minimization of Haz-
ardous Waste. Volumes I and II.
EPA/530-SW-86-033A. Office of
Solid Waste, U.S. Environmen-
tal Protection Agency. Washing-
ton, DC (Available from NTIS:
PB87-114336 & PB87- 114344).
4. U.S. EPA, 1986b. Waste Minimi-
zation Issues and Options.
Volumes I. II. and III. EPA/
530-SW-86-041. Office of Solid
Waste, U.S. Environmental Pro-
tection Agency. Washington, DC
(Available from NTIS: PB87-
114351, PB87-114369 and PB87-
114377).
5. Campbell, Monica
liam M. Glenn,
from Pollution
E., and Wil-
1982, Profit
Prevention.
Pollution Probe Foundation, 12
Madison Avenue, Toronto, Ontar-
io, Canada MR5 2S1.
6. Huisingh, Donald, Larry Martin,
Helene Hilger, and Neil Seld-
man, 1985, Proven Profits from
Pollution Prevention. Insti-
tute for Local Self Reliance,
2425 18th Street, NW, Washing-
ton, DC 20009, ISBN 0-912582-
47-0.
7. U.S. Congress, Office of Tech-
nology Assessment, September
1986, Serious Reduction of Haz-
ardous Waste: For Pollution
Prevention and Industrial Effi-
ciency. OTA-ITE-317 (Washing-
ton, DC: U.S. Government Print-
ing Office).
8. U.S. EPA, 1988, Waste Minimiza-
tion Opportunity Assessment
Manual. EPA/625/7-88/003.
Alternative Technologies Divi-
sion, Hazardous Waste Engineer-
ing Research Laboratory, U.S.
Environmental Protection Agen-
cy, Cincinnati, OH 45268.
241
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
242
-------
PAINT REMOVAL STRATEGIES
EFFECTIVE IN REDUCING
WASTE VOLUMES AND RISKS
Thcmas F. Stanczyk
Recra Environmental, Inc.
Amherst, New York
ABSTRACT
As new paint and surface coat-
ing formulations are developed,
it becomes increasingly dif-
ficult for paint and
maintenance personnel to remove
surface coatings in preparation
of product finishing. A number
of new paint stripping tech-
nologies have emerged to keep
up with advances in the paint
technologies, as well as, the
issues dealing with environmen-
tal risks, workplace safety,
rising disposal costs, new
multi-media control standards
and increased productivity
demands. Improvements in
chemical resistivity are
directly affecting chemical
usage rates, and the pollutant
loadings comprising air, water
and solid wastes. The paint
stripping inefficiencies are
dictating new strategies geared
at waste minimization and
optimum performance without
jeopardizing the workplace and
the environment. The strategic
evaluations are resulting in
controversy in the aircraft
industry over the use of new
stripper formulations versus
dry paint removal technologies.
The strategic options are
weighed in the context of paint
removal performance ef-
ficiencies and waste
generation.
INTRODUCTION
The paint and surface coatings
industries are recognizing
record demands for shipments of
new products employed by
numerous industries requiring
applications of architectural
coatings, product coatings and
special-purpose coatings.
Research endeavors have
revolutionized the quality and
performance of surface coat-
ings. A number of the major
plastics and automotive
materials suppliers, including,
but not limited to: PPG, BASF,
Dow, DuPont, Mobay and ICI
continue to invest millions in
improvements to the con-
stituents comprising the
formulations in an effort to
reduce volatile organics caus-
ing environmental and safety
concerns while improving ex-
terior appearance, durability,
chemical resistivity and over-
all applications.
243
-------
A lumber of paint applications
for example the re-painting of
aircraft, utilize a blend of
surface coatings that are
inferior in terms of chemical
resistivity. The chemical
resistivity properties of many
of the thermosetting acrylics,
catalyzed epoxies and
polyurethanes are creating
problems for individuals
responsible for correcting poor
quality finishes, removing
surface coating formulations,
and preparing surface sub-
strates for repainting. With
the aircraft industry re-
painting is a critical step in
the overall maintenance of an
aircraft displaying poor ap-
pearance and/or excess coatings
influencing fuel consumption.
Many of the conventional paint
removal formulations can remove
the new coatings at the expense
of using excess labor and
chemicals while creating in-
creased loadings of hazardous
pollutants in air emissions,
waste waters, and solid
residues.
There are a number of new
chemical formulations, some of
which have eliminated the use
of chlorinated solvents, which
have recently emerged in
response to many of the paint
removal issues. Their perfor-
mance is dictated by the
chemistry of the coatings, the
standard operating protocol in
terms of application, and the
properties of the substrate.
Several alternatives have also
emerged which utilize dry
systems as a means of breaking
the bonds holding the paint to
the substrate. Other
strategies have focused on
mechanical and application
improvements which suppress
volatile organics as a means of
reducing exposure and pollution
control requirements. These
new formulations and paint
removal systems are critical to
the reduction of environmental
and OSHA concerns.
PAINT CONSTITUENTS INFLUENCING
WASTE GENERATION
The quality and performance of
the constituents comprising
paints continue to be a prime
focus of many industrial re-
search endeavors. The
resulting paint formulations
directly, as well as in-
directly, influence the
quantities and characteristics
of wastes generated by paint
removal operations.
Before a paint removal system
is selected and a waste manage-
ment strategy is developed, it
is important to identify the
constituents comprising ex-
terior coatings and the
likelihood that their
properties will influence the
hazards inherent to waste by-
products.
Many volatile solvents used in
commercial paint formulation
are being phased out and, in
sane cases, totally eliminated.
The solvents generally do not
effect the quality of paint
residues removed by chemical
stripping, since the major
portion are volatilized during
the painting cycle. Among the
technologies improving
durability and chemical resis-
tivity of surface coatings as
well as minimizing air emis-
sions is the process of
polymerizing polyester coatings
for powder coatings. The oven
cured powders significantly
reduce many of the environmen-
tal hazards inherent in paint
244
-------
by the fact that the processes
do not emit volatile organics
and they lend themselves to
easy maintenance eliminating
overspray disposal problems.
The powdered coatings, while
not totaly applicable at this
time to the aircraft industry,
may create additional problems
with paint removal as a result
of their improved resistivity
properties.
The paint chemistry will
provide valuable insight into
the most practical strategy for
paint removal. Many of the
pigments, inorganics as well as
organics, can contribute to the
pollutant loadings typically
found with paint residues and
wash solutions. The age of the
coating can influence the
properties of the residual
pigments in terms of their
solubility in water, as well as
leaching solutions. Many of
the new formulations of pigment
display low water solubilities
and their impact in terms of
leachability, is minimal in
comparison with some older
paint systems; for example,
coatings using red lead. Many
of the surface coatings are
formulated with resins that are
resistant to an array of
aqueous and organic-based
chemicals.
For the most part, the resins
have good to excellent resis-
tivity to acids, alkalies,
solvents and oxidizers. As
such paint formulations will
directly influence the selec-
tion of an appropriate
stripper.
Considering the variables
associated with the resistivity
of surface coatings, the fol-
lowing factors should be
considered in terms of optimiz-
ing stripper performance
including; the type of film
formed, the thickness of the
coatings, the type of surface
applied, the primer used, the
type of pigment, the chemical
stripper used, dwell time,
temperature and method of
stripper application and the
age of the coating.
CHEMICAL FACTORS INFLUENCING
STRIPPING PERFORMANCE
The performance of
coatings/paints not only
depends on the resistivity
properties, but also to a
greater extent on their adhe-
sion properties. The paint
stripping formation has to take
into account both the
capability of distracting the
integrity of paint films and
the ability of destroying the
adhesive forces between coat-
ings and substrate. The
formation of adhesion bonds is
due to not only the close range
atomic interaction at the
interface, but also to inter-
diffusion at the interface on
electrostatic interaction.
Various types of interfacial
forces including chemical bonds
such as: ionic, covalent, and
metallic bondings, and inter-
molecular forces such as
hydrogen bonds, dipole-dipole,
dispersion, dipole-induced
dipole have been identified.
If the interfacial properties
between coating and substrate
can be altered, the adhesive
forces between coating and
substrate may also be weakened.
The weakening of the adhesive
forces will enhance the ef-
ficiency of paint removal.
Paint stripping technologies
generally rely on physical
and/or chemical mechanisms.
The physical methods mechani-
cally destroy the bondings and
abrade the coating from the
245
-------
substrate in addition to
utilizing differences in physi-
cal properties between coating
film and substrate to break
various bondings. The chemical
methods utilize the properties
of the strippers, such as
solvency, oxidation and swell-
ing to chemically attack the
integrity of the coating film,
while indirectly affecting the
interfacial properties which
can break the adhesive bonds.
There are numerous manufac-
turers and suppliers of
chemical strippers, generally
categorized as: alkaline, acid,
and solvent-based. Alkaline
strippers will generally break
the ester linkages, however,
they are not able to remove
pigments. As such, various
additives, including solvents,
are added to improve perfor-
mance. Acid strippers rely on
chemical destruction, i.e.,
oxidation or dehydration, of
the paint constituents. Acids
will generally solubilize the
pigments, open the coatings and
remove the surface oxides,
resulting in a destruction of
the adhesive bond of coating to
substrate. Inherent corrosion
hazards must be considered
prior to using acidic formula-
tions. The solvent-types
generally work by dissolving
the paint, however, redeposi-
tion is a problem.
Ifethylene chloride-based strip-
pers are generally viewed as
the most effective strippers
for removing polyurethane and
epoxy types of paint without
damaging the surface. As
illustrated on Figure 1,
methylene chloride will
penetrate the coating, swelling
it to a volume as much as ten
times of the original film.
Penetration is made possible as
a result of methylene
chloride's small molecular size
as well as its polarity. The
swelling of the film builds up
pressure on all directions and
the pressure can be relieved
only in the direction directly
away from the substrate.
As noted in Figure 1, the
tensile strength is developed
because of difference of swell-
ing characteristics between
coating film and substrate. To
promote the rate of film swell-
ing, cold strippers can be
enhanced by the addition of
various acids, alkalis, amines,
and special solvents or agents.
Acids may break the adhesive
bond by dissolving surface
oxides. Wetting agents (or
penetrants) can increase the
stripping rate by allowing more
rapid penetration of the film.
The paint stripping begins most
rapidly at the edge or at flaw
areas of the surface, where the
penetration is easiest. The
wrinkling and blistering will
proceed from these areas until
the entire film is effected.
If there are no such areas, the
wrinkling or blistering can
only proceed by a slower
penetration of the stripper
through the surface of the
film. One major drawback of
the methylene chloride stripper
is the evaporation loss of
solvents which not only makes
the stripping operation ineffi-
cient, but also results in
environmental pollution
problems. Although the
evaporation loss can be
retarded by using evaporation
retarders such as water or wax
seals, this loss just cannot be
eliminated completely. Upon
application of the evaporation
retarders, the wax crystallizes
out to form a film on the
surface, retarding the evapora-
tion of solvents. The epoxy-
baaed paints typically have a
246
-------
strong resistivity to alkaline-
based methylene chloride
formulations. Epoxies, like
the alkali-resistant coatings,
often respond to the acidic
paint removal formulations
which are able to break the
ether linkage.
In terms of projecting waste
generation, the solvent-based
formulations can contain sig-
nificant concentrations of
other inorganic and organic
constituents serving as cosol-
vents, activators, thickeners,
penetrants and evaporation
retarders . Some of the
specific constituents compris-
ing these formulations are
methanol, toluene, sodium
chromate, ammonia, bentonite,
metallic soaps, polyacrylate
esters, cellulose acetate,
ethyl cellulose and various
waxes. The volatility and
water solubility of these
constituents will vary sig-
nificantly. For example, some
industries use water to remove
large surface areas treated
with stripper. Depending on
the quantity and characteris-
tics of the product rinse
solutions it is conceivable
that the volatile fraction and
phenol can be detected at
concentrations varying by an
order of magnitude. The pol-
lutant loadings will have a
major effect on treatment
requirements.
Besides methylene chloride
based formulations, a number of
chemical manufacturers are
developing for industrial use
non-halogenated formulations
which are proving attractive in
terms of reducing risks and
disposal problems. The physi-
cal and chemical methods are
compared in the next section.
ALTERNATIVE STRATEGIES FOR
PAINT REM3VAL
Figure 2 is a summarized logic
flow chart depicting some of
the existing and new strategies
for removing paint from
aircraft.
Plastic media blasting,
cryogenic stripping. C0*> blast-
ing and laser beam paint
removal are among the new
technologies which rely on
physical paint removal
mechanisms eliminating the
usage of chemicals posing
potential environmental hazards
and risks to human health.
Some of the unique features
associated with each waste
minimization strategy are
summarized as follows:
1. Cryogenic techniques use
extremely cold tempera-
tures to cause paints to
be brittle, allowing for
effective debonding by
non-abrasive plastic
media. The technology,
which is principally
supplied by Air Products
Co., has been applied to
an array of paints includ-
ing powdered coatings.
This option is generally
used to remove paints from
equipment which has a
build-up of coatings on
the surface. Among the
documented advantages are:
its efficiency and
amenability to processing
at high throughputs; its
elimination of hazardous
chemicals; and the low
likelihood of substrate
damage. There are some
disadvantages which re-
quire evaluation
including: the inef-
ficiencies in removing
urethane and epoxies;
247
-------
mechanical limitations
decline with large surface
areas and thin-film coat-
ings; and the costs. The
reaction in potential air
emissions concerns is
significant.
2. Plastic media blasting
has been receiving in-
creased attention in terms
of paint removal amidst
controversy over potential
damage to structural
integrity. For several
years, PMB has generated
performance data substan-
tiating its use for paint
removal. The technology
uses plastic medium par-
ticles, activated by
compressed air to physi-
cally abrade the paint off
the substrate. This
option has demonstrated
wide applicability to
substrates not sensitive
to structural damage. The
processes are apparently
cost competitive with
chemical stripping
methods, however, there
are arguments disclaiming
this benefit. The process
does eliminate the use of
chemicals, however, the
resulting solid wastes
still require proper
disposition. Standard
operating controls are
very important in terms of
preventing structural
damage.
3. The laser paint stripping
technique is still in the
developmental stage. The
technology, using high
energy photons from tn«
laser, causes bond scis-
sors in the coatings with
significant increase in
volume causing the coat-
ings to be blown away from
the substrate. The
elimination of chemicals,
excess labor and equipment
prepartion are among the
advantages to this option.
Substrate damage, control
development, excess costs
and the lack of commercial
units are among the disad-
vantages .
Source control strategies can
focus on the development of new
paint stripping formulations or
in-plant source segregation
strategies that optimize the
reduction of waste volumes as
well as pollutant loadings.
There are considerations as-
sociated with source control
strategies, including but not
limited to:
- Alternative, non-
halogenated, paint
stripping formulations
may be more costly,
dependent upon heat,
and may not be as
effective in removing
some of the combined
coatings, in particular
urethane epoxies.
- Alternative commercial
formulations of paint
strippers which vary in
chemical content,
specifically methylene
chloride, do exist;
however, each formula-
tion requires
evaluation in terms of
stripping performance
and retardants effec-
tive in attenuating air
emissions.
248
-------
- The potential for
mechanically suppress-
ing and/or containing
volatile organic emis-
sions, reducing toxic
loadings and the mag-
nitude of air flow
requiring abatement.
- Modifying standard
operating protocol in a
manner that reduces
dwell time, while
improving stripping
performance efficiency.
- Modifying material
handling practices in a
manner that calls for
the removal and collec-
tion of paint residuals
before the units are
subjected to water
rinse applications.
The segregation alternatives
can have a major impact on the
characteristics and pollutant
loadings of the resulting waste
by-products. The trade-offs
associated with these source
segregation alternatives need
to consider the potential
likelihood of water soluble
stripper constituents being
attenuated in wastewater treat-
ment sludges or residuals,
precluding landfills as a
viable disposal option.
Among the commercially avail-
able substitutes are stripper
formulations replacing
methylene chloride with either
N-methyl-2-pyrrolidone (M-
Pyrol) or dibasic esters.
M-Pyrol is a versatile solvent
that has found applicability as
a prime constituest in non-
halogenated paint stripping
formulations. M-Pyrol provides
a number of properties which
are viewed as advantageous in
terms of reducing hazards
including; its high flash
point, low vapor pressure (0,29
nun Hg) selective solvency,
chemical stability, chemical
stability and biodegradable
properties. Among the disad-
vantages warranting further
evaluation are; the costs of
using the formulations
(generally higher in comparison
to conventional formulations),
slower penetration rates and
the potential requirement for
heat.
Dibasic esters is used as an
active ingredient in paint
formulations and industrial
cleaners. The product does not
require chlorinated solvents
and it has a flash point of
>100 C, a low vapor pressure,
low hazard potential and ideal
raise ibility properties with
other solvents. Among the
disadvantages are: its high
molecular weight; which hinders
penetration rates, and the fact
that it is not a universal
solvent for all hydrocarbons.
There are also a number of
commercially available strip-
pers, each offering variable
chemical formulations. The
variability in feedstock
characteristics will have a
direct bearing on the volumes
and characteristics of waste
by-products.
In reviewing the chemical
options it is important to keep
in mind the operations entailed
with paint removal. If a water
spray or rinse is employed for
this purpose there is a good
likelihood of detecting many of
the inorganic and organic
stripper constituents in the
resulting wastewater. The
concentration of methylene
chloride and phenol in the
249
-------
changes in hazardous waste
definition and treatment per-
formance, the practice of
conbining the resultant solid
and liquid wastes could cul-
minate in treatment residues
which are characteristically
more hazardous and difficult to
manage than the wastestreams
which were originally gener-
ated.
The paint stripping operations
relying on volatile solvents
are faced with a major portion
of their waste being released
as air emissions. Future
standards are expected to
regulate air emissions from
paint stripping operations,
most likely dictating air
abatement standards. The
pollutant loadings in air
emissions can be significantly
influenced by changes in
feedstock forumations, mode of
stripper application, dwell
time, paint removal practices
and mechanical, as well as
chemical suppressants. These
control strategies may reduce
loadings and the degree of
control, but they do not
eliminate risks and
liabilities.
There are a number of new
chemical formulations emerging
in this market. New standards
in the work place will magnify
industry's search for sub-
stitutes as long as
productivity and product
quality are not adversely
impacted. The performance
record of new (non-halogenated)
chemical formulations will
enhance the need for changes in
feedstocks since regulatory and
competitive edge factors remain
critical factors.
resulting rinsewaters could
range between 1000 - 5000 ppm
to 50 - 500 ppm respectively.
Segregation practices and
chemical usage rates can have a
direct effect on the resulting
pollutant loadings. In addi-
tion the resulting reduction in
loadings achieved with source
segregation can significantly
reduce capital needs for ex-
ample with air stripping
systems as well as with was-
tewater treatment systems. As
an illustrative example of how
the wastewater pollutants can
influence compliance status
Figure 3 provides a summarized
profile of paint stripping
rinse water characteristics.
The constituents are cross-
referenced with pertinent
regulatory listings (existing
and future) as a mechanism for
prioritizing risks and the
requirements for source sub-
stitution and/or segregation.
Malti-media transfer problems
associated with the rinse
process as well as the treat-
ment by-products of the
wastewater can be prevented or
minimized by instituting an
awareness program addressing
feedstock contents and
materials transfer.
SUMMARY
Data substantiates the fact
that the chemical content of
paint stripping feedstocks are
the primary source of con-
taminants in the multi-media
wastes resulting from opera-
tions dictating the removal of
various types of paint and
coatings. Without segregation,
typical standard operating
protocol can magnify environ-
mental concerns by increased
waste volumes and pollutant
loadings. With regulatory
250
-------
Alternate "dry" stripping
technologies are proving very
promising in that they
eliminate the reliance on
chemical formulations, thus
significantly reducing result-
ing waste volumes and
characteristics posing concerns
dealing with hazard and
mobility. Cryogenics and
plastic media are among the
technological advancements
warranting further evaluation.
Each strategy has distinct
advantages and, to a certain
extent, disadvantages which
require close examination from
an economic, regulatory and
performance perspective.
One of the prime factors that
will be quantified as part of
this assessment deals with the
reduction of pollutant loadings
and their associated risks and
liabilities. As each strategy
is weighed and prioritized, the
chemical content stands out as
one of the prime decision
factors over time.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
FIGURE 1
TENSILE
STRENGTH
BUBBLIN8
SUBSTRATE
PRIMER
BONDIM6 LAYER
HETHYLENE
CHLORIDE
SWELLING
OF THE
COATING
FILM
SWELL I NO
COATINe FILM
251
-------
FIGURE 2
SUMMARY OF OPTIONAL TECHNICAL STRATEGIES
FOR REMOVING PAINT
EXTERIOR SURFACES
REQUIRING
PAINT REMOVAL
DRY SYSTEMS
• PLASTIC MEDIA
• SAND BLASTING
• WALNUT SHELL BLASTING
• STEEL SHOT & GRIT
BLASTING
• C02 BLASTING
• CRYOGENICS
• HEAT TREATMENT
• LASER STRIPPING
WET SYSTEMS
• HALOGENATED
- methylene chloride Oased
- methyiene cniortdc/
phenol based
• NON-HALOGENATED
- m-pyrol
- DBE formulations
• WET/AQUEOUS BASED
- alkali wash
- baking soda blasting
- high pressure water
blasting
FI6URE 3
A PROFILE OF WASTEWATER PROPERTIES AND
APPLICABLE REGULATORY LISTINGS
PRIMARY
CONTAMINANT
TOLUENE
METHYLENE CHLORIDE
METHAHOL
ETHANOL
AMMONIA
CHROMIUM
HEXAVALENT CHROMIUM
(AS SODIUM OICHROMATE)
LEAD
ZINC
MOBILITY
WATER
0
1
4
4
3
NA
3
•
I
AIR
22mm
330mm
97mm
40 mm
>l itm
O
o
0
o
TOJtICITY
u
•
•
•
•
6
e
o
o
o
0
u
•
•
•
o
•
•
f)
•
o
CARCINOGEN
O
Protwtolt
O
O
O
O
O
Suaptctttf
O
OSHA
pa
(TWAI
200 MM
500 MM
200 MM
1.000 MM
SO MM
0.5 mf/m3
O.t mf/m3
S0uf/m3
0
APPLICABLE REGULATORY
LISTING GOVERNING
CONSTITUENT MANAGEMENT
M
3
.j
TCLP LISTER
FOOI-FOOS
WW/OTHER SO
A
B
C
O
O
o
o
o
o
M
u
TCLP PROPOSI
CHARACTERISE
D
E
O
O
O
F
O
F
O
EP TOXIC
O
o
o
o
o
F
o
F
o
M
f
•
•
O
o
o
•
o
•
•
&
3
• in
I
ii
•
•
•
o
•
•
0
•
•
a
I
O
•
o
o
o
o
o
o
o
0 - Insoluble
I - Slightly Svluklt
2 - Modtrittly Soluble
3 • Vtry Solicit
4 - Mljclble
O - Appllciblt
O - Not Appllc»tlH
A- I.l2/0.33m«/l
8 - 02/0 96 mfl/l
C • 0.2S/0.75 mf/l
0- 14.4 mf/(
E - B.B mfl/l
F - 5.0 mg/l
252
-------
TREATMENT AND RECOVERY OF HEAVY METALS FROM INCINERATOR ASHES
I.A. Legiec, C.A. Hayes, and D.S. Kosson
Rutgers, the State University of New Jersey
Department of Chemical and Biochemical Engineering
Piscataway, New Jersey, 08855-0909
ABSTRACT
High levels of potentially leachable toxic metals such as Pb, Cd, and
Cr, in the ash residues require consideration of treatment technologies
prior to disposal or utilization of the ashes in other applications.
Extraction studies revealed information on the leaching characteristics of
the ashes and on the ability of the extraction solutions to separate these
metals from the ash matrix. The kinetics of the Pb, Cd, and Cr extraction
were determined through a series of batch extraction studies. The specific
incinerator design and the equilibrium conditions strongly influenced the
ash matrix and the extraction kinetics. Recovery of the metals from the
waste extract solution was accomplished utilizing electrochemical plating
techniques. A laboratory scale pilot plant was designed to continuously
treat ash residues at a rate of 1 kg/hr through extraction and electro-
chemical recovery techniques. Initial pilot plant operational study results
are presented.
INTRODUCTION
Due to decreasing available
landfill space, the minimization of
the amount of solid materials requi-
ring disposal is imperative. The
incineration of municipal solid
waste (MSW) results in a 90% reduct-
ion in volume and 75% reduction in
mass, as compared to direct MSW
disposal [1]. However, heavy metal
concentrations extracted from MSW
ashes were found to exceed the
Extraction Procedure Toxicity Jest
(EP Tox) limits [2]. The EP Tox
test is the procedure established by
the United States Environmental
Protection Agency (USEPA) in testing
for leachable inorganic species
from solid wastes. The leachability
of heavy metals from MSW ash
residues must be considered prior to
disposal or utilization of these
materials.
Extraction studies employing
ash obtained from several resource
recovery facilities provided insight
to the leaching characteristics of
the ashes and the dependency on the
extractant solution composition
[3,4]. The anion and cationic
effect of the extractant salt solut-
ions were investigated, as well as
253
-------
the pH of the system. A 1 N NaCl
solution (aqueous), acidified with
HC1, was found to be the most effec-
tive in heavy metal extraction. The
kinetics of the leaching process
were defined and the electrochemical
recovery of the metals in solution
was investigated [5]. These studies
led to the design and operational
parameters of a laboratory scale
pilot plant, incorporating contin-
uous extraction, separation, and
electrochemical processes. The
results of the initial pilot plant
operation experiments are presented
in this paper.
Incinerator Designs
Two incinerators were sampled
for ash residues, one located in
Canada and one in Massachusetts.
The Canadian incinerator includes a
three tiered vibrating grate primary
combustion chamber and a secondary
combustion chamber. Solids are fed
to the combustion chamber using a
hopper and have about a three hour
residence period. Sixty percent of
the combustion air is provided as
underfire air, the remainder of the
air is introduced above the burner
beds. A full waterwall is provided
in the incinerator. A shell and
tube heat exchanger cools down the
exit gases and fly ash, and separat-
ion of the gases and fly ash occurs
in an electrostatic precipitator.
The Massachusetts incinerator
utilizes a baffled, three tiered
combustion chamber in which the
solids are periodically moved tier
to tier via a hydraulic ram. The
solid residue is quenched in the ash
pit and then removed, while the
gases and fly ash pass through a
secondary combustion chamber. The
ash is separated from the combustion
gases utilizing a charged gravel bed
and a bag house. Detailed descript-
ions of both incinerators are
available [4].
Kinetics Studies
A series of batch extraction
studies investigated the ability of
the extractant solution to separate
and remove the metals from the ash
matrix [5], Various ash residues
were investigated, the Canadian fly
ash (CF3) and the Massachusetts fly
ash (MF) results will be presented.
The extractant solution utilized was
1.0 N NaCl (aqueous), acidified with
HC1 to achieve an equilibrium pH of
3.0 in the ash-extract slurry. This
solution was found to be the most
efficient in earlier experimentation
[4].
Twenty four HOPE bottles con-
taining 10.0 g ash and 200.0 ml
extractant were placed on a rotary
shaker and removed at eight
different time intervals (in tri-
plicate), up to a steady state
period. These intervals were as
follows: 10, 20, 30, 45, 90, 180,
300 and 720 minutes. Separation of
the solid and liquid phases was
carried out through vacuum filtra-
tion. The resulting waste extract
was analyzed for heavy metals (Pb,
Cd, Cr) via atomic absorption spec-
troscopy, pH, and conductivity.
The kinetics of the leaching
process for CF3 and MF ash was inve-
stigated through observation of the
time dependence of the measured
variables. The CF3 extract pH
increased from an initial extractant
pH of 0.41 to a steady state equili-
brium pH of 2.83, while the MF
extract pH increased from 1.92 to
2.91 (Figure 1).
The lead extraction curve for
CF3 peaked at a reaction time of 10
minutes, and subsequently decreased
(Figure 2). Concentration levels in
the extract at this time period were
254
-------
id
P.
3.5 -
3 -
2.5 -
2 -
0.9
200 400
Time (mln)
D CF3 + MF
600
800
Figure 1 - Batch Kinetics Studies, pH Response in Extract
• n
* -a
6
M I
fl<
Pi
BOO
D CP3 + MF
Figure 2 - Batch Kinetics Studies, Pb Response in Extract
255
-------
69.1 mg Pb/L, or 1380 ug Pb/g ash.
The lead extraction process for CF3
incinerator ash was observed to be
dependent on the pH of the system,
based on comparison of the lead and
pH curves for the ash. The rate of
lead removal occurred at a much
faster rate than the neutralization
process for the ash. The MF lead
extraction followed more of a
diffusion controlled mechanism, with
maximum values of 367 mg Pb/L, or
7350 ug Pb/g ash at 720 minutes.
Electrochemical Recovery
Electrochemical methods offered
the advantage of the recovery of a
relatively pure metal in a usable
form; whereas conventional precipit-
ation methods generate a flocculant
which requires disposal. Electro-
deposition of metals does not
require the addition of any extra-
neous compounds to an already
complex solution.
A parallel-plate electro-
chemical cell was designed for lead
recovery. Lead was used as the
cathode in this cell. Cyclic volta-
mmetry was carried out using this
design on standards of known metal
concentration and compared to the
aqueous extracts to determine the
operational parameters of the lead
recovery process. The current
response observed in the cell
utilizing a 180 mg Pb/L standard
solution indicated a plating
potential of -0.6 V for lead in this
solution. Subsequent cyclic runs
performed on CF3 extract at a pH of
3.00 and a lead concentration of 175
rag/L exhibited a peak at this value,
indicating that lead may be plated
out of the extract solution.
In order to determine the
amount of metal recovered, timed
studies were performed on the lead
standard in a CF3 extractant solu-
tion. Lead concentration was
measured as a function of time, as
was the current (Figures 3,4). Both
the current and the lead concentra-
tion showed the same response.
This is indicative of the fact that
the amount of lead recovered was
controlled by the current in the
cell and reached a lower limiting
concentration of 110 mg Pb/L.
PILOT PLANT DESIGN AND OPERATION
A laboratory-scale pilot plant
was designed to continuously treat
MSW incinerator ash residues; refer
to Figure 5 for a simplified process
flow diagram. The ash is fed to the
continuous stirred tank reactor
(CSTR) via a feed hopper/screw pump
system at a rate of 1 kg/hr.
Extractant solution is fed to to the
CSTR at a rate of 20 kg/hr; this
flow rate maintains the 1 gram to 20
ml extraction ratio. A pH
controller system is utilized to
maintain an setpoint pH within the
CSTR; this optimum pH value was
previously determined from the
results of the kinetics studies [5].
The volume and pH maintained within
the CSTR are the controlling para-
meters for the residence time of the
solid-liquid reaction. The liquid
separation operates continuously
with recovery of both phases,
utilizing a 20 to 25 micron
cellulose filter. The treated solid
ash residues are recovered and
stored for further experimentation.
The ash extract is pumped through a
5 micron basket filter for further
filtration, passed through the
electrochemical cell and then
recycled. Electrochemical plating
of the lead in solution allows for
recovery of the metal in a
relatively pure form; this process
occurs in a multi-plated cell
equipped with polymer coated
titanium anodes (commercial product)
256
-------
s
n
u
K
a
u
100 -
130 -
140 -
130 -
100
TlMK (MINUTES)
Figure 3 - CF3 Extract Batch Plating, Pb Response
UJ
Cli
_« 1-
1 2@O
Figure 4 -
Til IE '.'.SEC>
CF3 Extract Batch Plating, Current Response
257
-------
s
1
ASH DELIVERY SYSTEM
RECYCLE
EXTRACTANT
I;XIRH;IANT
TANK
MAKEUP
TANK
SCRBI PUMP
CSTR
pll CONTROLLER I N IIC1
(aqueous)
fUKOE
CAi. RECOVER
CfLL
T
SOLID - LIQUID SEPARATION
WASTE EXTRACT
STREAM
FILTER
KEY:
SAMPLE POINT
Figure 5 - Ash Treatment Pilot Plant Process Flow Diagram
and lend cathodes. The majority of
the process equipment in contact
with the solid or liquid phases are
constructed from HOPE or PVC to
avoid interference with the metals
analyses as well as avoid corrosion.
All pilot plant processes were oper-
ated at room temperature.
Extraction Process
Initial investigations were
carried out to test the extraction
portion of the pilot plant,
excluding the recycle stream. The
CF3 incinerator ash residues were
utilized in these experiments. The
pH controller setpoint was set at a
pH of 1.59 to control the extraction
reaction at the optimum level. This
optimal level occurs at the peak of
lead extraction from the ash (refer
to Figures 1 and 2 for pH and lead
concentration curves). The pilot
plant was operated for 2 hours and
waste extract samples were obtained
prior to the electrochemical por-
tion. Subsequent atomic absorption
analyses of the waste extract stream
revealed high levels of extractable
lead in solution, matching the conc-
entration levels at the optimum
conditions in batch studies (Table
1).
Subsamples of the treated CF3
ash were obtained and assayed for
total metal content. One gram aqui-
lots were digested with 20 ml each
of nitric and perchloric acids [6].
Solids were removed using vacuum
filtration, and the filtrate was
quantitatively diluted to 100 ml and
analyzed for metal species; results
258
-------
Table 1. CF3 Continuous Extraction
Metal
Pb
Cd
Cr
Extract Cone.
rmz/L)
43.6
69.2
54.8
158.8
96.2
2.
2.
2.
16.
10.
< 1
< 1
< 1
< 1
< 1
Time
( minutes)
0
30
45
60
65
0
30
45
60
65
0
30
45
60
65
are presented in Table 2 and
compared to untreated CF3 ash
assays. The total lead content
decreased from 5750 ug/g ash to 3070
ug/g ash, resulting in a 47% removal
of lead from the ash matrix. These
results exhibited a 90% decrease for
cadmium. Other metals species were
not affected by the extraction
process; variances observed in the
data were due to the heterogeneous
nature of the ash solids.
Initial pilot plant experiment-
ation also utilized MF ash to
observe the continuous extraction
process. The pH controller setpoint
was set at a pH of 2.65 to control
the extraction reaction at 45
minutes (Figures 1 and 2). The
continuous extraction process was
carried out for 2 hours, and the
extract contained an average of 168
mg Pb/L, correlating to the results
Table 2 .
Metal
Cd
Cr
Cu
Ni
Pb
CF3 Total
(UE/E ash)
Before
Extraction
210
470
520
74
5750
Metal Assay
After
Extraction
20
470
610
110
3070
from the kinetics studies. The
total metal content of the treated
MF ash was compared to the untreated
ash residues, refer to Table 3. The
total lead content decreased from
9150 ug/g ash to 4460 ug/g ash,
resulting in a 51% removal of lead.
The total cadmium in the MF ash
matrix decreased by 81%, and the
total zinc content decreased by 56%.
Table 3.
MF Total Metal
(ug/g ash)
Assay
Metal
Cd
Cr
Cu
Ni
Pb
Zn
Before
Extraction
260
350
740
190
9150
18300
After
Extraction
49
170
680
170
4460
7980
Electrochemical Process
Batch lead plating experiments
were carried out with the pilot
plant electrochemical cell and
utilized an ash extract solution
spiked with a lead standard to a
known concentration. The waste
259
-------
extract was obtained through extrac-
"tions using a fly ash obtained from
another incinerator, Massachusetts
fly ash (MF). The MF extract was
spiked with lead standard solutions
to a concentration of 243.0 mg Pb/L.
The cell voltage was set at 0.6 V,
and after five minutes the concen-
tration decreased to 197.0 mg Pb/L,
indicating that lead plating is
achievable. Further experimentation
will investigate the continuous
plating process. Higher lead reduc-
tions are expected due to an
increased plating time (the cell
residence time is 12 minutes) and
increased lead concentrations in the
extract resulting from the recycle
stream.
CONCLUSIONS
The residual ashes from MSW
incineration contain high levels of
hazardous metals, and leachable
concentrations of these materials
often surpass the EP Tox test limits
as mandated by the USEPA. The
classification of these materials
may be considered hazardous,
warranting costly landfilling. This
led to the investigation of the
extraction and recovery of Pb, Cd,
Cr from the residual ashes. A
series of kinetics experiments were
carried out, developing the charact-
eristic extraction trends of the
various ashes. The extraction
kinetics of CF3 ash exhibited a peak
in the lead concentration curve at
an unsteady state time period of 10
minutes, indicating that the
ash/extractant reaction should be
controlled at these conditions in
order to remove a maximum quantity
of lead. Electrochemical experimen-
tation exhibited the potential for
metals recovery in a relatively pure
form through plating techniques.
Based on the results of these
investigations a pilot plant was
designed and constructed. Initial
experimentation indicated that the
extraction process was capable of
controlling the extraction reaction
at the optimum setpoint. Total
metals analyses performed on the
treated and untreated CF3 ashes
reveal a 47% reduction for lead and
a 90% reduction for cadmium. The
total lead content in the MF ash
matrix was reduced by 51%, while
total cadmium and zinc content was
reduced by 81% and 56%, respect-
ively. The electrochemical cell
plated lead out of solution at a
voltage of 0.6 V, and higher lead
reductions are expected from
increased plating time (the cell
residence time is 12 minutes) and
increased lead concentrations in the
waste extract during recycle opera-
tion.
Future experimentation will
include the treatment of various
other ash residues as well as
process optimization to maximize
lead and cadmium recovery, including
extract recycle. Treated ash
residues will be analyzed for leach-
able metals by EP Tox and TCLP. The
design and economic analysis of a
full scale ash treatment plant will
be investigated, also.
ACKNOWLEDGMENT
The work described in this
paper was not funded by the U.S.
Environmental Protection Agency and
therefore the contents do not
necessarily reflect the views of the
Agency and no official endorsement
should be inferred. This work was
funded in part by the New Jersey
Hazardous Substance Management
Research Center, Project INC1N-13.
260
-------
REFERENCES
1. Hinchey, M.D., and Bruno, J.L.,
"Where Will the Garbage Go?",
1988 update report from the New
York State Legislative Commiss-
ion on Solid Waste Management.
2. Clapp, T.L., Kosson, D.S., and
Ahlert, R.C., "Leaching Charac-
teristics of Residual Ashes
from the Incineration of
Municipal Solid Waste",
Proceedings Second Internat-
ional Conference on New Fron-
tiers for Hazardous Waste
Management. EPA/600/9-87/018F,
pp 1-8.
3. Clapp, T.L., Magee II, J.F.,
Ahlert, R.C., and Kosson, D.S.,
"Municipal Solid Waste Composi-
tion and the Behavior of Metals
in Incinerator Ashes", Environ-
mental Progress, in press.
4. Ontiveros, J.L., "A Comparison
of the Composition and Proper-
ties of Municipal Solid Waste
Incinerator Ashes Based on
Incinerator Configuration and
Operation", Doctoral Disser-
tation in the Department of
Chemical and Biochemical
Engineering, Rutgers, the State
University of New Jersey, May
1988.
5. Legiec, I.A., Hayes, C.A., and
Kosson, D.S., "Continuous
Recovery of Heavy Metals from
Incinerator Ashes", 10th
Canadian Waste Management Con-
ference. Winnipeg, Manitoba,
Canada, October 1988.
6. Methods of Soil Analysis. Part
2, 1982, 2nd ed., ASA, Inc.,
SSSA, Inc., Madison, Wisconsin,
7-8.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
261
-------
GOVERNMENT-PROVIDED TECHNICAL ASSISTANCE FOR
HAZARDOUS WASTE MINIMIZATION
by
Robert Ludwig, Jim Potter, David Hartley, Kim Wilhelm
California Department of Health Services
Sacramento, CA 95814
and
Lisa Brown
United States Environmental Protection Agency
Cincinnati, OH 45268
ABSTRACT
Waste minimization, particularly source reduction and recycling, provides
generators the mechanisms to reduce their hazardous waste generation, reduce
their waste management costs, and reduce long-term liability. Nonetheless, many
businesses remain unaware of the opportunities for waste minimization.
Governmental technical assistance programs have proven to be cost effective
methods for encouraging industry to minimize the generation of hazardous waste.
The California Department of Health Services (DHS) and local environmental health
programs have assisted generators in measurable reductions of hazardous wastes.
The most notable is the Ventura County program that achieved a 70% reduction
county-wide in the volume of land disposed hazardous waste over a two-year
period.
The Department also works closely with EPA through a three-year cooperative
agreement to provide waste minimization technical assistance to and funding for
generators. This cooperative effort encouraging waste minimization is continuing
through the Waste Reduction Innovative Technology Evaluation (WRITE) Program.
Federal, state, and local programs operate in unison. The federal and state
programs offer technical materials and financial support to local programs which
act as the principal generator contacts. These efforts demonstrate that
technical assistance programs can be successful within a regulatory program.
The Department and EPA act as an information and technology clearinghouse The
activities include: publishing "user friendly" waste minimization manuals and
fact sheets for generators to evaluate waste minimization options; sponsoring
waste minimization training seminars for industry and local government
regulators; operating a waste exchange; offering technology development grants
to hazardous waste generators; and maintaining a directory of financial resources
for waste minimization projects.
262
-------
INTRODUCTION
Hazardous Waste Minimization -- What
is it?
Hazardous waste minimization
reduces or eliminates the use of
hazardous materials and the
generation of hazardous wastes
resulting from the manufacture,
processing, and use of materials.
The overall goal is to eliminate
disposal of untreated hazardous
wastes to air, land or water. Waste
minimization strategies in order of
preference include: (1) source
reduction: elimination or reduction
of waste at the source, usually
within a manufacturing process; (2)
recycling: the use or reuse of a
waste material as an effective
substitute for a commercial product
or as an ingredient or feedstock in
an industrial process; and (3)
treatment: a process which
eliminates or reduces the hazardous
nature of the waste.
Hazardous Waste Minimization - - Why
do it?
The regulatory element driving
hazardous waste generators toward
minimization is the federal and state
land disposal restrictions, or Land
Disposal Bans. California's Land
Disposal Restriction Program has been
phasing out the disposal of specific
hazardous wastes since 1983. The
1984 Hazardous and Solid Waste
Amendments (HSWA) of RCRA prohibit
the continued land disposal of about
400 chemicals and hazardous wastes by
May 8, 1990. Although treatment
technology research and development
efforts reduced the volume and
toxicity of hazardous wastes,
treatment alone will not solve all of
the problems associated with the
generation of wastes.
Hazardous waste generators
should consider implementing waste
minimization, such as source
reduction or recycling, to reduce the
generation of hazardous wastes and
costs related to the storage,
transportation, treatment, disposal,
permitting, and taxation. Businesses
that invest staff time and money to
incorporate waste minimization
techniques will most likely decrease
the generation of hazardous wastes,
improve the working environment, and
be in regulatory compliance.
Hazardous Waste Minimization
Information - - Who has it?
In California DHS is the lead
agency for the coordination of waste
minimization efforts and
implementation of RCRA, CERCLA, and
other state-mandated requirements.
However, it is not the only state
agency responsible for environmental
management. Two other State agencies
responsible for hazardous wastes and
pollution control are the State
Water/Regional Water Quality Control
Boards and the Air Resources Control
Board. These media specific agencies
are currently working independently;
very little coordinated effort is
being made for multi-media hazardous
waste reduction.
Similarly, local entities exist
as counterparts to the State
regulatory agencies. Many county
environmental health departments
actually conduct the RCRA inspections
according to Memoranda of
Understanding (MOUs) from DHS.
Regional Water Quality Control Boards
regulate direct industrial discharges
and oversee Publicly Owned Treatment
Works (POTWs). POTWs, however, have
their own powers and are city-owned
or are special districts with their
own governing bodies. Under the Air
Resources Control Board, local air
263
-------
pollution control districts have
jurisdiction over air basins which
may cross county boundaries. These
local entities are autonomous and are
in charge of the local regulation of
stationary air pollution sources.
Most of the state and local
agencies pursuing waste minimization
are currently doing so apart from
other programs regulating a
different environmental medium.
There is a serious need to address
the fragmentation and lack of
communication that characterizes
these local efforts through
strategies such as formal MOUs,
informal agreements, integrated
inspections, and other cooperative
agreements.
G. PURPOSE
The purpose of this paper is to
provide information on federal,
state, county, and city sponsored
waste minimization programs in
California. Projects include (1)
EPA's Waste Reduction Innovative
Technology Evaluation (WRITE)
Program, (2) California's Hazardous
Waste Reduction Program, and (3) San
Diego and Ventura County's Waste
Minimization Programs, (4) City of
Los Angeles' Waste Minimization
Training Program, (5) California
Conference of Directors of
Environmental Health, and (6) the
Local Government Commission's
Guidelines for Hazardous Waste
Minimization. All of the following
projects received complete or partial
funding from the State of California
or the U.S. EPA.
D. APPROACH and RESULTS
Waste Reduction Innovative Technology
Evaluation (WRITE) Program
The EPA's Waste Minimization
Branch and DHS have introduced the
WRITE Program to promote preferred
waste minimization options for multi-
media pollution prevention. This
program will evaluate technical and
economical aspects of innovative and
operational technologies that reduce
the volume and/or toxicity of
hazardous wastes via source reduction
and recycling. The results of the
individual evaluations will be
consolidated into a state-of-the-art
review and be distributed at the
state and national level.
The EPA is also sponsoring a
national symposium to present
industry's waste reduction results.
This symposium, scheduled for June,
1990, will provide a national forum
to discuss and promote waste
reduction, encourage the use of
successful innovative technologies,
and acknowledge the participating
companies and their efforts to reduce
the generation of hazardous wastes.
California's Hazardous Waste
Reduction Program
The Hazardous Waste Reduction Program
is managed by the DHS' Toxic
Substances Control Division's
Alternative Technology Section (ATS).
Specific units within ATS provide the
basis for implementing waste
reduction and recycling strategies in
California. The Waste Reduction Unit
manages the state's Hazardous Waste
Reduction Grants Program and the
Waste Reduction Audit Program. These
two programs emphasize technology and
information transfer to the hazardous
waste generators and the waste
management community.
264
-------
The Resource Recovery Unit
focuses on promoting recycling and
resource recovery and operates the
California Waste Exchange, which
publishes the "Directory of
Industrial Recyclers" and the "News-
letter/Catalog." The Technology
Clearinghouse Unit acts as a focal
point for the dissemination of
current information on the successes
of source reduction and the
performance of recycling and
treatment technologies. The unit
sponsors technology transfer
activities such as seminars and
classes, as well as, writing and
distributing guidelines and fact
sheets. The unit's goal is to
translate technical literature and
information into publications and
presentations directed toward
specific industries, focusing on
small and medium sized businesses.
In February, 1989, a State
Hazardous Waste Reduction Award
Program acknowledging those companies
with operational innovative hazardous
waste reduction technologies was
started. Based on criteria related
to type of industry and waste,
technology, status of development,
and application to other industries,
the best of these will be publicly
recognized state-wide and a report
prepared for interested parties.
This program is being done in tandem
with EPA's WRITE Program. Those
technologies meeting the criteria of
the WRITE Program were further
evaluated. The results will be
available in a report and presented
at EPA's WRITE national symposium.
Ventura County Hazardous Waste
Minimization Program
In 1987, Ventura County's
Environmental Health Department
established a hazardous waste
minimization technical assistance
program to aid industry in reducing
dependency on land disposal. The
program identified the hazardous
wastes generated in the county in
order to prioritize waste reduction
efforts, provide waste reduction
consulting services for generators,
and assist companies in the
development and implementation of
waste minimization programs which
included source reduction, recycling,
and on-site treatment.
The waste minimization program
was established and strongly
supported by the County Board of
Supervisors. A "model plan",
entitled "Hazardous Waste Reduction
Guidelines For Environmental Health
Programs", demonstrated that a local
government can directly promote waste
reduction without hindering
established hazardous waste program
regulations. The guidelines offer a
variety of waste minimization
components that can be developed in
conjunction with new and expanded
hazardous materials program. One of
the more notable results is a 70 %
volume reduction of hazardous waste
being land disposed after 2 years.
To further encourage industry to
use waste reduction techniques, a
second report, "Ventura County
Environmental Health Hazardous Waste
Minimization Program Results and Case
Studies" was completed. This report
summarizes the methodology used by
the Program and describes what 55
companies did to reduce their waste
generation.
San Diego County's "Promote Landfill
Alternatives Now" (PLAN')
San Diego County's Department of
Health Services developed a
comprehensive hazardous waste
minimization program entitled
"Promote Landfill Alternatives Now!"
(PLAN). Various strategies were
developed to promote waste
265
-------
minimization and assist businesses
with implementing modifications. The
PLAN also served as a focal point for
the community to voice concerns
pertaining to hazardous waste
management, thereby enhancing the
level of communication between
government agencies, industry, and
the public.
Results of this program are the
development of industry-specific
questionnaires, workshops, and
booklets which summarize waste
reduction techniques. Another was to
train county staff how to conduct
industrial waste minimization audits
and how to incorporate the audits as
part of compliance inspections.
Third, a computer file was designed
to compile data from waste
minimization audits, local waste
information, and waste stream-
specific pollution control
alternatives. Lastly, a graduate
work-study program with a university
engineering department emphasizing
waste minimization was established to
assist industries.
Citv of Los Angeles Wasta
Minimization Program
The City of Los Angeles has
established a Hazardous and Toxic
Materials Project to ensure that City
Departments and industries promote
and practice hazardous waste
minimization. The program promotes
source reduction and recycling of
hazardous wastes to reduce discharges
to air, water, and land. One
potential result of this program is
the reduction of heavy metals and
other toxic wastes being discharged
to the City sewer system, as well as
the disposal in off-site landfills.
Waste minimization technical
assistance is also being provided to
industries through the City's Bureau
of Sanitation. This was accomplished
by first offering training to city
industrial waste inspectors and
sanitary engineers, as well as
industry personnel. The training
increases the level of awareness of
waste minimization and hazardous
waste requirements so that
information can be transferred to
industry personnel during routine
visits. Practical methods are also
being developed for inspectors and
industry to determine if a waste
minimization program was in fact
reducing the overall generation of
hazardous wastes without transferring
undesirable components of the
wastestream to a different medium.
These waste minimization methods
are being evaluated at 10 to 20
electroplating companies using
cadmium, nickel, lead or chromium in
their process. Technical assistance
is being provided by City staff and
an engineering consultant to
generators to determine the efficacy
of the methods and the extent of
waste minimization. A final report
detailing the results is expected in
May, 1990.
Local Government Commission
The Local Government Commission of
Sacramento (a non-profit, non-
partisan membership organization of
elected and appointed officials) has
produced three guidebooks designed to
help local governments establish
waste minimization programs.
Low Cost Ways to Promote
Hazardous Waste Minimization:
A Resource Guide for Local
Governments.
This guidebook describes why and
how local governments can set up
educational, technical assistance,
and regulatory outreach programs for
hazardous waste minimization. A
complete resource listing for 28 low
266
-------
cost activities are provided in
detail. The guidebook also provides
examples of successful local
government programs. A model
resolution for establishing an
educational program to assist
businesses in the cities and counties
is also included.
Reducing Industrial Toxic Wastes
and Discharges: The Role of POTWs.
This guidebook explains the
importance of Publicly Owned
Treatment Works (POTWs) and the
opportunities for promoting hazardous
waste minimization. POTWs can help
area firms significantly reduce their
toxic discharges to the sewer,
without transferring those same
pollutants to other media, by
developing educational, technical
assistance, and regulatory programs.
A model POTW resolution for
developing a hazardous waste
minimization program, examples of
established programs, and appendices
with useful information are included.
Minimizing Hazardous Wastes:
Regulatory Options for Local
Governments.
This guidebook describes the
program and regulatory framework that
can be used to promote multi-media,
multi-agency hazardous waste
minimization at the local level in
California. It explores and
identifies the role of direct
requirements, indirect regulatory
inducements, and positive incentives
for waste minimization. A model
resolution for developing a hazardous
waste minimization program at the
local level and appendices of related
information are included.
California Council of Directors of
Environmental Health
The needs and concerns of local
environmental health programs are
represented by an organization
entitled the California Conference of
Directors of Environmental Health
(CCDEH). CCDEH meets regularly to
address policy issues, develop
positions on proposed legislation,
and provide a forum for discussing
consistency in enforcement of
California's laws and regulations.
CCDEH recognizes the utility of
developing strong and effective waste
reduction programs and has
established a subcommittee of
individuals who are active in this
area. The CCDEH Waste Reduction
Subcommittee meets every two months
to discuss the latest developments in
waste reduction, resolve problems
associated with implementing waste
reduction programs at the local
level, and to exchange information
between local programs in order to
avoid duplication of efforts.
A current goal of the
Subcommittee is to assist counties
without waste reduction programs and
provide them with the information,
tools, and support to develop
effective programs. The
Subcommittee, in conjunction with the
Department's Technology Clearinghouse
Program, is currently developing two
one-day training sessions on the
basic concepts of waste reduction as
it relates to local enforcement
efforts. The ultimate goal of the
Subcommittee is to provide all
counties with information on how to
effectively implement. waste
reduction.
267
-------
REFERENCES
1. U.S. EPA, Risk Reduction
Engineering Lab., Waste Min. Branch,
1988, WRITE Pilot Program with State
& Local Governments, Cinn., OH, 46p.
2. Ventura County Environmental
Health, 1987, Hazardous Waste
Reduction Guidelines for
Environmental Health Programs, CA
Dept. Health Services (DHS), Toxic
Substances Control Div. (TSCD), Alt.
Tech. Sec. (ATS), Sacramento, CA,
45p.
3. Ibid, 1987, Hazardous Waste
Minimization Program Results and
Case Studies, DHS, TSCD, ATS,
Sacramento, CA, 80p.
4. Hanlon, D., 1988, Waste
Minimization Assessments and
Procedures to Estimate the
Effectiveness of Waste Minimization
Techniques in the Electroplating
Industry (Proposal), City of L.A. ,
Haz. & Toxic Material Project, Bd. of
Public Works.
5. Local Government Commission,
1988, Low Cost Ways to Promote
Hazardous Waste Minimization: A
Resource Guide for Local Governments
and Appendix B, Sacramento, CA.
6. Ibid., 1988, Reducing Industrial
Toxic Wastes and Discharges: The Role
of the POTWs, Sacramento, CA.
7. Ibid., 1988, Minimizing Hazardous
Wastes: Regulatory Options for Local
Governments, Sacramento, CA.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
268
-------
ROTARY KILN INCINERATION SYSTEMS:
OPERATING TECHNIQUES FOR IMPROVED PERFORMANCE
Joseph J. Santoleri
Four Nines, Inc.
Plymouth Meeting, PA 19462
ABSTRACT
Experience in the operation of rotary kilns goes back many
years with the thousands of kilns throughout the world. However,
much of this experience is in the cement, lime, and calcined
dolomite industries. In the past twenty to thirty years, rotary
kilns have been used in the incineration of municipal and
industrial wastes. The operating practices differ in that tne
industrial kilns are used to generate a quality-controlled product.
Flame size and shape, heat transfer by radiation and convection,
temperature distribution, and contact-time all play a critical part
in the quality of the end product. These kilns are normally 50 to
200 meters (150 to 600 ft.) in length.
Kilns used for incineration are typically batch fed with
solids of varying shape, size, and heat content. This provides
flexibility not available in other incinerator systems. These kilns
may also burn liquids, slurries, sludges and contaminated soils at
a continuous feed-rate. The Resource Conservation and Recovery Act
(RCRA) establishes the combustion performance required to obtain
an operating permit. Many existing kilns have been modified in
design and operating practices to allow the system to meet RCRA
standards. These modifications have included feed devices, seals,
lance design and location, controls, scrubber systems, monitors,
safeties, etc.
This paper covers the experience gained at several rotary kiln
installations burning hazardous wastes. This includes the
modifications in design and operation to minimize fugitive
emissions, temperature, pressure and stack emission upsets. This
has provided systems whose performance insures a safe environment
to the owner and the surrounding community.
269
-------
INTRODUCTION
Early development of the
rotary kiln started about 1877
in England. The first com-
mercial rotary kiln was the
result of the proven work of
American engineers in 1895, by
Hurry and Seaman of the Atlas
Cement Co. Those first kilns
were 45 cm (18 in) in diameter
and 4.5 meters (15 ft) in
length. Kiln sizes started to
explode in the 1960's when they
reached dimensions up to 6.5
meters (21 ft.) diameter and up
to 238 m (780 ft) in length.
The energy crisis of the
70's represented a blessing in
disguise in matters of kiln
design. This occured world wide
when modifications to preheaters
were made along with use of
alternate fuels. The major
breakthrough came in Europe
where precalcination was
successfully attempted in the
late 1960's using a very low BTU
bituminous shale as a component
of the kiln feed. As early as
1957, oil shale in slurry form
was used as a potential source
of energy in Canada (4).
Other wastes burned in
cement kilns have included
brines, aqueous metal-bearing
wastes, acidic wastes, lime-alum
sludges, halogenated wastes,
spent solvents and still bottoms
(2).
Although cement kilns will
accommodate a variety of mater-
ials both as fuel and feed in
the cement process, there are
limitations in the use of haz-
ardous wastes as there are
compounds not desirable in the
cement process. The experience
gained in this industry has led
to the acceptance of the rotary
kiln as a means of disposing
wastes, both hazardous and
non-hazardous by incineration.
Rotary kilns provide a
number of functions necessary
for incineration. They provide
the conveyance and mixing of
solids, a mechanism for heat
exchange, serve as a host vessel
for chemical reactions and
provide a means of ducting the
volatilized gases for further
processing. The kilns are
equally applicable to solids,
sludges, and slurries and are
capable of receiving and
processing liquids and solids
simultaneously (7).
WASTE DATA
Waste streams in the
process industries are numerous
in kind and therefore defy easy
definition. Disposal of these
wastes has become a serious
problem for the plant operator.
The following waste data must
be provided before a selection
of incinerator design can be
completed.
Waste Data Required for Design
Chemical Composition
Specific gravity
Heat of Combustion
Corrosivity
Ignitability
Reactivity
Moisture Content
Size Consistency
Slagging properties
(Temp., Eutectic Data)
The quantity of the waste
materials in total; that is,
solids, sludge, slurries,
liquids (and fumes), will
determine not only the total
270
-------
throughput in units of weight
(tons, kg, pounds) per unit of
time (min, hr, day, year), but
also the total heat duty
requirements of the systems.
INCINERATOR TYPES
The physical data will aid
in determination of the type of
incinerator designs that can be
considered. There are many
types that may be used to
incinerate hazardous wastes.
They are as follows:
Liquid Injection Incinerators
Rotary Kiln
Fixed Hearth (Two-Stage)
Mono-Hearth (Rotating)
Multiple Hearth
Fluidized Bed
Infra-Red Furnace
Rotary Reactor (Cascading Bed)
Molten Glass Process
Wet Oxidation
Many of these designs are
limited to a feed that is
prepared specifically for the
incinerator type, i.e., liquids
and slurries that may be pumped
and atomized. Others require
a limit to the physical
dimensions of the feed and
require pretreatment; i.e.,
shredding prior to introduction
to the transport system. Some
designs will handle only easily
transported sludges and soils.
The rotary kiln and fixed hearth
are ideally suited to large size
solid feeds such as bulk trash,
containerized process waste
solids, as well as contaminated
soils, sludges, slurries and
liquids. The fixed hearth
design requires multiple ram
feeders to expose the surface
of the waste materials to the
combustion and heat exchange
process. This paper will
highlight the rotary kiln
incinerator since it is the most
flexible of solids incinerator
designs today. Others (fluid
bed, rotary reactor, etc.) are
used for incinerating solids and
require extensive feed prepar-
ation and transport to optimize
the mixing, heat transfer, and
combustion which are considered
major advantages for these other
designs (Fig. 1).
One major feature of both
kiln and fixed hearth designs
is that most solid wastes must
be batch fed. Other systems can
accept wastes continuously into
the combustion zone; this
provides improved combustion
and emission control. This is
a singular disadvantage to the
operation and performance of
batch-fed incinerators.
However, experience gained in
the past few years driven
by the RCRA regulations has
minimized this disadvantage.
The standards developed by
U.S.E.P.A. via RCRA and now
implemented by most of the
States' Solid Waste Authorities
are as follows:
RCRA Standards (5)
1. 99.99% Destruction and
Removal Efficiency (DRE)
of the Principal Organic
Hazardous Constituents
(POHC)
2. 99% removal of Hydrogen
Chloride (HCl) or HCl
emission not exceeding 1.8
Kg/hr (4 Ib/hr)
3. Particulate emission
concentration of 180
mg/dscm (0.08 gr/dscf)
corrected to 7% O2.
Guidelines have been issued
271
-------
Guidelines have been issued
to permit regulators to insure
that these standards are being
met during the trial burns and
subsequent operation of the
system. These apply to the
carbon monoxide emission
monitoring which is the
surrogate used to establish
conformance to the 99.99% DRE.
Guidelines are also in final
stages regarding the metals
emissions issues. Standards
covering particulate emissions
may be lowered to 0.015-0.02
gr/dscf to insure metals
emission control. Many states
have established maximum
emission levels well below the
0.08 gr/dscf. These levels are
based on the "Best Available
Control Technology" (BACT)
demonstrated at operating fac-
ilities within the state.
ROTARY KILN OPERATIONS
PAST AND PRESENT
In order to conform to
the above standards, past
operating practices at many rot-
ary kiln incinerator facilities
have had to be modified. Many
incinerator systems under
"Interim Status" (Part "A"
permit) have been operating
with the following conditions
as normal practice.
Past Practices
1. Batch Feed System:
Manual (Under operator
control)
Cycle for load (Varied
from 15 to 30 min.)
2. Kiln Draft Pressure:
Manual (Under operator
control)
Positive pressures
created fugitive
emissions.
3. Combustion:
Air/fuel control (Under
operator control)
Auxiliary fuel (Manual)
Afterburner (Volume
undersized)
Temperature Control
(Manual)
Burner (Design or
Location)
Atomizer (Design or
Location)
4. Scrubbers:
Submicron particulate
efficiency (Poor)
PH control (Manual)
Corrosive atmosphere
overlooked in material
selection.
Many systems had no
Scrubbers.
These operating con-
ditions resulted in stack
emissions of odors and
particulate emissions in the
surrounding community. The term
"NIMBY" (Not in My Backyard) was
originated. The environmental
groups soon formed to block any
future installation of
incinerator systems.
Those facilities operated
as on-site (industrial plants)
or off-site (commercial disposal
operators) have made many
changes and improvements to
comply with RCRA requirements.
As stated above, most of these
were driven by the modifications
needed to complete the Part "B"
permit and the final trial burn
at the facility. At most plants
preliminary test burns were run
to determine existing
capabilities. These tests
(mini-burns) resulted in many
or all of the following design
and operating improvements. This
272
-------
allowed the system to meet the
final standards for the trial
burn (6).
Improvements
1. Feed System
Continous Feed:
Shredder
Lump and Cake Breakers
Weigh Belt Control
Feed Devices:
Auger
Screw Feed (Single and
Multiple Screws)
Belt Feeder
Elevator
Ram
Batch Feed:
Container Size Control
Volume
.Weight
Container BTU Control
Cycle Time Modifications
Bar Code for Control
Mechanical Improvements:
(Fig. 2)
Roller Conveyors
Elevators
Intermittent Ram Feeder
Holding Chamber
Guillotine Door Size
Reduction
2. Combustion System (Fig.3)
Combustion Air Control
Improved Seal Design
Air Flow Meters to
Primary/Secondary
Oxygen Monitoring at
Secondary Level
Auxiliary fuel burner
type/location
Heat input control by
waste types
Secondary Combustion
Chamber Design/Mixing
Atomizer Design Modifica-
tions/Locations
Temperature Monitors/
Number/Location
Maximum Temperature
Control Via Air/Water
3. Pollution Control
System (Fig.4)
Combustion Chamber Design
Control velocity and
carryover
Orientation to prevent
ash-buildup
Waste Injectors
Location and size to
minimize attrition
Ash System
Hopper size/bridging
Temperature control
slagging or freezing
Quench Design
Quench liquid control
Temperature of operation
Scrubber
Venturi Throat
Control materials of
construction
Absorber
Packing
Demisters
pH control
CASE STUDIES
Experience with retrofit
and modifications to liquid in-
jection incinerators has been
covered in detail. Many changes
were required in order to obtain
approval for the Part "B" permit
(6).As stated above, former
practices with solids
incinerators led to problems
which generated excessive
emissions of particulates as
well as products of incomplete
combustion (PICS). Solid waste
materials are often collected
in 55 gallon drums to minimize
handling and labor costs. These
drums vary in weight from 400
273
-------
pounds to as high as 900 pounds
depending on the density of the
materials. The physical size of
the drum establishes the feed
mechanism size; i.e., belt,
chute feeder, elevator, ram
feeder, and guillotine door.
This establishes the kiln
opening. The kiln inside
diameter is then established.
The two limits to kiln capacity
are as follows:
1. The maximum feed rate of
solids in pounds per hour.
2. Maximum heat release
rate in BTU per hour.
Both must be evaluated to
determine whether an existing
kiln is operating at its optimum
conditions. Combustion systems
are optimized when operated at
a steady feed rate with a uni-
form heating value of feed with
a combustion air rate that
maintains constant oxygen level
in the stack gases. Process
furnaces or boilers operate at
maximum efficiency; i.e., at
minimum excess air (Oj). Waste
burning systems do not run with
low excess air levels (1-3% O2).
Most liquid waste incinerators
operate with stack at 3 - 5% O2.
This insures against variations
of waste composition and heating
value. When burning solids,
waste rates and BTU values are
even more difficult to control.
However, it is possible to
operate with proper air controls
at levels of 6 - 9% 02
continuously.
Many existing kilns have
poor seal designs. This will
result in high excess air drawn
into both front and rear seals
during operation. Some systems
have experienced excess air
levels of 100 to 150%. A major
disadvantage created by the poor
seal problem was limiting the
heat release of the kiln. The
high velocity created by this
additional volume causes fine
particle entrainment into
downstream equipment. This
volume adds to the heat load of
the afterburner. Residence time
in the afterburner must be held
in order to meet RCRA standards .
The gas volume establishes the
physical size of the downstream
air pollution control system
and induced draft fan.
Air in-leakage also occurs
at the feed chutes and
guillotine doors. Reduced
leakage at these points,
decreases the velocity in the
kiln and the particulate
carryover. However with a fixed
heat input, the temperature
will increase in the kiln.
Combining the same total heat
input with additional inert
materials such as soils and
moisture, the same physical
chamber size can be modified to
process an increased daily
tonnage of materials. The
cooling by water addition from
the wet feeds or waste water
sprays results in a reduction
of total gas flow in the kiln.
This will minimize kiln gas
velocity and carryover. The
afterburner design must provide
the necessary turbulence and
temperature rise needed to
provide the 99.99% DRE of the
organic components in the waste.
The solid feed rate limits
are determined by several
factors. The physical dimensions
of the kiln (diameter, length
and slope) and the rotation
(rpm) establish the residence
time of solids. This must be
reviewed based on the volatility
274
-------
their surface heat absorption
rate from the temperature and
O, in the kiln gases. This heat
absorption rate is dependent
upon the physical nature of the
solids, the water content which
establishes the drying load,
and the inert materials (ash)
in the waste (Pig. 5). For good
heat transfer, volatilization
and combustion of the organics,
a solid volume fraction of 5 to
15 % is recommended. Typical
residence time of solids may
vary from 30 minutes to 2 hours.
Many kilns were operating
at loading rate cycles that
varied from 10 minutes to as
long as 30 minutes. The
resulting variation in kiln
pressure and temperature often
was out of control of the
operator. Puffing at the seals
would occur with PICS entering
the operating area. The
afterburner control also was
difficult to maintain.
Temperatures would follow the
variations created in the kiln.
Since the kiln heat release rate
would vary, and auxiliary firing
of the afterburner was by a
forced draft burner, there was
no means for controlling stack
C>2 levels. The result would be
large variations in stack
oxygen. After monitors were
placed on the stack for carbon
monoxide (CO) or total
hydrocarbons (THC), large spikes
would be observed many times to
the upper limit of the scale
(3000 - 5000 ppm). At the same
time, C>2 would drop to 0%.
The capacity of a system in
total heat release rate is
established by the volume of
combustion air provided in the
forced draft (FD) and induced
draft (ID) fans. The individual
burners are limited by the FD
fans. The total incineration
system (kiln and after-burner)
is limited by the I.D. fan
capacity. One must first
establish the level of O^ needed
for satifactory operation.
Having the I.D. fan data (ACFM
@ Temp.), one may calculate the
dry SCFM handled by the fan.
Based on the oxygen level for
the system, the excess air level
may be established (Fig.6).
The total BTU capacity for the
incinerator system may then be
determined by the following
formula:
BTU/hr =
(DSCFM x 6000)/Total Air
- where -
Total Air =
1 + excess air fraction.
Having established the maximum
heat release of the system, one
would now review the capacities
in both kiln and afterburner .The
kiln is the critical zone since
it is a batch fed combustor.
Certain wastes due to feeding
problems and the nature of the
waste must be fed continuously;
e.g., sludges, slurries and
waste water streams. The heat
release for these should be
relatively stable if proper
measures have been taken in the
storage and mixing areas. The
heat release from the batch fed
wastes will be dependent upon
the weight per charge, the BTU/
charge, and the relative
volatility in each charge. If
each drum is prepared
consistently, the values of
weight, BTU, and volatility wil 1
be steady. The result is a
cycled heat input dependent upon
cycle time and the BTU/charge.
Locations with the ultimate in
control of waste feed have
275
-------
experienced sudden energy
release from charges into the
kiln. This is followed by a
sudden increase in temperature
and a rapid decrease of O«
exiting the chamber. The total
volume of air in the kiln is
maintained at a fixed rate by
the draft created by the I.D.
fan. With the rapid depletion
of 02 in the kiln/ the stack gas
emissions drop in O, and
increase in CO and THC. It these
reach permit shutdown levels,
all waste feeds must be cutoff
per present regulations. In
past operations, continuous
monitors such as kiln draft, 02,
CO, etc. were not required and
waste cutoffs were manually
controlled by the incinerator
operator.
In order to minimize these
upsets, modifications were made
in the operations and controls.
In the case of continued spiking
due to high volatility (BTU) of
the waste feed, the charge size
was reduced and fed more often.
One incinerator was designed to
operate with stack gas
containing 10% O* average. The
original charge or 270 Ibs. with
a heat content of 1.66 MM BTU
was fed every 15 minutes. Note
from Pig. 7 the effect of
materials volatilizing in three
(3) to five (5) minutes. The
five minute volatilization would
reduce the Q^ to 4.5%. However,
with volatilization occuring in
three minutes, there was not
enough air to operate without
high levels of CO and PICS. By
reducing the charge size to 90
Ibs. (0.55 MM BTU) charged every
5 minutes, the system was able
to stay in control with either
a 3 minute or 5 minute
volatility material (Fig. 8)
(6).
By utilizing the oxygen
analyzer as a control device,
another incinerator system was
brought under control reducing
waste feed cutoffs. Fibre-packs
are fed at a fixed rate with the
feed rate based on BTU/ charge
(Fig 9). Occasionally, a "hot"
drum would enter the kiln. A
"hot" drum is one containing
free liquids. This resulted in
a sudden drop in 0% measured at
the exit of the after-burner.
In this case, the kiln steady
state firing rate was 30 MM
BTU/hr and the afterburner -
70 MM BTU/hr. The total heat
release rate from the system
averaged 100 MM BTU/hr. Normal
operation resulted in waste feed
cutoffs two to three times
daily. With all of the auxiliary
heat added by the waste liquids
into the afterburner, a waste
feed cutoff resulted in waste
solids continuing to burn in the
kiln with no additional heat
input to the afterburner. The
CO level would spike followed
by a prolonged time period with
high THCs in the stack gases.
With the I.D. fan operating at
fixed output, the kiln and
afterburner chambers would cool
rapidly increasing the THC
levels.
The first indication of a
problem is the dropoff of 0^
level. This is followed by the
CO spike. By monitoring the O^
level (6 to 7%) and using the
average - 6.5% as a control
point, a dropoff of 1.5% indi-
cates a problem may be in the
initial stage. Since the liquid
firing rate was 70% of the total
heat input, this was used as a
control to maintain 0^ level at
a fixed afterburner temperature.
The initial 1.5% 0, drop
triggered a reduction or liquid
276
-------
heat input by 50%. A continued
drop of the 02 level reduced
liquid heat input an additional
30%. This drops the total heat
input by the liquid to 24% from
the 70% level (70-.8x70).
Most often this would be
sufficient to keep the C^ level
from reaching the waste feed
cutoff point - 3%. This assumes
that the sudden heat input from
the kiln feed has increased from
a level of 30% to 76% (more than
2.5 times). In many cases. The
heat input from the volatile
solids would be less with other
waste streams entering the kiln
such as sludges and slurries at
a constant flow and BTU input.
This control modification has
reduced the number of waste feed
cutoffs from 02 trips with the
resultant spikes of CO and THC.
These control modifications are
needed to insure the environment
in the area of the incinerator
for the operators as well as
that for the surrounding
community is maintained.
Complaints of odors from the
"NIMBY" groups may be justified
based on past practices. The
modifications needed to improve
operating conditions and
essentially eliminate these
problems are possible. It re-
quires close observation of the
daily operating procedures, dis-
cussions with the operators and
review of the strip charts and
log books. Only then can one
determine that there are
problems and that modification
to operating procedures or
controls will eliminate the pro-
blem. In many cases however,
redesign of the basic hardware
will be needed to achieve the
control necessary (Fig.10).
SUMMARY
The cases described above
cover instances where
improvements have been made to
meet the standards established
by RCRA. The results have shown
an improvement in the efficiency
of the operation as well as
lower operating costs . One major
benefit has been in lowering
maintenance costs, especially
in refractory repair. Closer
control of heat input has
maintained more uniform
temperatures which has resulted
in increased refractory life.
It also reduces downtime and
provides higher utilization of
the equipment. Additional
capital expenditures were
necessary to make these modif-
ications. Training costs were
also higher. All of this has
resulted in operating plants who
pride themselves with systems
that provide not only increased
employment and revenues to the
local community, but also a
means of eliminating the
hazardous materials from the
environment forever. It has
allowed the rotary kiln to
become one of the most flexible
incinerator designs for all
hazardous waste streams. From
its original use as a process
furnace, it can now be found in
use at MSW plants, commercial
hospital waste disposal
facilities, and large commercial
hazardous waste sites. The
design most often selected for
many large on-site facilities
has been the rotary kiln, either
slagging or ashing. Many of the
mobile or portable units used
for Superfund (SARA) cleanups
have' been the rotary because of
its flexibility in handling the
variety of waste types and
physical shapes.
277
-------
Many rotary kiln
incineration systems have been
modified and tested for the RCRA
and TSCA permits in the past 10
years. This testing has shown
that rotary kilns can achieve
DREs well above 99.99% and often
> 6 - 9s. The incineration of
hazardous wastes in a rotary
kiln has become popular because
the designs, concepts, and the-
ories are well established and
proven in many solids
processing industries. The
effort to meet the standards
established by RCRA has enabled
the latest control and
instrumentation technologies to
be included in the design and
operation of all systems.
(1,3,8) It is extremely
important that the technical
community who understand the
results of these improvements
inform the citizens who oppose
siting of units about the
advantages of a properly
designed and operated incin-
eration system to effective
disposal of hazardous wastes.
REFERENCES
1. Bastian, Ronald E.
"Eastman Kodak Company Chemical
Waste Incineration" LSU
Conference on R.K.
Incineration - Nov. 1987
2. Chadbourne, J.F.
"Cement Kilns"
Sect. 8.5 of Freeman's "STANDARD
HANDBOOK OF HAZARDOUS WASTE
TREATMENT AND DISPOSAL",
McGraw Hill, 1988
3. Osborne, J. Michael
"3M Operating Experiences with
a High Temperature Kiln"
LSU Conference on Rotary Kiln
Incineration Nov. 1987
4. Peray,R.E.
"The Rotary Cement Kiln"
2nd Ed. Chemical Publishing
Co.,Inc. N.Y.,N.Y.- 1986
5. Resource Conservation and
Recovery Act. Standards for
Owners and Operators of Waste
Facilities: Incinerators. 40
CFR 264, RCRA 3004, Jan. 25,
1981, Rev. July 9, 1984
6. Santoleri, Joseph J.
"Mini-Burns - Critical to Trial
Burn Success"
APCA - Dallas,TX No.88-015.06
"Design and Operating Problems
of Hazardous Waste Incin-
erators"
ENVIRONMENTAL PROGRESS
(Vol.4-# 4)
Nov. 1985
7. Schaefer, C.F. and Albert,
A.A.
"Rotary Kilns"
Sect. 8.2 of Freeman's "STANDARD
HANDBOOK OF HAZARDOUS WASTE
TREATMENT AND DISPOSAL"
McGraw Hill, 1988
8. Williams, Gad L.
"Status of the Technology of
Rotary Kiln Incineration of
Hazardous Waste"
LSU Conference on R.K.
Incineration - Nov.1987
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
278
-------
Rotary
Kiln
Unit
SECONDARY
FIG. 1.
FIG. 2.
279
-------
FIG. 3 ROTARY KILN AND AFTERBURNER
HUH IHIET —
EMERGENCY VENT
FLUE DAS
IHIET TO QUENCH
OR BOILER
SIGHT PORT
..
RESIDUE CUNVEYOR-
FIG. 4. ROTARY KILN INCINERATOR
Air
(|uench/(*auslic
Solids
Aax. Fuel
Alter bur ner'
Rotary
Kiln
—A
Ash
Yenluri
Scrubber
I.D.
Fan
\
Stack
280
-------
KILN ROTATION 0.5-2.0 RPU
AIR V[T/ / / / / / / / REFRACTORY LINING
WASTE .
LIQUIDS/
FUELS
-DRYING—LTRAHSFCRMATION~p-«MBUSTIGN
INCINERATION -
SLAG
FIG. 5. ROTARY KILN PROCESSES
PERCENT O, VS. PERCENT EXCESS A1B
I ! : i \ 1
M.A. H04K. turn. »1«*c. 13/7«
HIHIS-nat.LIH9-lfllOOIPOMT
K01.LIHJ-E.H, LA.
Q "0 DATA. >-l_,r
I I I j I
«o *o too .130 144! i«o i«o
PEBCEHT EXCCS3 AIR
FIO. 8«. ROTAHY KILN OPERA71NQ CURVES
HEAT OUTPUT OF INCINERATOR. MM BTU/HR
V8.
8TACK FLOW WITH FIXED AIB-IN LEAKAOB
AND
STACK TCHP. • IOO F. SAT.
SO «O TO «0 to IOO 11O
HEAT OVTPUT, MM BTU/HR
FIQ. 8B. ROTARY MLM OPERATINO CURVES
281
-------
i* i?
§5. *£
X' "".m 5 r
< o * 2 o
, en in ~* "~
l u u
1 /
i _^
~— — V
~ = =£
3
~=~c
-=.r/
~S5>
i ii^^
~
L_J
o
«o
in
V
O'
(0
in
o
n
u
e
E
r
E
in
^-
SD
I/)
in
r
r
r-
£ £
a £
a
f»
e
E
en
u.
O
H-
u
.1 1.
D V
N tN
ail/niO MM '3SV313U 1V3II
, co v o «o *CN , to v o
CS tN
J 3
e ?
•c
UJ
en
2
E
IR
(D
r
c
SO
C
o
CM
i^j
r
in p
^ to
l_ O
«J
o
en
k^
e
^«
u
DC
a
K
B
r
(VI
m
m
. i
B
E S
£ I
e v
MM '3SV313U 1V3II
en
S
282
-------
FIG. 9.
TYPICAL KILN HEAT RELEASE
TOTAL HEAT RELEASE
<
<
'.u
UJ
HIGH BTU LIQUID WASTES
DRUMS & OTHER BATCH FED WASTES
AQUEOUS LIQUIDS, SLUDGES AMD OTHER PUMPA8LE
WASTE HOT SUITABLE FOR HEAT RELEASE CONTROL
5878
TIME ( MINUTES )
283
-------
-
1
-i
"1
1!
- SI
Si
<
i
CL
*-— rj
i
f
"^
— 1
I
— ••%.
Stf
nlS
«
S
i
i
i
»
-~--.
^<
^
\
— ~~
-
.^ n
>
i*
^
^
I
»
1^
S
<
\» IOUCTOB '
\\SLUOOI / ALKY
(OUCTOH /// !
JL/ALKYL //
3
»*•
—
«•,
^ z
* — a
o
z
7 J
K
j. V
re
* s!j
KO
-
-
-
-
-
to
8
S
8
m
s
5
S |
a *
* i
I i
8 •"
S
e
>
Mt
FIG. 10b. ROTARY KILN OPERATING DATA
? g g 2 8S8S8ES585S9388S852"0
I HII /ma nn
ODS Ml ZO 1N30H3J
-i
I §
i i
i
8
£ 8
|
E
;
,
I
>
a
w
[
n
x
3
O
w
w
ft
*"
^
a
o
z
a
n
n
*•>% I«««KJW w 0
JH C
aii/nxu nn
o ^ « t»
ODS Nl *0 ON30«3d
284
-------
THE USE OF OXYGEN IN HAZARDOUS WASTE INCINERATION
— A STATE-OF-THE-ART REVIEW
Min-Da Ho and and Maynard G. Ding
Linde Division, Union Carbide Industrial Gases Inc.
Tarrytown, New York 10591
Copyright X©) 1989, Union Carbide Industrial Gases Inc.
Revised June 1989"
ABSTRACT
The use of advanced oxygen combustion technologies in hazardous waste incineration
has emerged in the last two years as one of the most significant breakthroughs
among all the competing treatment technologies. Unlike most others, oxygen
combustion technologies can be easily retrofitted onto various existing
incinerators. The capacity of existing incinerators can typically be increased by
a factor of two or three, and destruction and removal efficiencies (DRE's) can
potentially be improved with such a retrofit.
For many years, industrial furnaces have used oxygen enrichment of the combustion
air and oxygen-fuel burners, but with conventional technologies a high oxygen
level generally poses problems. The flame temperature is high, leading to high
NOx formation and local overheating. Different technical approaches to overcome
these problems and their respective effectiveness will be reviewed. Previously,
commercial oxygen enrichment in incinerators was limited to a rather modest level
(less than 26% 02). This paper will review some of the recent commercial
applications of much higher oxygen enrichment levels in hazardous waste
incinerators.
The general characteristics of any oxygen enriched flame, the benefits that can be
anticipated, and the associated economic ramifications are explored in this paper.
Also reported are the results of recent EPA evaluations of two unique oxygen
combustion technologies: The Pyretron™burner by American Combustion, Inc. and the
LINDE®Oxygen Combustion System by the Linde Division of Union Carbide Industrial
Gases, inc.
285
-------
INTRODUCTION
In response to the Federal Resource
Conservation and Recovery Act (RCRA)
amendments of 1984 and Superfund
Reauthorization and Amendments (SARA)
of 1986, incineration is generally
considered to be the most permanent
solution of hazardous chemical waste
treatment. RCRA includes a statement
of national policy which emphasizes
that "reliance on land disposal of
hazardous waste should be minimized or
eliminated...land disposal should be
used as a last resort and should be
replaced in most cases by advanced
treatment, recycling, incineration and
other hazardous waste control
technologies." Incineration is a
proven technology tor treating a wide
range of materials including liquid,
solid, and semi-solid wastes such as
PCBs, solvents, organic residues,
halogenated hydrocarbons, pesticides,
herbicides, and laboratory waste.(1,2)
In addition, the continued discovery
of abandoned hazardous waste sites as
a result of Superfund investigations
has placed increasing pressure on the
U.S. Environmental Protection Agency
(EPA) to find alternate solutions for
treating and disposing of toxic and
hazardous wastes. The decreasing
availability of landfill sites and the
increasing public opposition to toxic
and hazardous waste transport have
added to the pressure.
EPA's regulations for incineration of
hazardous wastes require that the
system must achieve a Destruction and
Removal Efficiency (DRE) of at least
99.99 percent of the Principal Organic
Hazardous Constituents (POHCs) present
in the waste and at least 99.9999
percent for dioxiri and PCB
contaminated wastes. High excess air
levels are generally used to ensure
that the incinerators meet these high
performance standards.
The use of oxygen or oxygen-enriched
air ixi place of air for incineration
can improve the overall performance
and efficiency of chemical waste
incinerators, and reduce the overall
cost of the system. As oxygen
replaces part or all of the air for
incineration, the nitrogen portion is
reduced in both the oxidant and the
flue gas. Hence, the volume of the
oxidant and the flue gas are reduced
per unit of waste processed. In
addition, the concentration of oxygen
in the fuel-oxidant mixture is
increased.
EFFECTS OF OXYGEN ENRICHMENT
ON GENERAL COMBUSTION
CHARACTERISTICS
Oxygen enrichment (21-100%) of air
reduces the amount of nitrogen present
as a diluent in the reaction of fuel
and oxygen. The rate of combustion
reaction usually increases
significantly with oxygen enrichment
due to the higher partial pressures of
both oxygen and fuel and the resulting
higher equilibrium temperature. This
higher reaction rate is one of the
main reasons for the following changes
in the combustion characteristics^):
- higher flame speed
- lower ignition temperature
- wider flammability range
- higher blow-off velocity
gradients
- higher adiabatic flame
temperature
For incineration applications it is
necessary to evaluate how these
changes affect flame stability, flame
temperatures and safety of the
process.
Flame Stability
In general, a higher flame speed will
improve flame stability and create a
286
-------
more intense and shorter flame. It
is, however, difficult to
quantitatively predict the actual
effects on industrial burner flames
which are post mixed and turbulent.
The lower flammability limit (lean
limit) of a fuel and air mixture is
little influenced by oxygen enrichment
of the air. This is expected since
excess oxygen in the lean limit is
considered to act as a heat sink
similar to nitrogen(A). The higher
flammability limit (rich limit), on
the other hand, is extended
substantially with oxygen enrichment,
as shown in Table 1. For example, the
higher flammability limit of methane
is increased from 14% to 61% by going
from air to pure oxygen. For all the
fuels cited, the ratio of the limits
expands greatly.
For burner applications, wider
flammability limits generally
correlate with greater flame
stability. The change in the upper
limit allows the combustion of the
fuel or waste to begin even in a
highly fuel-rich mixture.
Stability of a premixed flame can be
measured in terms of the critical
velocity gradients at blow-off limits.
"Blow-off" is the condition of a.
burner flame where the flow velocities
of the gases forming the combustible
mixture exceed the burning velocity
everywhere in the flow field. In such
conditions, the combustion wave is
driven back from the burner and loses
its stable "anchor" in relation to the
burner face. As much as 100-1000 fold
increases in blowoff velocity
gradients were measured when pure
oxygen was used instead of air (7) •
The dramatic increases in the blow-off
velocity gradients with oxygen
enrichment are considered to improve
flame stability for wide turn down
ranges of firing rate. A higher
blow-off limit is also advantageous in
designing high velocity burners with
good flame stability.
Adiabatic Flame Temperature
The adiabatic flame temperatures
increase significantly with oxygen
enrichment due to the reduction of
nitrogen which acts as a diluent in
combustion. The flame temperature
increases by as much as 100°F for a 1%
increase in oxygen concentration for
low enrichment levels. The rate of
increase in the flame temperature
decreases gradually with the
enrichment level and tapers off at
high enrichment levels.
There are two main reasons for this
phenomenon: (1) the amount of
nitrogen eliminated with each unit
percent oxygen enrichment diminishes
with increasing enrichment level, arid
(2) endothermic dissociation of C02
and H70 becomes increasingly
significant at high temperatures.
Table 2 lists adiabatic flame
temperatures of selected fuels with
air and oxygen at stoichiometric
ratios.
Adiabatic flame temperature is the
maximum flame temperature attainable
under an ideal condition. The actual
temperature of an industrial burner
flame is significantly lower than the
adiabatic flame temperature due to
radiative heat loss and turbulent
mixing with surrounding colder furnace
gases.
It is important to recognize that the
adiabatic flame temperature simply
provides an upper limit in the
attainable flame temperature and that
higher flame temperature is not
essential in increasing heat transfer
in a furnace with oxygen enrichment.
Special burners and oxygen enrichment
techniques have been developed and
287
-------
applied in industries to increase heat
transfer in a furnace without creating
higher flame temperatures that might
cause a local overheating
problem( 13-J5). Further discussion on
furnace heat transfer and flame
temperature is given later in this
paper.
Oxygen Safety
An important safety consideration
resulting from higher reactivities and
lower ignition temperatures in an
oxygen enriched atmosphere is the
material compatibility for oxygen
service. Many materials which do not
ignite in air can ignite in oxygen
enriched atmospheres. Extensive
studies have been conducted to
evaluate metals, sealing materials and
lubricants for oxygen service (B), and
practical guidelines have been
reported for piping (9,10),
compressors, and pumps (11).
From the viewpoint of combustion
safety, the potential to create a
flammable mixture in a furnace prior
to startup would increase if both the
fuel and oxygen enriched air leak
simultaneously. Thus, the prevention
of accidental accumulation of oxygen
enriched air in the furnace becomes an
important design consideration for an
oxygen-enriched combustion system.
ADVANTAGES OF USING OXYGEN
The main advantages of using oxygen
for incineration can be summarized:
(1) the fuel consumption, if
supplemental fuel is required, is
lowered primarily due to the reduced
sensible heat loss to the flue gases;
(2) the throughput of the incinerator,
which is normally limited by the air
blower capacity, the gas residence
time and the size of the flue gas
cleaning system when using air, can be
significantly increased; (3) the HRE
can potentially be improved due to the
higher oxygen concentration in the
fuel-oxidant mixture and longer
residence time; (4) pollution control
of the reduced flue gas is less costly
and more effective; and (5) control of
"puffs", as indicated by CO
excursions, is achievable.
Fuel Savings with Oxygen
For the incineration of low BTU wastes
such as aqueous waste and contaminated
soil, very significant amounts of
auxiliary fuel are consumed. In such
a case, flue loss is usually the
biggest single source of heat loss for
a high temperature incinerator. In a
typical waste incinerator fired with
natural gas and cold air, about 50-70%
of the higher heating value of the
fuel is lost as sensible heat in the
flue gas.
By replacing combustion air with
oxygen, the corresponding reduction of
nitrogen in the flue gas lowers the
sensible heat loss dramatically. In
Figure 1 the fuel required to provide
1 MM BTU of available heat to a
furnace is plotted as a function of
flue gas temperature for ambient air,
enriched air, and oxygen. The
"available heat" can be defined as the
gross quantity of heat released within
a combustion chamber minus the
combustion flue gas loss. For this
example, the fuel is methane and the
oxygen concentration in the flue gas
is 6 percent by volume. As the flue
gas temperature increases, the fuel
requirement to provide a given amount
of energy above that temperature level
increases, and the difference between
air and oxygen as oxidants becomes
greater. For example, as the
temperature level increases from
1800°F to 2400°F, the fuel requirement
to obtain 1 MM BTU of available energy
increases from 2.2 to 6.1 MM BTU for
288
-------
air, and from 1.2 to 1.4 for oxygen.
As a result, the fuel savings using
oxygen increases substantially as the
operating temperature of the furnace
increases. Shown in Fig. 2 are the
specific fuel savings by using oxygen
enrichment in terms of MM BTU fuel
savings per ton of oxygen, as a
function of flue gas temperature and
excess oxygen level.
Note that when the concentration of
excess oxygen in the flue gas is
increased for both the air and oxygen
combustion systems to the same level,
fuel efficiency of the air system
deteriorates much faster than that for
the oxygen system. Consequently, fuel
savings by switching from air to
oxygen becomes greater when the
concentration of excess oxygen in the
flue gas is higher. In addition, when
the system throughput is increased
with oxygen enrichment, further fuel
savings can be realized.
Throughput Increases
Oxygen enrichment has been.
successfully used for throughput
increases in a broad range of
industrial furnaces(3,16-19).
Production increases of 10-20% are
typically possible with a few percent
increase in oxygen concentration for
most furnaces.
The extent of throughput improvements
possible for a particular incinerator
depends on the nature of the
incinerator limitations. Some of the
typical limitations are listed in
Table 3. The most common limitations
are those related to the capacity
limitations of fuel and air supply
systems and the flue handling system
including the air pollution control
devices. Oxygen enrichment is very
effective in overcoming these
limitations due to the reduction of
the volume of oxidant and flue gas for
the same fuel input and higher
available heat to the furnace. Such
benefits are especially significant
for a low BTU waste which requires
auxiliary fuel input. Shown in Figure
3, as an example, is the relative flue
gas volume as a function of oxygen
enrichment to obtain the same
available heat.
In the incineration of medium BTU
waste (2000-8000 BTU/lb) with oxygen
enrichment, by reducing and in some
cases eliminating the use of auxiliary
fuel, the thermal capacity of an
incinerator can be dedicated to the
combustion of hazardous waste instead
of auxiliary fuel. Throughput
increase in this manner can often be
achieved. Even for incineration of
high BTU waste where auxiliary fuel is
not required, a specific flue gas
volume reduction can be achieved with
oxygen enrichment in conjunction with
the use of waste water injection.
The dust carryover problem, a common
process limitation, is related to
particle size, characteristics of the
particulate matter, and the
aerodynamic patterns within the kiln.
Although the improvement by using
oxygen can not be accurately
predicted, the lower kiln superficial
gas velocity should be beneficial in
reducing the dust carryover. For
example, dust carryover problems in
cement, lime and hazardous waste
rotary kilns have been effectively
alleviated by the lower flue gas
volumes resulting from oxygen
enrichment(17,19,20).
In many cases the capacity of an
incinerator is limited by the
mechanical design of the system.
Mechanical modifications must be made
in order to overcome such limitations
before oxygen enrichment can be
useful.
289
-------
For the incineration of low-BTU
wastes, such as contaminated soil, the
heat transfer rate to the heat load
may be a rate-limiting factor. High
temperature oxygen/oxygen-enriched
flames have been successfully applied
to certain glass melters and kilns to
increase heat transfer to strategic
areas in the vessels. In most waste
incinerators such as a rotary kiln,
however, a high temperature flame can
cause overheating of the refractory
walls and possible slagging problems.
On the other hand, an extremely high
temperature flame is not essential to
achieve a higher overall heat transfer
rate. Gas radiation from hot
combustion products to the surrounding
refractory walls and re-radiation to
the heat load is the primary mode of
heat transfer in most high temperature
furnaces. The intensity of gas
radiation is not only a function of
gas temperature and concentrations of
C02, H20, and soot, but also is
strongly influenced by the volume of
the radiating gas. The volume of a
flame is usually a small fraction of
the entire furnace. Thus, in a
radiation dominant furnace with
temperature limitations, the preferred
condition for productivity improvement
is to increase the average gas
temperature of the heat transfer zone
by enhancing the temperature
uniformity, rather than by a localized
increase in flame temperature. The
above principle should be applied
intelligently to maximize the benefit
of oxygen-enriched combustion.
Performance Improvements
It is sometimes argued that the high
flame temperature achievable with
oxygen enrichment is conducive to
higher DREs due to improved oxidation
kinetics. However, studies have shown
that with a temperature above 2000°F,
the combustion reaction is limited by
the rate of mass transfer processes of
oxygen and toxic molecules (i.e.
atomization, evaporation and mixing),
rather than the kinetic rates, the
contribution of extremely high
temperatures being quite limited.
Therefore, a flame with high momentum
and moderate temperature may be most
suitable for incinerator applications.
EPA studies have also shown that
well-run conventional air-based
incineration systems achieved very
high DREs at temperatures between
1800°F to 2200°FU,21.).
On the other hand, EPA studies and
pilot scale tests also show that the
performance of incinerators could
deteriorate significantly during some
upset conditions (or so called
"failure modes")(22,23). Oxygen
enrichment can alleviate many of the
failure modes. One of the important
failure modes is the occurrence of
flameout. This failure mode can
clearly benefit from oxygen enrichment
which improves flame stability. Poor
atomization, low combustion
temperature arid slow evaporation of
liquid waste have been cited as
important failure modes (24).
Although atomization of waste depends
mostly on the burner system design,
oxygen enrichment has been shown to
improve evaporation of liquid waste by
raising the intensity of the flame and
therefore improving the burnout
efficiency of waste (25).
Transient Emissions Control
Another important failure mode is
transient emissions (puffs) from
incinerators. When high-BTU wastes
are fed into rotary kiln incinerators
in an intermittent mode, the transient
combustion behaviors of these
materials create unsteady releases of
combustible gases which may
momentarily exceed the oxygen supply
to the incinerator. These temporary
oxygen-deficient conditions can cause
290
-------
the release of products of incomplete
combustion (PICs) which are often
called "puffs". These "puff"
phenomena have raised public concerns
recently and have been the subject of
research projects sponsored by the EPA
(26,27).
It has been suggested that the higher
partial pressure of oxygen in the
combustion chambers available with
oxygen enrichment, can alleviate the
temporary oxygen-deficient conditions.
This approach, the subject of a. recent
EPA study reporting mixed results, is
discussed later in this paper (28).
In addition, it should also be noted
that too high an oxygen level is not
only inefficient, but may also cause
high NOx emission level (28).
Therefore, oxygen enrichment, per se,
may not be an ideal solution to the
transient puff problem.
On the other hand, advanced process
control techniques utilizing advanced
process sensors and dynamic oxygen
injection can reduce puff occurrences
significantly (20). Properly designed
computer-control algorithms can
automatically adjust the amount of
oxygen according to unforeseen changes
in the heating value of the waste.
Such benefits are more difficult to
achieve with conventional air systems.
The main reason for this difficulty is
that the critical process variables of
temperature, residence time and oxygen
feed are inter-dependent when
combustion air is used. For example,
an increase in the excess oxygen level
in the combustion chamber would carry
enough associated nitrogen to lower
the temperattire and the residence time
of combustion gases and possibly cause
the loss of kiln vacuum.
If oxygen is used in place of air for
excess oxygen level control, and
auxiliary fuel and/or dynamic water
spray is used to control incinerator
temperature variations, the above
process variables can be controlled
independently without adversely
affecting the others.
TECHNICAL CONSIDERATIONS IN
THE USE OF OXYGEN IN
INCINERATION
For many years industrial furnaces
have used oxygen enrichment of the
combustion air and oxygen-fuel
burners, but with conventional
technologies a high enrichment level
may pose problems. The flame
temperature is typically high, leading
to potentially high NOx formation and
local overheating.
Various techniques exist for
introducing oxygen into industrial
furnaces. The selected technique
depends on the desired results and the
present limitations of the furnace.
Oxygen enrichment can be achieved
quite inexpensively through routine
enrichment of the combustion air.
However, general enrichment techniques
are typically limited to a level of 5
percent (26 percent oxygen in the
oxidant) due to increased flame
temperatures.
Oxygen-enriched combustion can also be
accomplished by strategically
injecting the oxygen into the furnace
using either lances or oxy-fuel
burners. Undershot lancing of the
existing air-fuel burners is a
beneficial technique for significant
production increase in most rotary
kilns. The advantage of this method
over general enrichment is that only
the segment of the flame facing the
solid bed (load) is enriched. The
bulk of the main flame shields the
refractories from the high flame
291
-------
temperature produced at the point of
oxygen impingement.
Oxygen-fuel burners, which offer
greater flexibility in heat
distribution, are applicable for
production increases as high as 100%.
Conventional oxygen-fuel burners also
tend to produce a high temperature
flame and high NOx formation. They
are used quite commonly in
applications such as auxiliary burners
in electrical arc furnaces for scrap
melting, and recently in glass melting
furnaces.
In order to overcome the disadvantages
associated with conventional oxygen
combustion technologies noted above,
the patented LINDE® Aspirator Burner
(or "A" Burner) was developed by the
Linde Division of Union Carbide
(31,32). The key feature of the "A"
Burner is that the furnace gases are
aspirated into the oxidant jets prior
to mixing with the fuel, as explained
in references 32 and 33. By
maintaining sufficient distance
between high velocity oxygen jets and
the fuel flow, enough of the furnace
gases can be aspirated into the oxygen
jet prior to mixing with the fuel so
that the resulting flame temperature
can be reduced to a value equivalent
to an air flame temperature. Gas
mixing and recirculation within the
furnace are accomplished by the mixing
effect of the high velocity oxygen
jets, which results in a uniform
temperature distribution within the
incinerator.
Nitrogen Oxides
The nitrogen oxides (NOx) emissions
due to thermal fixation of ambient
nitrogen are strongly dependent on the
flame temperature. They depend not
only on the particular burner design,
but also on the furnace temperature,
the post-flame oxygen partial
pressure, and the nitrogen content of
the furnace gases. A recent study
sponsored by the Department of Energy
(DOE)(35) obtained extensive NOx data
on various oxygen-enriched combustion
conditions. The burners tested
represented both conventional
air-fired designs and oxygen/fuel
burners designed primarily for very
high oxygen levels. The four burners
tested were:
o Bloom® Engineering Hot Air
Burner
o Maxon Kinemax® Burner
o Maxon/Corning Oxytherm® Burner
o LINDE "A" Burner
The Bloom Burner and the Maxon Kinemax
Burner are conventional air-fired
designs. These two burners were
selected for this project as
representative of a large number of
current industrial furnace
installations. These burners are not
necessarily optimized for
oxygen-enriched air firing or low
emissions. They are typical examples
of industrial burners with non-water
cooled refractory burner tiles. The
Oxytherm Burner was developed jointly
by Maxon and Corning Glass for the
application of oxygen/fuel combustion
in glass furnaces. The burner is also
a non-water cooled refractory design
with a specially designed refractory
exit which shapes the flame while
withstanding high flame temperatures.
The Oxytherm Burner tested was
designed for operation with
essentially pure oxygen, that is,
90-100 percent oxygen, as the oxidant.
The design basis for the LINDE "A"
Burner was discussed earlier.
Referring to Figure 4, this DOE study
showed that for conventional air
burners NOx emissions increased
sharply (up to 2.2 Ib NOx/MMBTU) as
the level of oxygen enrichment was
increased in the range of 35-50
292
-------
percent oxygen. However, the LINDE
"A" burner s~howed low NOx emissions
for the entire range of 35-100 percent
oxygen. The lowest NOx emissions were
achieved at 100 percent oxygen (0.01
to 0.03 Ib/MM BTU) due to the low
partial pressure of nitrogen in the
furnace. However, for incinerators
under even a slight vacuum, close to
100% oxygen would be very difficult to
achieve due to the inevitable air
infiltration.
More recently, in the EPA Mobile
Incinerator trial burn tests using the
LINDE "A" Burner (20), NOx emissions
were 0.07 to 0,18 Ib/MMBTU. In a
separate pilot test using the
Pyretron(TM) Burner by American
Combustion, Inc., EPA reported NOx
levels that averaged between
approximately 1.1 to 2.6 lb/MMBTO.
More details are given in a later
section on these results.
Flame Temperature and Furnace
Temperature Distributions
In the DOE study, the different
burners demonstrated distinct axial
radiant heat flux distributions which
changed by varying degrees as the
level of oxygen enrichment was
increased. Linde's "A" Burner
demonstrated the ability to vary its
flame patterns and thus heat flux
distributions by employing
interchangeable oxygen nozzles.
Due to the varying characteristics of
waste feeds in hazardous waste
incinerators, the control of
temperature distribution is both
challenging and important. Local hot
spots can cause potential refractory
damage, and more frequent clinker
formation or slag buildup (slagging),
which requires eventual shutdown of
the furnaces for slag removal. These
slags are masses of ash material which
have deformed and fused together.
This is a condition which occurs
approximately between 1800°F and
2500°?; however, the presence of low
melting point ash and metals in the
waste feed may cause the formation of
eutectic solutions with substantially
lower melting temperatures. Suspended
solid particles can also be melted in
a hot flame and later redeposited onto
the refractory wall.
The geometric configuration of an
incinerator, e.g. a long and narrow
rotary kiln, may also present a
special challenge in heat
distribution. Therefore, the ability
of an oxygen-enriched burner system to
control both its flame temperature and
the furnace temperature distribution
in a hazardous waste incinerator is
critical in order to achieve the
maximum benefits.
RECENT TECHNICAL AND
COMMERCIAL DEVELOPMENT
While oxygen combustion is used more
commonly in the metal industries, its
use for full-scale hazardous waste
incineration is now emerging.
ENSCO/Pyrotech tested the use of
oxygen enrichment up to 50% 02 in an
early liquid injection mobile
incinerator for the destruction of
PCBs (3-6). This system used a
water-cooled burner block to withstand
the high flame temperature (over
4000°F). Extremely high NOx emissions
(over 5000 ppm) and burner maintenance
problems were cited as deterrents to
the use of oxygen. In addition,
ENSCO's original incentive for using
oxygen was to achieve a high flame
temperature, but they found that such
temperatures were not necessary to
achieve a 99.9999% destruction and
removal efficiency (DEE) (3_7).
According to a recent review paper
(6), ThermalKEM (formerly Stablex) in
293
-------
Rock Hill, South Carolina, has been
using oxygen enrichment on their (35
MM BTU/hr) Gonsumat dual chamber
incinerator since 1986 by enriching
the combustion air to the upper
chamber (afterburner) to the 25-26% 02
level. Reportedly, this technique was
successful in eliminating high CO
excursions caused by batch loading of
solid wastes.
Tn Japan, a bench-scale oxygen burner
was developed by Nippon Sanso K.K. to
incinerate pure PCBs at about 5 Ib/hr.
Excellent destruction efficiency was
reported (>99.9999%) with a
temperature more than 2700°F (38).
In Europe, during 1985(39) Union
Carbide Europe (UCE) in conjunction
with Studiecentrum voor Kernenergie
(SCK) in Mol, Belgium successfully
replaced an air-oil burner with an
oxygen-oil burner in a small high
temperature slagging incinerator
treating radioactive and hazardous
waste, doubling the capacity. A
scale-up unit of a similar design will
be operational in 1989.
In addition, UCE has been assisting a
customer in France since July 1987 in
using oxygen enrichment at a large
commercial rotary kiln incinerator.
Originally the slow evaporation of
water from the liquid wastes led to a
long and lazy flame which extended the
length of the combustion chamber.
Oxygen enrichment of up to 24% is used
to shorten the flame length
significantly. It also helps to
stabilize the burner operation despite
product inconsistencies and
variations. A 100% production
increase of liquid waste with low to
medium heat content (900 to 5000
Btu/lb) was effected from the use of
oxygen enrichment (25).
Recently, the U.S. EPA has evaluated
separately two modern oxygen
combustion technologies using
dissimilar approaches: the LINDE
Oxygen Combustion System by the Linde
Division of Union Carbide Industrial
Gases Inc. and the Pyretron burner(30)
by American Combustion Inc. (ACI).
The Pyretron burner was tested earlier
in a bench scale rotary kiln simulator
at the EPA Research Triangle Park
(RTP) Laboratory. It was later tested
in a pilot scale rotary kiln
incinerator at the EPA Combustion
Research Facility (CRF) in Arkansas.
Test results for both programs have
been recently reported. The pertinent
conclusions about oxygen combustion
with the ACI burner from the RTP test
were(2_8):
(1) It is hard to draw
conclusions regarding the individual
effects of oxygen enrichment on
transient puffs due to the confounding
effects of temperature, oxygen partial
pressure, and total oxygen feed rates.
(2) Oxygen enrichment may cause
unacceptable emissions of NOx.
Concentrations as high as 1500 ppm
(corrected at 7 percent oxygen) were
reported for higher oxygen enrichment
levels (up to 30%).
In the CRF test of the ACI burner
conducted in late 1987(29), the waste
stream selected for the test was a
mixture of waste material from the
Stringfellow Superfund site and
decanter tank tar sludge (listed waste
K087). The resulting waste stream had
a heat content of approximately 8600
BTU/lb. The base test using air
burners had a feed rate of 105 Ib/hr.
The maximum feed rate using Pyretron
burners in both the kiln and the
afterburner was 210 Ib/hr. Water
injection was used to control
temperature in the kiln. In all tests
DREs exceeded 99.99%. The effect of
the Pyretron on transient emissions
could not be directly ascertained.
Transient emissions, as measured by CO
294
-------
spikes, were obtained during operation
of both the air burner and the
Pyretron burner. However, the pattern
of these emissions showed too much
variability to conclusively state
whether the Pyretron reduced transient
emissions. NOx data have recently
been reported for these tests (45).
NOx emission levels from this
operation of the American Combustion
burner system averaged 725 ppm to 1753
ppm at about 9% CO- and 15% 00
(approximately
NOx/MMBTU).
1.1 to 2.6
A LINDE Oxygen Combustion System was
installed in the EPA Mobile
Incineration System (MIS) in June,
1987 to replace an air burner on the
rotary kiln (20,40). The LINDE pure
oxygen burner normally supplied over
60% of the overall oxygen in the kiln,
the rest coming principally from the
kiln air leakage. The capacity of the
modified MIS was more than doubled
with the use of oxygen. The original
maximum throughput of dioxin
contaminated soil was about 2000
Ib/'hr. The system capacity was easily
increased to 4000 Ib/hr as confirmed
by certified verification tests. Based
on these test results, a RCRA Part B
permit was extended and modified for
the EPA MIS. Also, trial burn tests
of the unit with PCBs and other RCRA
listed POHCs in solid and liquid
matrixes showed DREs sxirpassing EPA
standards at solid waste feed rates of
about 4000 Ib/hr.
According to the trial burn
results(41), NOx emission levels from
the Linde oxygen system averaged
between 54.6 and 138.3 ppm at about
15% C02 and 7% 02 (0.07 to 0.18 Ib
NOx/MMBTU) which compare favorably
with the data from the previous air
system levels obtained in the 1985
trial burn(42) (between 126-166 ppm at
about 11% C02 and 7% 02 or 0.19 to
0.235 Ib NOx/MMBTU). Also for the
incineration of solid wastes with high
heat content, kiln "puffs," as
measured by CO spikes, were virtually
eliminated with the help of Linde"s
proprietary oxygen
feedforward-feedback system. It was
found that the tendency of the rotary
kiln to slag was not aggravated from
the use of the LINDE Oxygen Combustion
System when the flame pattern was
adjusted correctly. This experience
has shown that when the system is
operated with a good understanding of
the process and waste feed
characteristics, the occurrence of
slagging is minimal (43).
The EPA MIS, after its modifications,
was used to decontaminate more than 5
million pounds of soil from several
dioxin-contaminated sites in southwest
Missouri. EPA has found that it is
both more economical and more reliable
to use the modified Mobile Incinerator
equipped with the LINDE Oxygen
Combustion System than the previous
air system (43). Since the EPA Mobile
Incinerator is to date the only
incinerator in the U.S. with a "RCRA
permit to incinerate dioxin
contaminated waste, EPA used the
modified MIS to incinerate over 2
million pounds of brominated sludges
contaminated with dioxin from sites in
southwestern Missouri(43,44).
Recently, the LINDE Oxygen Combustion
System has also been demonstrated
successfully in a transportable
incinerator with a system capacity
many times greater than the EPA MIS.
Results of this installation will be
published at a later date. Also, the
Army Corps of Engineers recently
awarded a $52 million incineration
contract for a major Superfund site to
a contractor using the LINDE
Technology.
295
-------
ECONOMICS
With all the technical benefits, the
success of oxygen combustion
technologies also depends on the
economic impact of using oxygen. The
principal economic benefit from oxygen
combustion is derived from the very
significant throughput improvement.
The large fixed portion of daily
incinerator operating costs (typically
$10,000 to $30,000 per day) is spread
over a much larger quantity of waste
processed. For example, for
mobile/transportable incinerators a
doubled throughput can reduce the
allocated incineration cost of
contaminated soil by typically $100 to
$500 per ton of waste, while the cost
of oxygen required is typically less
than $50 per ton of waste incinerated.
In addition, whenever supplemental
fuel is required for incineration, the
fuel savings by using oxygen can
offset the cost of oxygen and often
show a net cost savings based on fuel
savings alone. The economics of using
oxygen to save fuel, of course, depend
on the relative cost of fuel and
oxygen. For example, at the EPA MIS a
remarkable specific fuel savings of 50
million Btu per ton of oxygen used was
demonstrated (20). Assuming a No. 2
fuel oil cost of $0.60 per gallon or
$4.40 per million BTU, the break-even
oxygen cost is $220 per ton of oxygen.
The actual cost of oxygen ranges from
about $50 per ton of oxygen produced
by a large on-site facility to about
$120 per ton for delivered liquid
oxygen.
CONCLUSION
Recent technology innovations have
demonstrated the significant
advantages of using oxygen in the
field of hazardous waste incineration.
Common concerns associated with
conventional oxygen combustion such as
local overheating and high NOx
emission are valid. Advancements in
combustion technology to overcome
those problems have been demonstrated
in extended field operations.
While the general principles discussed
in this paper would apply to most
situations, the best method to employ
oxygen varies with each individual
incineration system as well as with
the profiles of the waste feed. The
application of sophisticated
technological know-how is often
required. The process economics also
change from site to site and need to
be analyzed on a case by case basis,
although frequently the benefits of
oxygen far exceed its cost if used
intelligently. The increasing use of
oxygen combustion in commercial
incineration applications speaks well
for its economic and technical
attractiveness.
296
-------
LITERATURE CITED
(1) Oppelt, E. T., "Incineration of
Hazardous Waste," APCA Critical
Review, 80th Annual Air Pollution
Control Association, June 1987.
(2) "Hazardous Waste Incineration - A
Resource Document," The ASME
Research Committee on Industrial
and Municipal Wastes, January
1988.
(3) Kobayashi, H., "Oxygen Enriched
Combustion System Performance
Study," Phase I, Interim/Final
Report, Vol. I, Union Carbide
Corporation, Tarrytown, DOE
Contract No. DE-AC07-85ID12597.
(4) Coward, H. F. and Jones, G. W.,
"Limits of Flammability of Gases
and Vapors," U. S, Bureau of
Mines, Bulletin 503, 1952.
(5) Zabetakis, M. G., "Flammability
Characteristics of Combustible
Gases and Vapors," Bulletin 627,
Bureau of Mines, U. S. Dept. of
the Interior, Washington, D.C.
1976.
(6) McGowan, T.F., "The Use of Oxygen
in Industrial Incinerators,"
Proceedings of the International
Conference on Incineration of
Hazardous, Radioactive and Mixed
Wastes by the University of
California at Irvine, San
Francisco, California, May, 1988.
(7) Lewis, B. and Von Elbe, G,,
"Combustion, Flames and
Explosions of Gases," Academic
Press, Inc., 1961.
(8) Werley, B. L. (Editor),
Flammability and Sensitivity of
Materials in Oxygen Enriched
Atmospheres." -ASTM Special
Technical Publication 812, 1982.
(9) Union Carbide Corporation, Linde
Division Publication L-5110,
"Guidelines for Design and
Installation of Industrial
Gaseous Oxygen Piping
Distribution Systems."
(10) Compressed Gas Association,
Pamphlet G-4..4, "Industrial
Practices for Gaseous Oxygen
Transmission and Distribution
Piping Systems."
(11) deJessey, L., "Safety in Oxygen
Pipeline Systems," Proceedings,
Air Separation Plant and Oxygen
Safety Symposium, Compressed Gas
Association, Arlington, VA.,
1971.
(12) Boraelburg, H. J., "Efficiency
Evaluation of Oxygen Enrichment
in Energy Conversion Processes,"
Report No. PNL-4917, for U. S.
Department of Energy, Washington,
D.C., December 1983.
(13) Kobayashi, H., and Anderson, J.
E., "Fuel Reduction in Steel
Heating Furnaces with a New
Oxygen Combustion System: The
Linde "A" Burner System,"
presented at the American Iron
and Steel Institute, Technical
Symposium No. 9, "Energy
Conservation in the Steel
Industry," Pittsburgh,
Pennsylvania, April 28, 1983.
(14) Browning, R. A., "The Linde "A"
Burner System: Demonstrated Fuel
Savings and Even Heating With
100% Oxygen," presented at AISE
Annual Convention, Pittsburgh,
Pennsylvania, September 23-26,
1985.
297
-------
(15) Walsh, L. T., Ho, M., and Ding,
M. G., "Demonstrated Fuel Savings
and Uniform Heating with 100%
Oxygen Using Linde's "A" Burner
in a Continuous Steel Reheat
Furnace," presented at the 1986
Industrial Combustion Technology
Symposium, American Society for
Metals, Chicago, Illinois, April
29-30, 1986.
(16) Gaydos, R. A., "Oxygen Enrichment
of Combustion Air," J. of the PCA
RSO) Laboratories, Vol. 7, No. 3,
49-56.
(17) Wrampe, P. and H. C. Rolseth,
"The Effect of Oxygen upon the
Rotary Kiln's Production and Fuel
Efficiency: Theory and
Practice," Transactions on
Industry Applications, Vol.
1A-12, No. 6, November/December
1976.
(18) Burfield, L. J. and Petersen, M.,
"Oxygen Enrichment of Continuous
Reheat Furnaces to Increase
Productivity and Reduce Fuel
Consumption," International
Conference on Process Control and
Energy Savings in Reheat
Furnaces, Sweden, 1985.
(19) Mason, D. R., Rolseth, Harold C.,
"Calcining Petroleum Coke in
Oxygen Fired Rotary Kilns," Union
Carbide Corporation and Reynolds
Metals Company, 1985.
(20) Ho, Min-Da and M. G. Ding, "Field
Testing and Computer Modeling of
an Oxygen Combustion System at
the EPA Mobile Incinerator,"
JAPCA Vol. 38, No.9, September,
1988.
(21) Lee, K. C., Incineration of
Hazardous Waste, Critical Review
Discussion Paper, JAPCA, Vol. 37,
No. 9, September 1987.
(22) Oppelt, F. T., "Hazardous Waste
Destruction," Environmental
Science and Technology, Vol. 20,
No. 4, 1986.
(23) Trenholm, D., "Assessment of
Incinerator Emissions During
Operational Transients,"
International Symposium on
Hazardous and Municipal Waste
Incineration, AFRC, Nov. 2-4,
Palm Springs, CA
(24) Mulholland, J. A. and R. K.
Srivastara, "Influence of
Atomization Parameters on Droplet
Stream Trajectory arid
Incineration, USEPA, Cincinnati,
Ohio, May 1987.
(25) Lauwers, E., Union Carbide
Europe, Private Communication,
March 1988.
(26) Linak, W. P., J. D. Kilgroe, J.
A. McSorley, J.O.L. Wendt and J.
E. Durn, "On the Occurrence of
Transient Puffs in a Rotary Kiln
Incinerator Simulator, Part I,"
JAPCA, VOL. 37, No. 1, Jan. 1987.
27) Linak, W. P., J. D. Kilgroe, J.
A. McSorley, J.O.L. Wendt and J.
E. Durn, "On the Occurrence of
Transient Puffs in a Rotary Kiln
Incinerator Simulator, Part II,"
JAPCA, VOL. 37, No. 8, Aug. 1987.
(28) Linak, W. P., et al, "Rotary Kiln
Incineration: The Effect of
Oxygen Enrichment on Formation of
Transient Puffs During Batch
Introduction of Hazardous
Wastes," Proceedings of the 14th
Annual Research Symposium on Land
Disposal, Remedial Action,
Incineration and Treatment
Hazardous Waste, USEPA,
Cincinnati, Ohio, May, 1988.
298
-------
(29) Staley, L.J., and Moxtrnighan,
R.E., "The SITE Remonstration of
the American Combustion Pyretron
Oxygen-enhanced Burner," JAPCA,
Vol. 39, No.2, Feb. 1989.
(30) Gitman, G.M., U.S. Patent No.
4, 622 .,007, "Variable Heat
Generating Method and Apparatus,"
November 11, 1986.
(31) Anderson, J. E., U. S. Patent
Nos. 4,378,205 and 4,541,796,
"Oxygen Aspirator Burner and
Process for Firing a Furnace,"
March 29, 1983, September 17,
1985.
(32) Anderson, J. E., "A Low NOx, Low
Temperature Oxygen-Fuel Burner,"
Proceedings of the American
Society of Metals, 1986 Symposium
on Industrial Combustion
Technologies, Chicago, Illinois,
April 29, 1986.
(33) Ho, Min-Da and Ding, M. G.,
"Proposed Innovative Oxygen
Combustion System for the
Incineration of Hazardous Waste,"
Hazardous Materials Management
Conference & Exhibition/West,
December 3-5, 1986, Long Beach,
CA.
(34) American Combustion, Inc.,
Technical Bulletin 11-R,
"Optimizing Oxygen Utilization
for Continuous Reheat Furnaces,"
Norcross, Georgia, July 1986.
(35) Abele,, A. R., Y. Kwan, S. L.
Chen, L. S. Silver and H.
Kobayashi, "Oxygen Enriched
Combustion System Performance
Study," 9th Industrial Energy
Technology Conference, Texas A&M
University, Houston, TX, Sept.
14-18, 1987.
(36) Acharya, P., "Incineration of
Hazardous Waste on a Mobile
System," Symposium of American
Flame Research Committee, Tulsa,
Oklahoma, Oct. 1986.
(37) Acharya, P., "PCS Trial Burn in a
Modular, Movable Incinerator,"
Proceedings of Second
International Conference on New
Frontier for Hazardous Waste
Management, Pittsburgh,
Pennsylvania; September 27-30,
1987.
(38) Hirano, T. and T. Imayashi, "The
Incineration of Polychlorinated
Biphoriyl (PCBs) with Oxygen,"
IOMA Broadcasting,
September--October, 1986.
(39) Vanbrabant, R., and N. V.
deVoorde, "High Temperature
Slagging Incineration of
Hazardous Waste," Proceedings of
Second International Conference
on New Frontier for Hazardous
Waste Management, Pittsburgh,
Pennsylvania; September 27-30,
1987.
(40) Gupta, G. D., et al, "Operating
Experiences with EPA's Mobile
Incineration System," Int'l
Symposium on Hazardous and
Mxmicipal Waste Incineration,
AFRC, Nov. 2-4, Palm Springs, CA.
(41) King, G. et al, "Demonstration
Test Report for Rotary Kiln ,
Mobile Incinerator System at the
James Denney Farm Site, McDowell,
Missouri," Envirespon^e, Inc.,
Edison, New Jersey, January 1988.
299
-------
(42) Mortensen, H., et al,
'^Destruction of
Dioxin-Contaminated Solids and
Liquids by Mobile Incineration,"
USEPA report, EPA Contract
#68-03-3255, Hazardous Waste
Engineering Research Laboratory,
Cincinnati, Ohio, April, 1987.
(43) Ho, M. et al, "Long-Term Field
Demonstration of the LINDE Oxygen
Combustion System Installed on
the EPA Mobile Incinerator,"
Proceedings of the 15th Annual
EPA Research Symposium, U.S. EPA,
Cincinnati, Ohio, April 1989.
(44) Hazel, R. H., "Recent Activities
with EPA's Mobile Incinerator,"
Second Annual National Symposium
on Incineration of Industrial
Waste, San Diego, California,
Sponsored by Toxcontingency
Company, Houston, Texas, March,
1988.
(45) Waterland, L. and J. Lee, "SITE
Demonstration of the American
Combustion Pyretron Oxygen
Enhanced Burner System," Final
Report, EPA Contract 68-03-3267,
Cincinnati, Ohio.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
300
-------
Table 1
COMPARISON OF FUEL FLAMMABILITY LIMITS
IN AIR AND OXYGEN (5,6)
Flammability Limits (vol% fuel)
Lower
Fuel
Hydrogen, H2
Carbon monoxide, CO
Ammonia, NH3
Methane, CHA
Methylene Chloride
Vinyl chloride
Air
4
12
15
5
*
4
o.
4
16
15
5
16
4
Higher
Air
74
75
28
4
*
22
0,
94
94
79
61
66
/O
Not flammable in air
Table 2
APPROXIMATE FLAME TEMPERATURES OF SELECTED
FUELS AT STOICHIOMETRIC RATIOS (12)
Ail-
Fuel
Acetylene
Carbon monoxide
Heptane
Hydrogen
Methane
K
2600
2400
2290
2400
2210
4220
3860
3660
3860
3520
Oxygen
3410
3220
3100
3080
3030
5680
5340
5120
5080
4990
Table 3
EFFECTIVENESS OF OXYGEN ENRICHMENT FOR
THROUGHPUT INCREASE
Limitations
Effectiveness of
0-3/Q., -Enrichment
THERMAL CAPACITY
Air Blower
Fuel Supply
Burner
Incinerator Draft
Air Pollution Control System
Combustion Gas Residence Time
DUST CARRYOVER
MECHANICAL SOLID PROCESSING CAPACITY
HEAT TRANSFER
Effective
Effective
No Effect
Conditionally Effective
301
-------
Ill
(AHH HIS
lN3W3ainO3d 13f1d
302
-------
CM
3
0
III
5
I
III
i?5
I
0
1
ill
2
8
CM
-I
IL
e
•n
UJ
c
g
O
111
p
N39AXO
'ONIAVS
303
-------
IU
o
I
8 8 §
8
UJ
X
o
ID i-
w z
LU
J.V3H aiaxnivAV awvs aod ouva
svo anid
304
-------
2.5 -
Figure 4
SUMMARY OF NOX EMISSIONS
FUEL: NATURAL GAS EXCESS O2 = 2% (DRY)
SOURCE: DOE STUDY
FURNACE TEMP(°F)
BURNER 2000 2300 2700
5O 60 7O
O2 lnOXIDANT(°/o)
80 9O 100
305
-------
IN
FLUE GAS CLEANING BY
WET SCRUBBING AND CONDENSATION
A SPECIAL WASTE INCINERATION PLANT
J.-D. Herbell, P. Luxenberg, D. Ramke
Gesellschaft zur Beseitigung von Sondermull in Bayern
Munich, FRG
ABSTRACT
Since 1976 "Gesellschaft zur Beseitigung von Sondermilll in Bayern
mbH" (GSB) (Corporation for the Disposal of Toxic Waste in Bavaria) has
been operating an incineration plant for the disposal of solid, liquid
and pasty special waste in Baar-Ebenhausen, W. Germany. The combustion
takes place in 2 rotary kilns and 1 secondary combustion chamber common
to both. A waste heat boiler is installed downstream for heat recovery.
In 1987 an indirect flue gas cooling (condensation) system coupled
with a wet electrostatic precipltator was added to the existing flue gas
cleaning system which until then consisted of a 2-stage dry electrostatic
predpitator and a 3-stage wet srubber.
The present report describes the operational
this combined cleaning process which is patented.
results and merits of
INTRODUCTION
The special waste incineration
plant being operated since 1976 by
"Gesellschaft zur Beseitigung von
Sondermiill In Bayern mbH" (GSB)
consists of a bunker for solid waste
(900 m3), 4 bunkers for pasty waste
(4x100 m3) and storage tanks (650
m3) for liquid special waste. Inci-
neration takes place in two rotary
kilns at approx. 1000 °C, and one
common secondary combustion chamber
designed for temperatures up to
1400 °C. The maximum continuous heat
load 1s around 52 GJ/h. In the heat
recovery boiler located downstream,
the flue gases are cooled down to
approx. 280 °C, thereby producing
abt. 30 t/h of steam (270 °C,
28 bar) which is used for the gene-
ration of nearly 2 MW of electric
power.
Originally the flue gas clean-
ing system consisted of a two-stage,
dry-type electrostatic precipitator
for dust separation and a two-stage
wet scrubber neutralized by soda
lye (NaOH). With this equipment it
was possible to conform to the
emission limit values laid down in
the German Clean Air Act of 1975.
In the meantime a three-stage
scrubber has been installed, the
first stage of which is designed as
a venturi part for acid processing
without neutralization. The two
subsequent stages (radial-flowtype)
operate on neutral processing with
soda lye. Limestone (CaCOs) and
lime milk Ca(OH)2 are used for ex-
ternal neutralization of the dis-
charge from the acid scrubber
stage.
306
-------
PURPOSE
The Ebenhausen incinerator
being sited close to a military
base, the maximum waste gas stack
height had to be limited to 28 m.
Without corrective measures, which
were taken in the event, this would
have meant polluting the adjacent
area with inadmissibly high emis-
sions. On top of that the German
Clean Air Act of 1986 laid down even
more stringent emission values which
- after a grace period - had to be
complied with by existing plants,
too. As a result, GSB was forced to
retrofit additional flue gas clean-
ing facilities for its incinerator
in line with the statutory require-
ments.
APPROACH
The aim of retrofitting was to
minimize the amount of fine dust
particles as carrier of heavy metals
and toxic organic compounds passing
through the existing flue gas clean-
ing system including mist eliminator
after the wet scrubber. (Original
limit value for dust emission:
50 mg/m3, present: 30 mg/m3,
referred to standard conditions,
dry, 11 % 02). In particular the
mercury emission rate had to be
reduced considerably.
After an extensive study of
the technical possibilities and
after performing a series of tests
[1-2], the following process was
selected for further reducing the
residual pollutants:
- Cooling down the steam-saturated
flue gases after the scrubber with
the formation of mist as a result
of the microscopic dust particles
in the gas providing nucleii for
the condensation of steam.
- Passing of this mist into a wet
electrostatic precipitator and
separation.
- Reheating with hot air and remo-
val of the dry flue gases.
This process combination is
protected by letters patent [3].
Following tests in which a
partial gas stream was cooled down
from 65 °C to 20 °C after the
scrubber and considerable amounts
of heavy metals including mercury
were found in the condensate, a
pilot plant was installed at the
Ebenhausen facility. By means of a
tubular cooler the influence of
temperature reduction through di-
rect and indirect flue gas cooling
was examined and the efficiency of
condensation compared with that of
the wet-type electrostatic precipi-
tator. These investigations re-
vealed that formation of the mist
by condensation was of vital im-
portance to the final cleaning of
the flue gas.
Finally, cooling down by
about 15 °C proved to be the best
compromise between operational
economy and collection efficiency
of mercury, utilizing the latter
as a reference parameter for the
determination of plant efficiency
as also other heavy metals and the
noxious HC1 and HF gases. Under
these conditions the condensate
produced could be absorbed by the
scrubber water circuit.
In the test the collection
rates for mercury ranged between
55 % and 80 %, for the heavy metals
cadmium, chromium, copper, iron,
nickel, lead and zinc, as well as
for arsenic between 90 % and >99 %,
and for hydrofluoric acid between
94 % and 96 %. Thus the collection
efficiency to be expected for the
overall cleaning system, comprising
dry electrostatic precipitator,
307
-------
scrubber and condensation with wet-
type electrostatic precipitator, was
80 - 92 % for mercury and more than
99.9 % for the other parameters.
In contrast to the collection
efficiencies of the other heavy
metals - which were consistently
high irrespective of the initial
concentration - precipitation of
mercury was subject to wide fluc-
tuations. On the whole, initial con-
centrations of mercury of upto
0.6 mg/m3 were reduced to values
below 0.2 mg/m3 in the clean gas to
comply with the requirements of the
1986 Clean Air Act.
PROBLEMS ENCOUNTERED
Retrofitting of the indirect
flue gas cooling (condensation)
system and wet electrostatic preci-
pitator to the existing flue gas
cleaning system at the Ebenhausen
incinerator had to be performed
taking due account of the following
requirements:
- Integration into the existing flue
gas cleaning system without im-
pairing its proven service relia-
bility, in the place available,
and within a period of 3 weeks
(scheduled downtime for maintenan-
ce), with the possibility of re-
storing the original state (by-
pass) for retrofitting.
- No production of additional solid
waste or effluents.
- Replacement of the I.D. fan by a
system designed for new pressure
conditions.
- Reasonable operating costs.
- High service reliability.
RESULTS
Process Description
The flue gas cleaning system
retrofitted by Lurgi at the end of
1987 at investment costs of 7.2
million DM is shown schematically
in Fig. 1.
About 77,000 m3/h (standard,
wet) flue gas at appr. 280 °C flow
from the two-stage dry electrosta-
tic precipitator into the venturi
inlet of the three-stage wet scrub-
ber. 87,000 mVh (standard, wet)
flue gas at 65 °C leave the scrub-
ber and are cooled by 15 °C in two
carbon tubular coolers operated in
parallel. Here 10 mVh condensate
are produced, while 77,000 m3/h
(standard, wet) of flue gases at
50 °C are led into the downstream
wet electrostatic precipitator,
where the mist drops formed by
cooling are separated (appr. 0.5
m3/h). The flue gas is heated up
with hot air to 75 °C and exhausted
into the atmosphere via an induced
draft fan and a stack.
The coolers are operated with
700 m3/h of circulating water which
is recooled in a tubular forced-air
heat exchanger. Given the location
of the incinerator, river water as
cooling medium was ruled out as was
the use of an evaporative cooler.
The condensate from cooling
and the electrostatic precipitator
together with 1.5 m3/h spray water
for wetting the electrodes in the
wet electrostatic precipitator is
taken to the lower scrubber water
circuit. During flue gas cooling
10 m3/h evaporate from the upper,
first scrubber water circuit. Fur-
thermore 6 m3/h are withdrawn from
this acid water circuit for exter-
nal neutralization and heavy metal
precipitation. The total water ba-
lance is kept constant by adding
308
-------
4 m3/h fresh water to the lower wa-
ter circuit.
The energy consumption of the
retrofitted third flue gas cleaning
stage is 200 kW on an average, and
the pressure drop is 22 mbar.
Emissions
Messrs. Ecoplan performed the
acceptance tests of the retrofitted
flue gas cleaning system in May,
1988. The measuring points shown in
Fig. 1 are marked A-D. 4 series of
measurements were taken under the
following operating conditions:
- Measurement series 1:
normal pollutant-load in raw gas
feeding at app. 120 % flue gas
volume
- Measurement series 2:
normal feeding at 100 % flue gas
volume
- Measurement series 3:
increase of noxious substance
at appr. 120 % flue gas volume
- Measurement series 4:
increase of noxious substance
at appr. 100 % flue gas volume.
Each series consisted of six
half-hour measurements. The results
taken from the measuring report of
Ecoplan are compiled in table 1. The
results will be published in detail
elsewhere [4].
In summary it can be stated
that the flue gas leaving the dry
electrostatic precipitator is still
dust-laden. The acid noxious gases
correspond to the load of the pollu-
tants in the raw gas. On account of
the temperature, mercury at measu-
ring point A is in the gas phase.
In the three-stage wet scrub-
ber, besides HC1 and HF (not mea-
sured), a part of dust, and, to a
certain extent, S02 are separated.
Whereas in the acid scrubber stage
the mercury content is reduced con-
siderably, a proportion of aerosols,
volatile mercury compounds and me-
tallic mercury are still left be-
hind [5]. The condensation stage is
the governing factor in the further
elimination of mercury. The micros-
copic dust particles are also re-
moved with the condensate.
Finally the wet electrostatic
precipitator proves to be an effi-
cient fine cleaning system:
In all the emissions dust is no
longer traceable. Thus heavy metals
are no longer present in the solid
phase. HC1 and HF are far below the
limit values (50 mg/m3 and 2 mg/m3
respectively). Total mercury is se-
parated with an efficiency of 86 %
to 98 %, and the remainder is in
elementary form. The other heavy
metal contents measured are in the
gas phase as wel 1.
Further emission measurements
were performed by TUV Bayern (Tech-
nical Inspectorate of Bavaria) in
June 1988 with a series of 6 to 8
half-hour measurements. Table 2
shows the average values obtained.
The S02 measurements are part-
ly unsatisfactory and thus point to
the system limits. For this parti-
cular parameter the guarantee
values (<100 mg/Nm3 at a raw gas
load up to 1000 mg/Nm3) have not
been achieved. Summed up for all
measurements, the efficiency of
S02 separation ranges between 48 %
and 70 %, and thus is not satisfac-
tory at high raw gas concentra-
tions. In the meantime it has been
proved [6,7] that S02 can also be
separated in the acid pH phase, and
that the oxygen dissolved in the
scrubber water is the limiting fac-
tor for S02 oxidation to sulfate
in the acid range as well. There-
fore, S02 emissions can be mini-
mized by an optimum control of
excess oxygen, the temperature,
309
-------
the pH value and the mass transfer
areas (large surface due to mist
formation through condensation).
In the meantime Gesellschaft
fiir Arbeitsplatz- und Umweltanalytik
GbR haS proved in addition that or-
ganic emissions, too, can be mini-
mized with the effectively flue gas
cleaning combination described
above. The mean collection rates
achieved were as follows:
Penta- und hexachlorobenzenes 23 %
Sum of penta- to deca-
chlorinated biphenyls 33 %
Sum of tetra- to octa-
chlorinated dibenzofurans 67 %
Sum of 11 polycyclic aromatic
hydrocarbons 83 %
Sum of tetra- to octa-
chlorinated dibenzo(p)dioxins 93 %
Sum of 9 polycyclic aromatic
hydrocarbons (without the more
volatile fluoranthene and
pyrene) 99.5 %
REFERENCES
1. von Beckerath, K., 1985, Schwer-
metalle werden auch als Aerosole
emittiert, Entsorgungspraxis 6,
p.331;
1986 a, Schwermetall im Konden-
sat, Entsorgungspraxis 2, p. 102,
1986 b, Abscheidung von Reststof-
fen aus Miillverbrennungsanlagen
mit nachgeschalteten Rauchgas-
reinigungsanlagen, In: EF-Verlag
fur Energie- und Umwelttechnik,
Berlin, p. 361
2. Herbell, J.-D., Luxenberg, P.,
1987, Nachriistung der Rauchgas-
reinigung der Sondermiillverbren-
nungsanlage Ebenhausen, In: Be-
richte aus Wassergutewirtschaft
und Gesundheitsingenieurwesen
Nr. 74, Techn. Universitat Miin-
chen
3. von Beckerath, K. Luxenberg, P.,
West-German letters patent
35 20 885; Europ. Patent Appli-
cation 8610 7913.5
4. von Beckerath, K., 1989,
Entsorgungspraxis. under pre-
paration
5. Braun, H., Metzger, M., Vogg,
H., 1986, zur Problematik der
Quecksilberabscheidung aus
Rauchgasen von Miillverbrennungs-
anlagen, Mill 1 und Abfal 1, Vol,
2, pp. 62-71; Vol 3, pp. 89-95
6. Romer, R., Wunder, R., 1988,
(unpublished)
7. Bb'rger, G.-G., Jonas, A., 1988,
(unpublished)
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
310
-------
01
u
c
ro
CO
1
3
C
O
•r—
4.J
ra
in
c
0)
•a
c
o
u
-C
CO
CM
t— t
o
+J
ro
S_
O
Q.
ra
OJ
S- ra 4-1
•r- 3 ro
U 2
a.
!~ 3 J=
(U 1 Ul
4-> O) 0)
3 ra <4-
E "-^ '
S-
i 1U
"°"^
n
E
C
ro
n>
O)
ro
05
C
•t-
4J ra jQ 4J ra
O 4-1 2 W) U 4-1
a> -i- t- ra
-------
Table 1
Average values of 6 half-hour measurements each
Emission
[mg/m3n dry]
Dust
S02
HF
HC1
C total
Hg total
Hg particles
Hg (0)
Hg volatile
Zn particles
Zn volatile
Pb particles
Pb volatile
V wet
[m'n/h]
V dry
02
C02
1st series
A
68,2
114
344
5317
6,3
0,068
<0,01
0,07
3,53
1,63
1,12
0,18
B
57
85
-
-
-
0,096
<0,01
0,035
0,062
-
-
-
-
C
31,2
41
-
-
-
0,07
<0,01
0,03
0,028
-
-
-
-
D
<3
38,2
0,8
4,0
3,6
0,04
<0,01
0,035
0,012
<0,1
1,32
<0,1
1,12
2nd series
A
94,5
292
69
3002
11,4
0,31
<0,01
0,30
1,2
1,0
0,6
0,13
B
53
128
-
-
-
0,11
<0,01
0,04
0,07
-
-
-
-
C
28
112
-
-
-
0,07
<0,01
0,02
0,04
-
-
-
-
D
<3
97
0,3
9,6
5,4
0,02
<0,01
0,02
<0,01
<0,1
2,1
<0,1
0,58
Random measurements at the start of each series
76200
56500
11,1
6,3
74900
57800
11,3
6,3
64900
56500
11,2
6,3
83500
75200
13,1
5,2
68100
57800
10,8
6,4
61790
47900
12,8
5,8
62720
54680
12,2
6,0
66800
60900
13,4
5,0
Emission
Dng/m3n dry]
Dust
S02
HF
HC1
C total
Hg total
Hg particles
Hg (0)
Hg vol.
Zn particles
Zn volatile
Pb particles
Pb volatile
V wet
Dn'n/h]
V dry
02
C02
3rd series
A
131
705
520
4974
8,4
1,83
<0,01
1,82
7,6
0,7
3,7
<0,1
B
81
561
-
-
-
0,59
<0,01
0,06
0,52
-
-
-
-
C
54
407
-
-
-
0,34
<0,01
0,02
0,20
-
-
-
-
D
<3
310
0,7
2,5
3,8
0,04
<0,01
0,04
<0,01
-------
Table 2 Emission mean values of dust constituents [mg/m3 n], dry, 11 % 02
Hg total
Cd
Tl
0,123
0,00003
< 0,00009
Limiting value
^ = 0,2
As
Cr (IV)
Co
Ni
Se
Te
0,00045
0,00036
0,00009
0,00009
0,00045
0,00009
=1
Pb
Cu
Mn
V
Sn
Sb
Cr (VI)
CN-
F-
0,00171
0,00027
0,00018
0,00009
0,00045
0,00126
0,00036
0,00045
0,00478
313
-------
EVALUATION OF MECHANISMS OF PIC FORMATION IN LABORATORY EXPERIMENTS:
IMPLICATIONS FOR PIC FORMATION AND CONTROL STRATEGIES
IN FULL-SCALE INCINERATION SYSTEMS
by: Philip H. Taylor, Barry Dellinger, and Debra A. Tirey
Environmental Sciences Group
University of Dayton Research Institute
Dayton, Ohio 45469
and: C. C. Lee
U.S. EPA Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
ABSTRACT
It is generally agreed that products of incomplete combustion (PICs) are produced from hazardous
waste incinerators burning chlorinated hydrocarbons (CHCs), although to what degree they are
formed and their effect on human health and the environment remain controversial issues.
Laboratory studies indicate that the high-temperature thermal decomposition behavior of CHCs is
highly complex and generally involves the formation of many intermediate species before thermo-
dynamic equilibrium is achieved. In this manuscript, a quantitative description of these processes
for the simplest of CHCs, the chlorinated methanes, is presented. Specific emphasis is given to PIC
formation reaction channels for chloroform and carbon tetrachloride, which are often observed in
full-scale systems. Potential PIC formation and control strategies for these compounds and their
decomposition products are presented based on the rate-limiting reaction channels observed.
INTRODUCTION
Control of emissions of toxic organic compounds
from incinerators burning hazardous materials
is one of the major technical and sociological
issues surrounding the further implementation
of thermal destruction as awaste disposal alter-
native [1]. Laboratory, pilot, and full-scale stud-
ies have produced data indicating that properly
designed and operated facilities can achieve
the destruction of toxic organic waste feed com-
ponents to environmentally acceptable levels
[2-5]. However, total organic emissions from
the incinerator effluent are seldom fully charac-
terized. This has led to both scientific and public
concern over possible emissions of toxic prod-
ucts of incomplete combustion (PICs) from the
incineration of toxic materials under less than
ideal conditions [6,7].
Dellinger et al. [8,9] have presented data that
strongly suggests that the relative destructive
efficiencies (DEs) of principal organic hazard-
ous constituents (POHCs) are controlled by
high-temperature, thermal decomposition
kinetics in the post-flame regions of incinera-
tors. Dellinger et al. [8,9] and Tsang [10] have
suggested that poor mixing of waste and oxy-
gen in these regions is responsible for emis-
sions of undestroyed POHCs [8-10]. This result
may be extrapolated to PIC formation proc-
esses, as theory and experiment indicate that
oxygen-starved conditions are responsible for
most PIC emissions since the rate of POHC de-
struction is significantly slowed and the rate of
PIC formation is substantially increased [11,12].
As a first step towards developing guidelines for
predicting full-scale PIC emissions, PIC forma-
tion and destruction mechanisms for simple
chlorinated hydrocarbons (CHCs) are being
evaluated using precisely controlled oxygen-
starved laboratory flow reactor experiments [13].
Results indicate that the high-temperature ther-
mal decomposition behavior of CHCs is highly
complex and generally involves the formation of
many intermediate species. A series of elemen-
tary reactions, which often constitute a radical
chain mechanism, are believed to control the
314
-------
nature and concentration of intermediates in
this kinetically limited system. Dellinger et al.
[12-14] have previously discussed in a qualita-
tive manner general mechanisms of waste de-
struction and PIC formation for simple CHCs,
e.g., chlorinated alkanes, and more complex
CHCs, e.g., chlorinated benzenes, PCBs, etc.
In this manuscript, a more quantitative descrip-
tion of these processes is presented for the sim-
plest of CHCs, the chlorinated methanes. Spe-
cific emphasis is focused on PIC formation
pathways for chloroform (CHCI3) and carbon
tetrachloride (CCI4), which are often observed
as emissions from full-scale tests [4,5]. Poten-
tial PIC formation and control strategies for
these compounds and their decomposition
products are also presented based on the rate-
limiting reaction channels observed.
EXPERIMENTAL DESCRIPTION
The laboratory results presented in this manu-
script were generated using the Thermal De-
composition Analytical System (TDAS) [15,16].
The thermal decomposition unit of the TDAS
consists of a high-temperature fused silica
tubular reactor in which a flowing gas stream,
exhibiting a segregated flow pattern, is exposed
to temperatures as high as 1100°C for mean
residence times of 0.5 to 6.0 s. Reactor design
ensures that samples experience a square-
wave thermal pulse with a very narrow, near-
Gaussian, residence time distribution. Heated,
fused silica transfer lines (250°C) connect the
insertion chamber to the reactor and the reactor
to the effluent analysis system. The analytical
function of the TDAS is conducted by a HP
5890B programmed temperature gas chroma-
tograph and a 5970A Mass Selective Detector.
Data reduction is achieved with a HP 59970
ChemStation and the accompanying system
software which includes an on-line NIH-EPA
mass spectral library.
RESULTS
The high-temperature, gas-phase, thermal de-
composition behavior of chloromethane (CH3CI),
methylene chloride (CH2CI2), chloroform
(CHCI3), and carbon tetrachloride (CCI4) were
evaluated under the following conditions: resi-
dence time = 2.0 s, fuel/oxygen equivalence
ratio = 3.0 (oxygen-starved), reactor concentra-
tion = 2 x 10'5 moles/liter, temperature range =
300-1000°C [14]. The decomposition profiles of
CH3CI, CH2CI2, and CHCI3 were accurately
modeled using the reaction pathways and ki-
netic rate parameters shown in Table 1. The
composition of the radical pool which governs
bimoleculardestruction pathways was estimated
using a pseudo-equilibrium calculation^ ap-
proach [17].
Three-center concerted HCI elimination is the
principal decomposition pathway for CHCI3,
accounting for greater than 99% of the decay at
the 99% DE level (600°C). HCI elimination and
Table 1
Chlorinated Methane Decomposition Rate Parameters3
Cmpd
CH3CI
CH2CI2
CHCI3
CCI4
AH°300(kcal/mole)b
C-CI
82.8
80.0
77.6
70.3
HCI
Elim.
83.6
67.3
49.6
H abst.
(Cl)
-3.0
-4.0
-7.8
log k300
C-CIC
(1/s)
15.40-82.5/2.3RT[22]
16.42-79.7/2.3RT
16.19-77.4/2.3RT
16.30-70.0/2.3RT
HCI Elim.d
(1/S)
14.0-90.0/2.3RT
14.1-69.0/2.3RT
1 4.3-54.5/2.3 RT [25]
H abst. (Cl)[23]
(cm3/mol-s)
13.5-3.3/2.3RT
13.4-3.0/2.3RT
12.8-3.3/2.3RT
Footnotes:
a Pathways from left to right are unimolecular C-CI bond fission, unimolecular 3-center HCI elimination,
and bimolecular H abstraction by Cl.
b Thermodynamic data obtained from reference 18.
c Except where noted, Arrhenius pre-exponential coefficients were calculated by transition state theory. Arrhenius
activation energies were calculated by subtracting 0.5RT [24] from the reaction enthalpy.
d Except where noted, Arrhenius pre-exponential coefficients were calculated by transition state theory. Arrhenius
activation energies were estimated from a limited database of 3-center HCI elimination reactions.
315
-------
C-CI bond rupture represent approximately 95%
and 3% of the destruction of CH,CI2 at the 99%
DE level (780°C), respectively. Due to the rela-
tive strength of the C-CI bond in CH3CI as
compared to the other chloromethanes, unimol-
ecular C-CI bond rupture and concerted HCI
elimination make relatively small contributions
(42% and <1 %, respectively) to the decomposi-
tion of this compound at the 99% DE level. As a
result, H atom abstraction by Cl atoms make a
significant contribution (57.5%) to the CH3CI
decay at the 99% DE level (910°C).
The thermal decomposition profile for CCI4 could
not be accurately modeled by initiation proc-
esses involving C-CI bond fission and Cl ab-
straction reactions. The experimental data
suggested that a high-temperature POHC ref-
ormation pathway was operative for this mole-
cule. As the following paragraphs demonstrate,
construction of radical chain mechanisms which
are consistent with thermochemical considera-
tions can be used to identify which reaction
pathways are responsible for PIC formation and
POHC reformation at elevated temperatures.
Under identical experimental conditions, we
have also evaluated the yields and stability of
PICs produced from the thermal decomposition
of the chloromethanes [14]. Results indicated
that CHCIg and CCI. produced significant yields
of chlorinated products (cf. Figures 1 and 2)
while CH3CI and CH2CI2 produced primarily low
molecular weight hydrocarbons and HCI. Since
the chlorinated products are of greatest interest
due to their potential toxicity, we will now focus
our attention on determining the rate-determin-
ing pathways which produce these products
from the higher chlorinated methane precur-
sors.
Inspection of Figure 1 indicates that the major
initial organic reaction products from the CHCI3
decomposition are tetrachloroethene (C2CI4),
CCI4, and hexachloroethane (C2CI6). The initial
formation of C2CL is consistent with a mecha-
nism involving HCI elimination followed by re-
combination of dichlorocarbene biradicals
(CCI2). However, the formation of the remaining
products indicates that a more complex set of
reactions must also be kinetically significant. A
preliminary elementary reaction kinetic mecha-
nism (including reaction enthalpies and esti-
mated kinetic rate parameters) which qualita-
tively accounts for the major products observed
from CHCL (and CCI4) decomposition is given
in Table 2. The initiation, chain branching, trans-
fer and termination reactions were selected
based on the observed products and thermo-
chemical considerations as outlined by Benson
[18].
As already mentioned, the main initiation reac-
tion forCHCI3 is concerted HCI elimination (R1).
However, (R2) is also important as this chain
branching reaction produces trichloromethyl
radicals (CCI3) and an exponential increase in
the radical concentration. Two of the major
products, C2CI6 and CCI4, are produced directly
by reactions involving CCI3 radicals, (R11) and
(R3), respectively. A quasi-steady-state analy-
sis indicated that Cl atom and C2CL radical con-
centrations were small compared to the con-
centration of CCI2 and CCI3 radicals at tempera-
tures less than 650°C [18]. The Cl concentra-
tion becomes slightly greater than the CCI3 con-
centration (~50%) at higher temperatures. CCI2
recombination (R10) and CCI3 recombination
(R11) are the most significant termination reac-
tions below 650°C while CCI + Cl (R12) and Cl
+ CI + M(R13) are mostsignifcantathighertem-
peratures.
Figure 3 presents a comparison of the rates for
the potential pathways of C2CI6 consumption
(R6, R7, and R8). At low temperatures, Cl
abstraction by CCI3 radicals (R8) dominates. At
higher temperatures, C-C bond fission (R6) and
C-CI bond fission (R7) become more kinetically
significant.Further analysis indicates that C2CI,
formation is dominated by recombination of
CCL radicals (R10) at low temperatures and
(R8) followed by decomposition of the pen-
tachloroethyl (C2CI5) radical (R9) at higher tem-
peratures (> 600°C). The C2CI4 production rate
decreases at temperatures greater than 650°C
due to an increase in (R6) versus (R7) and (R8).
This is in reasonable agreement with the experi-
mental data in Figure 1. Additional comparison
of the relative rates of CCI4 formation indicates
that, for temperatures less than 575°C, CCI4 is
produced by both Cl abstraction by CCL radi-
cals (R3.R8) and radical recombination (R12).
316
-------
10
10
200 300
400
900 1000 1100
500 600 700 800
TEMPERATURE (°C)
Figure 1. Thermal behavior of CHCI3 and its associated products. See text for experimental conditions.
10
10 -H . 1 • 1 • 1 • 1 •—r
200 300 400 500 600 700 800 900 1000 1100
TEMPERATURE (°C)
Figure 2. Thermal behavior of CCI4 and its associated products. See text for experimental conditions.
317
-------
Table 2
Preliminary CHCI3 (and CCI4) Decomposition Mechanisms
Rxn Type Reaction
1. initiation CHCI3 -> CCI2 + HCI
2. branching CHCI3 + CCI2 --> CCI3 + CHCl2
3. transfer CHCI3 + CCI3 <-> CCI4 + CHCI2
4. initiation CCI4 -> CCl3 + Cl
5. transfer CHCI3 + Cl <--> CCI3 + HCI
6. initiation C2CI6 -> CCI3 + CCI3
7. initiation C2CI6 ~> C2CI6 + Cl
8. transfer C2CI6 + CCI3 <--> C2CI5 + CCI4
9. transfer C2CI5 <-> C2CI4 + Cl
1 0. termination CCI2 + CCI2 -> C2CI4
1 1 . termination CCI3 + CCI3 -> C2CI6
1 2. termination CCI3 + Cl -> CCI4
13. termination CI + CI + M->CI2 + M
Footnotes:
A H° a
" n 300
49.6
10.8
6.9
70.3
-7.8
69.1
70.0
-0.3
17.2
-117.4
-69.1
-70.3
-58.0
log kb Ref.
14.3-54.5/9° 25
13.0-14.0/9 d
13.0-10.0/9 d
13.2-5.0/9 d
16.3-70.0/9 13
12.8-3.3/9 23
11.6-11.3/9 23
17.5-68.8/9 19
16.8-69.7/9 19
13.0-5.0/8 d
12.7-5.0/8 d
15.8-16.2/8 23
12.2-1.0/9 23
13.0 d
12.7 23
13.8 23
14.9 23
a Reaction enthalpies calculated from data available in references 1 8 and 26.
b Units of reaction enthalpies and rate constants are cm3, mole,
C 6-2.303RT.
d Estimated from a large kinetic data base for similar reactions;
10 "7 -J ' i.i.
—. 10 "* - ^***
CO i .*•—**
o 10 ~9 - ^^•'"***
§ • /
o 10 -10 i //
£ i S /
W n S^ /
p 10 ] s's'
Z 10'12r 'S^
s I ^^
H • S^
< 10 -13 i
Ul :
cc :
10 -14
450 500 550
s, kcal.
see references
,
.*•*""""
^^
/^^
' .--*
/
600
18,23, and 27.
^^
\
-^-""" :
[
'•
':
;
r
•^— — Re :
— R7 r
• — R8 :
650 700
TEMPEERATURE (°C)
Rgure 3. Comparision of reaction rates of C2CI6 consumption from CHCI3 thermal decomposition. See
Table 2 for a description of the CHCI3 decomposition mechanism and rate parameters.
318
-------
At higher temperatures, (R12) dominates.
The mechanism of CCI4 decomposition (see
Table 2) producing C2CI6 and C2CI4 as the major
products is largely a subset of the CHCIg mecha-
nism. C-CI bond rupture (R4) is the initiation
step. A quasi-steady-state analysis for tem-
peratures less than 800°C indicated that Cl
atoms and CCI3 radicals were in highest con-
centration with the CI:CCI3 ratio ~ 1.5. This ratio
is expected to be maintained at higher tempera-
tures.
Figure 4 presents a comparison of rates for
three different pathways of C2CI6 decomposi-
tion (R6, R7, and R8). The results are similar to
those for CHCI3. At low temperatures, Cl ab-
straction by CCI3 radicals (R8) dominates.
Analysis of the reaction rate of an analogous
bimolecular reaction, Cl abstraction by Cl atoms,
indicated that this reaction was kinetically insig-
nifcant at all temperatures. At higher tempera-
tures, unimolecular decomposition of C2CI6 be-
comes faster than bimolecular attack. Analysis
of the relative rates of C-C bond fission (R6) and
C-CI bond fission (R7) indicates that (R6) is
nearly an order of magnitude faster at all tem-
peratures. The C2CI4 production rate thus de-
creases at temperatures greater than 650°C
due to an increase in (R6) versus (R7) and (R8).
This is in reasonable agreement with the experi-
mental data in Figure 2.
A major difference between the CHCI3 and CCI,
mechanisms is the nature of the initiation and
chain branching step. Formation of HCI fol-
lowed by rapid reaction of CCI? with the parent
compound reduces the probability of CHCI3 ref-
ormation. For CCI4, no chain branching step is
possible. Furthermore, formation of CCI4 (R8)
and the reversibility of initiation (R12) compete
with the consumption of CCI4 by other routes.
The importance of these reactions is evidenced
by the slow decomposition of CCI4 as illustrated
in Figure 2.
DISCUSSION
In this manuscript, we have not considered de-
struction pathways for C2CI4. In research being
conducted concurrently, we are investigating
reaction channels leading to the higher molecu-
lar weight products shown in Figures 1 and 2,
i.e., hexachlorobutadiene (C4CL), and hexa-
chlorobenzene (C6CI6) [12,13,19]. These prod-
ucts have also been observed in molecular
beam sampling of premixed trichloroethylene/
oxygen/argon flat flames [20]. We feel that the
10 -°
« 10"9
-------
rate of trichlorovinyl (C2CI3) radical addition to
unsaturated species such as C.CI4 may be rate-
limiting. We have recently conducted Quantum
RRK calculations [21] demonstrating that C2CI3
addition to C2CI4 producing C4CI6 is much more
rapid than the hydrocarbon reaction analogue
at combustion temperatures. We are currently
analyzing similar molecular growth reaction
channels in an attempt to identify the
mechanism(s) leading to chlorinated aromatics
from simple chlorinated hydrocarbons.
The results of this analysis indicate that knowl-
edge of the steady-state radical concentrations
is an important parameter in determining the
relative importance of various reaction path-
ways leading to PIC formation. For CHCI3, radical
recombination reactions dominate the forma-
tion of C2CI6 and C2CI4 at lower temperatures
and CCI4 at higher temperatures. For CCI4,
recombination reactions are also responsible
for C2CI6 formation at low temperatures and the
reformation of the parent compound at higher
temperatures. For both mechanisms, radical-
molecule reactions dominate the formation of
C2CI4 and higher molecular weight products at
higher temperatures.
Full-scale testing of several hazardous waste
incinerators and industrial boilers cofiring haz-
ardous wastes indicate that chlorinated alkanes
and chlorinated alkenes are frequently observed
PICs [4,5].With the exception of chloromethane,
laboratory testing indicates that these com-
pounds are unstable at high temperatures
(> 900°C), even under oxygen-starved condi-
tions. Thus, one may hypothesize that recombi-
nation of polyatomic and atomic species are
Important PIC formation pathways downstream
of the combustion zones of full-scale combus-
tors. This is due to the relative rates of radical-
molecule and radical recombination reactions
at these lowertemperatures. Most radical-mole-
cule reactions require a significant source of
thermal energy for reaction to occur. Radical re-
combination reactions, on the other hand, are
primarily dependent only on the collision fre-
quency (some recombination reactions may
exhibit very small negative temperature de-
pendencies).
To reduce the concentration of PICs in the efflu-
ent of such systems, a modification of the waste
feed composition, or a reduction in residence
times at low temperatures may be effective PIC
control strategies. A reduction in the Cl content
of the waste or, more practically, thorough mixing
of waste streams to remove high Cl pockets,
should diminish the probability of formation of
perchlorinated, polyatomic, reactive species in
the high-temperature zones. Consequently, the
probability of formation of high molecularweight
perchlorinated species at high temperature will
be reduced. Reformation of CCI4 and other
chlorinated alkanes at the exit of the high tem-
perature zone will also be limited due to the low
reactive species concentrations.
Numerical integration of the ordinary differential
equations listed in Table 2 is currently being
conducted to verify the importance of the afore-
mentioned PIC formation pathways. In the near
future, the high-temperature PIC formation
mechanisms of other CHCs will also be evalu-
ated in a similar manner. One of the objectives
of these investigations is to determine the rela-
tive importance of radical-molecule and radical
recombination reaction pathways leading to the
formation of intermediate products. Concentra-
tions of radical species from equilibrium and
reaction kinetic analyses at various tempera-
tures are also being compared. If a simple rela-
tionship between radical concentrations com-
puted by these two different approaches can be
found, it may be possible to develop an algo-
rithm based on pseudo-equilibrium to predict
the nature and yields of stable PICs produced
by recombination reactions in the low tempera-
ture zones of full-scale systems.
ACKNOWLEDGMENTS
This research was partially supported by the
US-EPA under cooperative agreement CR-
813938.
REFERENCES
1. Oppelt, E. T., 1987, JAPCA, Vol. 37, pp.
558-586.
2. Lee, K. C., 1987, JAPCA, Vol. 37, pp. 1011-
1017.
320
-------
3. Bellinger, B., 1989, in Hazard Assessment
of Chemicals-Current Developments, J.
Saxena, ed., pp. 293-337.
4. Trenholm, A., P. Gorman, and G. Junge-
laus, 1985, "Performance Evaluation of
Full-scale Incinerators," EPA-600/2-84-
181 a-e, US-EPA, Cincinnati, OH.
5. Castaldini, C., Unnash, S., and Mason, H.
B., 1985, "Engineering Assessment Re-
port-Hazardous Waste Cofiring in Indus-
trial Boilers," EPA-600/2-84-177, US-EPA,
Cincinnati, OH.
6. EPAScience Advisory Board, 1984, "Report
on the Incineration of Liquid Hazardous
Waste by the Environmental
Effects,Transport, and Fate Committee."
7. Hershkowitz, A., 1987, Technol. Review,
pp. 26-34.
8. Dellinger, B., Rubey, W.A., Hall, D.L, and
Graham, J.L., 1986, Hazard. Waste Haz-
ard. Mater., Vol. 3, pp. 139-150.
9. Dellinger, B., Graham, M., and Tirey, D.,
1986, Hazard. Waste Hazard. Mater., Vol.
3, pp. 293-307.
10. Tsang, W., 1986, "Fundamental Aspects
of Key Issues on Hazardous Waste Incin-
eration," ASME Publication 86-WA/HT-
27.
11. Graham, J.L., Hall, D.L, and Dellinger, B.,
1986, Environ. Sci. Technol., Vol. 20, pp.
703-710.
12. Taylor, P.H. and Dellinger, B., 1988, Envi-
ron. Sci. Technol., Vol. 22, pp. 438-447.
13. Tirey, D. A., Taylor, P. H., and Dellinger,
B., 1989, "Products of Incomplete Com-
bustion from the High Temperature Py-
rolysis of the Chlorinated Methanes," Sym-
posium on Emissions from Combustion
Processes: Origin, Measurement and
Control, in press.
14. Dellinger, B., Taylor, P. H., Tirey, D.A.and
Lee, C. C., 1988, "Pathways of PIC For-
mation in Hazardous Waste Incinerators,"
14th Annual Hazardous Waste Research
Symposium, EPA/600/9-88/021, US-EPA,
Cincinnati, OH, pp. 289-301.
15. Rubey, W. A., 1980, "Design Considera-
tions for a Thermal Decomposition Ana-
lytical System," EPA 600/2-80-098, US-
EPA, Washington, D.C.
16. Rubey, W. A. and Games, R. A., 1986,
Rev. Sci. Instrum., Vol. 56, pp. 1795-1798.
17. Reynolds, W. C., 1986, "STANJAN Equilib-
rium Program, Version 3.0," Department
of Mechanical Engineering, Stanford
University, Stanford, CA.
18. Benson, S. W., 1976, Thermochemical Ki-
netics, Second Edition, John Wiley and
Sons, New York, NY.
19. Dellinger, B., Taylor, P. H., Tirey, D.A. and
Lee, C. C.,1989, "High Temperature Py-
rolysis of C2-Chlorocarbons,"Proceedings
of the 198th National ACS Meeting, Sym-
posium on Incineration, Miami Beach, FL,
September 10-15.
20 Senkan, S. M. and Chang, W-D., 1989,
Envir. Sci. Technol., Vol. 23, pp. 442-450.
21. Dean,A. M., 1985,J. Chem. Phys., Vol. 89,
pp. 4600-4608.
22. Weissman, M. and Benson, S. W., 1984,
Int. J. Chem. Kinet., Vol. 16, pp. 307-333.
23. Kondratiev, V. N., 1972, "Rate Constants
of Gas Phase Reactions,"NSRDS COM-
72-10014, National Bureau of Standards,
Washington, D.C.
24. McMillen, D. F. and Golden, D. M., 1982,
Ann. Rev. Phys. Chem., Vol. 33, pp. 493-
532.
25. Schug, K. P., Wagner, H. G.g., and Zabel,
F., 1979, Ber. Bensenges Phys. Chem.,
Vol. 83, pp. 167-175.
26. Stull, D. R., Westrum, E. F., and Sinke, G.
C., 1969, Chemical Thermodynamics of
Organic Compounds, John Wiley and
Sons, New York, NY.
27. Kerr, J. A. and Moss, S. J., eds., 1981,
"Handbook of Bimolecular and Termol-
ecular Gas Reactions, Volume II," CRC
Press, Boca Raton, FL.
321
-------
Disclaimer
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency peer and administrative review policies and approved for
presentation and publication.
322
-------
COST EFFECTIVE REMEDIATION THROUGH VALUE ENGINEERING
: Patrick F. O'Hara
Kenneth J. Bird
William C. Smith
Paul C. Rizzo Associates, Inc.
Pittsburgh, PA 15235
ABSTRACT
The recommended remedial action alternatives for a site, whether in
the form of a Record of Decision (ROD) or consent order and agreement,
allow a certain degree of flexibility in which several design options
may be permissible as long as the intent of these decisions is
maintained. A value engineering (VE) review allows the A/E to explore
that flexibility. A VE review is a less time consuming and structured
process than a formal VE study.
A VE review examines the major cost components of recommended
alternatives to determine if:
• More cost-effective remedial actions can be
employed, thus achieving the same level of
remediation at a lower overall cost.
• The components of the remedial actions are
adequate from an engineering and regulatory
standpoint and, if not, what the cost impact of
meeting these criteria would be.
• The costs presented in the ROD for specific
alternatives are realistic and what the
potential impact of a more accurate estimate
might be.
This paper presents the advantages and disadvantages of VE reviews
for environmental remedial design projects. Although VE reviews are
routinely performed for government-lead remediation projects, they are
not routinely used in the private sector. This paper, through examples,
makes a case for more use in the private sector.
323
-------
INTRODUCTION
The process by which the sites
having environmental problems are
remediated is somewhat complex and
is contingent upon many variables.
Principal variables are:
• Is the site on the National
Priorities List (NPL)?
• Is the site being remediated in
cooperation with or under the
auspices of a regulatory agency?
• Is the remedial action the
responsibility of a single
entity or a group of entities
acting in concert?
This paper explores the role of
value engineering in determining
cost effective remedial programs. A
characteristic common to all
remedial programs is that they tend
to cost more than those funding the
programs care to pay. A goal of all
remedial action programs should be
to provide maximum environmental
benefit in terms of public health
risk reduction, environmental
impairment reduction, and future
resource utility for the most
reasonable cost. Remedial actions
comprise a gammet of studies,
technologies, and traditional design
and construction practices.
Value engineering is hereby
defined as the process by which one
assesses the goals of a specific
remedial action program, and in a
very broad sense, evaluates and
assesses a variety of approaches to
fulfill those goals. Value engi-
neering considers both capital costs
and long-term costs associated with
a variety of remedial approaches and
provides a consistent presentation
of cost information such that the
appropriate decision makers for a
given project are presented with the
data necessary to select the optimal
remedial alternative.
Value engineering sometimes
results in studies to acquire data
that is necessary for minimizing or
even controlling overall project
costs. Value engineering reviews
are, in themselves, relatively
inexpensive and tend to cost less
than one percent of the overall
project budget. Sometimes studies
and design changes which result from
value engineering reviews are
significant in terms of their
expenditures, however, these more
expensive studies and assessments
are only undertaken when the
potential cost savings of the
overall remedial effort are large in
comparison to the cost of the study.
The following sections of this
paper describe the remedial process
for both NPL and non-NPL sites, and
the value engineering process
itself. This paper also presents
two case histories of the adaptation
of that value engineering process to
real-world remedial programs and the
resulting cost savings obtained via
this process.
THE REMEDIAL PROCESS
Sites are assessed for cleanup
through a variety of mechanisms.
Many sites that have been placed on
the NPL have been assessed and
scored by the U.S. EPA and
cooperating state agencies. Other
sites have been placed on individual
state priority lists and slated for
cleanup under the auspices of the
appropriate state agencies. Many
cleanups come about as a result of a
property transfer or simply an
assessment performed on a voluntary
basis by the owner or occupant of a
given property. How the remedial
process progresses is somewhat
contingent upon who discovered the
site, whether or not it is on the
NPL, whether or not it is on a state
324
-------
priorities list, and whether or not
a regulatory agency is involved in
the process.
If a site is on the NPL, its
remediation is normally accomplished
through a process which involves a
remedial investigation/feasibility
study (RI/FS). During the RI/FS,
the site is investigated and a
series of alternatives, including a
no-action alternative, are assessed
from the standpoint of both public
health risks and economics. The
relative advantages and
disadvantages of these alternatives
are assessed in a manner that
permits decision makers (normally
the U.S. EPA Regional Administrator
and staff members) to readily review
environmental benefits achieved
versus expenditures for a series of
alternatives. This portion of the
RI/FS process is actually quite
analogous to a value engineering
study, however, there is normally
insufficient data available during
the RI/FS process to assess in depth
the specifics of design/remedial
construction.
For sites that are not on the
NPL, there is often no formal
mechanism for assessing the
environmental benefits achieved
versus expenditures in the selection
of the given remedial alternative.
If a regulatory agency is involved
in the remedial process for a non-
NPL site, negotiations amongst the
parties which will fund the study
and the appropriate agencies are
generally undertaken informally
until a series of remedial actions
are agreed upon or mandated.
For remediations which are not
subject to regulatory scrutiny, the
owner or other responsible party is
in the position of making an
assessment regarding the degree of
risk reduction (environmental
benefit), the amount of expenditures
which may be employed, and often
assess for themselves the measures
most responsive to their individual
needs. It is in this situation that
a value engineering approach is
oftentimes most beneficial, as
owners or other responsible parties
certainly have a vested interest in
achieving: Da specified amount of
environmental benefit/risk reduction
at the least possible cost and/or,
2) achieving a maximum amount of
environmental benefit/risk reduction
at a given cost.
The points at which the value
engineering process can be applied
to the evaluation of remedial
alternatives are a function of each
project. The benefits of value
engineering can be achieved at
different points in the remedial
planning process.
THE TRADITIONAL VALUE ENGINEERING
PROCESS
The agency which is most
familiar with the value engineering
process is the U.S. Army Corps .of
Engineers. The Corps has
traditionally applied value
engineering concepts to both its
design and construction projects
throughout the past ten years. The
traditional value engineering
process is outlined in the U.S. Army
Corps of Engineers' ENG Form 3986-R,
dated September 1985. The process
is described as a five-phase process
consisting of the following:
• Phase I - Information
Acquisition; In this phase,
pertinent facts are established,
the goals of the remedial
process are clearly defined, the
framework of the most likely
remedial alternatives is
established, and an appropriate
325
-------
team is assembled to perform the
study. The key questions which
are raised during this phase
regarding each component of the
series of alternatives
identified are:
- What is it?
- What does it do?
- What must it do?
- What does it cost?
These questions are answered for
the various components of the
remedial alternatives in
question and the information is
recorded.
Phase II - Critical Review; A
group of individuals critically
reviews the information recorded
in Phase I. The techniques of
this critical review are:
- The use of creative thinking.
- The elimination of regulations
as a design constraint.
— The elimination of unnecessary
and redundant components.
- The simplification of existing
components and approaches.
- The modification and
combination of alternatives
into a program or series of
programs.
Key techniques employed in this
phase are:
- The use of good human
relations.
- Mot permitting the most vocal
and outspoken person on the
value engineering team to
dominate conversation.
- Requiring all participants to
answer the following question
for each alternative or
component which has been
established in Phase I: "What
else will perform the basic
functions?11.
Again, as in Phase I, the
results of the discussions are
recorded in a systematic manner.
1 Phase III - Analysis; At this
point, each series of remedial
alternatives is criticized as an
entity. Combinations of
alternatives have been
established and the advantages
and disadvantages of those
combinations are assessed in
Phase III. The techniques which
will be utilized are:
- The use of realistic cost
references.
- Critical review of the ability
of each series of alternatives
to fulfill project goals.
- Solicitation of an expert
opinion outside the group, if
necessary.
The key questions to answer
during Phase III analyses are:
- What does each feasible series
of alternatives cost, both on
a capital basis and an O&M
basis?
- Will each series of alterna-
tives realistically perform
the project objectives, i.e.,
satisfaction of regulatory
constraints, realization of
true environmental benefit,
etc.?
During Phase III, we are
assessing which alternatives
will fulfill a basic level of
project acceptability from a
remedial standpoint and the cost
of each of those alternatives.
Phase IV - Critical Review; The
best alternatives assessed in
Phase III, in terms of cost effec-
tively complying with project
goals, are critiqued again using
the following techniques:
326
-------
- Specifics of each alternative
are looked at in more detail
in terms of design
practicality and compliance
with regulatory constraints.
- Evidence supporting the
suitability of each series of
alternatives is gathered and
critically assessed, i.e.,
regulatory precedent, past
project successes/
failures, etc.
- Through refining the aspects
of each series of remedial
alternatives, modifications
are made to the alternatives
as though the value
engineering team were spending
its own money in performing
the remedial program.
At end of Phase IV, after
establishing the specifics of a
series of remedial alternatives
and evaluating how each series
of alternatives may be employed
as economically as possible, the
key question to ask is; "Will
these alternatives now meet all
necessary project requirements?"
This question should be answered
in writing for each of the
remaining alternatives at the
conclusion of Phase IV.
• Phase V - Presentation; Phase V
is a specific presentation, in
writing, of the series of
alternatives which have resulted
from the previous four phases.
This presentation is as follows:
- Identification of the remedial
program evaluated.
- Brief summary of the problem.
- Description of early
impressions of the remedial
alternatives.
- Cost of original alternatives.
- The results of the critical
reviews of early phases on
those alternatives.
- Identification of alternative
approaches.
- Cost data associated with
those approaches.
- Advantages and disadvantages
of each group of remedial
alternatives.
- Sketches of proposed
approaches.
- Problems and costs associated
with implementation of each
alternative.
- Assessment of cost savings
which could be realized by the
implementation of
alternatives.
- A summary statement
recommending the most
appropriate alternative.
This process, as designed and often
implemented by the U.S. Army Corps
of Engineers, represents an
innovative framework for assessing
remedial alternatives. It was
originally developed for non-
remedial construction projects but
has been employed by the Corps and
by other parties in the assessment
of environmental remediation
projects.
CASE HISTORIES
The following case histories
are the personal experiences of the
authors and those of others in the
authors' organization. These case
histories highlight how the value
engineering process, as previously
described, was applied to specific
environmental remediation programs
and the resulting changes in
remedial design and remedial
construction from the process.
Caae History No. 1
Case History No. 1 involves the
remediation of an NPL site. The
project was a federal lead project
in which an RI/FS had been, done in
327
-------
the early 1980s. The Record of
Decision or selection of the most
appropriate remedial alternative
resulting from the RI/FS process was
issued in early 1985. The proposed
remedial alternative for the site
included a variety of remedial
techniques which included excavation
and removal of highly-contaminated
waste materials, retrofit of the
groundwater/leachate collection and
treatment system, a surface water
management program, and a 26 acre
cap for materials which were not
highly contaminated and, based upon
the Record of Decision, would remain
in place.
One aspect of the Record of
Decision was the specification of
the type of cap which would be used
to prevent surface water and
precipitation from infiltrating into
solid waste and municipal refuse.
Based upon the data evaluated during
the RI/FS process, which did indeed
look at the environmental benefit
achieved versus, in a qualitative
manner, the amount of environmental
benefit and risk reduction achieved,
a specific cap design had been put
forth. This cap design, however,
was not critically reviewed during
the RI/FS process to the degree it
would have been during a value
engineering study. Individual
components of the cap, with respect
to their ability to fulfill project
objectives versus their capital and
O&M costs, were not assessed in
significant detail. During the
remedial design process, a remedial
design contractor was asked to
perform an overall value engineering
study using the techniques
previously described.
During the design process the
actual configuration of the capping
system as well as the individual
materials used in the capping system
were assessed. During this review,
additional data was solicited on the
cost of the various materials for
the design as specified in the
Record of Decision. Cost and
performance characteristics were
also solicited for alternative
materials and configurations which
had the ability to fulfill
performance objectives as well as or
better than the materials selected
and identified in the Record of
Decision.
As a result of this value
engineering review, it became
apparent that the substitution of a
synthetic geomembrane for a two-foot
thick clay cap component, as
depicted in the Record of Decision,
had both performance advantages as
well as significant potential cost
savings. It was therefore
recommended that the plans and
specifications for this project be
prepared with provisions for both a
clay capping system and a
geomembrane capping system and that
the construction contractors be
required to provide quotes on both
alternatives.
Upon receipt of the competitive
bids for the overall remedial
program, all eight bidders indicated
that they would install the
synthetic capping system at a lower
price than the capping system stated
in the Record of Decision. The low
bidder on the project had a price
for the synthetic capping system
which was $790,000 less than the
capping system depicted in the
Record of Decision.
The project was awarded with
the government selecting the
alternative of the synthetic capping
system. Actual savings to the
government for construction are
$790,000. The actual cost of
undertaking the overall value
engineering study, specific portions
328
-------
of the value engineering study
related to capping system compon-
ents, and the preparation of an
alternative design specifications
and bid sheet was less than $50,000.
Net savings to the government are
approximately $740,000.
Case History No. 2
At another NPL site, the Record
of Decision mandated that a leachate
trench be constructed around the
periphery of a land disposal area.
The basis of this Record of Decision
included data indicating groundwater
contamination in the vicinity of the
landfill, which was to be
controlled, collected, and treated
as part of the remedial action
process. The perimeter collection
trench was defined as an alternative
and its estimated depth was based
upon several monitoring wells
constructed in the vicinity of the
landfill. These monitoring wells
indicated that groundwater impacts
had indeed occurred at the landfill
periphery. These wells, however,
were constructed as open—hole
monitoring wells, which at some
points extended to a depth in excess
of 70 feet. In addition, it was
apparent that these wells
intersected several hydrogeologic
zones such that samples from these
weils were actually a mixture of
groundwater from several zones. The
relative contribution of these zones
to the contamination was impossible
to determine.
During remedial design,
problems were encountered with the
Record of Decision collection
option. Simply implementing the
Record of Decision without critical
review would have resulted in
leachate collection/groundwater
collection trenches being installed
at depths of up to 70 feet around
the periphery of this landfill, and
that the installation be performed
through 50 feet of hard rock at some
points. Capital costs associated
with the installation of this system
are significant, the techniques for
rock excavation are somewhat limited
due to the desirability of not
furthering fractures in the bedrock
system.
During an initial value
engineering review, performed in
accordance with the techniques
described previously, it was noted
that the Record of Decision
recommended collection trenches to a
depth of approximately 70 feet based
upon data obtained from monitoring
wells at that depth. However, it
had not been established that the
contamination extended to a depth of
70 feet. The manner in which the
monitoring wells were constructed
made it impossible to determine with
any degree of precision the
hydrogeologic zones which actually
exhibited contamination.
It was, therefore, recommended
that the existing monitoring wells
be properly abandoned and that they
be replaced by wells that monitor
discrete hydrogeologic zones such
that the vertical extent of
contamination could properly be
assessed. A rational assessment of
the vertical extent of contamination
was imperative in properly designing
a groundwater/leachate collection
trench that would fulfill the goals
of this project.
A subsurface investigation
program and monitoring well
installation program are currently
in progress involving the
installation of wells at certain
critical locations which will be
capable of obtaining samples from
vertically discrete hydrogeologic
zones such that the vertical extent
of contamination can properly be
evaluated. Should this program
329
-------
indicate that the preponderance of
contamination is actually from zones
substantially more shallow than 70
feet, as is currently believed, a
remedial design will be undertaken
which will have a leachate
collection/groundwater collection
system that is substantially more
shallow than 70 feet and may in fact
be constructed primarily in soil as
opposed to rock. Should this
approach be justified through the
investigative programs currently
underway, potential construction
savings are approximately $2.7
million. The cost of performing
these investigations and undertaking
these studies is $92,000. The
potential cost savings are obviously
highly significant with respect to
the expenditures incurred to gather
this information.
panacea regarding cost effective
remedial planning. However, based
upon real-world experiences at real-
world sites achieved in the last
three years, the potential for
significant cost savings has indeed
been demonstrated. These savings
could be accomplished both at sites
on which little regulatory
involvement is anticipated up to and
including NPL sites for which the
Record of Decision has already been
signed. The rational engineering
design and remedial planning process
provides ample opportunities for the
incorporation of value engineering
techniques at a variety of points in
the remedial planning process, from
initial project conception through
remedial construction and
implementation.
SUMMARY
Using value engineering
techniques, as originally developed,
fostered and propagated by the U.S.
Army Corps of Engineers, remedial
planners are afforded the
opportunity to achieve highly-
significant cost savings with
relatively limited expenditures.
Two case studies have been presented
in which the sum total of savings
approach $3.5 million and the sum
total of expenditures to achieve
those savings is less than $140,000.
It is not the intent of this
paper to indicate that the value
engineering process in itself is a
ACKNOWLEDGMENTS
The authors gratefully
acknowledge the input of their
colleagues at Paul C. Rizzo
Associates, Inc., several clients,
in particular the U.S. Army Corps of
Engineers, for encouraging the use
of these techniques and being •
willing to both fund the value
engineering studies and to endorse
their results, even after Records of
Decision have already been signed.
REFERENCES
U.S. Army Corps of Engineers, ENGR
Form 3986-R, September, 1985. "Value
Engineering Work Book".
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
330
-------
THE IMPORTANCE OF EXPOSURE PATHWAYS
IN TOXIC SUBSTANCE CONTROL:
A CASE STUDY FOR TCE AND RELATED CHEMICALS
H. C. Yeh and W. E. Kastenberg
Mechanical, Aerospace and Nuclear Engineering Department
University of California
Los Angeles, CA 90024-1597
ABSTRACT
Most current studies for the transport and sorption of toxic contaminants
have focused on chemical, physical, and biological activity in the saturated
zone of groundwater systems. However, volatile pollutants that can readily move
between the aqueous and vapor phases in saturated and unsaturated zones have made
the prediction of their fate and transport very difficult. In the soil
environment, some categories of groundwater pollutants are recognized as
biotransformable and/or biodegradable. Risk assessment on such a family of
chemicals has not been thoroughly studied, except for some individual members
of the family. In this paper exposure pathway analysis and its impact on risk
assessment for a whole chemical family will be treated.
Various computer models for pollutant transport and transformation in the
multimedia environment have been investigated with particular emphasis on the
groundwater system. An exposure pathway analysis based on multimedia
environmental transport models is developed, and will be used for risk
assessment/management. A case study of Tricloroethene (TCE) and its related
chemicals (in a biotransformable sense) has been conducted. The GEOTOX,
FEMWASTE, and BI01D computer codes have been used for the transport and
biodegradation calculations so as to determine the environmental concentrations,
the most important exposure pathways and a subsequent health risk assessment.
Several interesting findings have been explored and will be presented in this
paper.
INTRODUCTION
The growing concern with
potential problems posed by toxic
substances in the environment has led
to pollution control legislation at
the state and federal level. Such
regulation should be based on
assessment of the relative risks
associated with emission and the costs
of controlling them. Because of the
complexity of the environmental
transport/transform mechanisms
regarding the emission of toxic
substances, the determination of
control costs as well as the
assessment of health risks are
generally difficult and are highly
331
-------
uncertain.
Toxic pollutants, when released
to the environment, are distributed
among environmental media (the air,
water, and soil), and subsequently
enter the human body through various
exposure pathways. The potential
health effects of these contaminants
will depend on the dose received as
a result of the exposure. In order
to assess the risk, a thorough
understanding of how the contaminants
behave in various media and their
possible pathways to humans is
necessary.
TCE and related chemicals (in a
biotransformable sense) such as vinyl
chloride(VC),perchloroethylene(PCE),
and l,l-dichloroethylene(DCE) were
selected for the present study because
PCE and TCE are major industrial
solvents used for degreasing metal
parts and electronic components. They
are among the most common volatile
organic compounds(VOCs) detected in
groundwater and reported all over the
world.
PURPOSE
A framework for risk assessment
regarding toxic wastes is being
developed by the risk assessment group
at UCLA with support from the National
Science Foundation Engineering,
Research Center. Research underway
has as its focus, the improvement and
development of options for toxic waste
control and methods for their
evaluation. One objective of the risk
assessment group is to develop models
and methods to address exposure
pathways. The methodology being
developed is applied to an exposure
assessment of trichloroethylene(TCE)
and its related chemicals. The case
study represents a 'typical' situation
and demonstrates the developed
methodology. The results of this
study can then be used as a risk
management tool to provide information
concerning site remediation.
APPROACH
Risk Assessment
In general, prediction of the
following is necessary in order to
assess the risks due to toxic wastes:
a) Chemical and physical state of the
source, its rate of release and its
magnitude.
b) Dispersion of the pollutant from
the source in air, ground water, and
other media given the meteorologic
and hydrogeologic conditions of site.
c) Assessment of pollutant exposure
pathways, ultimate human uptake, and
predicted health effects.
The risk assessment and risk
management methodology being developed
can be broken into three major parts:
1. Source/receptor (risk) assessment
-- source and site characterization,
selection of appropriate transport
and transformation models in the
multimedia environment, and
development of exposure pathway and
dose/response models for health risk
assessment.
2. Option generation -- based on the
source/receptor assessment and
available technologies such as
biodegradation, air stripping, pumping
with hot patch, etc., recommend
options for remediation, mitigation
or interdiction.
3. Risk management -- use goal-driven
criteria for ranking alternatives
which includes uncertainty in models
and data, probability of success of
332
-------
the technologies being applied, and
socio-economic and socio-political
considerations.
A flow chart of the risk
assessment/management process is
depicted in Figure 1.
Exposure Pathways Analysis
In this paper the major tool used
for exposure assessment is the GEOTOX
code [1] which is based on simplified
multimedia modeling. Therefore, a
composite multimedia modeling
approach, using independent single-
medium models for each medium, couples
inputs and outputs in order to provide
the necessary interactions between
models. For example, BI01D [2] is
used for the biodegradation
calculation in the unsaturated zone,
and FEMWASTE [3] is used for the
transport calculation. This approach
serves to validate the simplified
models employed in the GEOTOX code.
However, a large number of model
parameters are required and often
unavailable. A best-estimate method
has been employed to resolve this
problem. The resulting environmental
concentrations are used to determine
the most important exposure pathway
based on rate calculations for each
pathway (i.e. inhalation, dermal
contact and ingestion).
There are eight compartments used
in GEOTOX to model the environment.
They are air, air particle, upper
soil, lower soil, ground water,
surface water, biota, and sediment.
Exposure is expressed as the rate at
which a quantity of material comes in
contact with the human system and is
estimated from various pathways based
on daily intake per unit body weight
averaged over the population. The
general model is expressed as:
E = 1/70 SUM Ii(t), i = 1, p
where E = child, adult, or lifetime
exposure; p = number of pathways; I±
= the daily intake by pathway i. The
exposure pathways considered in GEOTOX
are inhalation, drinking water
ingestion, biota ingestion, meat and
dairy product ingestion, fish
ingestion, soil ingestion, and dermal
absorption. For simplicity, the
exposure rate for each pathway is
calculated by multiplying each
exposure factor and the concerned
environmental concentration^) . For
example, the drinking water exposure ,
rate is determined by groundwater and
surface water pollutant concentrations
and an exposure factor which relates
the age group and averge daily water
intake. More detailed information can
be found in McKone's report [1] .
A Case Study -- TCE and Its Related
Chemicals
There are two situations of
interest for PCE, TCE, DCE, and VC
discharges from a hypothetical site
located in the San Diego region. They
are described as follows:
1) PCE and its bio-decayed daughter
TCE are treated in a coupled sense.
This means that they are related by
a bio-decay relationship in the
simulation.
2) PCE, TCE, DCE and VC are treated
as separate chemicals, i.e. no bio-
decay relation is considered.
A similar case study for TCE alone
was conducted by Cohen and Ryan [4]
for the prediction of environmental
concentration. A six-compartment
multimedia model was used to simulate
transport behavior in the San Diego
basin.
PCE, TCE, DCE and VC undergo
anaerobic transformation [5, 6, 7] in
the unsaturated soil water zone. The
bio-decay relationships between each
333
-------
chemical compound and their bio-decay
half-life are shown in Figure 2. The
important physical and chemical
properties of PCE, TCE, DCE and VC are
listed in Table 1. Viruses are
believed to be solely responsible for
the biodegradation reactions. The
distance viruses can travel, depending
on the nature of the soil and other
site-specific factors, was reported
to be as far as 67m vertically and
408m horizontally from land
application [8]. Thus, the media that
should be considered are upper soil,
lower soil, groundwater, and sediment.
This family of biodecay product and
its effect on exposure pathway and
health risk is considered. Better
information can then be provided for
overall health risk assessment and
management. Also, the environmental
concentrations predicted by Cohen and
Ryan and some field measurements [9,
10] are used to compare the results
obtained using the GEOTOX simulation.
The environmental settings, shown
in Table 2, are compatible with Cohen
and Ryan's compartmental system for
the purpose of easy comparison. The
removal and transformation rates in
some compartments for PCE, TCE, DCE
and VC, based on a best-estimate
approach are shown in Table 3 and 4
respectively. The San Diego region
is assumed to be 400 km square. The
atmospheric height and depth of the
water are 700m and 10m respectively.
The average temperature is taken as
20 degree C, the average humidity is
7.8e-6 kg/L, and the yearly average
wind speed is 5 m/sec. The partition
coefficients for PCE, TCE, DCE, and
VC are summarized in Table 5. The
source strength is chosen to be the
same as that in Cohen and Ryan's work
[4] in the air compartment, which is
1.54xlO'9 mole /hour/m3.
Our approach is to obtain
preliminary estimates of exposure to
TCE, PCE, DCE, and VC contaminants
and will be easily adapted to site-
specific assessments.
PROBLEMS ENCOUNTERED
When PCE or TCE undergoes
anaerobic biodegradation they will
produce daughter products including
various dichloroethene(l,1, 1,2 cis,
1,2 trans) and vinyl chloride which
have been known as possible
carcinogenic agents (probable human
carcinogen [11, 12]). To simulate
their behavior in the environment,
each daughter product requires the
solution of an additional equation
where the decay sink of the parent
chemical equation provides a source
term in the daughter product equation.
However, in reality it is possible for
PCE, TCE, DCE, and VC to metabolize
in biota and transform to other toxic
products [13] . This leads to a higher
removal rate in biota and a compound-
ing effect on exposure and health
impacts. Such a metabolic enhancement
of removal of contaminants can be
modeled if the metabolic rate
constants are estimated correctly.
Also, in the event of rain, scavenging
will have a significant effect on the
concentration in the upper soil and
surface water compartments especially
when a significant emission of TCE,
DCE, or PCE occurs. A further
investigation of these issues will be
conducted in the near-term.
Vapor phase diffusion and
sorption of the common volatile
pollutant TCE, has been studied by
Marrin [14] and Peterson [15].
However, inclusion of gas phase
transport has not been found in the
available multimedia transport codes.
Also, one known deficiency in the
GEOTOX model is its simplified
modeling of intermedia transport
processes of non-particulate
pollutants. A proper correction of
334
-------
this problem and the ignored pathway
will be investigated. Guidelines for
modification of the existing computer
code will be provided in the
continuation of present research.
RESULTS
The results of the simulation in
terms of environmental concentration
and exposure rates are presented in
Tables 6 through 8 for two simulated
situations. The environmental
concentrations of TCE predicted by
GEOTOX with and without considering
the effect of biodecay are compared
to Cohen and Ryan's results and other
studies Table 6. A comparison of
exposure rates with and without
considering biodecay for adult, child
and lifetime average are shown in
Tables 7 and 8. To compare relative
exposure and ranking, the reference
safe dose corresponding to a lifetime
risk of 1 in a million is used. The
relative ranking for PCE, TCE, DCE,
and VC is shown in Figure 3.
Discussion
The environmental concentrations
predicted by the GEOTOX code have less
than an order magnitude deviation from
field measurement as listed in Table
6. Clearly, further refinement of
each compartment model will require
more studies of intermedia mass
transport processes and multimedia
field data.
As a result of this study, we
conclude that vapor transfer of PCE,
TCE, DCE, and VC in various media such
as upper soil, surface water, lower
soil, and ground water are the most
important transport routes. This is
due to their high Henry's law constant
which accounts for higher mobility in
the vapor phase and a strong tendency
towards the air compartment. In the
exposure analysis, we found that
drinking water exposure rate is lower
for PCE(one order of magnitude) and
TCE(two orders of magnitude) when
biodegradation is considered. As for
DCE and VC, there is little change
for both cases. Inhalation and
ingestion of contaminated drinking
water and food are the principal
routes for all chemicals under
consideration, which partially
coincide with a previous study by EPA
[11, 12]. Biodegradation has only a
secondary effect on exposure rate and
health risk for medium and high
Henry's law constant chemicals, as can
be seen from Table 7 and 8. To
properly assess the risk due to low
Henry's law constant chemicals such
as pesticides, it is necessary to
consider biodegradation to more toxic
chemicals.
The aid of various computer
programs including BI01D, GEOTOX, and
.FEMWASTE has made the prediction of
the exposure rate and subsequent
health impact possible. These
programs are useful tools for the
study of exposure pathways. They can
provide critical information for
applying remediation, mitigation or
interdiction measures and can be used
as a basis for the selection of dose-
response models in subsequent risk
assessments.
335
-------
ACKNOWLEDGMENTS
This work was sponsored in part
by the University of California, Los
Angeles, Engineering Research Center,
under a grant from the Nation Science
Foundation on Hazardous Substance
Control.
REFERENCES
1. T. E. McKone, "Methods for
Estimating Multi-Pathway Exposures
to Environmental Contaminants,"
AD UCRL-21064, Lawrence Livermore
National Laboratory, June 1988.
2. P. Srinivasan and J. W. Mercer,
"BI01D:A Code for One-dimensional
Modeling of Biodegradation and
Sorption in Contaminant
Transport," GeoTrans, Inc. VA,
1987.
3. G. T. Yeh and D. S. Ward,
"FEMWASTE: A Finite-Element Model
of Waste Transport Through
Saturated-Unsaturated Porous
Media," ORNL-5601, Oak Ridge
National Laboratory, Oak Ridge,
Tennessee, 1981.
4. Y. Cohen and P. A. Ryan,
"Multimedia Modeling of
Environmental Transport:
Trichloroethylene Test Case,"
Environ. Sci. Technol. 19, 412-
417, 1985.
5. P. R. Wood, R. F. Lang and I. L.
Payan, "Anaerobic Transformation,
Transport, and Removal of Volatile
Chlorinated Organics in Ground
Water," In Ground Water Quality,
C. H. Ward, W. Giger, and P. L.
McCarty, editors, Wiley-
Interscience Publications, New
York, pp. 493-511.
6. R. D. Kleopfer, D. M. Easley,
B. B. Hass, Jr., and T. G.
Deihl, "Anaerobic Degradation
of Trichloroethylene in Soil,"
Environ.Sci. Technol. Vol. 19,
No. 3, 1985.
7. G. Bario-Lage, F.Z. Parsons, R.
S. Nassar and P. A. Lorenzo,
"Sequential Dehalogenation of
Chlorinated Ethenes,"
Environ.Sci. Technol. Vol. 20,
No. 1, 1986.
8. B. H. Keswick and C. P. Gerba,
"Viruses in Groundwater,"
Environ. Sci. Technol. 14: 1290-
1297, 1980.
9. C. R. Pearson, G. McConnel,
Proc. R. Soc. London, Ser B
1975, 189, 305.
10. C. Su, E. D. Goldberg, In
"Strategies for Marine
Pollution Monitoring,"
Goldberg, E. D., Ed., Wiley:
New York, 1976.
11. Staff Final Report, "Health
Assessment Document for
Trichloroethylene," EPA/600/8-
82/006F, July 1985.
12. Staff Report, Ambient Water
Quality Criteria for
Tetrachloroethylene," EPA 440/5-
80-073, October 1980.
13. B. B. Fuller, "Air Pollution
Assessment of Trichloro-
ethylene, : EPA Report MTR-7142,
PB-256-730, 1976.
14. D. L. MarrinandG. M. Thompson,
"Gaseous Behavior of TCE
Overlying a Contaminated
Aquifer," Ground Water, Vol.
25, No. 1, 1987.
15. M. S. Peterson, L. W. Lion, and
C. A. Shoemaker, "Influence of
336
-------
Vapor-Phase Sorption and Diffusion
on the Fate of Trichloroethylene
in an Unsaturated Aquifer System,"
Environ. ScL & Technol. , vol. 22,
No. 5, 1988.
16. R. C. Reid, T. K. Sherwood,
Properties of Gases and Liquids,
3rd ed., McGraw-Hill: New York,
1977. ' .. '
17. C. R. Wilke, P. Chang, AIChE J.,
1955, 1, 264-270.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
337
-------
Table 1: Physical Properties of PCE, TCE, DCE, and VC
PCE
TCE
Chemicals
Physical Properties
molecular weight 165.83
Henry's law const (torr/(mole/l) 17240
organic carbon part. coef. koc 220.52
diffusion coef. in air (m2/s) 7.0xlO~6
difussion coef. in water (m2/s) 8.834xlO"10 9.71xlO'10 1.09xlO"9 1.25xlO"9
bioconcentration factor 124.8 37.45 54.28 15.81
131.4
8837
42.45
DCE
96.94
99840
70.53
VC
62.5
17610
13.04
7.78xlO"6 9.94xlO"6 1.07xlO"5
Note:
1. bioconcentration factor is defined as:
bcf = (ppm in fish meat)/(ppm water).
2. air diffusion coefficient is calculated by Hirschelder formula
at 25 degree C [16].
3. water diffusion coefficient is obtained from Wilke and Chang's
work [17].
338
-------
Table 2: Environmental Settings
Landscape : San Diego Ecoregion
area in km2: 400
height of the air compartment (m): 700
humidity (kg/L): 7.8xlO~6
wet deposition scavenging efficiency: 0.8
yearly average wind speed (m/s): 5.0
precipitation onto land (cm/yr): 72.2
precipitation onto surface water (cm/yr): 0.76
total surface water runoff (cm/yr): 3.0
land surface runoff (cm/yr): 2.3
atmospheric dust load (micro-gm/m3): 61.5
deposition velocity of atmospheric particles (m/d): 334
evapotranspiration from soil (cm/yr): 41.9
evaporation from surface water (cm/yr): 1.17
thickness of the A soil horizon (m): 0.26
bulk density of the soil in the A horizon (kg/L): 1.3
water content of the soil in the A horizon (kg/L): 0.45
volumetric air content in the A horizon (L/L): 0.03
mechanical erosion rate (kg/km2/yr): 3.06xl05
irrigation from ground water (cm/yr): 1.0
thickness of the B soil horizon (m): 2.4
water content of the soil in the B horizon (kg/L): 0.28
bulk density of the soil in the B horizon (kg/L): 2.0
volumetric air content in the B horizon (L/L): 0.02
groundwater inventory (kg/km2): 2.Ie09
porosity of rock in the ground water zone (L/L): 0.3
density of rock in the ground water zone (kg/L): 2.33
fraction of the total surface area in surface water: 0.015
average depth of surface waters (m): 6.0
suspended sediment load in surface water (kg/L): 0.0034
deposition rate of suspended sediment (kg/mz/yr) : 150.0
thickness of the sediment layer (m): 0.05
bulk density of the sediment layer (kg/L): 1.5
porosity of the sediment zone: 0.2
resuspension rate from the sediment layer (kg/m2/yr): 150
ambient environmental temperature (k): 293
biota dry mass inventory (kg/km2): 3.IxlO7
biota dry mass production (kg/kmz/yr) : 1.6xl06
biota dry mass fraction: 0.33
boundary layer thickness at air/soil interface (m): 0.02
boundary layer thickness at water/air interface (m): 0.02
boundary layer thickness at sediment/water interface (m): 0.02
fraction organic carbon in the upper soil zone: 0.024
fraction organic carbon in the lower soil zone: 0.001
fraction organic carbon in the groundwater zone: 0.00287
fraction organic carbon in the sediment zone: 0.02
339
-------
Table 3: Removal Rates of PCE, TCE, DCE, and VC
Chemical
Compartment
air
air particle
upper soil
lower soil
ground water
surface water
sediment
PCE
TCE
DCE
VC
2.40X10"1
l.OOxlO"4
2 . 94xlO"2
2 . 04xlO"4
2.93xl(T3
5.48xlCT3
2 . 944xl(T2
2.40X10"1
l.OOxHT4
l.eiOxlO"2
i.eixio-4
i.eioxio"3
8.33xlO"2
1.610xlO"2
4. 48x10-*
4.48xlO-2
1.307xlO"2
1.307xlO"2
1.307xlO-2
8.67xlO"3
1.307xlO-2
3.38X10"1
3.38xlO-2
l.OxlQ-7
l.OxlO"7
l.OxlO"7
8.66xlO'3
8.66xlQ-6
Note:
Removal rate constants (I/day)
Table 4: Transform Rates of PCE, TCE, and DCE
Chemical
Compartment
upper soil
lower soil
ground water
sediment
PCE
TCE
DCE
2 . 94X10'2
1 . 22x10-*
2.93xlO"3
2.94xlO"2
1. eiOxlO'2
1.61xlO"3
i.eioxio-5
i.eioxio"2
1.307xlO"2
1.307xlO"3
1.307xlQ-4
1.307x10-2
Note:
Transform rate constants (I/day).
340
-------
Table 5: Partition Coefficients of PCE, TCE, DCE, and VC
Chemicals
Compartment Interface
Air/Water
Soill/Water
Soil2/Water
Rock/Ground water
Sediment/Surface water
Biota/Soill
Meat fat/Diet
Milk fat/Diet
Fish/Water
PCE
TCE
DCE
VC
0.09429
5.292
0.2205
0.06328
5.954
0 . 9448
0.0094
0.0094
124.8
0.04833
1.019
0.0425
0.01218
1.146
4.9088
0.0051
0.0051
37.45
5.461
1.693
0.0705
0.02024
1.904
2.954
0.00318
0.00318
54.28
0.9632
0.313
0.01304
0.00374
0.352
0.16
0.00845
0.00845
15.81
Note:
1. Temperature is 293 degree K.
2. SOIL1 -- upper soil.
3. SOIL2 -- lower soil.
341
-------
Table 6: Comparison of Environmental Concentrations of TCE
Compartment
AIR
PMAIR
BIOTA
SOIL1
SOIL2
GWTR
SWTR
SDMT
GEOTOX*
1.558xlO'4
2.86X10'13
2.29xlO'2
4.670xlO~3
l.mxlO"5
1.57xlO'5
4.023xlO'4
3.078xlO'3
GEOTOX**
1.556X10"4
2.858xlO"13
2.289xlO"2
4.663xlO"3
1.697xlO"3
4.553xlO"3
2.611X10'2
2.912xlO'2
COHEN & I
io-2
/
/
4xlO"2
/
/
6xlO"2
7xlO"2
f LIVERPOOL LA JOLLA
6.4xlO"3 7.8xlO"3
3.3X10"1 7xlO~2
Note:
1. * considering biodecay effect
2. ** no biodecay effect
3. PMAIR -- air particle
SOIL1 -- upper soil
SOIL2 -- lower soil
GWTR - - ground water
SWTR - - surface water
SDMT -- sediment.
342
-------
Table 7: Exposure Rates(/w Biodegradation)
Chemical: PCE
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
Chemical: TCE
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
Chemical: DCE
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
Chemical: VC
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
adult exposure
(mg/kg-d)
9.62E-02
6.60E-05
3.74E-05
3.34E-05
2.68E-05
2.23E-08
1.65E-08
9.64E-02
adult exposure
(mg/kg-d)
4.90E-02
5.97E-06
4.26E-05
9.30E-06
6.99E-07
. 3.86E-09
2.86E-09
4.90E-02
adult exposure
(mg/kg-d)
1.16E-02
4.26E-06
8.58E-07
1.35E-06
3.83E-07
1.36E-10
1.01E-10
1.16E-02
adult exposure
(mg/kg-d)
3.21E-01
3.28E-04
1.80E-04
1.01E-04
7.88E-06
3.11E-09
2.31E-09
3.22E-01
child exposure
(mg/kg-d)
1.80E-01
1.36E-04
1.78E-04
6.75E-05
3.39E-06
1.58E-07
6.48E-08
1.81E-01
child exposure
(mg/kg-d)
9.16E-02
1.23E-05
2.02E-04
1.88E-05
8.86E-08
2..74E-08
1.12E-08
9.19E-02
child exposure
(mg/kg-d)
2.18E-02
8.78E-06
4.08E-06
2.73E-06
4.85E-08
9.69E-10
3.97E-10
2.18E-02
child exposure
(mg/kg-d)
. 6.01E-01
6.75E-04
8.57E-04
2.03E-04
9.99E-07
2.21E-08
9.06E-09
6.02E-01
lifetime average
(mg/kg-d)
1.08E-01
7.60E-05
5.75E-05
3.83E-05
2.34E-05
4.17E-08
2.34E-08
1.08E-01
lifetime average
(mg/kg-d)
5.51E-02
6.88E-06
6.54E-05
1.07E-05
6.12E-07
7.22E-09
4.06E-09
5.52E-02
lifetime average
(mg/kg-d)
1.31E-02
4.91E-06
1.32E-06
1.55E-06
3.35E-07
2.55E-10
1.44E-10
1.31E-02
lifetime average
(mg/kg-d)
3.61E-01
3.77E-04
2.77E-04
1.15E-04
6.90E-06
5.83E-09
3.27E-09
3.62E-01
343
-------
Table 8: Exposure Rate(/wo Biodegradation)
Chemical: PCE
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
Chemical: TCE
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
Chemical: DCE
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
Chemical: VC
Pathway
Inhalation
Drinking Water
Biota ingestion
Meat/dairy ingstn
Fish ingestion
Soil ingestion
Dermal absorption
Total
adult exposure
(mg/kg-d)
9.63E-02
4.89E-04
3.75E-05
3.41E-05
1.70E-04
2.23E-08
1.65E-08
9.71E-02
adult exposure
(mg/kg-d)
4.89E-02
4.38E-04
4.25E-05
9.65E-06
4.54E-05
3.85E-09
2.86E-09
4.94E-02
adult exposure
(mg/kg-d)
1.16E-02
8.68E-06
8.57E-07
1.36E-06
1.30E-06
1.36E-10
1.01E-10
1.16E-02
adult exposure
(mg/kg-d)
3.21E-01
3.24E-04
1.80E-04
1.01E-04
7.84E-06
3.11E-09
2.31E-09
3.21E-01
child exposure
(mg/kg-d)
1.80E-01
1.01E-03
1.78E-04
6.89E-05
2.15E-05
1.58E-07
6.49E-08
1.82E-01
child exposure
(mg/kg-d)
9.15E-02
9.02E-04
2.02E-04
1.95E-05
5.75E-06
2.73E-08
1.12E-08
9.27E-02
child exposure
(mg/kg-d)
2.17E-02
1.79E-05
4.07E-06
2.74E-06
1.65E-07
9.68E-10
3.97E-10
2.18E-02
child exposure
(mg/kg-d)
6.01E-01
6.67E-04
8.57E-04
2.03E-04
9.93E-07
2.21E-08
9.06E-09
6.02E-01
lifetime average
(mg/kg-d)
1.08E-01
5.63E-04
5.76E-05
3.91E-05
1.49E-04
4.17E-08
2.34E-08
1.09E-01
lifetime average
(mg/kg-d)
5.50E-02
5.04E-04
6.53E-05
1.11E-05
3.97E-05
7.21E-09
4.05E-09
5.56E-02
lifetime average
(mg/kg-d)
1.31E-02
9.99E-06
1.32E-06
1.55E-06
1.14E-06
2.55E-10
1.43E-10
1.31E-02
lifetime average
(mg/kg-d)
3.61E-01
3.73E-04
2.77E-04
1.15E-04
6.86E-06
5.83E-09
3.27E-09
3.62E-01
344
-------
Figure 1: Risk Assessment/Management Flow Chart
POLLUTANT SOURCE
CHARACTERIZATION
I
EXPOSURE
PATHWAYS
DOSE
RESPONSE
f
REMEDIAL
OPTIONS
GENERATION
ENVIRONMENTAL
TRANSPORT/TRANSFORM
POLLUTANT
CONCENTRATION
EXPOSURE
RATES
HEALTH
EFFECTS
TOXICOLOGY
PHARMACOKINETICS
RISK
CHARACTERIZATION
EVALUATION
OF SOCIO-ECONOMIC
SCOCIO-POLITICAL CONSEQUENCE
REMEDIALN
ACTION
345
-------
Figure 2: Biodecay Relationship
Reductive Dehalogenation:
H
Cl
/
Cl
Cl { Cl
Cl
\
c =
cr
Cl
PCE
= 34 day
TCE
TV= 43 day
\
o ~~
H
X
^
\
H * Cl Cl
Xc = -X x
Cl
x
Cl
1.1
DCE
' \ / \
Cl H H H
1,2 trans 1.2 cis TJ,= 53 day
\
H I H
\ = ^
H
Cl
vc
346
-------
Figure 3: Relative Ranking of PCE, TCE, DCE, and VC
6.
5.0-
4.0-
3.0-
2.0-
1.0
FCE
T
TCE
VC
347
-------
INTERNATIONAL PERSPECTIVES ON CLEANUP STANDARDS FOR CONTAMINATED LAND
Robert L. Siegrist, Ph.D., P.E.
Institute for Qeoresources and Pollution Research
N-1432 Aas-NLH, Norway
ABSTRACT
A critical but extremely difficult task associated with cleanup of
contaminated land has been assessing the significance of contamination and the
degree of cleanup required. To assist the Norwegian government with its
newly evolving program for dealing with contaminated land, a study was
undertaken to identify approaches utilized internationally. A key part of
this study focused on the use of predetermined standards, guidelines and
criteria (PSQCs). This approach has been controversial and criticized for
various reasons including lack of consideration of site-specific factors and
insufficient (eco)toxicological data to establish PSQCs for a comprehensive
list of contaminants. Nevertheless it was found that PSQCs are viewed as an
important part of an overall program for dealing with contaminated land.
There is a clear desire and need for PSQCs to streamline the assessment and
cleanup of non-catastrophic sites, facilitate soil protection programs and
encourage redevelopment of old industrial sites. However, it is recognized
that PSQCs will not obviate the need for consideration of site specific
factors nor risk assessment and risk management approaches. PSQC approaches
have been utilized for years in several countries, most notably The
Netherlands. Other nations have developed or are considering similar PSQC
approaches.
INTRODUCTION
Norway has long been regarded as
a pristine nation with majestic
mountains and enchanting fjords.
Unfortunately, during the past few
years an increasing number of old
waste sites and parcels of
contaminated land have been
discovered. For example, an old
waste site was recently discovered
near Oslo during railroad-related
construction activities. The site
had been used for dumping and burning
of flarrmable liquids in an effort to
reduce fires at a municipal landfill.
Approximately 10,000 m3 of soil
contaminated with solvents ultimately
were excavated and properly disposed
of.
Since Norway derives most
(> 80*) of its potable water from
surface water, concern over ground
water pollution has so far been
limited. However, there is
recognition of potential hazards to
public health and the environment via
other pathways. While there is
little question that contaminated
sites exist in Norway, little is
known about the nature and extent of
the problem. National inventories
have recently been initiated
including industrial branch surveys
and old waste site surveys.
348
-------
Discoveries of abandoned waste
sites and contaminated land, often
related to construction activities,
have necessitated prompt action by
regulatory authorities. As in the
rest of the world, a critical but
extremely difficult task has been
assessing the significance of
contamination at a particular site
and determining the extent of cleanup
required.
PURPOSE
A study was undertaken to
identify the approaches used for
establishing cleanup goals for
contaminated land . and the
technologies employed to achieve
those goals. The information derived
was to assist the Norwegian
government in the development and
implementation of a newly evolving
program for assessment and cleanup of
contaminated land. A key aspect of
this study concerned the perspectives
toward and current application of
"predetermined standards, guidelines
and criteria" (PSQCs). A synopsis of
this aspect of the work is given
below while details may be found
elsewhere [1].
APPROACH
Information for this work was
gathered by several means. The
international literature was surveyed
by computerized and manual
techniques. Personal inquiries were
made to responsible agencies and
individuals in ten countries with
both well-established and relatively
new programs for dealing with
contaminated land (United States,
Canada, England, The Netherlands,
West Germany, France, Denmark,
Sweden, Finland and Norway).
Personal site visits were made where
appropriate and feasible to gather
firsthand information. The
information gathered from all sources
was reviewed and sunrmarized.
PROBLEMS ENCOUNTERED
At the onset, the subject of
this study was recognized as a
complex one, intertwined not only
with government policies and programs
for dealing with contaminated land,
but also with those for environmental
protection in general. It was
accepted that efforts to gather and
review all relevant literature and
contact all knowledgeable agencies
and individuals would be futile.
Rather attempts were made to gather
representative current information.
This was made somewhat difficult,
since the issue of cleanup standards
is a diverse and dynamic one on a
local and national level, even in
nations with apparently well
established programs (e.g. USA, The
Netherlands, West Germany).
RESULTS
Contaminated Land Programs
There are many different names
used to refer to what might generally
be defined as "contaminated land",
but few formal definitions exist.
Perhaps the most widely held is that
put forth by NATO/CCMS [2]:
"Land that contains substances that, when
present in sufficient quantity or
concentrations are likely to cause harm
directly or indirectly to humans, the
environment or on occasions to other targets."
Approaches to dealing with
contaminated land have generally
evolved in response to both
industrial site redevelopment as well
as improper waste management. In
most cases, one or more notorious
incidents has stimulated public
attention and inquiry eventually
leading to an awareness of the nature
349
-------
and extent of the problem. In
response, legislation and regulations
were enacted and policies and
programs evolved. Some nations
embarked on cleanup campaigns almost
a decade ago (e.g. USA, The
Netherlands) while others have begun
in earnest only recently (e.g.
Norway, France, Canada). The level
of concern and program development
may be related to many factors
including population density,
industrialization and reliance on
ground water for drinking water
(Table 1). in seme nations the
programs have been incorporated into
broad soil protection programs. The
clearest example of this is The
Netherlands where powerful national
laws were enacted in 1983 and 1987
(The Soil Protection Act) [3,4].
West Germany initiated a conceptually
similar program in 1985 [5].
The nature and extent of the
problem with contaminated sites
varies widely between different
nations (Table 2). Typically as
concern grows, national inventories
are initiated and cleanup programs
are formalized. While the number of
sites identified can be large, the
number remediated and restored to
productive use can be relatively
small (i.e. typically
Early remediation efforts
typically involved excavation and
offsite treatment or landfilling.
More recently there has been
increased interest in onsite and in
situ technologies such as vapor
extraction, leaching and
bioremediation. The situation in The
Netherlands is somewhat unique in
that there have been for years,
numerous plants dedicated solely to
treatment of contaminated soil.
There are thermal, extraction and
biological treatment plants with a
total annual capacity of nearly 0.5
million m3 [6]. Similar plants have
been or are being implemented
elsewhere (e.g. Denmark, Germany).
Approaches to Establishing Cleanup
Goals
The approaches used to establish
cleanup goals vary widely both within
and between countries (Table 3). In
a critical review of the USA
Superfund program, the U.S. Office of
Technology Assessment identified
seven different approaches [7]: (1)
ad hoc, (2) site-specific risk
assessment, (3) national goals for
residual chemicals, (4) cleanup to
background or pristine levels, (5)
best available technology or best
engineering judgment, (6) cost-
benefit approach, and (7) site
classification. While these
approaches are based on practices in
the USA, they represent fairly the
state of practices internationally.
Regardless of the approach utilized,
it is clear that science and
engineering are but one part of the
picture, with site-by-site decision
making often heavily influenced by
economic, social and political
forces.
There has been considerable,
sometimes heated, debate over which
approach to establishing cleanup
goals is the "best approach" [e.g.
7-11]. One aspect of this debate has
centered around the establishment and
application of PSGCs. There are
advantages and disadvantages to this
approach as will be outlined later.
Given first is a review of the
current attitudes toward and use of
PSGCs in several countries around the
world.
350
-------
Table 1. Area and population densities of selected nations [28].
Country
USA
Canada
United Kingdom
Netherlands
W.Germany
France
Denmark
Sweden
Finland
Norway
Statistic Date
Population
(mi 1 1 ions)
247.5
25.3
56.6
14.7
60.2
55.8
5.1
8.4
5.0
4.2
1989
Area
(103 sq.mi.)
3640
3852
94.2
15.8
96.0
220.7
16.6
173.7
130.1
125.2
Population Density
(capita/sq.mi.)
68
6.6
601
931
627
253
305
48
38
34
Urban Population
(*)
79
76
92
88
fUft
86
77
84
85
61
80
1980-1986
Table 2. General characteristics of "contaminated land".
Site Discoveries
Country Site Number and Concern
USA
Canada
England
Nether-
lands
23000 ('87) with 900 ('87)
nat. priority (NPL) sites.
Total unknown.
300 estimated.
6060 ('86).
35000 ('85) with 5400 req.
inrmediate action.
West
Germany
France 453 ('87) with 82 serious.
Denmark
Sweden
Finland
1599 ('88). Estimate 9000
potential sites.
3800 ('85) old waste sites
with 500 est. of concern.
Site Remediation
App. ttExample Methods
Sites Cannonly Used
130 NPL
?Non-NPL
Few
>500
380
95
30-60
Few
Total unknown. 1200 landfills Few
with 378 with haz.wastes,
112 need immediate action.
Few
ref.
Norway Total unknown.
Excavat ion/1andf ill 7,8
Incineration
Insitu treatment
Excavat ion/1andf ill 9
Isolation/capping 10,24
Excavation/1andfi11
Excavation/treatment 6
by thermal .washing
Excavat i on/1andf i11
Encapsulation 18
Excavation/1andfi11
Excavation/landfill 17
Encapsulation
Solidification
Excavation/landfill 18,19
Incineration
Onsite treatment
Excavation/1andf ill 20
Incineration
Encapsulation
Excavation/landfill 21
Seme incineration
Some 1andfarming
Excavation/landfill 22
351
-------
Table 3. Approaches to establishing cleanup goals for contaminated land and
use of predetermined standards, guidelines and criteria.
Country
Approaches to Establishing Cleanup Goals
sjtef O-e- Superfund) use applicable, relevant and appropriate
federal and state requirernents where available and formalized site-specific
risk assessment methodologies. For npn-NPL sites, procedures vary widely by
State and government jurisdiction and include multiples of generic criteria
andbackground levels as well as site-specific formalized risk assessment
Only Quebec has a formalized approach where a comprehensive list of generic
criteria adapted[from the "Dutch List" is used for initial guidance and
screening with site-specific risk assessments as appropriate. [9,16]
national system. National guidance on "Trigger Concentrations" for seme
contaminants cormonlv'found on industrial sites ccmnonly considered for
redevelopment (e.g. old gas works). [10,24]
,, approach , control by provincial governments (Lander). Use of
st " consideration given to local conditions. West German
i po,1l?y«of maintaining soil "multi-functionality". Generic criteria
levels) for evaluating significance of pollution enacted in 1983 (often
referred to as the "Dutch List". Reference values for good soil quality (new
A-level) enacted in 1987. Contaminated land must be cleaned up to multi-
functional quality (A- level) unless it is shown to be technically or
financially unfeasible or environmentally harmful. [3,4,6,11-14]
.. ^ ,. .
Guides/Threshold values for soil contamination now under development based
on soil protection policy initiated in 1985. [5,18,25]
Fr
No
ranee
national approach
control by local governments. Use qualitative risk
iw nuviwiKi.1 ^Hh" Wf*'-!v **"•»•«*'« "r i«j^^»i guverTiiMrrbs. use qualitative riSK
assessments. If pollution by natural substances, must reference background.
Development of standards for soil pollution now under consideration. [17,18]
national approach, control by local governments. Use "Dutch List" for
general guidance and screening as well as existing Danish standards where
available. Final decision on particular site based on site-specific
considerations. Formalized risk assessment methods now under development.
[19»26]
jto nationalI approach. Limited experience to date. Use generic criteria (e.g.
'Dutch List") if available for initial guidance but site specific decision
based on local factors including technical, political, economic and
psychological. [20]
Finland
W2. na?1?na2..apiroa
-------
Predetermined Standards. Guidelines
and Criteria
The desire and need for PSQCs
specific to contaminated land were
evident in most of the ten countries
considered. Readily available,
comprehensive PSQCs were viewed as
essential to facilitating initial
site review and screening. There are
demands for unequivocal cleanup
criteria, often put forth by owners,
developers and future users of
contaminated land. Equally evident,
however, was a strong appreciation
for the difficulties and potential
problems of establishing and
implementing PSQCs and the belief
that there must be site-by-site
flexibility in setting final cleanup
goals.
The first country to establish a
national, comprehensive set of PSQCs
for contaminated land was The
Netherlands [3,11-13]. In 1983 a
national act was promulgated which
put forth the concept of "multi-
functionality" for soil and included
criteria for assessing the
significance of soil and ground water
contamination and guiding site
assessment and cleanup ("Dutch List",
Table 4)[3]. In support of a broad
soil protection policy, reference
values were recently enacted for a
"good soil quality" [12]. All of
these criteria were never intended to
be standards, but rather guideline
values for deciding upon the
necessity for carrying out (further)
investigations and risk assessments
[3,14]. In practice however, the
criteria have been implemented as if
they were in fact standards, in parts
of The Netherlands and elsewhere
(Table 3). In 1988, the province of
Quebec, Canada promulgated their own
similarly comprehensive list of
criteria, based in large part on the
"Dutch List" [15].
Other national and provincial
government agencies have also
established PSQCs in the form of
acceptable limits for soil
contaminants. These have different
names including "Trigger
Concentrations" (England! Table 5),
"Cleanup Guidelines" (New Jersey,
USA; Table 6), and "Guide/Threshold
Values" (West Germany). While far
less comprehensive than the Dutch or
Quebec lists, they are intended to
serve as guidance in site assessment
and cleanup. In many cases the
criteria are given with reference to
a proposed land use (e.g. Table 5).
In most cases, the PSQCs are not
legal standards, but rather guidance
criteria intended to be used with due
consideration of site specific
factors. The exceptions seem to be
for a few notorious substances such
as polychlorinated biphenyls (PCBs),
some polycyclic aromatic hydrocarbons
(PAHs) and dioxins.
Even in those jurisdictions
where PSQCs have not been formulated
specifically for cleanup of
contaminated land, reference is
commonly made to existing PSQCs, such
as the Dutch List (Table 3). There
is also direct use or adaptation of
existing national or international
standards, often developed under
programs and legislation unrelated to
contaminated land. Examples include
drinking water standards, ambient
water quality standards, storm water
runoff limits, limits on sewage
sludge application to agricultural
lands, occupational air quality
standards, ambient air quality
standards, and so forth. Notably,
in several instances, ground water
quality standards have been set
roughly equivalent to drinking water
standards (e.g. Wisconsin, USA,
Denmark, The Netherlands). In some
cases, reference to existing
353
-------
Table 4.
Soil and groundwater criteria used in The Netherlands for assessing
the significance of contaminated land ("Dutch List") [3].»
Component
1. Metals
ium
Soil (mg/kg dry matter) Qroundwater (ug/L)
A-Level B-Level C-Level A-LevelB-LevelC-Level
1(
Nicke
Motypdenum
Cadmium
Selenium
Barium
jleroury
2.
Inorganics
T"
m — -— j- - J^l^G^3
rto€.
ttota
Ctota
(as P)
3 .Arcmat i c Compounds
UnyTbenzene
Toluene
200
ota? Arcmatics
plic Hydrocarbons
Anthracene
!-enantnrene
irlouranthene
I'yrene v o
5enzo(a)pyrene. 0
Total Polycyclics 1
5..Chlorinated Hvdrocarl
zi' *
8:!
inera
20
50
i^
7
5
18
18
oropenzenes Hnd.
6. Pesticides
Pollutants
---- , —yoroTuran
Pyridine
Tetrahydroth iof ene
Cyclonexanone
St
FM«! (gasoline)
M
800
8:?5
0.1
mplies
e, values are guidelines not "standards". A- level implies unpolluted
'"I'SLlJff PoJTuCjon present and further investigation reauiredTC-leye
significant pollution present and cleanup required (bacK to A^-leveT);
354
-------
Table 5. Tentative "Trigger Concentrations" used in England [24].
Compound
Planned Land Uses
Trigger Concentrations
Threshold Action
(mg/kg air-dried soil)
Selected Inorganic Contaminants
Domestic gardens, allotments
Parks, playing fields open space
Domestic gardens, allotments
Parks, playing fields open space
Domestic gardens, allotments
Parks, playing fields open space
Domestic gardens, allotments
Parks, playing fields open space
Domestic gardens, allotments
Parks, playing fields open space
Domestic gardens, allotments
Parks, playing fields open space
Domestic gardens, allotments
Parks, playing fields open space
Boron (wat.sol.) Any uses where plants are grown
Copper Any uses where plants are grown
Nickel Any uses where plants are grown
Zinc Any uses where plants are grown
Arsenic
Cadmium
Chromium
(Hexavalent)
Chromium
(Total)
Lead
Mercury
Selenium
10
40
3
15
25
600
1000
500
2000
1
20
3
6
3
130
70
300
To
To
To
To
TO
To
To
To
To
To
To
To
To
To
To
To
To
be
be
be
be
be
be
be
be
be
be
be
be
be
be
be
be
be
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
developed
Contaminants Associated with Former Coal Carbonization Sites
Poly aromatic
Hydrocarbons
Phenols
Cyanide (free)
Cyanide (comp.)
Thiocynanate
Sulphate
Sulphide
Sulphur
Acidity (pH)
Domestic gardens, allotments, pi ay areas 50
Landscapes, buildings, hardcovers 1000
Domestic gardens, allotments 5
Landscapes, buildings, hardcovers 5
Domestic gardens, allotments, landscapes 25
Buildings, hardcovers 100
Domestic gardens, allotments 250
Landscapes 250
Buildings, hardcovers 250
All proposed uses 50
Domestic gar dens, allotments, landscapes 2000
Buildings 2000
Hardcovers 2000
All proposed uses 250
All proposed uses 5000
Domestic gar dens, allotments, landscapes 5
500
10000
200
1000
500
500
1000
5000
None
None
10000
50000
None
1000
20000
3
i All proposed values are tentative and/or preliminary requiring regular
updating. All values are for concentrations determined on "spot" samples. If
all values are below the threshold concentrations, site may be regarded as
unoontaminated for these contaminants and development may proceed. Above the
thresholds, remedial action may be needed. Above the action concentration,
remedial action will be required or the form of development changed.
355
-------
Table 6. Cleanup guidelines used in the State of New Jersey, USA [9],
Substance
Soil
Qroundwater
Chromium
Zinc
Lead
Copper
Arsenic
Cadmium
Selenium
Nickel
Barium
Mercury
Silver
Total Volatiles
Volatiles plus Base Neutrals
Total Hydrocarbons
Petroleum Hydrocarbons
(ppm)
100
350
100
170
20
3
20
100
-
-
-
1
-
100
100
(ppb)
50
50
_
50
10
10
1000
2
50
100
1000
standards has been formally
incorporated into a waste site
cleanup program (e.g. USA Superfund
program).
DISCUSSION
Based on the information derived
in this study, it became apparent
that there were numerous potential
advantages and disadvantages of a
PSQC approach to establishing cleanup
goals (Table 7). Approaches
employing PSQCs are not claimed to be
the best approach for setting cleanup
goals, but rather a necessary part of
an overall program for dealing with
contaminated land. PSQCs facilitate
national or regional soil protection
programs and encourage redevelopment
efforts for contaminated land. In
this context they may be used for
both initial screening and
contamination assessment as well as
for determination of final cleanup
goals. For contaminated sites of
national and regional significance
(e.g. uncontrolled hazardous waste
sites involving large concentrations
or amounts of highly toxic materials)
such a PSQC approach will usually not
be appropriate. While PSQCs may
provide an early indication of the
extent of the problem, site-specific
risk assessments will likely be
needed and risk management decisions
will necessarily have to be made.
The use of PSQCs appears to be
gaining favor in many nations,
especially for use in preliminary
assessment of the significance of
contamination and the potential
extent of cleanup. The Netherlands
has used this approach for more than
5 years to remediate several hundred
sites [6,13]. Recently, the Province
of Quebec in Canada, issued a
similarly comprehensive list of soil
and ground water criteria, in large
part adapted from the "Dutch List"
[15]. Establishment of similar lists
is also under consideration in other
356
-------
Table 7. Example advantages and disadvantages to the use of predetermined
cleanup standards, guidelines and criteria.
Advantages
o Speed and ease of implementation.
o Similar sites would be handled in a similar manner.
o Useful for initial assessment of significance of contamination.
o A priori information facilitates planning and action.
o Encourages developers to undertake decontamination and restoration.
o Potential consistency with strategies for environmental standards.
o Reality of contaminated land made easy for layman.
o Facilitate environmental audits of industrial sites.
o Facilitates monitoring/permitting of operational industrial sites.
o Can be used for performance assessments of soil treatment plants.
o Implies non-negotiability and reduces local political influences.
Disadvantages
o
o
Seme important site-specific considerations cannot be accounted for.
Standards, guidelines and criteria are not formulated for many toxic
substances of concern. Existing standards formulated under other
programs are not necessarily appropriate for contaminated land.
PSQCs imply a level of understanding, knowledge and confidence which
likely does not exist.
Once PSGCs are established, site-specific flexibility may be difficult.
jurisdictions (e.g. Alberta, Canada
[9,16], West Germany [5], France
[17]).
It appears that PSQCs represent
an important component of an overall
program to deal with soil protection
and contaminated land. The challenge
is to develop scientifically well-
founded PSQCs specific to
contaminated land which are
consistent with other laws and
regulations and supported by various
concerned parties (e.g. scientific
and engineering ccnmunity, regulators
and politicians, environmental and
citizens groups).
ACKNOWLEDGMENTS
The study reported herein was
conducted by the Institute for
Oeoresources and Pollution Research,
Aas, Norway, with sponsorship in part
by the Norwegian State Pollution
Control Authority. Gratefully
acknowledged are the agencies and
individuals in Scandinavia, Europe
and North America without whose
contributions this study could not
have been accomplished.
Correspondence may be addressed to
the author at 4014 Birch Avenue,
Madison, Wisconsin, 53711, USA,
telephone 608 238-7697.
REFERENCES
1. Siegrist, R.L. 1989. International review
of approaches for establishing cleanup
goals for hazardous waste contaminated
land. Final res. rept. to Norwegian State
Poll. Cont. Agency by Inst. for
Georesources and Pollution Res., Aas-HLH,
Norway. 68 p.
2. Smith, H.A. 1988. An international study
on social aspects etc. of contaminated
land. In: K. Holf, W.J. van den Brink,
F.J. Colon (eds.), Contaminated Soil '88,
Kluwer Acad. Publ., London, pp. 415-424.
357
-------
3, Moan, J.E.T., J.P. Cornet and C.W. Evers.
1986. Soil protection and remedial
actions: criteria for decision making and
standardization of requirements. In: J.W.
Assink and W.J. Vandenbrink (ed.).
Contaminated Soil. Martinus Mijhoff Publ.,
Dordrecht, Netherlands, pp. 441-448.
4. Hoen, J.E.T. 1988. Soil protection in The
Netherlands. In: K. Wolf et al. [see 2.],
pp. 1495-1503.
5. Bachmann, Q. and D.F.W. von Borries. 1988.
Soil protection and abandoned hazardous
waste sites. In: K. Wolf et al. [see 2.1,
pp. 1549-1554.
6. Van Drunen, T.S.Q. and F.B. deWalle. 1988.
Soil pollution and reuse of cleaned-up
soils in The Netherlands. Proc. Conf.
Soil, The Aggressive Agent. Oct. 1988.
IBC Technical Services Ltd., IBC House,
Canada Road, Surrey, England.
7. U.S. Congress, Office of Tech. Assessment.
1985. Chapter 4: Strategies for setting
cleanup goals, In: Superfund Strategy,
Report OTA-ITE-252. pp. 103-121.
8. Kavanaugh, H. 1988. Hazardous waste site
management: water quality issues.
Colloquium by the Water Sci. and Tech.
Board, U.S. Hatl. Res. Council. Feb.,
1987. Natl. Academy Press, Washington,
D.C. pp. 1-10.
9. Richardson, Q.H. 1987. Inventory of
cleanup criteria and methods to select
criteria. Unpubl. rept., Industrial
Programs Branch, Environ. Canada, Ottawa,
Ontario. 46 p.
10. Beckett, M. 1988. Current policies in the
U.K. and elsewhere. L.U.T. Shortcourse on
Contaminated Land. Sept., 1988. 12 p.
11. Vegter, J.J., J.H Roels and H.F. Bavinck.
1988. Soil quality standards: science or
science fiction. In: K. Wolf et al. [see
2.], pp. 309-316.
12. Bayinek, H.F. 1988. The Dutch reference
values for soil quality. Hin. of Housing,
Physical Planning and Environ.,
Leidschendam, Netherlands. 9 p.
13. DeBruijn, P.J. and F.B. deWalle. 1988.
Soil standards for soil protection and
remedial action in The Netherlands. In:
K. Wolf et al. [see 2.], pp. 339-349.
14. Vegter, J.J. 1988. General secretary for
Tech. Comm. on Soil Prot., Min. of
Leidschendam, Netherlands. Persona]
commun., 16 Dec. 1988.
15. Anonymous. 1988. Contaminated sites
rehabilitation policy. Gouvernement du
Quebec, Mim'stere de L'Environnement,
Direction des substances dangereuses.
Sainte-Foy, Quebec, Canada. 43 p.
16. Lupul, S.L. 1988. Branch Head, Industrial
Wastes Branch, Alberta Environ., Waste and
Chemicals Div., Edmonton, Canada.
Personal commun., 8 Dec.1988.
17. Qoubier, R. 1988. Inventory, evaluation
and treatment of contaminated sites in
France. In: K. Wolf et al. [see 2.], pp.
1527-1535.
18. Palmarck, H. et al. 1987. Contaminated
land in the European Communities:
Summarizing Rept. UBA-FB. Comm. European
Communities, Brussels. 216 p.
19. Keiding, L.M., L.W. Sorensen and C.R.
Petersen. 1988. On investigation and
redevelopment of contaminated sites.
Proc. Conf. Impacts of Waste Disposal on
Groundwater. Aug. 1988, Copenhagen.
Danish Water Council.
20. Von Heidenstam, 0. 1988. Swedish Natl.
Environ. Prot. Board, Stockholm. Personal
commun., 10 Nov. 1988.
21. Assmuth, T. et al. 1988. Assessing risks
of toxic emissions from waste deposits in
22
23
24
25
27
Finland. In:
1137-1146.
Johannsen, J,
Cont. Auth.,
K. Wolf et al. [see 2.],
pp.
Housing, Physical Planning and Environ., 28
1988. Norwegian State Poll.
, Oslo. Personal commun., 9
Aug. & 6 Dec. 1988.
U.S. Environ. Prot. Agency. 1987.
Hazardous Waste System. Office of solid
wastes and emergency response, Washington
D.C. pp 3-7.
ICRCL. 1987. Guidance on the assessment
and redevelopment of contaminated land.
ICRCL 59/83 (2nd ed.), Dept. of Environ.,
London. 20 p.
Franzius, V. 1988. Fed. Environ. Agency,
Umweltbundesamt, Berlin. Personal commun.,
26 Oct. 1988.
26. Sorensen, L.W. 1988. Waste Sites Office,
Natl. Agency for Environ. Prot.,
Copenhagen. Personal commun., 5 Oct.
Assmuth, T. 1988. Tech. Res. Office, Natl.
Board of Waters and Environ., Helsinki.
Personal commun., 8 Nov. 1988.
Hoffman, H.S.(ed.). 1989. World Almanac
and Book of Facts. Pharos Books, N.Y.
358
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. Hie contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
359
-------
THE STATUS OF HAZARDOUS WASTE MANAGEMENT IN TAIWAN, R.O.C.
Larry L. G. Chen
Deputy Administrator
Environmental Protection Administration
Government of the Republic of China
1, Hsiang-Yang Rd., Taipei, R.O.C.
ABSTRACT
A large quantity of industrial waste is produced daily in Taiwan, R.O.C..
A 1985 survey found that the amount of waste generated equalled approximately
30 million tons per year. Hazardous waste represents 9.7% of this total.
Based on statistics from this same 1985 survey, 72% of the factories disposed
of their waste without intermediate treatment. Since most methods used for
treatment of hazardous wastes were implemented incorrectly, the proper treat-
ment of such waste has become the focal point of environmental protection in
Taiwan. From July, 1987 the "short-term program for industrial waste control"
has had as its first priority the control of toxic, infectious and corrosive
hazardous waste. At the same time, a registration system for permission,
reporting and results inspection for hazardous wastes is being developed. An
industrial waste exchange and reclamation system is also being developed. It
is predicted that a complete hazardous waste management program can be
developed within the next four years.
INTRODUCTION
In the past forty years, owing
to successful economic policies and
the hard work of our people, Taiwan
has experienced an economic miracle
known as the "Taiwan Model". While
our economic growth has won the
praise of many countries, serious
environmental pollution has been
closely following the footsteps of
economic development. A 1985 survey
found industrial waste totalled
approximately 30 million tons per
year.(l) Hazardous waste represented
9.7% of this total. The survey
investigated 1,600 factories which
are potential sources of hazardous
waste. Among them, 72% of the factories
disposed of their waste without inter-
mediate treatment. Table 1 shows
present intermediate treatment methods
for industrial waste. Table 2 & 3
indicate types and quantities of
waste. 28.767, of waste was dumped in
landfills, 52.15% was recycled and
reused, 13.917., was not disposed of
(ie. it was stored, etc.), 0.027o was
dumped into the ocean and the method
of handling 4.987o was unkno'-m. The
majority of businesses in Taiwan are
small or medium size companies.
These companies have difficulty
handling their hazardous waste. They
360
-------
can not afford the huge investments
required to build facilities and
would have trouble finding trained
professionals to operate facilities
if they had them. Laying responsi-
bility for hazardous waste treatment
on the individual company is not a
realistic approach. Therefore,
building an effective hazardous
waste control program, solving
emergent hazardous waste issues,
constructing waste treatment faci-
lities and planning a waste exchange
system are our current emphases.
Table 1. Present intermediate
treatment methods for
industrial waste in
Taiwan.
MethodPercentage
Physical treatment 4.03
Chemical treatment 5.68
Biological treatment 0.75
Incineration 8.03
Other thermal treatment 1.03
Others 7,87
Untreated 72.33
PURPOSE
An Amendment to the Waste
Disposal Act was approved in November
1985 in an effort to solve the
hazardous waste problem and prevent
toxic chemicals from contaminating
the environment. Two years later, a
high level Environmental Committee
was organized by coordinating several
divisions of the central government.
A four year "short-term program for
industrial waste control" has been
developed.(1) It projects four goals
from July 1987 to June 1991.
1. Give highest priority to
toxic, corrosive, and infectious
industrial "waste. 2. Set up a
record-reporting and self-monitoring
system for industries to control main
pollution sources. 3. Build hazardous
waste treatment facilities and disposal
sites to prevent illegal dumping. 4.
Plan and promote the waste exchange and
reuse systems.
APPROACHES
We have studied our own unique
environmental problems as well as
similar problems experienced by industri-
alized countries, such as the U.S.,
Japan, and several European countries,
in order to draw guidelines for our
environmental programs. Several criteria
have been included to manage hazardous
waste effectively. They are:
1. A solid legal basis. 2. Allow-
ance standards for environmental pol-
lution. 3. Equipment for data process-
ing, monitoring and composition analysis.
,4. Permit procedures for storage,
treatment and disposal (TSD_)f acilities.
5. Effective facilities and management
system. 6. Sufficient personnel and
funds. 7. Promotion of waste minimi-
zation strategies such as: the recycling
and reuse of industrial waste, improve-
ments in the manufacturing process, and
waste exchange.
According to the above consider-
ations, we are planning to implement
the following control measures:
Control Program:
Four systems have been introduced
to regulate industries and hazardous
waste treatment organizations. They
are: a permit-recording system, a
planning-permission system;a /recording
system, and an inspection system.
Figure 1 shows the framework of these
control Systems. All information
gathered in these systems is incorporated
into a Chemical Substance Management
System in order to help build a nation-
wide network of information useful in
the handling and prevention of chemical
disasters.
361
-------
Table 2. Eighteen Catagories for Industrial Waste. Data taken form survey
of 1623 factories. Note that each factory may have several
different types of waste.
Waste Type
Industrial Waste
No. of Average Monthly
factories Volume
Hazardous Waste
No. of Average Monthly
factories Volume
ash 78
concentrated 137
dust
metal slag 507
mineral slag 43
process slag 92
sludge from 364
wastewater
treatment
oil/water 66
mixture
waste oil 57
waste acids 147
waste base 82
waste container 323
waste paper 446
waste solvent 151
refinery slag 49
waste plastics 242
waste rubber 47
waste fiber 66
others 611
Total 1623
Table 3. Volume
Toxic substance No.
Tons
month-factory
24.10
38
5.72
88.12
130.50
57.69
47.17
1.89
38.86
35.83
11.11
2.72
8.80
37.86
3.22
1.14
21.93
8.28
45.15
of toxic industrial
c?f factories Waste
63
115
476
39
71
292
60
48
136
72
295
312
131
23
178
37
50
503
1462
Tons
month- factory
28.16
35.68
5.90
94.27
40.24
67.31
51.87
7 1 8
t- . -LC?
38.78
35 OS
+J -J m \J ~S
11.01
1.95
8.52
4 70
*-r . / w
i 07
J- • \J 1
24.23
8.18
36.93
waste in Taiwan
volume
Percentage
Tons /Month/factory 7«
Hg
Cr
Pb
As
Cd
Zn
Cu
Cyanide compound
Fluorinated compound
PCBs
Org-P compound
Org-Cl compound
Metal &
Metal compound
Other organic
26 2.62
88 7.87
109 4.69
6 3.13
25 43.89
194 10.10
527 6.30
28 8.70
25 11.74
2 0.0001
72 30.47
131 2.95
503 18.46
1702 26.22
0.01
1.03
0.70
0.04
1.61
3.00
5.07
0.37
0.45
3.21
0.46
14.18
69.78
362
-------
Hazardous Waste Treatment Facilities:
Because most companies on
Taiwan do not have huge investments
or the technical know-how required
to build hazardous waste treatment
facilities, the government will
first set up such facilities to act
as models. The government will also
provide incentives to encourage
private companies to move into
the hazardous waste treatment busi-
ness. The EPA and the Industrial
Development Bureau:. (IDE), (under the
Ministry of Economic Affairs) are
coordinating companies located
inside industrial parks in building
waste treatment facilities within
those parks. Since the treatment
facilities will be located within
the parks where waste is produced,
the cost and risk of transporting
waste for treatment will be reduced.
It is also expected that locating
the treatment facilities in the
industrial parks will be more..accept-
able to a public loathe to have such
facilities "in their back yard". In
addition, EPA and IDE are coordinating
companies which produce similar kinds
of hazardous waste in cooperating
with each other by, for example,
jointly funding the construction of
their own treatment facilities.
Decreasing the 'Effect of High Impact
Hazardous Waste on -the Environment:
. Several years ago, a hazard to
human health arose when rice-bran
oil was contaminated by Polychlo-
rinated biphenyls (PCBs). Most
recently, a Cadmium-rice accident
was caused by industrial effluent
containing cadmium flowing into rice
fields. These accidents caused a
great wound to our society and there
are still other potential wastes
threatening the public health. It is
estimated Taiwan has 1000 tons of
PCBs, 2000 tons of the Cd-sludge and
120,000 tons of mercury-containing
sludge. " These dangerous pollutants
are the first priority of our work.
In addition, we are planning to inci-
nerate infectious waste on a regional
basis. Treatment of infectious hospital
waste is our first target.
Promoting Waste Minimization Strategies:
Our short-term targets are the
recycling and reuse of large amounts of
waste, including: pesticide containers,
waste oil, rubber tires, solvent, .and
plastics. Information on waste will be
published periodically. Data on the
quantity and quality of waste will be
provided to industries to promote
recycling and reuse.
Other control measures we are actively
on :include:
1. Developing legislation. 2.
Establishing an information databank.
3. Promoting information exchange among
government divisions. 4. Carrying out
research and development. 5. Promoting
education and training programs.
PROBLEMS ENCOUNTERED
The inherent complex property of
hazardous waste pollution forms a.
difficulty challenge for us. Hazardous
waste has various forms, is distributed
widely, and huge quantities are generated
at a high rate. Proper treatment demands
a tremendous amount of investment as
well as technological know-how. •• The
following issues are encountered in the
process of hazardous waste control:
Insufficent Regulations:
Before 1985 there was only the Waste
Disposal Act to regulate the treatment
of solid waste, but the contents of the
Act were over simplified. The Act only
required that a plant must take care of
the treatment of its own waste. This
act did not regulate treatment facilities
or specify control measures. Regulations
relating to hazardous waste were non-
existent. It was not until 1985 that
the the Waste Disposal Act was amended
363
-------
to include the regulation of hazardous
waste.
Lack of Accurate Data on Hazardous
Waste:
We conducted a survey of hazard-
ous waste in 1985, however the
actual amount of hazardous, waste is
still unclear. There are approxiately
80,000 factories in the Taiwan area.
Complete data about their waste and
treatment methods is still unavail-
able. _Tab_les_2_&_3_ show the results
of the 1985 survey, roughly indicat-
ing the different kinds and amounts
of hazardous waste in Taiwan. Without
sufficient information, nationwide
planning for hazardous waste control
is handicapped. The Industrial
Technology Research Institute is
presently working on a more compre-
hensive nationwide survey.
Adequate Treatment and Storage
Facilities are Unavailable in Both
the Public and Private Sector:
The progress of environmental
protection in the ROC is similiar to
that of other industrialized coun-
tries. In the early stages, only
waste water and air pollution issues
were addressed. Hazardous waste was
not a major concern, resulting in a
shortage of treatment facilities.
Shortage of Skilled Personnels and
Funds:
The Environmental Protection
Bureau, the forerunner of the present
EPA, was not a cabinet level depart-
ment in the central government. At
the time of its existence, the
public did not have a good sense of
environmental protection. As a
result, there was a serious shortage
of manpower and funds for environment
protection in both the central and
local governments. The management of
hazardous waste involves a high
degree of scientific knowledge,
specially trained professionals and a
tremendous amount of investment. It was
not until August 22, 1987 that the
formation of the EPA brought with it a
huge increase in the amount of finances
and personnel for environmental pro-
tection.
RESULTS
In August 22, 1987 the Environmental
Protection Administration (EPA) was
organized, with Dr. Eugene Chien as
administrator. Since then, environmental
protection has moved into a new era.
The number of personnel has increased
from 124 in the Environmental Protection
Bureau era to 284; the budget has
increased from US$164.7 million (1987)
to US$ 11.7 million (1989). US$ 2.4
million has been invested for hazardous
waste control in 1989 and US$ 37 million
is estimated to be invested in 1990.
Since the implementation of the "short-
term program for industrial waste
control" we have achieved the following:
Legislation.
We are actively formulating and
amending laws in order to build a sound
foundation for environmental protection
affairs. The new Waste Disposal Act
includes: the definition of and standards
for hazardous waste; different treatment
procedures for solid and hazardous
waste; regulations for the import and
export of toxic substances, etc.
Standards for hazardous waste are
listed in Table 4. We have several
newly proposed regulations based on the
Waste Disposal Act. These regulations
govern: the storage, cleaning and
treatment of industrial wastes; facility
standards; public and private waste
disposal organizations; the setting up
of industrial waste treatment'facilities;
the certification of public and private
waste disposal technicians; and the
incineration and landfill-dumping of
hazardous waste.
Control System
364
-------
The EPA has declared that
hazardous waste is the top priority
of this year's work. The framework
for waste control, as presented in
Figure 1, has already been organized.
We have set aside as top priorities
for special improvement and monitor-
ing programs fourteen industries:
the smelter, oil-refinery, petro-
chemical, dye and its intermediate
substances, Ti02 and related products,
asbestoes, coal & related products,
metal surface treatment, textile,
tannery, waste recovery, nickel,
cadmium, lead and mercury battery,
acid-alkaline, pesticide, and en-
vironmental pesticide industries.
Laboratory and hospital waste have
become additional targets recently.
Table 4 lists the extraction test
limits for hazardous waste.
Treatment Centers
The EPA is planning to set up
waste treatment demonstration centers
in northern and southern Taiwan. (2)
Planning stages have been completed
and funds for the first phase of the
project have been obtained from the
1987 and 1990 fiscal year budgets. A
treatment center for the concentrated
solution of electroplating is being
planned and is expected to be in
operation by June 1990. (3) The Waste
Disposal Act allows the organization
of private waste teatment businesses,
treatment facilities , landfills or any
qualified TSD facilities.. Several'com-
panies have already planned to move
into the field.
Recycling and Reuse
The Industrial Technology
Research Institute organized the
first Waste Exchange Information
Center (WEIC) in November. 1987.(4)
WEIC provides information on waste
for companies which produce or use
waste. Up to the present, 292 facto-
ries have registered their infor-
mation, 486 exchange cases have been
recorded and a total of 40,000 tons
waste have been exchanged.
Table k. Standards for hazardous
waste extraction tests
Toxic chemical substance Extractable Content
Limit (mg/1)
Organic mercury
Mercury
Lead
Cadmium
Chromium
Chromium (VI)
Copper
Zinc
Arsenic and arsenic
compounds
Cyanide compounds
Pesticides
1. organic phosphorus
2. carbarmate
3. organic chloride
2,3,7,8, tetrachloro dioxine
Asbestos
Chlorinated solvents
NO
0.25
5.0
0.5
10.0 :
2.5
15.0
25.0
2.5
5.0
2.5
2.5
0.5
NO
_
-
-
-
-
-
-
-
-
-
-
-
-
-
15
1
REFERENCES
1. EPA, R.O.C., 1988, Short-term
Control Measures of Hazardous Waste,
EPA-043770030.
2. EPA, R.O.C., Report on the Plann-
ing of a Hazardous Waste Treatment
Demonstration Facility, 1988, EPA-
044770026.
3. Industrial Development Bureau,
R.O.C., Planning of an Industrial
Waste Treatment Facility, 1988,
Unpublished.
4. Union Chemical Laboratory, Indus-
trial Technology Research Institute,
Waste Exchange Communication, 1987.
365
-------
1
Solid
Waste
^
hazai
hazardous
wast e
4, Standards of (Act 2)
(Act 17)
(Act 19,34)
> Conditional
# record needed
* Record needed
(Act 15)
(Act 15,16)
Figure j Framework of hazardous waste control program
366
-------
CHEMICAL WASTE MANAGEMENT IN HONG KONG
M.J.Stokoe
R.W.Jordan
R.Tong
Environmental Protection Department
Hong Kong Government
ABSTRACT
In Hong Kong, the control of chemical wastes is
provided for in the Waste Disposal Ordinance. The enabling
regulations of the Ordinance are presently being drafted and will
be enforced in the near future. Presently, because of the lack of
legislative control together with a general lack of knowledge on
chemical wastes and the unavailability of suitable treatment
facilities, the majority of the chemical wastes generated are
being discharged into the sewers or drains. In order that the
control regulations can function effectively, it is decided that
a Chemical Waste Treatment Centre (CWTC) has to be provided by
Government to ensure that the proper treatment facilities are
available to the industry in the first place. As the majority of
the chemical waste producers in Hong Kong are small generators,
it is envisaged that most of these waste generators will have to
rely on the CWTC for the proper treatment of their chemical
wastes. The CWTC will also provide a waste collection service to
collect and transport the chemical wastes from the industrial
establishments to the CWTC. The waste generators are required to
provide sufficient interim storage for their waste prior to their
collection.
Because of the scarcity of land in Hong Kong, most of
the small chemical waste generators are located in multi-storey
industrial buildings on a shared occupancy basis. The storage
space available in these small establishments is generally
limited and is, as a rule, of second priority to production
areas. The current study attempts to identify the potential
problems that may arise due to the requirement for industries to
provide interim storage of chemical wastes, and to provide a set
of basic solutions to alleviate the problem.
367
-------
INTRODUCTION
Over the past decade, the
economy of Hong Kong has experienced
tremendous growth. The GDP having
risen by an average 10% per annum.
This is largely due to the perfor-
mance of its export-oriented manufac-
turing industry. As a result of the
increased industrial activities,
large quantities of chemicals are
consumed, which, in turn, results in
a large amount of chemical waste.
Chemical wastes are also generated
from non-manufacturing industries,
mainly as residues from the storage
of materials and damaged or unwanted
products. According to a recent study
commissioned by the Environmental
Protection Department (EPD), about
100,000 tonnes of chemical wastes
were generated in Hong Kong in 1987.
These include waste acids, alkalis,
etchants, toxic metals bearing
wastes, organic solvents and oily
wastes. Table 1 is a breakdown of the
estimated waste arisings of the dif-
ferent types of chemical wastes (3).
In Hong Kong, the control over
chemical wastes is provided in the
Waste Disposal Ordinance. The ena-
bling regulations of the Ordinance
are presently being drafted and will
be enforced in the near future, sub-
ject to the completion of the legis-
lative procedures. Presently, because
of the lack of legislative control
coupled with a general lack of expert
knowledge of chemical wastes and the
lack of suitable treatment
facilities, only a limited quantity
of chemical wastes are disposed of by
approved means, such as at a Govern-
ment operated codisposal landfill or
in house treatment plant. The
majority of the chemical wastes are
simply discharged into sewers or
drains. In order that the proposed
Waste Disposal (Chemical Waste)
Regulations can function effectively,
a Chemical Waste Treatment Centre
(CWTC) will be provided by Government
to ensure that the proper chemical
waste treatment facilities are avail-
able in the first place for use by
industry. As the majority of the
chemical waste generators are small
establishments, it is envisaged that
most of the waste generators will
have to rely on the CWTC to properly
dispose of their wastes. It is in-
tended that a design-construct-
operate contract will be awarded to a
consortium with world recognised ex-
pertise in the design, construction
and operation of centralised chemical
waste treatment facilities. The
treatment facilities to be provided
will include oil/water separation
units, chemical/physical treatment
units and a high-temperature in-
cinerator. The contractor is also en-
couraged via the contract terms to
establish material recovery units as
optional additions to the CWTC. The
CWTC operator will also provide a
collection service to collect and
transport chemical wastes from the
waste producers' premises to the
CWTC. The waste producers will be
required to provide interim storage
of their waste prior to the collec-
tion.
PURPOSE
As in many other cities the
368
-------
majority of chemical waste producers
in Hong Kong are small generators. An
analysis of the size distribution of
the potential waste generators indi-
cated that about 85% of the in-
dustrial establishments employ less
than 20 workers, and about 95% employ
less than 50 workers(5). Available
statistics (1) indicated that about
90% of the chemical wastes generators
are located in multi-storey in-
dustrial buildings. According to a
recent survey conducted by EPD on in-
dustrial buildings in the Kwai Shing
area (a heavily industrialised
district) (4), a typical multi-user
industrial building could hold up to
680 factories, and the average floor
areas per establishment could range
from 25 to 2800 square meters. These
waste producing establishments rep-
resent the majority of the chemical
waste producers in Hong Kong. Storage
areas are always secondary to produc-
tion and storage of chemical wastes
will be given an even lower priority.
The safe storage of chemical wastes
before their collection by the CWTC
operator will be a difficult issue
complicated by the requirement for
the safe storage of incompatible
chemical wastes.
The objective of the current
study is to identify the potential
problems that may arise during the
operation of the CWTC in respect of
the chemical waste storage in the
waste producers' premises, so that
practical solutions could be worked
out beforehand to alleviate the
situation.
APPROACH
There are about 50,000 in-
dustrial establishments in Hong Kong
(2). Of these, about 11,500 in-
dustrial establishments are con-
sidered to be potential chemical
waste generators. In order to ac-
curately assess the potential
problems that may arise when the
Waste Disposal (Chemical Waste)
Regulations are enforced and the
chemical waste generators are forced
to provide interim storage for their
wastes, an industrial survey was com-
menced in summer 1988. The survey
concentrated on a number of target
industries that are considered to be
potential chemical waste generators.
These target industries include
electroplating, electronic, fabri-
cated metal, electrical machinery,
transport equipment, non-ferrous me-
tal products, spray painting and op-
tical equipment.
The survey was conducted in
the following manner:
Questionnaires were first sent to
selected industrial establishments to
inquire into their chemical waste
generation pattern. Specific ques-
tions were included to probe into the
practical problems related to waste
storage. Follow up visits were then
conducted to the selected estab-
lishments by appointment. During the
site visits, the field officers con-
ducted detailed inspection/interviews
with the responsible personnel of the
industrial establishments on the
issue of chemical waste generation
and present methods of disposal. Ob-
servations were made in respect of
369
-------
the ability of the establishment to
provide interim storage of chemical
wastes. In many case, the field of-
ficers had to assist the interviewees
to accurately complete the question-
naire. In all cases, the field of-
ficer would go through the production
processes with the interviewees to
understand and identify the quality
and quantity of the chemical waste
streams. The information obtained
were then analysed with a view to
identify the potential problems as-
sociated with the interim storage of
chemical wastes and the feasible
solutions. Appendix 1 is a sample
completed questionnaire from a typi-
cal electroplating factory.
PROBLEMS ENCOUNTERED
The questionnaire-visit ap-
proach was considered useful as it
was noted that many of the inter-
viewees had little knowledge on the
chemical wastes they generated and
their proper management. The follow
up visit and the field officers' as-
sistance in the completion of the
questionnaire is considered vital in
attaining accurate information for
the subsequent analysis.
Another problem encountered was
related to the availability of
storage space. Most of the chemical
waste generators are small estab-
lishments located in multi-storey in-
dustrial buildings where shared
tenancy on the same floor is a common
feature. The solutions proposed must
be practical in terms of utilisation
of space, affordable in terms of
financial requirement, and acceptable
in terms of the safety standpoint.
RESULT
So far, questionnaires were
sent to a total of 87 industrial es-
tablishments, 74 questionnaires were
returned or collected. Follow up
visits were conducted to 37 estab-
lishments. Of the 37 establishments
visited, 23 generated appreciable
amount of chemical wastes. A number
of the chemical waste generators,
notably the printed circuit boards
(PCB) manufacturers and the
electroplators, generate more than
one waste stream. Some of the wastes
generated were found to be incom-
patible, such as cyanide waste and
waste acids from a typical
electroplating factory. In such
cases, different waste containers,
and, depending on the waste gener-
ation pattern, physically separated
interim storage areas would be
required.
A number of points were observed in
the survey:
(a) Due to the shortage in factory
space, storage areas were given very
low priority in the utilisation of
available area. In many cases,
storage areas for chemicals con-
stitute only about 0.5 to 3 percent
of the total floor area. In some ex-
treme cases, there were no
"designated" storage area for chemi-
cals in the factories at all. (see
Plates 1 & 2). But much of the dif-
ficulty could be alleviated via bet-
ter planning in terms of production
processes and factory layout.
370
-------
(b) In many cases, the chemical waste
generators have little knowledge
about the general properties of the
raw chemicals that they are using
(except, perhaps, for the role of the
chemical in the production process),
and even less so for the chemical
wastes generated. It was noted that
in many cases, different chemicals,
some of which were incompatible, were
seen stacked together at the same
place, (see Plates 3 & 4). This ob-
servation calls for the need for an
extensive programme to educate the
chemical waste generators on the
properties and hazards of chemical
wastes (and chemicals) and the proper
ways to manage these wastes. It also
points to the need for comprehensive
Codes of Practice on the proper
methods to handle, store and dispose
of chemical wastes. Also, any
proposed solution for the interim
storage problem has to be simple and
straight forward in order that the
waste generators can easily comply.
(c) The rate of chemical waste gener-
ation could be highly variable with
respect to the type of industry as
well as the actual production
processes adopted in a particular
factory. In most cases, only small
quantities of chemical wastes are
generated during the processes and
the discharge frequency could be as
low as once per half year (when the
spent chemicals in a process bath are
discharged). For PCS and electroplat-
ing factories, however, substantial
quantities of different chemical
wastes could be generated each day.
The discharge rates and frequencies
would have serious implications on
the collection requirements for the
CWTC operator as well as the interim
storage requirements for the gener-
ators.
(d) Most of the establishments inter-
viewed were unwilling to disclose
financial information, but from the
information gathered, the capital set
up cost of a typical chemical waste
generating small "flatted factory"
(employing less than 20 operatives)
could be in the range of 300 thousand
to 6 million Hong Kong dollars (i.e.
about 40 to 800 thousand US dollars).
Assuming normal plant life of 7
years, the amortised capital cost
would be in the range of 43 to 860
thousand Hong Kong dollars Per annum.
Similarly, the available range of an-
nual turnover rate is 1.9 to 9.6 mil-
lion Hong Kong Dollars.
From the survey findings, EPD
is in the process of designing a num-
ber of basic plans for the interim
storage of chemical wastes for the
reference of (small) chemical waste
generators. In addition, general
"collection plans" are being con-
sidered to provided guidance for the
CWTC operator on the collection of
the chemical wastes. It is considered
that the recommendations should be :
(a) Simple, so that the waste gener-
ators can easily and accurately fol-
low.
(b) Practical, so that the recommen-
dations can be put into practice.
Care should be taken to avoid onerous
and financially unaffordable arrange-
ments which could act as disincentive
deterring waste generators from
making use of the CWTC service to
371
-------
solve the chemical waste disposal
•problems.
(c) Notwithstanding (a) and (b), the
recommendations must provide adequate
safety 1n practice.
Considering the above con-
straints, it is recommended that :
(a) In all cases, waste generators
should review carefully their exist-
ing production processes to identify
opportunties for waste, reduction or
material recovery.
(b) The CWTC operator will provide
the containers for the interim
storage of chemical waste. This will
ensure that only the correct types of
containers are used. A reverse-
milkman service should be provided in
which the CWTC operator will collect
chemical wastes from the factories at
the required frequency while at the
same time provide adequate empty con-
tainers for the interim storage of
chemical wastes generated before the
next collection.
(c) The waste generators will be
required to provide simple interim
storage for the chemical wastes-
generated. No permanently constructed
stores are specified but different
types of wastes (especially incom-
patible wastes) should be stored
separately. It is proposed that cor-
rosive wastes or non-flammable toxic
wastes should be stored in bunded
areas with impermeable bunds and
floor to avoid spillages and
seepages. Impermeable floor areas can
be created by chemical treatment of
existing floor or the use of large
drip trays made of appropriate
materials grouted to the floor. For
flammable wastes, free standing
safety storage cabinets up to
required safety specifications could
be used if permanent stores are not
available or impractical.
(d) To fully utilise the vertical
headroom within the storage areas,
purpose built multi-tier racks could
be used for the storage of chemical
wastes. For small generators, small
waste containers such as 25-litre
jerricans instead of the standard
200-litre drums are recommended for
easier handling and better
flexibility in stacking, and as a
result reduce the area requirements.
Case studies of typical multiple
wastes generators (such as
electroplating factories) adopting
the above recommendations indicated
that both the set up and the amor-
tised cost for the installation of
the above recommended storage ar-
rangements are much less than
HK$10,000 per annum - far less than
their annual turnover rate.
(e) To ensure safe storage of the
chemical wastes and the efficient
operation of the CWTC, it is further
recommended that maximum chemical
waste storage limits should be set to
limit the total quantities of dif-
ferent chemical wastes allowed to be
stored in the industrial premises
before collection. At the same time,
a collection "trigger" level should
be set. When the chemical waste under
interim storage reaches the "trigger"
level, the waste generator may call
upon the CWTC operator to collect the
waste. This will ensure that suffi-
372
-------
cient quantities of chemical waste
are available at the time of collec-
tion while at the same time the
chemical waste generators are respon-
sible for providing an acceptable
level of storage space for their
wastes.
(f) All storage areas as well as the
containers have to be properly
labelled. Preferably, the containers
should have colour codes to assist
easy identification of their content.
It is envisaged that the above recom-
mendations could form the basis of a
series of practical but safe solu-
tions to assist chemical waste gener-
ators to provide interim storage for
their chemical wastes before their
ultimate collection by the CWTC
operator. In view of the wide varia-
tions in the industry types involved
and their production practices, it is
considered that the current survey
should be further extended. At the
same time, consultation and education
programmes should be organised via
meetings and Codes of Practice issued
to the industrial sectors.
REFERENCE
(1) Anon. 1987a, directory of Hong
Kong Industries, 1987, Hong Kong
Productivity Council.
(2) Anon. 1987b, Names of Buildings
1987, Rating and Valuation Depart-
ment, Hong Kong.
(3) Anon. 1988, Environment Hong
Kong 1988 - A Review of 1987, En-
vironmental Protection Department,
Hong Kong.
(4) Chan, E., 1989, Industrial Sur-
vey of Kwai Shing in Kwai Tsing Dis-
trict, Environmental Protection
Department, Hong Kong.
(5) Unpublished industrial survey
data, Environmental Protection
Department, Hong Kong.
373
-------
Table 1
CURRENT AND LIKBLT FUTURE CHEMICAL HASTE ARISIHGS
Waste type
Acid
Alkali
Copper containing waste solution
Zinc containing Haste solution
Nickel containing waste solution
Other metal salts containing waste solution
Cyanide containing waste solution
Hon-chronluo bearing oxidizing agents
Chromium bearing oxidizing agents
Haloganated solvents
Hon-halogenated solvents
Phenols and derivatives
Polymerization precursor and production wastes
Mineral oil
Fuel oil
Oil/water mixtures
Pharmaceutical products
Mixed organic compounds
Mixed Inorganic compounds
Miscellaneous chemical wastes
Interceptor & Treatment Plant sludge
Tank cleaning sludge
Tar, asphalt, bitumen and pitch
Tannery wastes
Printing wastes
Dyestuff wastes
Plating bath sludge
Paint wastes
Haste catalysts
MARPOIi Annex I (a)
MARPOL Annex 11 (b)
Total (rounded up)
1987
20,000
35,000
12,640
13
120
1,200
100
10
55
1,300
1,500
2
40
5,600
50
12,000
1
130
70
30
40
1,000
140
400
90
70
10
640
4
5,000
600
98,000
1992
22,000
42,000
19,150
13
140
1,300
130
11
59
1,700
1,800
2.2
42
5,700
51
13,000
1
140
74
32
42
1,000
143
400
93
59
11
700
4
H/A
N/A
110.000
1997
25,000
50,000
25,160
14
160
1,400
160
12
68
2,000
2,100
2.4
44
5,900
53
13,000
1
150
78
35
44
1,000
146
400
94
52
12
750
4
N/A
H/A
130,000
(a) Oily wastes arising from the application of annex I of the International
Convention for the Prevention of Pollution from Ships (MARPOL Convention).
(b) Chemical wastes arising from the application of Annex II of MARPOL Convention.
374
-------
Plate 1 A Typical Small Scale Electroplating Factory
in Hong Kong. (Note the Storage of Different
Types of Chemicals in the Foreground)
Plate 2 A Medium Scale Factory (Electroplating)
(Note the Stacking of Chemicals along
the Process Line)
375
-------
Plate 3 Storage of Different Chemicals at the Same
Place - A Typical Scene in Many Factories
Plate 4 Another Shot Showing the Storage of Different
Chemicals at the Same Place. (Note the presence
of Cyanide Containing Salt (the Green Drum) and
Chromic Oxide (the Red Drums) )
376
-------
I. COMPANY PROFILE
Appendix 1
Company Name :
ife &
Address :
Tel No. :
044 «£ I
w m A
Contact Person ;
Position :
Total Floor Area
FT*
C?
Workforce : • (i) Management
(ii) Production
Category of D.G. Store :
(if any)
Location of D.G. Store :
(within/outside factory)
(SIC
: 583435
7
377
-------
^£ 21 DTP if?
.II. PRODUCTION DETAILS :
, . 2iiffl(St$i&jlfcS) .
(a) Products Ctype and Quantity) :
(b) Production Processes : (please identify chemicals used and waste
streams)
Material^ ^ j| ±M ffi ) Process (
waste (
Wc«.
j Cu-ttuao,
/it ^8-
ffc Ju X 6^J '^ $• (MM )
(c) Raw Materials (Types of materials )
0
ttCt.
^j,
1 I*)
378
-------
II. (d) Chemical Consumption ( fff /fj -ft f$ ml ± M $5 >>-
Chemical
Averag*
Consumption
Rate/Hontii
I
Dilution
Pnctor
of Storage
D.G. Store
I.ock«rs
Anvuhrrc
tnpt'lp f nrtorv
'
Sire of
ConCfl tn<» rs
AGIOS
ALKALIS
METAL SALTS
CYANIDES
11.1 *£
SOLVENTS
ififi'l
AOIIESIVES
OILS Hll ft]
DETERGENT
DYES ;S
PETROLEUM
FLAMHABLES
PRINTING INK
4
y <2&b
Jfttsi*\cM>
JZ_
'£k_£
^t£nfcjtw
^L
fe
Water Consumption : __^___
*f tick vh*r* apprnprtare
m/month
379
-------
lit. IA..II- limiKHATII.ll !- Ifi \ \ f'£ t't t I
(M| l.i>|ul3C JH
I/
^/
.
tM^poKf wlch Sold co
solid
munlc Ipa 1
1
dealers
j|^iU^f
M&WKIH-IA
Dispose after
t reaCmcnc
(Please specify
Creatmenc
method)
f l4ii?Wt )
Kemnrks
p- ^^ (/v)-rii
P^f°, lrfc*T^
X
"1 *^T" -rtAyTL^-^^0-
^tfijrrt.-v^i
d.^T^-rtui.
V
v^
/tu^y^' ,^^Q
tfs<
\
* pltaie tick wher* appropriate
380
-------
III. (Continued)
U,J Solid/Sludges U
-------
IV. Are there any facilities for waste recycling? If yes, please
provide details :-
V. Area usage within your factory :-
Percentage %
Office
i fc * DD0 ± ft &
Storage area for chemicals
Storage area for unfinished products/tools/etc.
& ?1 ± ffl ife
Production Area
JJ~\ VfO / .rf*. HLi
:- (i) Preparation/assembling
(ii) Occupied by production machinery/
equipment (excluding (i))
)
V.
382
-------
VI. Investment Figures :- (These data are used to assist us in
understanding the profile of your industry, data will be kept
confidential)
(a) Estimated investment cost required
for a similar new factory :
HUG
m c
(c) Typical monthly operating cost :-
(including all charges/expenses)
7"
wi »,
-------
AEROBIC MTNERALIZAITaT OF
ORGANIC CONTAMINANTS BOUND ON SOIL FINES
Robert C. Ahlert, PhD, PE, Dist. Prof.
David S. Kbsson, PhD, Asst. Prof.
Chemical & Biochemical Engineering
Rutgers University
P.O. Box 909, Piscataway, NT 08855
and
John E. Brugger, PhD, Project Scientist
Risk Reduction Engineering Laboratory
USEPA, Edison, NJ 08817
ABSTRACT
The goal of the overall research program, a part of which is
discussed in this paper, is to demonstrate a sequence of
aerobic/anaerobic microbial process steps for degradation of
cxMTtaminated soil fines and slurries of soil fines. Toward this end,
it must be possible i) to assay individual organic species and total
contaminant organic carbon in soils of varying properties, ii) to
separate whole soils into fractions according to particle size, and
iii) to assay [as in i) ] reactor slurries containing suspended soil
particles, microbial culture and dissolved, dispersed and sorbed
organic cxsntaminants and metabolites. These techniques are required
to define the nature of the contamination, devise operating
conditions to facilitate microbial contact, and assure complete
mineralization of target csontaminations and "clean" residuals.
The first major section of this paper describes the development of
analytical methodology for whole soil and soil fractions,- in
parallel, techniques for mixing/homogenizing, fractionation and
extraction used in sample preparation are discussed. It has been
possible to separate soil fines and some "bulk" organic matter. A
large part of the total organic chemical contamination is due to
sorption and physical pore interactions with the fine particle
fractions [clay minerals and humic substances] of whole soil.
A second section describes microbial degradation experiments.
Systems and procedures for microbial reactions were designed and
implemented to accomodate the properties and behavior of target
substrates. Both shake flask and fermentation reactions are being
carried out on slurries of soil fines. Low molecular weight
polynuclear and chlorinated aromatic hydrocarbons are readily
biodegraded.
384
-------
INTRODUCTION
Distillation bottoms and sludges
from benzene-toluene-xylenes [BTX]
production were impounded for
several decades. The production
process consisted of the catalytic
cracking of naphtha, in the presence
of fuming sulfuric acid, and
distillation. Therefore, lagoon
contents include naphtha-related
compounds, distillate residues and
compounds resulting from reactions
of these species and sulfuric acid.
Possible contaminants include, but
are not limited to, simple aromatic
species, polynuclear aromatic
hydrocarbons [PAHs], phthalates, as
well as sulfonated derivatives of
these compounds. Many of these
species have slight solubility in
water and/or an affinity for some of
the constituents of soil and have
migrated into and through the soil
immediately adjacent to the lagoon.
During this study, soil samples were
obtained from the containment area
surrounding the lagoon. The
impoundment has been designated a
CERCLA-NPL site; it exceeds ten
acres in extent and contains an
estimated 100,000 cubic yards of
residues. The contents of the
lagoon have separated into several
distinct layers that include, in
bottom-to-top sequence, a solid
mixture of organic and inorganic
substances, a tar-like layer, a
layer of viscous organic matter, and
a floating aqueous layer.
ANALYTICAL APPROACH
Initially, soil samples are mixed
and homogenized. The resulting
material is air-dried and sieved
through a 3-cm brass screen to
remove debris, rocks and gravel;
this procedure also breaks up macro-
agglomerates. A second sieving,
with a 5-mm screen, improves
homogeneity, enhances mixing and
helps toward analytical
reproducability. Direct solvent
extractions of homogenized,
contaminated soil utilize methanol,
cyclohexane, or methylene chloride.
Methanol has relatively high
polarity, cyclohexane is a model
cyclic compound, and methylene
chloride is a moderately polar,
volatile compound with broad solvent
capabilities.
Gas chromatography [GC] is used to
identify and quantify compounds in
soil extracts. USEPA Test Methods
602 and 610 are employed in these
analyses. Method 602 is used to
assay aromatic species in GC column
effluent with a photoionization
detector [PID] in series with an
electrolytic conductivity detector
[ELCD]. Method 610 is employed to
detect PAHs and phthalates,
utilizing a flame ionization
detector [FID]. Standard solutions
are assayed in sequence with solvent
solutions to match retention times
for compound identifications. Peak
areas are used-to construct standard
curves and provide a basis for
determination of contaminant
concentrations. Compound
concentrations are calculated from
both PID and FID output to check
analytical consistancy.
Experimental Methods
Soil contaminant levels were
initially estimated to fall between
2 and 5 % on a dry weight basis.
Direct solvent extractions are
carried on varying masses of soil
with the goal of limiting
contaminant concentrations to about
100 mg/L in extract solutions. This
target concentration was adopted
to avoid overloading the detectors.
Soil masses varying from 0.15 to
0.38 g are extracted with a fixed
volume of solvent.
Duplicate amber serum bottles, each
385
-------
with label, septum and aluminum cap, 1 pg when used in the detection of
a V»a iifA-i rtKft/4 i.i-S 4*U-\Mx*4-'l-l*»*nr''^^rtrt _ i_ i _ _ • • • • . __
are weighed with a Mettler PE3600
balance. Bottles are 100 ml in
volume. Soil is added until target
weights are attained; approximately
50 ml of methanol, cyclohexane or
methylene chloride is added to both
bottles. Bottles are sealed with
the septa [Teflon-coated neoprene]
and reweighed. Experimental errors
include the small discrepancies in
obtaining target soil masses and
measuring solvent volumes. In
general, these are accounted for in
concentration calculations. A
second form of the experiment is
carried out to facilitate compound
identification. It is the same in
all respects, except that 20 g of
soil are added; since quantification
is not desired, duplicates are not
performed.
Extraction vessels are shaken for
approximately one hour. This time
was found to be adequate in earlier
studies; however, it assumes that
only readily reversible, high-rate
sorption processes are involved.
Higher energy binding processes and
sorbate trapped by capillary forces
would not participate in such short-
term partitioning. Extract
solutions are filtered through 0.2-
um MSI Cameo II 25-mm disposable
syringe filters, into duplicate 5-mL
chlorinated compounds or to verify
PID results. Sample size is 2 uL.
Method 610 is applicable to PAHs and
phthalates. It utilizes a 1.8-m
long by 2-mm ID glass GC column
packed with 100/200 mesh Chromosorb
W-AW-DCMS, coated with 3 % OV-17.
Oven temperature is held at 100°C
for 4 min; a 8°C ramp takes the oven
to a final temperature of 280°C.
The FID has detection limits of 10
to 100 pg. Sample size is 5 uL.
Chemical Oxygen Demand [COD] is
determined for some contaminated
soil samples. This procedure is
identical to that described in
Standard Methods". The COD has some
value for comparison with carbon in
identified species, to estimate
extraction and identification
efficiencies.
Analytical Results
Extracts generated in Experiment
12888 were distinctly different in
color. After filtration, the
cyclohexane extract was translucent
orange, methylene chloride gave an
opaque brown liquid, and the
methanol solution was clear and tan.
This appeared to be evidence for
susbtantial variation in extraction
efficiency. In addition, the high
-. **• --/ - •- — — -,[-..»,«„«„ ,Mi- V* i i i v* i t^i iv*jr • in auuitsiuii)
vials. Samples are stored at 4°C, polarity of methanol leads to
4° ^ *** •! M 4 VM £»WK>».B.*1_J.,?*I': J_ •_._•! • . -. _ _ __
to minimize volatilization losses,
and enclosed to exclude light and
avoid photolytic chemical reactions,
destruction of soil aggregates;
thus, methanol is capable of
extracting contaminants held in
micro- and macro-pores by
capillarity and interfacial
tension. Solvents can be compared
on the basis of the mass of
Method 602 utilizes a 1.8-m long by
2-mm ID stainless steel GC column
packed with 100/200 mesh „.. _ W1 „„„ ,,,D
Supelcoport, coated with 5 % SP-1200 naphthalene extracted. No
and 1.75 % Bentonite-34. Oven naphthalene was extracted by
temperature is held at 50°C for 2 cyclohexane; methanol and methylene
min; a 6 C ramp takes the oven to chloride extracted 1,089 and 1,403
2? tor ? final Per1od of 23 min. mg/kg dry soil, respectively. Thus,
The PID has detection limits of 1 to for naphthalene, solvent power
10 pg for unsaturated carbon bonds varied considerably.
found in aromatic compounds. The
ELCD has detection limits of 0.1 to The PID sees cyclohexane and
386
-------
impurities in methylene chloride.
The ELCD detects chlorinated
compounds and is overloaded by
methylene chloride solutions. The
consequence is chromatogram baseline
fluctuations and large peak area
integration inaccuracies. Methanol
was the only solvent suited to
Method 602. Methylene chloride was
used in conjunction with Method 610.
Eight major organic compounds were
identified and quantified in
methanol solution. In order of
decreasing concentration [mg/kg],
they are: naphthalene - 1,090; 1,2-
dichlorobenzene - 360; toluene -
150; xylene isomers - 145; benzene -
113; and, ethyl benzene - 28. The
eight species account for 64 % of
the total peak area of
chromatographic responses.
Benzene, toluene and the xylenes are
primary products of naphtha
distillation. GC residence times
for standard solutions and extract
solutions, with PID detection,
varied less than 0.004 sec for this
group. Naphthalene is also a major
component of naphtha; residence
times differed by 0.004 sec.
Ethyl benzene [EB] is formed by
catalytic reaction of benzene with
ethylene, an olefin found in
industrial naphtha. In BTX
production, sulfuric acid is the
catalyst. EB residence times varied
by 0.008 sec.
The appearance of 1,2-
dichlorobenzene was signaled by the
PID and verified by the ELCD, a
halogen-specific detector.
Chlorinated compounds are not
normally found in naphtha nor are
they produced by sulfuric acid
catalysis. The presence of this
compound may indicate disposal to
the impoundment from another
manufacturing source or a spill
clean-up activity. Two substantial,
unidentifiable peaks were
encountered with Method 610; neither
was observed with Method 602. These
peaks correspond to compounds that
are believed to be sulfonated
aromatic hydrocarbons; operating
temperatures for Method 602 preclude
elution of such higher boiling
species. Naphthalene sulfonic acids
represent compounds of higher
molecular weight and boiling point,
requiring increased GC oven
temperatures [Method 610] and
extended residence times.
Standard solutions included seven
purgable aromatic compounds in
methanol [see Table la] and fifteen
PAHs in a 50:50 methanol:methylene
chloride mixed solvent [see Table
lb]. Gas chromatograms are obtained
by Method 610 for the PAH standard
solution and methanol extract,
respectively. Similarly, are GC
outputs are obtained by Method 602
for the purgable aromatic standard
and the methanol extract,
respectively. Naphthalene is run
with a modified version of Method
602.
COD has been measured for several
soil samples. This assay is used to
determine the total oxygen required
to fully oxidize all reduced
species; it does not distinguish
between contaminant organic carbon
and hydrogen, soil organic matter,
and metals in reduced or partially
oxidized states. The COD of sieved,
air-dried, unextracted soil
corresponds to about 50 mg C/kg
soil. In comparison, the total
organic carbon [TOC] associated with
the eight quantified contaminant
species cited above is approximately
1.6 mg C/kg soil or 3.2 % of the COD
carbon equivalent. This is not a
reflection of extraction efficiency.
However, CODs run on soil samples
taken at points remote to the
disposal lagoon average close to 50
mg C/kg soil. Thus, extraction
efficiency with methanol is probably
relatively good.
387
-------
Table la
Purgable Aromatic Standard Mixture [602-M1
fin Methannl)
Compound
Benzene
Toluene
Ethyl benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-Di chlorobenzene
1,4-Di chlorobenzene
Concentration fma/l)
2000
2000
2000
2000
2000
2000
2000
Table Ib
Polynuclear Aromatic Hydrocarbon Standard Mixture [610-M1
nn 50:50 MethanolrMethvlene ChloHdp)
Compound
Acenaphthene
Fluoranthene
Naphthalene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Chrysene (93 %)
Acenaphthylene
Anthracene
Benzo(g,h,i)pyrene
Fluorene
Phenanthrene
Dibenzo(a,h)anthracene
Indeno(l,2,3-cd)pyrene
Pyrene
Concentration (mg/l)
1000
200
1000
100
100
200
100
100
2000
100
200
200
100
200
100
100
388
-------
MICROBIAL APPROACH
The solubilities of PAHs in aqueous
media decreases with increasing
molecular weight; for four rings or
more, saturation concentrations are
very low and difficult to measure.
Within this context, rates of
aerobic metabolism of PAHs are
insignificant. In addition, this
class of aromatic compounds has a
high affinity for several
constituents of natural soils, i.e.,
soil organic matter [humic
substances] and clay minerals. PAHs
are lipophilic, i.e., prefer
sorptive association and/or
dissolution in organic rather than
hydrophilic phases. In addition,
many multi-ring molecules can
migrate into stable positions inside
clay mineral structures, either
diffusing between laminae in mica-
like structures or entering the
crystal lattice directly. The
latter process is a form of
clathration. These preferred soil
components are, in general, the
smaller particle fractions of the
soil system. Thus, it is possible
to introduce PAHs into a microbial
systems in slurry form, with
substrate(s) bound to particulate
matter.
A slurry form of bioreactor has a
number of possible thermodynamic
phases present, including: aqueous
medium with dissolved substrate and
nutrients, substrate bound to soil
particles, suspended single and
clustered cells, substrate emulsions
and colloids, and cells attached to
the exterior and macro-pore surfaces
of dispersed particles. Sampling
and analytical methodology are
difficult throughout the design and
implementation of biodegradation
experiments.
An earlier Section dealt with the
problems of accurate chemical
contaminant identification and
quantification for soils prior to
chemical or biochemical reaction.
Often tightly-bound PAHs must be
extracted from a soil mass, with
uncertain efficiency or recovery.
Method verification is critical.
Extract solution is assayed by GC or
High Performance Liquid
Chromatography [HPLC]. Multiple
phases must be sampled and
extracted, during and after
biochemical transformations. The
possible appearance of intermediate
or final metabolites, i.e.,
incomplete mineralization, adds to
the complexity of mass balances for
substrate species. COD is a useful
tool for quasi-continuous monitoring
of the progress of slurry-type
bioreactor systems. In well-aerated
aqueous systems, metals are not a
serious factor. However, variable
cell mass and natural soil organic
matter severely limit this approach.
Before slurry reaction can be
undertaken, whole soil must be
reduced to several particle size
fractions. Contaminant PAHs favor
finer particles, thus, fractionation
is a useful way of concentrating
these compounds prior to degradation
experiments. A reproducible method
for soil classification has been
developed and has been demonstrated
to lead to the desired concentration
of target substrates, as follows.
Whole soil is separated into three
phases: a tar-like organic phase, a
coarse sandy phase and an aqueous
suspension of soil fines. The
latter is suited to slurry reactor
experiments. Experiments consist of
small-scale studies in shake flasks,
performed in matrix formats, and
fermentation studies in reaction
vessels of larger volume.
Background
The rates of microbial assimilation
of PAHs have been demonstrated to be
functions of solubility, molecular
389
-------
weight, number of six-member rings,
degree and type of substitution, as
well as environmental conditions,
such as temperature, pH and oxygen
concentration. The solubility of
unsubstituted PAHs,in water, drops
sharply as the number of rings
increases. It rapidly diminishes to
levels that are too low to support
significant biological activity; see
Table 2 for data. Compounds of six
or more rings have vanishing
solubility in water.
The number and type of substituents
on or in a PAH molecule have a
marked influence on solubility. The
solubilities of phenols, nitrogen
heterocyclics, polynuclear polyols,
sulfonates and other mono- and polv-
substituted PAHs are often
significantly higher than the basic
hydrocarbons. Therefore, substituted
compounds are more likely to be
observed as solutes in contaminated
groundwater. Also, surfactants
increase PAH solubility. However,
these compounds complex or "react"
with the high molecular weight
polynuclear species to create a
composite hydrophilic exterior. The
result is either a stable emulsion,
colloidal suspension or micro-
dispersion; it cannot properly be
classed as dissolution in the
thermodynamic sense of a homogeneous
liquid phase. Sodium laurylsulfate
increases the solubility of 2- to 7-
ring PAHs by 2 orders-of-magnitude
or more.
Biodegradation of 2- to 3-ring PAHs
by pure microbial cultures has been
demonstrated; naphthalene,
phenanthrene and anthracene have
been shown to be assimilated
quantitatively. Higher molecular
weight compounds, i.e.,
benzo(a)anthracene and
benzo(a)pyrene, can be degraded to
simpler intermediates in the
presence of supplementary carbon
sources or cometabolites, i.e.,
biphenyl and succinate.
Bacteria concentrate, grow, and form
bioslimes in aqueous boundary layers
at liquid/liquid and liquid/solid
interfaces. Organic cosolvents can
transport PAHs to such interfaces
and increase the rate of
biodegradation. There is no
information to support microbial
metabolism of solid PAHs; similarly,
there is little data on the
bioreaction of PAHs sorbed on
nonreacting surfaces.
Experimental
i) Apparatus
Aerobic biodegradation experiments
are carried out in 60-mL Ehrlenmeyer
flasks, on a laboratory shake
device, or in 3-L [working volume]
fermentation vessels. The shaker
studies are arranged in matrix
format with the following common
composition: 30 ml soil slurry, 20
ml inoculum and 10 ml nutrient
medium [see Table 3]. The
composition of the inoculum is
varied to provide a basis for the
evaluation of substrate volatility
losses, reactor surface wetting and
sorption onto biomass as mechanisms
of substrate disappearance. The 20
ml "inoculum" is live culture; in
sterile controls, the 20 ml of seed
is replaced by autoclaved culture or
deionized water. Controls are
intended to illustrate the extent to
which volatilization and inorganic
surface sorption influence sub-
strate fate. Given that biomass
rendered unviable by autoclaving
retains substantial sorption
capacity, autoclaved culture is
designed to investigate this loss
pathway. Also, this control
provides a zero-time or baseline
measurement for carbon or oxygen
demand. The pH of the composite
solution is adjusted to 7.15 by
addition of a mixture of solid
390
-------
potassium monobasic and dibasic
phosphates.
Fermentation studies are carried out
with a working volume of 3 L
prepared in the same ratio of 3:2:1
for soil slurry:inoculum:nutrient
medium as in the shake flask
studies. The reactor is sparged
with air at the rate of 5 to 6
L/min; it is stirred at 300 rpm.
The pH is maintained at 7.15 by
periodic additions of 0.25M sodium
hydroxide regulated by a pH
controller. Samples are taken at
24-hr intervals, to monitor the
course of reaction. Separate
studies utilizing deionized water
are used to evaluate volatilization
losses.
ii) Soil Slurry
Soil slurry is prepared by a
sequence of homogenization,
extraction, and fractionation steps.
This procedure is designed to create
a suspension of soil fines that
displays minimum variation from
batch-to-batch. Whole soil samples
are homogenized by passing the air-
dried material through a 5-mm
screen, quartering the resulting
solids cone through the apex,
segregating the quarters, and
sieving each quarter to form a new
cone. Soil is sieved three times. A
prescreening with a 3-cm sieve
removes rocks and miscellaneous
debris and, also, serves to break-up
larger clumps of packed soil.
Homogenized soil [84 g on a dry
basis] is extracted with 350 ml of
water at pH 7. Extraction separates
the contaminated soil into three
phases: a tar-like [smell, sticky,
viscous, etc.] organic phase
corresponding to 0.65 - 0.75 % of
the initial dry mass; a mixture of
larger, heavier particles [sand];
and, an aqueous supension [slurry]
of fine soil particles. The aqueous
suspension is separated from the
settleable solids by screening
through a 10-micron sieve; filtered
solids are washed with 650 ml of
water to dilute the filtrate slurry
to 1 liter and the final
concentration. The fractionation
procedure has been found to retain
approximately 65 % of the initial
dry mass of soil on the sieve. The
final slurry of fine particles is
stable and does not show any
evidence of settling under
experimental conditions.
iii) Inoculum
The aerobic inoculum is obtained as
waste sludge from the
Somerset/Raritan Valley Sewage
Authority.
iv) Nutrient Medium
The nutrient medium is prepared from
a conventional recipe for aerobic
cultures and does not employ a
primary or supplementary carbon
source; see Table 3. The soil
slurry supports microbial activity
without the use of a co-substrate;
none is added.
v) Analytical
The progress of substrate conversion
is monitored by COD and GC analyses.
Samples are prepared for the latter
-by extraction with methylene
chloride at 8 ml of solvent for 20
ml of slurry. The remainder of the
assay is carried out in accordance
with USEPA Method 610. A 1.8-m long
by 2-mm ID glass column is packed
with a stationary phase consisting
of 100/200 mesh Chromosorb W-AW-DCMS
coated with 3 % OV-17. Oven
temperature is held at 100°C for 4
min; a 8°C/min ramp increases the
temperature to 280°C. An FID is
used to determine residence times
and peak areas. Gas pressures are
70, 40 and 65 psi for nitrogen,
391
-------
hydrogen and dry air, respectively.
A 5-uL sample is injected into the
EC.
Results
Table 4 summarizes moisture contents
and COD analyses for whole soil and
the fractions generated by repeated
sieving, fractionation and
extraction. COD determinations are
referred to 1 kg of air-dried whole
frM'^ The COD ba"lances vary about
15 % for a typical set; this is a
consequence of an error of at least
± 5 % in this measurement. Slurry
diluted with wash water, has a COD
of 7.1 g 02/L.
When an active microbial inoculum is
combined with a slurry of soil
fines, a lag period of approximately
6 hr is observed. Acid production
in shake flasks and fermentations,
and carbon dioxide generation by
fermentations, are not observed
until after the lag phase. There is
no loss of COD, as might accompany
volatilization or sorption. It is
assumed that COD attributable to
biomass remains unchanged during the
experiment, i.e., growth is
negligible. The reduction of COD in
the flasks inoculated with live
cultures is indirect evidence for
substrate mineralization, as opposed
to physical uptake (sorption) by the
biomass. The results of a shake
flask matrix study are summarized in
I able 5. Flasks contain a working
volume of 60 ml; total reaction time
is so hr. The slurry has an initial
son fines concentration of 30 g/L
?™ a C°D °f 7-° g 02/L; inoculum
COD is 4.75 g 02/L.
Table 6 sumarizes results of a
larger scale fermentation study.
Reactor working volume is 3 L and
reaction time is 68 hr. The slurry
has an initial soil fines
concentration of 45 g/L and a COD of
7.5 g 02/L; as in the shake flask
illustration, inoculum COD is 4.75
g 02/L. Initial reactor COD was
calculated to be 15.9 g 0?; an
experimental determination gave 18.9
g 02. The difference is probably
measurement error, due to the
several phases present, i.e., cells,
contaminated fines, suspended tarry
material and dissolved PAHs. A
volatilization loss study was
carried out with the fermentor. The
COD of air sparged slurry did not
change in 4 days; it remained at 3.1
g 02/L.
Figures 1 and 2 are gas
chromatographs for the fermentation
liquor after 21 hrs and at the end
of the experiment. The
characteristics of the soil slurry
without medium or culture added are
described in Figure 3. The contents
of the fermentor and the original
soil slurry were extracted with
methylene chloride; the volume ratio
was 5:2 for aqueous
suspensionrsolvent. Figures 1 and 2
show definite declines in the number
and size of peaks, especially those
corresponding to low molecular
weight PAHs.
CONCLUSIONS
1) Bench-scale shake flask studies
are performed with slurries of soil
fines and mixed microbial seed.
COD, corrected for the presence of
the inoculum growing at a trivial
rate, is reduced by 50 % in
approximately 80 hours. Similarly,
TOC is measured on settled
[filtered] aqueous phase and remains
low throughout. The latter assays
are a reflection of limited
hydrocarbon solubilities.
2) Larger-scale fermentations are
carried out in 3-liter, stirred,
air-sparged reactors. Inoculum and
nutrient medium are mixed with
slurry. Biodegradation is monitored
by assays on samples of aqueous
392
-------
dispersions and measurement of
carbon dioxide generation rates.
COD reductions exceed 84 % in 68
hours.
3) Whole soil is separated into size
fractions to characterize
contaminant distribution by soil
constituent type and particle size.
It is possible to separate whole
soil into larger, settleable
particles [primarily sand and silt],
slurried fines, a clarifiable
aqueous phase and a bulk organic
phase. Recovery of initial whole
soil COD in the slurry and a tar-
like organic phase is nearly
quantitative.
4) Analytical techniques have been
developed and demonstrated with
whole soil, soil fractions, slurries
of fines and filtered liquids.
These techniques are essential to
the identification of contaminant
species and quantification of
individual and total contaminant
concentrations. Assays are
necessary to define initial and
intermediate conditions and to
demonstrate that contaminant
destruction [mineralization] by
microbial reaction is effective and
approaches completeness.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
393
-------
Compound Mol . Wt. Solubility fua/U #
Naphthalene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)pyrene
Dibenzanthracene
128
154
166
178
178
202
202
252
278
Table 3 - Medi
Constituent
(NH4)2S04
MgS04.7H20
FeCl3.6H20
MnS04.H20
CaCl2
31,700
3,200
2,000
1,300
73
260
140
4
>1
urn Composition
Concentration Ting/Li
1,500
100
0.5
10
7.5
of Rinqs
2
3
3
3
3
4
4
5
5
Table 4 - Moisture Content and COD
Sample Type Composition T%1 COD To Oo/kgl
Whole Soil
Water
Large Particles
Soil Fines
Tar-like Residue
SIurry
16
53
31
89.6
0
trace
19.4
84.5
394
-------
Table 5 - Typical Shake Flask Matrix Study
Initial COD [mg 02]
Final COD [mg 02]
COD Change [%]
COD Change [%]
(corrected for inoculum)
Final Dissolved TOC
[mg c]
[mg C/L]
COD Equivalent [mg 02]
Live
Inoculum
310
204
-34
-50
8
131
18
Autocl aved
Inoculum
310
318
nil
nil
8
131
18
No
Inoculum
210
240
+14
+14
8
133
18
Table 6 - Fermentation Study
COD [mg 02]
COD Change [%]
COD Change [%]
(corrected for inoculum)
Final Dissolved TOC
[mg C] 318
[mg C/L] 106
COD Equivalent [mg 02] 848
21 hrs
13200
30
39
46 hrs
10200
46
61
68 hrs
6900
63
84
315
105
840
288
96
768
395
-------
Figure 1
MIA \ CONfCHSATll ANALYSIS
i.83
a.n
1 42
.98
.98
.94
.85
.8*
1 .3?
11.34
11.85
12.T9
14.39
18.11
18.46
It. 77
£9.12
2«.45
21.38
TOTAL A*CA •
•W.TIFLIE* •
3-12.44
1*24.38
4129*93
1781.33
3*77.37
2176.97
1719.91
1383.36
1346.23
29*3.36
1384.3*
3*31.91
71338.69
1
HH
HH
HH
Hh
HH
HH
HH
HM
HH
HH
HH
HH
HH
HH
HH
3 4«~
1.929
3.393
2.SS9
3.293
4.228
2.?«t
8.367
2.379
14.J72
3.772
T.657
3.9*2
2.403
2.213
1.382
3.972
1.933
7.873
Figure 2
WtA % COWIWSATEB ANALYSIS
RT
9.94
18.49
11.36
11.9*
12.88
13. U
is!rs
28.14
21.31
TOTAL APEA •
ARtA
438.84
3338^47
€71.43
2131.2*
333. ?i
2*!. 54
553.88
386.93
697.79
969.12
12:74. *e
TY«
ev
VV
VV
w
PV
VV
VV
VV
VV
V3
4f>tA ';
3.372
29. 9«*
7.138
ir.593
2.74S
S.Ifl
3.T31
7.9»9
Figure 3
396
-------
ENVIRONMENTAL FATE MECHANISMS INFLUENCING BIOLOGICAL DEGRADATION OF
COAL-TAR DERIVED POLYNUCLEAR AROMATIC HYDROCARBONS IN SOIL SYSTEMS
John R. Smith, Ph.D.
David V. Nakles, Ph.D.
Donald F. Sherman
Remediation Technologies, Inc.
Pittsburgh, Pennsylvania
Edward F. Neuhauser, Ph.D.
Niagara Mohawk Power Corporation
Syracuse, New York
Raymond C. Loehr, Ph.D.
David Erickson
Department of Civil Engineering
The University of Texas at Austin
Austin, Texas
ABSTRACT
Biodegradation is a technically viable and cost effective approach for the reduction and
immobilization of polynuclear aromatic hydrocarbons (PAH) present in contaminated soils and sludges
associated with coal-tar derived processes. While it is widely reported and accepted that PAH
biodegradation in soil systems does occur, the specific controlling mechanisms are not entirely
understood. One common observation among published reports is that the more soluble, lower
molecular weight PAH compounds are biodegraded to a greater extent than the less soluble, higher
molecular weight PAHs.
The rate and extent to which PAHs are removed from soil/sludges is influenced by the combined
and simultaneously occurring effects of volatilization, sorption and biological oxidation. The degree to
which each of these three environmental fate mechanisms occurs is mainly influenced by the
physical/chemical characteristics of the contaminated media, the physical/chemical characteristics of the
specific PAH compounds, and the design and operation of the particular biological treatment process.
Past researchers have interpreted reductions in PAH soil concentrations via biological treatment
using first-order biodegradation kinetic rates. Recent work by the co-authors indicates that sorption and
biological oxidation are interconnected phenomena and that PAH reductions are attributable to a
combination of these effects. For PAH compounds to be susceptible to biological oxidation, it is
postulated that they must be available to the bacteria by diffusing into the bulk aqueous liquid present
in the macropores of the soil matrix. If this desorption/diffusion does not occur, then ft follows that
biodegradation may be impeded.
Under this scenario, PAH desorption from the solid material and subsequent aqueous phase
diffusion may be the rate- limiting steps and, hence, control the rate and extent of both volatilization
and biological oxidation.
397
-------
INTRODUCTION
The viability of biological treatment processes
rests on the ability to degrade and immobilize
soil contaminant concentrations to acceptable
levels through biological and chemical transfor-
mations. At the same time, volatile emissions and
leachate from the treatment unit must be con-
trolled below levels that would cause public
health or environmental concern. Using the
results of laboratory and field experiments, in
conjunction with transport and fate analysis of
soil contaminants, a biological treatment process
can be designed and operated to: (i) maximize
residue degradation and immobilization, (ii)
minimize release of dust and volatile compounds,
(iii) minimize percolation of water soluble com-
pounds, and (iv) control surface water run-on
and runoff. The applied material can be liquid,
semi-solid, or solid. Detailed discussions of
relevant processes are given elsewhere [1, 2, 3,
4,5, 6, 7].
Polynuclear aromatic hydrocarbons (PAH) are
neutral, non-polar organic compounds consisting
of two or more fused benzene rings in linear,
angular, or cluster arrangements. Due to the
acute and chronic toxicity primarily associated
with the lower molecular weight PAHs, and the
potential carcinogenicity associated with the
higher molecular weight PAHs, the United States
Environmental Protection Agency (U.S. EPA) has
designated sixteen PAHs as being environmen-
tally important and representative of PAHs as a
class of compounds. The sixteen are cited in
Table 1 where they have been grouped on the
basis of the number of aromatic rings comprising
a particular compound. These are the same PAH
compounds which are included in EPA's Priority
Pollutant List [8]. While many other PAH isomers
exist, this EPA list will be given focused attention
in this paper.
Due to their toxicity at specific concentrations
and the potential carcinogenicity of some higher
ring PAHs, remediation of contaminated soils
and sludges to achieve PAH reductions is often
required. To this end, biological treatment has
been widely researched and utilized to treat soils
and sludges containing PAHs. Soils and sludges
contaminated with coal-tar derived PAHs are
associated with the industrial processes of cre-
osote preservative wood treatment, coal tar
distillation, coke manufacturing and manufactured
gas plants. While it is widely reported and
accepted that biodegradation of PAHs does
occur, individual PAH compounds and isomers
vary widely in their susceptibility to biodegrada-
tion with the specific controlling mechanisms not
entirely understood [5, 6, 7, 9, 10]. One com-
monly cited observation indicates that the more
soluble, lower molecule weight PAH compounds
are generally biodegraded at a faster rate and to
greater extent than the less soluble, higher
molecular weight PAHs. This paper presents
information which begins to elucidate the fate
mechanisms which appear to influence the rate
and extent of PAH reductions achieved through
biological processes. Focus is given to both
applicable theory and results of laboratory and
field biological treatment applications. While
PAHs are specifically addressed, the concepts
presented may also be applicable to other or-
ganic compounds, e.g. volatile aromatics, pen-
tachlorophenol and PCBs.
TWO RING
Naphthalene
THREE RING
FOUR RING
Acenaphthene *Benzo(a)anthracene
Acenaphthylene *Chrysene
Anthracene Fluoranthene
Fluorene *Pyrene
Phenanthrene
FIVE RING
SIX RING
*Benzo(b)fluoranthene *Benzo(g,h,i)perylene
*Benzo(k)fluoranthene *lncleiTC<1,2,3cd)pvrene
*Benzo(a)pyrene
*Dibenzo(a,h)anthracene
NOTE: *lndicates potentially carcinogenic compound by EPA (11)
TABLE 1. POLYNUCLEAR AROMATIC HYDROCARBONS ON EPA'S PRIORITY
POLLUTANT LIST
398
-------
PURPOSE
The use of biological treatment processes for
the remediation of PAH contaminated soil/sludges
must be based on sound scientific and engineer-
ing principals. These principals must be thor-
oughly understood and manipulated to achieve
final clean-up concentrations within specified time
periods. This understanding also serves to justify
clean-up concentrations on the basis of risk
assessment considerations.
Environmental Fate Mechanisms
Based on a review of available literature, Sims
and Overcash [5] and the U.S. EPA [8] cite
volatilization, sorption, and biological oxidation as
the three primary environmental fate mechanisms
influencing PAHs in the environment. While
photolysis, chemical oxidation, and bioaccumu-
lation of PAHs may occur, they are not con-
sidered to be significant relative to the other
three. Sorption refers to the combined and
simultaneously occurring processes of adsorp-
tion/desorption and diffusion within the soil
matrix.
On this basis, Figure 1 schematically illus-
trates the manner by which PAH removal from
soils is most likely influenced by the combined
and simultaneously occurring effects of volatiliza-
tion, sorption, and biological oxidation during
biofogica! treatment in a soil-water system. Of
these, the latter two fate mechanisms are cited
as generally being the more predominant. These
three fate mechanisms are in turn influenced by
the physical/chemical characteristics of the
particular PAH compound, the physical/chemical
characteristics of the particular media, (e.g., soil,
sludge, water), and the particular biological
treatment system design and operation. The
overall process depicted in Figure i models PAH
biodegradation as a water-based process in-
fluenced by chemical partitioning among the
solid, air, and water phases. In this model,
biological oxidation can occur only if a particular
PAH compound within or on a soil particle (Cs)
desorbs and diffuses into the bulk water phasse
(Ct). Once in solution, volatilization (C ) can also
occur. While some volatilization of only the lower
molecular weight, 2- and 3-ring PAHs may occur,
it is not considered a major fate mechanism in
most biological treatment processes [12]. In
many instances, PAH desorptioh and diffusion
into the bulk water phase may be the rate-limiting
steps controlling both volatilization and biological
oxidation; desorption and diffusion are generally
considered separate processes. This model is
supported by Annokkee [13] who cites that the
biodegradation reaction is rate limited primarily by
diffusion of the organic material to the surface of
the soil particles.'.
AIR PHASE
C,
XT'
"WATS' PHASE
J SOLID /
/ PHASE /\ /
f M /£ } „ ,., , /»
ADSORPTION/
DESORPT1ON
V VOLATILIZATION
1 O CO2 * HjO + CsH/QjN
BIOLOGICAL
OXIDATION
NOTE, BIODEGRADATION IS A WATER BASED PROCESS
NOTE: Cs =» Adsorbed Chemical Concentration,
C = Gas Phase Chemical Concentration,
C, = Liquid Phase Chemical Concentration
Figure 1. Role of Desorption/Diffusion in PAH Biodegradation
399
-------
With specific reference to sorption processes,
Figure 1 conceptualizes a soil matrix as a collec-
tion of porous, water-stable aggregates which are
loosely associated with one another. Only a
single soil aggregate is illustrated in Figure 1.
The aggregates consist of organic and inorganic
(e.g., clay) fractions which are ionicly bound
together by metal cations such as aluminum.
The size of the individual aggregates range from
less than 1 to 250 microns in diameter. Soil
water can exist in the macropores between the
aggregates (bulk soil water) or within the micro-
pores of the individual aggregates themselves
(pore water).
It is believed that the micropores of the in-
dividual aggregates are large enough to permit
some chemicals to move into, out of, and within
the aggregate, but they are not sufficiently large
to permit microorganisms to enter. Hence, for
biodegradation to occur, the chemicals must
migrate to the bacteria which exist at the surface
of the aggregate or in the bulk soil water (macro-
pores). The organic contaminant can be present
in the soil system bound to the soil organic/inor-
ganic fractions, in aqueous solution (either in the
micropore water or the bulk soil water), or as a
free hydrocarbon phase. As such, migration to
the bacteria requires some combination of the
adsorption/desorption, dissolution, and diffusion
processes. Desorption and dissolution are the
mechanisms by which the contaminant enters the
solution and diffusion is the mechanism which
governs its movement in the aqueous phase. It
is generally believed that for PAHs, adsorption/-
desorption and diffusion are critical parameters.
Dissolution is not considered as important since
the presence of the free phase hydrocarbon in
the aggregate is unlikely due to the inability of
hydrocarbons to enter the micropores of the
aggregate.
Much research on PAH sorption processes
has been done, and continues to be performed
in soil/water systems. General conclusions of this
work are that the lower molecular weight PAHs
(i.e., 2- and 3-ring) have a tendency to desorb
off soils to a greater extent and at a faster rate
than the higher molecular weight PAHs (i.e., 4-,
5- and 6-ring). Due to their inherent physi-
cal/chemical properties, the concentrations of
lower molecular weight PAHs in aqueous solution
can be on the order of the part per million (ppm)
range while the higher molecular weight PAHs
are typically in the range of part per billion (ppb)
to part per trillion (ppt). These differences in
solubilities result in more extensive and faster
rates of biodegradation for the lower molecular
weight PAHs.
In the past, some researchers have attributed
these differences in rates to the higher molecular
weight PAHs being more recalcitrant to biodegr-
adation than the lower weight PAHs. However,
research work presented in this paper begins to
support the premise that the lower bio-
degradation rates observed for higher molecular
weight PAHs may be due more to their lack of
bioavailability in solution at ppm concentrations.
Specific discussions related to the three environ-
mental fate mechanisms of sorption, biological
oxidation and volatilization of PAHs in soil-water
systems are given elsewhere [5, 8, 12, 14-17].
The significance of the hypothesis presented
is amplified by the fact that soil aggregates which
may have been in contact with contaminants for
50 years or longer will contain PAHs which may
have penetrated deep into individual aggregates.
An equally long period of time may be required
for the same PAH compounds to desorb and to
become available to biological oxidation treat-
ment. Thus, treatment processes which rely on
the availability of the contaminant in the bulk
liquid phase will most likely be rate-limited by this
desorption/diffusion process. This conceptual
model may help to explain why there is such
variability reported with regard to biodegradation
rates and achievable treatment levels. This
model relates to soil desorption/diffusion
processes in an aqueous environment without the
aid of chemical amendments (e.g., surfactants)
which may serve to enhance soil desorption/diffu-
sion of contaminants and thus enhance con-
taminant biodegradation. Additionally, naturally
secreted biosurfactants may also serve to in-
fluence soil desorption/diffusion processes which
are not accounted for in the conceptual model.
APPROACH
Biological treatment of contaminated soils and
sludges associated with coal-tar derived proces-
ses has been evaluated by the co-authors
through laboratory microcosms and a pilot-scale
field plot. The results of this work imply that both
the rate and extent of PAH soil reductions at-
tributed to biodegradation may be directly related
400
-------
to the extent of PAH desorption from the soils
and diffusion within the soil water matrix.
Data related to two separate studies are
presented. One study relates to laboratory soil
microcosms and the second to pilot-scale field
work simulating land treatment. The focus of the
studies performed was to evaluate the capability
of soil biodegradation under proper environmental
conditions (e.g., pH, temperature, nutrients)
without the aid of chemical amendments (e.g.,
surfactants or supplemental organics).
Laboratory Soil Microcosms
The technical feasibility of biodegrading PAHs
in contaminated soil from a Manufactured Gas
Plant (MGP) site was evaluated in soil microcosm
reactors. Two phases of work were performed.
During one phase, the microcosms were loaded
with 75 percent by weight contaminated MGP site
soil and 25 percent uncontaminated soil known
to contain PAH degraders. This uncontaminated
seed soil was added to augment the soil microor-
ganisms present, to dilute potentially toxic com-
pounds which might be present, and to buffer
the soil mixture. A total of 15 individual microco-
sms were established. Each microcosm con-
sisted of a 400 ml glass beaker containing 60
grams of a soii mixture and a dedicated glass
stirring rod. The microcosms were incubated at
35°C, moisture was controlled near 40 percent by
weight (i.e., 80 percent of field capacity), and the
pH remained near 7.6 for the study duration.
Nitrogen and phosphorus were not specifically
monitored for, but previous work with the seed
soil showed that supplemental nutrient addition
was not required to achieve PAH soil biodegrada-
tion. At each sampling event, three microcosm
reactors were sacrificed for PAH analysis. Tripli-
cate analyses were performed for statistical pur-
poses. The triplicate samples were taken initially
and at two, four, eight and twelve weeks of
operation.
A second phase of the soil microcosm work
examined the biodegradation of supplemental
naphthalene applied to both contaminated MGP
site soil and uncontaminated soil. The micro-
cosms were operated and maintained similar to
the first phase of work with the exception that 10
grams of a uncontaminated soil mixture was
added which included 1 gram of seed soil. The
naphthalene was added to the soil in each
microcosm in a 0.1 ml solution of acetone con-
taining 10 mg/ml naphthalene. At each sampling
event, triplicate microcosm reactors were sacrific-
ed for PAH analysis. Samples were taken initially
and at weeks one, two, three and four during
operation.
All PAH soil analyses were performed by EPA
Method 8310 (HPLC) as specified in the third
edition of SW-846 [18].
Pilot-Scale Bioremediation Test Plot
The second study cited involved the opera-
tion of a 12 ft. x 50 ft. pilot-scale bioremediation
plot treating creosote-contaminated soil for
approximately a 57 week period.
Contaminated soil was applied to achieve an
initial benzene extractable content of approxi-
mately 4 percent by dry weight. Initially, agricul-
tural manure equivalent to 10 tons/acre and
agricultural fertilizer equivalent to 2 tons/acre
were applied giving a C:N ratio ranging between
25:1 to 50:1. After approximately 57 weeks of
treatment, the C:N ratio ranged between 15:1 to
20:1 indicating a large reduction in soil organic
carbon compared to nitrogen, thus indicating
biological activity. Agricultural lime, equivalent to
2 tons/acre, was initially added and this served to
maintain the soil pH between 6 to 7 during the
entire treatment period.
The bioremediation plot was initially loaded
with contaminated soil in September; 1985, with
operation being performed through October,
1985. Operation basically ceased from Novem-
ber through April due to cold weather. Operation
resumed in May, 1986 and continued through
October, 1986.
During operation of the treatment plot, soil
moisture was maintained near 80 percent of the
field capacity and tilling was performed at an 8
inch depth bi-weekly for mixing and aeration
purposes. Triplicate composite samples were
taken initially and at weeks 5, 20, 31, 35, 42, 46,
51 and 57 during the operating period. Week 20
corresponds to a sampling event in February,
1986. In addition, to moisture and pH as needed
for operational monitoring purposes, the samples
were also analyzed for PAHs by EPA Method
8310 [18].
401
-------
PROBLEMS ENCOUNTERED
Due to the heterogeneous nature of the soils,
replicate soils analyses for the same soil did
exhibit some variability. Additionally, PAH volatili-
zation was not measured, thus reductions in soil
concentrations attributed to biological oxidation
may be slightly overstated. However, these
missing data are probably not critical due to the
low volatility of PAHs as cited by the US EPA [5,
6], Quantification of soil PAHs was also compli-
cated by interfering compounds present in the
sample extracts which produced a rising baseline
which could have contributed to the wide varia-
tions in PAH soil concentrations which were
measured.
RESULTS
Laboratory Soil Microcosms
Results of the laboratory microcosm study are
given in Figures 2 and 3. Data plotted are
averages of the triplicate samples analyzed.
Figure 2 illustrates a situation where no
statistically significant PAH soil reductions were
measured over a three month treatment period
with the microcosms operated under the proper
environmental conditions of pH, oxygen, moisture
and nutrients. Data are presented for a sum-
mation of all sixteen PAHs as well as the in-
dividual ring groupings cited in Table 1. Water
extracts of the soil showed no detectable PAH
concentrations which agree with the premise that
if PAHs do not desorb and diffuse into the bulk
liquid phase, they are unavailable for biological
oxidation.
The data given in Figure 3 further supports
the desorption/diffusion hypothesis. Data is pre-
sented for the degradation in microcosm reactors
of contaminated ,MGP site soil and uncontam-
inated soil both of which had been spiked with
naphthalene added in an acetone solution. The
acetone volatilized with the naphthalene left on
the soil. As shown, the majority of the naph-
thalene applied to the uncontaminated soil was
reduced within a one week period with less than
detectable soil concentrations at week three. The
naphthalene added to the contaminated soil was
reduced to a level near what was originally
measured in the soil before the naphthalene
addition. Statistically, there is no difference
between the baseline naphthalene soil concentra-
tion of approximately 26 mg/kg and the con-
centrations at weeks 1, 2, 3, and 4 for the MGP
soil with naphthalene applied. These results
imply that the freshly applied naphthalene was
rapidly mineralized while the naphthalene as-
sociated with the original MGP soil was not
susceptible to biological oxidation. Thus, the
effect of soil aging appears to play a significant
influence on PAH soil sorption and biological
oxidation processes.
Pilot-Scale Bioremediation Test Plot
Figure 4 presents summary data for the pilot-
scale test plot. The mean values given for the
triplicate sample analyses show rather good
reduction for all PAH ring groupings with total
PAH reduced approximately 97 percent from an
initial soil concentration of approximately 6,660
mg/kg to 176 mg/kg. The data also show that
the fastest rate of soil reductions occurred within
the first month of treatment with the soil con-
centrations somewhat leveling off after this. This
leveling off between weeks 5 and 20 corresponds
to cold weather operation. At week 31 (May,
1986), a second reduction in soil PAHs was
measured.
In agreement with the previously presented
hypothesis, the greatest extent of reduction
occurred with the 2-, 3- and 4- ring compounds
and relatively less for the 5- and 6- ring.
Corresponding to the PAH ring groupings cited
in Table 1, a summation of 2- and 3-ring PAHs
were reduced from an initial mean soil concentra-
tion of 2,540 mg/kg to 6 mg/kg, 4-ring PAHs
were reduced from 374 mg/kg. to 45 mg/kg, 5-
ring total PAHs were reduced from 310 mg/kg to
88 mg/kg, and 6-ring total PAHs were reduced
from 70 mg/kg to 37 mg/kg. These data do
show significant reductions of all sixteen PAHs
and document that the extent of biodegradation
decreases with increasing ring number and
molecular weight. This observation can be
partially explained by the fact that the PAH
aqueous solubilities decrease and the affinity for
desorption from solids decrease as molecular
weights (i.e. ring number) increase. These two
factors are the primary reasons why the PAH
mean half lives reported in the literature for land
treatment of contaminated soil generally increase
with increasing molecular weight [9,10].
402
-------
1X1
CJ
z
0
o
ICO
o 2-Ring
A 3-Ring
a 4-Ring
* 5-Ring
+ 8-Ring
o Total PflHs
I Z 3
TIME (Months)
Figura 2. Laboratory Microcosm PflH Results
3= I/Contaminated MSP Sail: Uncantaninoted Seed Sail)
\
\ o Z - Ring * 5 - Ring
. \
\ A 3 - Ring + 8 - Ring
. i 0 4 - Ring o Total PflH*
J \
\ '
A \
10 30 30' 40 ' SO' 8
TIME (Weeks)
Figure 4. PflH Reduction in Pilot-Scale Sioremediation Test.
403
-------
Summary
Comparison of the data presented in this
paper begins to support the premise that for
coal-tar related PAHs to degrade in soil systems,
they must desorb from the solid matrix and exist
in aqueous solution. During the initial period of
treatment, desorption of PAHs from the soil may
have been rapid, decreasing with time. It is also
possible that enzymes secreted by the bacteria
may serve to aid the PAH desorption process
over time. Volatilization of some of the 2- and 3-
ring PAHs may have also occurred with little or
no volatilization of the 4-, 5- and 6-ring PAH com-
pounds anticipated. This dynamic process of
desorption/diffusion, volatilization and biological
oxidation was operative until no further significant
desorption of PAHs occurred as evidenced by no
further reductions measured in PAH soil con-
centrations with time. While this premise may
explain the observed phenomenon of PAH soil
concentrations leveling off at certain concentra-
tion plateaus, it does not explain the mechanisms
by which the PAHs become so incorporated into
the soil material that desorption and subsequent
biological oxidation does not occur. It also does
not explain why treatment plateaus obtained vary
from soil to soil. Perhaps aging of the soil
material may have an effect, as well as the
characteristics of the coal-tar source that is
associated with the contaminated material.
The data presented in this paper should be
considered very preliminary in nature; however it
does begin to support the complexity of the
interactions between PAHs and soils and the
effects of such interactions on the susceptibility
of PAHs to biodegradation in soil systems. For
this reason, future research work should focus on
better understanding of the mechanisms affecting
such interactions. Specifically, the effects of
physical/chemical soil and waste characteristics
on the mechanisms discussed should be inves-
tigated. Through a better understanding, biologi-
cal treatment of PAH contaminated soils and
sludges can be applied in a scientifically sound
and acceptable manner. This especially relates
to the issue of risk assessment in that if soils and
sludges are bioremediated to levels where PAHs
are so bound to and incorporated within the soil
aggregates, they may no longer represent an
environmental risk in terms of both leachate
migration to groundwaters and public health
exposure issues.
REFERENCES
1. Loehr, R. C., Malina, J. F., Land Treatment -
A Hazardous Waste Management Alterna-
tjye, Water Resources Symposium Number
Thirteen, Center for Research in Water
Resources, Bureau of Engineering Research,
College of Engineering, The University of
Texas at Austin, Austin, Texas, 1986.
2. American Petroleum Institute, Land Treat-
ment. 1220 L Street, Northwest Washington,
D. C. 20005.
3. Gas Research Institute, Management of
Manufactured Gas Plant Sites - Volume IV
Site Restoration. Chicago, IL, GRI-87/0260.4,
October, 1987.
4. Koppers Company, Inc., The Land Treat-
ability of Creosote/Pentachlorophenol
Wastes. Pittsburgh, PA. 15218, August,
1985.
5. Sims, R. C. and Overcash, M. R., "Fate of
Polynuclear Aromatic Compounds (PNAs) in
Soil-Plant Systems," Residue Review. Vol.
88, pp. 1-88, 1983.
6. Sherman, D. F., Loehr, R. C. and Neuhaus-
er, E. F., "Development of Innovative Biologi-
cal Techniques for the Bioremediation of
Manufactured Gas Plant Sites,' Proceedings:
International Conference on Phvsicochemical
and Biological Detoxification of Hazardous
Waste. Atlantic City, New Jersey, May 3-5,
1988.
7. Sherman, D. F., Stroo, H. and Bratina, J.,
"Degradation of PAHs in Soils Utilizing En-
hanced Bioremediation," Proceedings: IGT
Symposium in Gas. Oil, and Coal Biotech-
nology. New Orleans, Louisiana, December
5-7, 1988.
8. National Technical Information Service, Water
Related Environmental Fate of 129 Priority
Pollutants. Versar Incorporated, Springfield,
VA PB80-204381, EPA 440/4-70-0296, Dece-
mber, 1979.
9. McGinnis, G. D., Borazzani, H., McFarland, L
404
-------
K., Pope, D. F., Strobel, D. A., 'Chara-
cterization and Laboratory Treatabilitv
Studies for Creosote and Pentachloro-
phenol Sludges and Contaminated Soil.
Mississippi Forest Products Utilization
Laboratory, Mississippi State University,
Mississippi State, Mississippi,
EPA/0600/2-88/055, January, 1989.
10. Ryan, J. R., Smith, J. R., "Land Treatment of
Wood Preserving Wastes", Proceeding of
the National Conference on Hazardous
Wastes and Hazardous Materials, Hazardous
Materials Control Research Institute, Wash-
ington, D. C., 1986, pp. 80-86.
11. United States Environmental Protection
Agency, Superfund Public Health Evaluation
Manual. Government Printing Office, Wash-
ington, DC, EPA/540/1-86/060, October 1,
1986.
12. Smith, J. R., Middleton, A. C., Fu, J.,
"Environmental Processes Influencing Biode-
gradation of Polynuclear Aromatic Hydro-
carbons", Paper presented at Hazardous
Materials Control Research Institutes Biore-
mediation Conference on Genetically
Engineered or Adapted Microorganisms in
Hazardous Waste Treatment, Washington,
DC, November 30-December 2, 1988.
13. Annokkee, G. J., "Research on Decontamin-
ation of Polluted Soils and Dredging Sludges
in Bioreactor Systems at TNO," Assessment
of International Technologies for Superfund
Applications. EPA/540/2-88/003, Washington,
D. C., September, 1988.
14. Nakles, D. V., Smith, J. R., Treatability Pro-
tocol for Screening Biodegradation of Heavy
Hydrocarbons in Soil", Paper presented at
Hazardous Materials Management Confer-
ence and Exhibition. Atlantic City, New Jer-
sey, June 13-15, 1989.
15. Smith, J. R., Adsorption/Desorption of Poly-
nuclear Aromatic Hydrocarbons in Soil-Water
Systems, Technology Transfer Seminar on
Manufactured Gas Plant Sites.. April 19-21,
Pittsburgh, PA, 1989.
16. Smith, J. R. and Weightman, R. L, "Co-
Treatment of Manufactured Gas Plant Site
Groundwaters with Municipal Wastewaters,"
Final Topical Report to the Gas Research
Institute. Chicago, IL, Contract No. 5086-254-
1350, August, 1988.
17. Gibson, D. T., Microbial Degradation of
Organic Compounds. Microbiological Series,
Volume 13, Marcel Dekker, Inc., New York,
NY, pp. 197-252.
18. U.S. Environmental Protection Agency. Test
Methods for Evaluating Solid Waste (Third
Edition). USEPA/SW-846, Washington, DC,
1986.
Disclaimer
Ihe work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
405
-------
DESIGN CONSIDERATIONS FOR FIXED-FILM, AEROBIC,
MICROBIOLOGICAL DEGRADATION OF HAZARDOUS WASTE
Marleen A. Troy and Wesley O. Pipes
Department of Civil Engineering
and The Environmental Studies Institute
Drexel University
Philadelphia, PA 19104
ABSTRACT
Microbiological processes are currently being investigated
as potential treatment methods for the remediation of contaminated
waters at hazardous waste sites. There is sufficient information
available documenting that biodegradation of hazardous wastes is a
feasible treatment method, however, many details of process design
still need to be determined.
This paper describes a study where two fundamental design
questions were addressed: 1) What is the minimum depth of the
material supporting the bacteria required for the removal of a
contaminant? and 2) What is the effect of hydraulic loading on
bacterial removal efficiency? The study was conducted in the
laboratory using fixed-film biological reactor columns.
Pentachlorophenol (PCP)'and other chlorophenols were used as the
test contaminants. Dechlorinated Philadelphia tap water was used
as the source water. The concentration of the contaminants added
to the source water ranged between 200 and 1000 \ig/1. The columns
were packed with Ottawa sand (20-30 mesh). Incremental depths of
sand were used for individual columns. The columns were operated
in a downflow mode and the flow rates through the columns were
varied from 0.5 gpm/ft2 to 2 gpm/ft2. A high percent removal of
the contaminants was achieved. The data indicate that biological
activity through a biofilm process in a small column for the
removal of trace amount of pollutants is possible, thereby allow-
ing for the complete cleanup of water from a contaminated area.
INTRODUCTION
Microbiological processes
have been extensively studied
and used for the treatment of in-
dustrial and domestic wastes.
The same processes may be
adapted for the remediation of
contaminated ground water from
hazardous waste sites. This
adaptation is necessary because
the design is for a funda-
mentally different situation.
With conventional industrial
waste or domestic waste treat-
ment, the water to be treated is
continuously flowing, neces-
sitating the biological unit
operation to be run contin-
uously all year long. A con-
taminated ground water will
usually be of a finite volume,
allowing for intermittent or
406
-------
seasonal treatment as well as
permitting the luxury of being
able to control the hydraulic
loading rate to the system.
Another important differ-
ence between the two situations
is the nature of the material to
be treated. With the conven-
tional treatment process the
material to be degraded is
usually available as a primary
carbon source and usually occurs
in mg/1 quantities. The
presence of the material as a
primary carbon source will allow
a shorter acclimation period and
will permit a much shorter start-
up time. With the water from a
hazardous waste site, the
contaminants may be only present
in jJ.g/1 quantities, which will
not be a sufficient carbon
supply to sustain the cells.
Therefore, a primary carbon
source will be required, and
should be present in the ground
or surface water as dissolved
organic matter. The primary
carbon source needs to be in a
quantity such that the micro-
bial cells can sustain them-
selves, however, an acclimation
period - (a time where the
microorganisms adapt their
metabolisms to the breakdown of
specific contaminants) may be
necessary.
Because of the differences
outlined above, conventional
designs for biological treatment
systems may not be able to
handle the special requirements
for a hazardous waste site re-
mediation. However, an at-
tached/entrapped growth system
may be modified for the special
needs of a hazardous waste site.
The ability of microorganisms to
attach and colonize a surface
may afford them adequate contact
time with the contaminants,
access .to necessary primary
nutrients and, under some
circumstances, protect them from
toxic conditions. The term
biofilm (or fixed-film) is used
to describe the attached micro-
organisms. The biofilm process
can be brought about either JJQ.
situ or as an above-ground pump
and treatment scheme, and either
alone or in conjunction with
other treatment process op-
erations .
PURPOSE
In this study two fundamen-
tal design questions were
addressed: 1) . What is the
minimum depth of biofilm support
material required for the
removal of a contaminant? and
2). What is the effect of
hydraulic loading rate on
bacterial removal efficiency?
The study was conducted in
the laboratory using "fixed-
film" biological reactor
columns. Pentachlorophenol
(PCP), 2,4,6 trichlorophenol and
2,4 dichlorophenol were used as
the test contaminants. These
compounds were chosen because of
their documented biodegrad-
ability, their occurrence as
contaminants in both soil and
water, their nonvolatile nature
and their ability to be detected
in |J.g/l (ppb) quantities.
APPROACH
Apparatus
Two types of .column sys-
tems were used. For the
preliminary (Phase I) studies,
borosilicate glass columns were
407
-------
J
Model
Contaminated
Vater
0.5
I!
1.0
i
„
ili
•:•:
:•:•
•:•:
ijjij
i!
i
=:=
*••'
:*:
ill
ill
2.
::•:
•::
j:
l|
•:•
;•;
&
jjjj
•:•:
;•;•:
0
lil
:
•
j
j
•i
j!
j|
il
2.0
MIS
jijijiji
!:!-I !
il
!•!• i
:|:| i
|j:|| ;
ijijij i
Ijljljl i
1
Phase I
-IT
pump
Phase II
Figure 1. Schematic of the experimental apparatus. For the phase
I studies 0.5 ft. incremental depths were used. One inch depths
were used in the phase II studies. N I S - No Initial Seed.
used. These columns were 3.0
feet in length and had a 2.5
inch internal diameter. They
were initially cleaned by acid
washing (50 % v/v HC1) and
rinsed five times with deionized
water. After air drying, they
were autoclaved for 15 minutes @
121°C. The columns were covered
with aluminum foil in order to
prevent any light induced
reactions. Figure 1 is a
schematic of the experimental
apparatus.
Ottawa sand (20-30 mesh)
was used to pack the columns.
The sand was initially rinsed
five times with deionized water
and allowed to air dry. After
drying, the sand was autoclaved
for 15 minutes @ 121°C. After
drying, the sand was aseptically
added to individual columns at
incremental depths of 0.5 ft. to
2.0 ft.
For the Phase II studies,
smaller columns were used.
These smaller columns were made
of teflon, and were 4.75 inches
in length and had an internal
diameter of 1.6 inches. The
columns were initially acid-
washed (50 % v/v HC1) followed
by 5 deionized water rinses. The
columns were allowed to air dry
and were then autoclaved for 15
minutes @ 121 °C. Sand,
prepared as described above was
added to the individual columns
to a depth of one inch. The
columns were also covered with
aluminum foil when in use.
A Cole Farmer peristaltic
pump was used to transport the
contaminated water. This pump
408
-------
was capable of delivering water
to each column at rates ranging
from 0.5 to 2.0 gpm/ft2. Silicon
tubing was used to connect the
pump to the columns. All other
tubing or fittings were made of
teflon, glass or stainless
steel.
Model contaminated water
for all experiments was made by
passing Philadelphia Tap Water
through Calgon Filtrasorb F400
Activated Carbon to remove
chlorine. The effluent from the
activated carbon was checked
daily for chlorine removal (Hach
field kit) and the influent and
effluent to the column was
monitored for total organic
carbon (nonpurgeable organic
carbon) to insure an adequate
primary carbon source. The
water was stored in 50 gallon
Nalgene tanks and allowed to
equilibrate to ambient tem-
perature (20-23°C) before use.
These tanks were covered to
prevent light infiltration.
The contaminants used were
PCP, 2,4,6 trichlorophenol and
2,4 dichlorophenol from Aldrich
Chemical (Milwaukee, Wisconsin).
All three compounds were added
to the reservoirs to give a
final concentration between 500
and 1000 ixg/1 of each compound.
Seed Organisms
The bacteria used for the
initial inoculation of some of
the columns consisted of a
suspension containing an
Arthrobacter (ATCC 33799) and
five isolates obtained from soil
from the EPA Region III Haver-
town, PA PCP Superfund site.
These isolates demonstrated the
ability in laboratory culture to
use PCP as a sole carbon source.
The six bacteria were in-
dividually grown in batch
culture for seven days at 25° C
with shaking in a mineral salts
medium amended with 500 M-g/1
PCP. Five grams of mud from the
Schuykill River in Philadelphia,
PA were also inoculated into the
PCP-mineral salts medium in the
same manner.
Startup and Operation
For the Phase I studies, a
total of five columns were used.
Four columns filled with sand at
O.5 ft incremental depths were
inoculated with the Seed bac-
teria in the following manner:
The columns were initially
filled with contaminated water.
Five ml from each batch culture
(35 ml total) were aseptically
added to each column. The final
density of each culture was
between 2 x 1Q5 and 1 x 10"?
CFU/ml. Each mixture in each
column was then recirculated
through each column operated in
a down-flow mode for twenty-four
hours. Next, the columns were
allowed to sit undisturbed for
twenty four hours. The columns
were then ready to go on-line. A
fifth column packed with the
same depth of sand as the
deepest seeded column was not
seeded, but was treated in the
same manner otherwise. The
columns were operated in a
downflow mode. For the phase II
studies the columns were ini-
tially started and then operated
in a similar manner.
Analytical Methods
Three parameters, dis-
solved oxygen (DO), pH, and
temperature were determined for
the influent and effluent of
each column each day. DO was
409
-------
monitored using a YSI Model 57
portable oxygen meter. pH was
monitored using an Orion
Research Model 399A pH meter.
Organic Analyses
The concentrations of the
chlorophenolic compounds were
measured using a Varian High
Pressure Liquid Chromatograph
(HPLC). The HPLC was used with
acetonitrile-water as the mobile
phase, and a Varian MicroPak MCH-
10 reverse phase column as the
stationary phase. A 1 ml sample
volume was used and the com-
pounds were characterized with a
UV detector at a wavelength of
254 nm. A detection limit of 50
p.g/1 was obtained for each
compound.
Total organic carbon (TOG)
analysis for non-purgeable
organic carbon (NPOC) was
performed on a Xertex, Dohrman
Carbon Analyzer (DC-80) accord-
ing to Standard Method (1)
procedures.
Heterotrophic Counts
Instantaneous grab samples
were collected in sterile 250 ml
Nalgene polypropelene bottles at
the influent and effluent of
each column. Serial dilutions
were made in sterile physiologi-
cal saline, and 0.1 ml aliquots
were spread-plated onto R2A agar
(3) in duplicate. The plates
were incubated in the dark for
seven days at 25° C.
PROBLEMS ENCOUNTERED
One problem that was en-
countered was the development of
an acclimated PCP degrading
population in the source water
reservoirs. This was overcome
by weekly draining and manually
cleaning and scrubbing of the
tanks.
Another problem concerned
the effect of temperature on
acclimation time. Low temper-
atures (< 20° C) inhibited the
establishment of an acclimated
microbial population within the
columns. This was overcome by
running all experiments between
20 and 23° C.
RESULTS
Phase I Studies
Results from the Phase I
studies showed that after an
initial acclimation period,
complete removal of PCP and
2,4,6 trichlorophenol was ob-
tained at all depths at a flow
rate of 0.5 gpm/ft2. 2,4
dichlorophenol was completely
removed from columns with depths
of sand 1.5 ft and greater.
Seeding of the columns
with acclimated organisms also
did not appear to have a
positive influence on removal of
the compounds. Removal of the
compounds occurred at about the
same times in both the seeded
and unseeded 2.0 ft. column:
approximately 150 hours for PCP,
170 hours for 2,4,6 tri-
chlorophenol and 200 hours for
2,4 dichlorophenol.
The other parameters
measured during the course of
the Phase I experiments showed
the temperature for the run
ranged between 21 and 23° C, the
pH varied between 7.5 and 7.7
for both the influent and ef-
fluent of all columns, and the
410
-------
D.O. depletion between the
influent and effluent averaged
about 2.0 mg/1. The hetero-
trophic plate counts from the
influent to all the columns
ranged from 4.0 x 104 to 2.4 x
105 CFU/ml (colony-forming-
units) . The counts from the
effluent from all of the columns
ranged from 2.0 x 104 to 1.4 x
105 CFU/ml. The influent consis-
tently had higher counts than
the effluent.
Two interesting observa-
tions were made during the phase
I study. Mechanical difficul-
ties were experienced which
caused the system to be shut
down for short periods during a
preliminary run, resulting in a
'pulsing' effect of the con-
taminated water onto the
columns. It was observed that
this action appeared to aid in
the establishment of an ac-
climated population within the
columns. It was also noted that
in all of the columns a
'biofilm' was observed that
extended approximately !/4 inch
in depth in the sand from the
top of the column. The 'biofilm'
was a brownish-olive green in
color and appeared to be evenly
distributed over the 1/4 inch
depth.
Because' of the results in
the Phase I study which showed
that apparent removal of the com-
pounds was possible with six
inches of sand and that the
biofilm only penetrated ap-
proximately 1/4 inch, a smaller
depth of sand (1 inch) was used
for the next series of experi-
ments (phase II).
Phase II Studies
Table 1 shows the length
of time required to establish an
acclimated culture for the de-
gradation of the three chloro-
phenolic compounds in an inch
depth of sand. As can be seen
from the table, there was no ad-
vantage to the seeding process,
as both columns behaved iden-
tically. PCP was the compound
most rapidly adapted to degrada-
tion by the populations within
the columns. 2,4,6 trichlor-
phenol and 2,4 dichlorophenol
were more slowly acclimated for.
The pattern of TOG removal
(measured as NPOC), as shown in
Table 1, by the populations in
both columns was erratic and
varied from day to day. This
phenomena was consistently ob-
served for all experimental runs
for all operating conditions and
is in agreement with observa-
tions made in other studies (2).
The effect of hydraulic
flow rate on percent removal of
contaminant was next studied.
These runs consisted of operat-
ing the system for seven days at
the designated flow rate with
daily sampling of the influent
and effluent. Seven days was
chosen because of the observa-
tion that equilibrium removals
were obtained within the columns
between three and five days. The
flow rates used were 0.5, 1.0
and 2.0 gpm/ft2.
Figure 2 illustrates re-
moval efficiencies vs flow rate
for the three compounds. Each
point represents the average
removal efficiency at equilib-
411
-------
Table 1. Percent Removal (influent vs effluent) for the three
chlorophenolic compounds used in the Phase II study.
S- seeded; NIS- no initial seed;* - No apparent removal
Elapsed Percent Removal of
Time 2,4 2,4,6
(hrs .) dichlorophenol trichlorophenol PGP NPOC
0
51.5
95.5
146.0
193.5
215.0
239.0
260.5
287.5
308.5
336.5
384.0
S
18
*
4
*
2
24
31
11
*
*
*
*
HIS
5
*
*
1
*
10
17
*
*
10
28
15
fi
*
5
23
19
*
29
*
*
32
*
*
24
HIS
11
*
*
6
*
32
*
27
34
35
54
17
fi
*
*
13
53
49
47
54
54
71
68
74
72
HIS
*
*
15
32
47
60
71
71
92
92
90
92
S
49
16
6
*
*
13
28
42
57
9
9
45
HIS
*
*
34
6
11
5
26
27
*
*
27
22
100
30.
80.
*« 70.
g c
£ § 60.
"I
^ a 5i
C 4*
41 S
Cu o
.6
• - 2,4 dichlorophenol
• - 2,4,6 trichlorophenol
- pentachlorophenol
.8 1 1.2 1.4 1.6
Hydraulic Loading Rate (gpm/ft )
1.8
Figure 2. Percent removal of each chlorophenolic compound vs
hydraulic loading rate (gpm/ft2) .
412
-------
rium which was at least 72 hours
for both columns. The removal
efficiency for all three com-
pounds did not differ at loading
rates of 0.5 gpm/ft2 but
decreased at a loading rate of
2.0 gpm/ft2. These results
suggest that through an increase
in the depth of biofilm support
material, high loading rates can
be handled and complete removal
of the contaminants is possible.
The data also show a
consistent pattern of PGP being
the contaminant most readily
degraded, followed by 2,4,6 tri-
chlorophenol and 2,4 di-
chlorophenol. These chloro-
phenolic compounds have proven
to be biodegradable by both pure
cultures and mixed natural
microbial populations and com-
plete mineralization of the
contaminants was observed in
most situations (4). However,
in this case the rate of disap-
pearance of the compounds varies
suggesting an effect by loading
rate as well as other contribut-
ing environmental factors. This
factor will prove to be impor-
tant for design at hazardous
waste sites, because in almost
all situations the waste is
composed of a variety of com-
pounds . The results from this
study indicate that even though
the compounds may be from the
same characteristic group, in
this case the chlorophenols,
different removal rates at all
depths and with all loading
rates were observed. This
factor will entail designing the
biological system based on
'target compounds,' or more
ideally using the biological
process in conjunction with
other physical-chemical opera-
tions to selectively remove all
contaminants of interest.
The results from this
study indicate that there is the
potential for promoting biologi-
cal activity through a biofilm
process in a very small area for
the removal of trace amounts of
contaminants, thereby allowing
for the complete clean-up of a
contaminated area. Work is
continuing to establish a more
quantitative relationship be-
tween flow rate, depth of sand
and microbiological activity
within the depth of sand on the
removal efficiency of trace a-
mounts of contaminants.
REFERENCES
1. . American Public Health
Assoc. 1985. Standard
Methods for the Enum. of
Water and Wastewater, 16th
ed. A.P.H.A., Inc.
Washington, B.C.
2. Maloney,S.W., K.Bancroft,
W.O. Pipes and I.H.Suffet.
1984. Bacterial removal
on sand and GAG. Journ.
Environ.Engineer.,110:519-
533.
3. Reasoner, D.J. and E.E.
Geldreich. 1985. A new
method for the enumeration
and subculture of bacteria
from potable water. Appl.
Environ. Microbiol.49:1—7.
4. Rochkind-Dubinsky,M., G.S.
Sayler and J.W. Blackburn.
1987. Microbiological De-
composition of Chlori-
nated Aromatic Compounds.
Marcel Dekker. New York,
New York.
413
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
414
-------
APPLICABILITY OF STEM! STRIPPING TO ORGANICS REMOVAL FROM WASTEWATER STREAMS
Benjamin L. Blaney
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
ABSTRACT
In the past five years the U. S. Environmental Protection Agency has
studied the effectiveness of steam stripping as a treatment technique for
removing organics from aqueous waste streams. This paper presents the data
obtained from field tests of steam strippers at seven industrial faciliteis.
The effectiveness of steam stripping for removing different types of organics
from wastewaters is discussed.
INTRODUCTION
Two components of the Hazardous
and Solid Waste Amendments of 1984
provide incentives to industry to
reduce the organic contentof wastes.
This can be done using organic
removal treatment technologies, such
as steam stripping. The Amendments
require that the U.S. Environmental
Protection Agency (USEPA) develop
land disposal restrictions for
hazardous wastes. These restric-
tions are intended to protect human
health and the environment from
releases of toxic compounds into the
groundwater or to the air from land
disposal facilities. The restric-
tions included limitations on the
concentration of hazardous constitu-
ents in land disposed wastes. In
addition, the Amendments require
that emissions of volatile organics
from hazardous waste treatment,
storage and disposal facilities
(TSDF) be reduced. Emissions
reductions are intended to minimize
the air pollution problems, such as
ozone formation, which result from
TSDF operation.
Wastewaters form a large per-
centage of the hazardous waste
streams generated in the United
States. For example, in 1981, 13.7
billion gallons of the 14.6 billion
gallons, or 92%, of the hazardous
waste disposed of in the nation went
to deep-well injection or surface
impoundments - disposal processes
that are typically used for liquid
wastes (1). Of the 398 million
gallons of solvent (i.e. F001-F005)
wastes disposed of in 1981, 266
million gallons, or 65%, were
aqueous streams contaminated with
less than 1% organics or solids (2).
The Agency is also studying
ways to reduce emissions from indus-
trial and municipal wastewater
treatment facilities under the Clean
415
-------
Air Act. Since open storage tanks
and activated sludge and other
aerated treatment processes are
ccranonly used for managaement of
organics in wastewaters, the poten-
tial exists for significant emis-
sions from these facilities. Plants
D, F and G in this paper were sampled
by the USEPA as part of this program.
A number of treatment technol-
ogies have been used to remove
organics from wastewater streams.
These include steam stripping, bio-
degradation, carbon adsorption,
solvent extraction and chemical
destruction techniques, such as
W/ozonation. Steam stripping
offers several advantages over the
other available technologies. It
can be used to recover the organics
separated from the stream. It per-
forms wall in removing halogenated
aliphatics; conpounds that are not
readily removed by carbon adsorption.
And, it does not usually require the
solvent recovery steps inherent in
solvent extraction. For these
reasons, steam stripping is expected
to see increased use in wastewater
cleanup.
This paper provides a summary
of the results from seven field
tests that have been performed by
the USEPA to determine the effec-
tiveness of steam strippers being
used for wastewater treatment at
industrial facilities. The results
demonstrate the wide variety of
organic compounds which this tech-
nology can efficiently remove from
aqueous waste streams.
Steam Stripping
Stripping is a physical separa-
tion unit operation in which dis-
solved compounds are transferred
from a liquid into the gas phase.
The driving force for mass transfer
is provided by the concentration
gradient between the two phases. In
steam stripping, live steam is used
as the gas phase. The steam both
heats the liquid, which enhances the
rate of mass transfer, and carries
the volatilized compounds away from
the liquid (3).
Steam stripping can be per-
formed in the batch or continuous
mode. In the batch process, waste
is charged to a boiler and steam is
injected directly into the waste.
The injection of live steam both
heats the waste to volatilize low
boiling components and creates
turbulence in the waste, thus
increasing the rate of volatiliza-
tion. The gases which are condensed
from a steam stripper will contain
water along with the more volatile
organic components of the waste in
the form of a two phase mixture.
This mixture is decanted and the
organic component is drawn off for
reuse or disposal. In a continuous
steam stripping column, as shown in
Figure 1, waste flows down a column
while steam flows up. The column is
designed to promote heat transfer
from the steam to the waste, to
cause turbulence in the waste and to
create a large waste surface area.
All of these properties promote
transfer of volatile components from
the waste to the gas phase. Differ-
ent liquid-roper equilibria exist in
the column, with the highest rela-
tive concentration of the most
volatile components found at the
top. The separation of different
volatile constituents may be en-
hanced by refluxing a portion of the
condensate.
Steam stripping is generally
used to separate insoluble or
slightly soluble compounds from
water. These compounds are readily
stripped because they have a large
air to water equilibrium coefficient
which increases as the tetnperature
416
-------
UJ
CO
u.
HI
O
to
O
to
O)
Q.
Q.
to
O)
CO
0)
417
-------
of the waste is raised by the steam.
In addition, a high percentage of
.the mass of the solvent in the con-
densate will be in the organic phase;
only a small portion of the stripped
organics are left in the condensed
steam (4).
The aqueous phase of a steam
stripper's condensate must be dis-
posed of, however, and this require
ment must be carefully considered in
deciding whether stripping is an
appropriate recycling technique. In
a few instances, the small amount of
water need not be separated from the
organic phase and the entire con-
densate can be recycled (5). Gener-
ally, however, the aqueous conden-
sate must be treated to a point
where it can meet permit conditions
upon discharge to either surface
waters or a municipal sewer. While
this can be accomplished through a
separate process, the problem is
usually handled by recycling the
aqueous condensate to the influent
stream of the stripper. This
requires a steam stripping unit that
is oversized to handle the
additional load.
PROCESS DESCRIPTION
Table 1 provides general steam
stripper process information for
each of the seven plants sampled by
the USEPA. Note that the waste
streams being stripped at these
facilities are not necessarily
classified as RGRA hazardous wastes.
Both Plant A and. B manufacture
mono-carbon chlorinated solvents.
At Plant A, the steam stripper is
used to treat wastewater generated
from the production of methylene
chloride, carbon tetrachloride and
chloroform. The wastewater at this
plant consists of equipment wash
water and rainfall collected from
diked areas around the plant. Waste-
water is pretreated for removal of
solids and any immiscible organic
phase in a decanter prior to strip-
Table 1. Selected steam stripper operating parameters.
Plant
Identifier
A
B
C
D
E
F
G
Wastewater Flow
Rate n/min)
41.5
21.0
852
2,390
499
110
30.5
Volatile Organic
Loadim Rate (kcr/1)
14.6
4.6
292
286
19.0
5.29
0.395
Feed/Steam
Ratio fkcr/ka}
9.6
10.5
28.8
NA
14.7
7.1
1.4
NA - Not available.
418
-------
ping. The stripper column is packed
with 2.5 cm (1-inch) saddles and
processes 42 1/min (11 gpm). The
stripper effluent, after cooling by
a heat exchanger, enters a holding
tank. If the organics analysis of
this effluent meets NPDES discharge
limits, it is pH adjusted and dis-
charged to a river. Condensate from
stripping is phase separated in a
decanter. Ihe aqueous phase is
returned to stripper feed holding
tank and the organic phase is
recycled to the manufacturing
process (5).
Facility B is a chemical manu-
facturing plant which produces a
number of chemical intermediates
that are used by the cosmetics,
chemical and agricultural industries.
Ihe facility operates several pro-
cesses including a methyl chloride
steam stripping process. The waste
routed to this steam stripper con-
sists of methylene chloride, water,
salt and organic residue. Feed is
pumped from storage tanks through a
preheat exchanger to the top of the
stripper at approximately 20 1/min
(5 gpm). The stripping tower is 20
cm (8-inches) in diameter and con-
tains 3.3 meters (10 feet) of 1.6 cm
(5/8-inch) Pall rings. The treated
effluent flows through a heat ex-
changer and ultimately to the local
publicly-owned treatment works
(POIW). The overhead vapors are
liquified in a water-cooled con-
denser and separated by gravity in a
decanter. The aqueous phase is
recycled to the top of the stripping
column while the lower layer of
methylene chloride is stored for
reuse (6).
Plant C produces 1,2-dichloro-
ethane (EDC) and vinyl chloride
monomer (VCM). Wastewaters from the
EDC/VCM production operation and
from other parts of the plant,
including storm water runoff, are
treated by steam stripping. The
feed rate to the stripper is in the
range of 760 to 950 1/min (200-250
gpm). There is no pretreatment of
this stream and the feed stream con-
tained 1.4 g/1 filterable solids
during USEPA field tests. As a con-
sequence, the column contains trays
instead of packing and both the
column and heat exchangers must be
backwashed periodically. The
effluent from the steam stripper
passes through a heat exchanger and
is then sent to a wastewater treat-
ment system for treatment of resid-
ual, nonvolatile organics. The
condensate could be phase separated
by decanting, but at this facility
the complete aqueous/organic mixture
is recycled directly to the manufac-
turing process (5).
Plant D is a chemical manufac-
turing facility producing
chlorinated hydrocarbons. Steam
strippers treat wastewater generated
by various chlorinated hydrocarbon
production units operated at the
facility. Scrubber blow down
streams, aqueous reactor equipment
streams and pad water are collected
from around the facility and pumped
to settling tanks prior to stripping.
The one exception is wastewater from
the VCM production unit which is
pumped directly to the stripper. In
the settling tanks insoluble
organics are separated from the
aqueous stream, which is then fed to
the strippers. Total suspended
solids concentrations in the feed
were low, ranging from 6.4 to 65.6
mg/1 over three days. Two waste-
water strippers operated in parallel
are used for treatment and waste
reclamation as NPDES treatment units.
Approximately 2,400 1/min (650 gpm)
of wastewater enter the stripper.
Plant E is an explosives manu-
facturing plant with process waste-
water streams that are predominately
419
-------
red water and white water. These
streams pass through decanters where
the oils are separated from the pro-
cess. There is no other pretreat-
ment of the stream and some fouling
of the feed preheater results,
although plant personnel reported
less than one percent down time for
the unit. The steam stripper is
packed with 2.5 cm (1-inch) diameter
stainless steel rings and had a feed
rate of approximately 500 1/min dur-
ing USEPA tests. The effluent from
the steam stripper passes through a
heat exchanger and then through one
of two carbon adsorption beds. The
carbon served as a polishing step
removing residual organics from the
waste stream. After pH adjustment,
the effluent stream is discharged to
a river. The condensate is phase
separated in a decanter. The
aqueous phase is returned to the
stripper feed tank, while the
organic phase is routed to an
organics slop sump (5).
Plant F is a chemical manufac-
turing facility which uses steam
stripping for treatment and material
reclamation from wastewater streams
generated by both production units
in one process area of the plant.
Toluene is the principal compound
being recovered. Waste streams are
pretreated by either a primary
decanter (for toluene removal) or an
evaporator (for solids thickening)
and then flow to the steam stripper
feed decanter where additional sepa-
ration of insoluble organics is
obtained. The flow rate to the
steam stripper is 60 to 150 1/min
(15-40 gpm). Total suspended solids
concentration ranged from nondetect-
able (<4.0 mg/1) to 8.5 mg/1 during
USEPA field tests. The column is
2.5 feet in diameter and contains 20
sieve trays. The condensate from
the first stage condenser is
recycled to the feed decanter. The
steam stripper bottoms are dis-
charged through the plant's process
sewer to the plant's on-site waste-
water treatment plant.
Plant G is an agricultural
chemical manufacturing facility.
The primary source of wastewater
for this plant's water layer steam
stripper is a jet collection system
which is used to pull a vacuum on
various process refining stills and
reactor recovery stills. A scrubber
decant pot and periodic reactor
washes also contribute to wastewater
in the stripper feed. The waste-
waters are pumped to a decanter tank
for liquid organics and sludge sepa-
ration and removal. Flow rates to
the steam stripper averaged 31 1/min
(8 gpm) during the test period.
Total suspended solids content of
the feed for this stripper was not
monitored. The stripper column is
60 inches in diameter and contains
14 Glitsch valve trays spaced one
foot apart. Nitrogen is used to
maintain the column pressure at 5
pounds per square inch. Condenser
overheads are pumped back to the jet
collection pot. The stripper
effluent is discharged through the
facility's process sewer to the
wastewater treatment plant.
STRIPPER PERFORMANCE
Table 2 presents performance
data for each of the 7 strippers
discussed above. Results are based
on one of three days of collecting
data at each facility.
The performance data demon-
strates that continuous steam strip-
ping achieves efficient (>95%)
removal for a range of compounds,
typically yielding concentrations of
individual compounds of less than
1.0 ppm. (An exception, Plant E,
found it more cost-effective to
achieve these levels by a combina-
420
-------
Table 2. Steam stripper organic removal effectiveness.
Pollutant
Plant A
Chloromethane
Methylene chloride
Chloroform
Carbon tetrachloride
Trichloroethylene
1,1, 2-Trichloroethane
Total VOC
Plant B
Methylene chloride
Chloroform
Carbon Tetrachloride
Total VOC
Plant C
1, 2-Dichloroethane
Chloroform
Benzene
Carbon tetrachloride
Chlorobenzene
Chloroethane
1, 1-Dichloroethane
1 , 1-Dichloroethene
1, 2-Dichloroethene
Methylene chloride
Tetra-chloroethene
1,1, 2-Trichloroethane
Trichloroethene
Vinyl chloride
Total VOC
Plant D
1, 1-Dichloroethene
1 , l-Dichloroethane
trans-1, 2-Dichloroethane
Chloroform
1 , 2-Dichloroethane
1,1, 1-^Tr ichloroethane
Trichloroethene
1,1, 2-Trichloroethane
Inlet
Concentration
ppnw
33
4,490
1,270
55
5.6
5.3
5,860
3,600
52
<2.3
3,654
5,630
271
0.27
1.7
0.38
9.6
11
4.7
8.9
1.2
1.4
7.5
4.8
8.4
5,960
25.4
61.7
68.5
181
974
44.6
45.0
156
Outlet
Concentration
pprtw
<0.005
<0.011
<0.006
<0.005
<0.005
<0.005
<0.037
<0.19
5.3
<0.17
<5.6
0.097
9.6
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<9.8
<0.041
<0.045
<0.227
1.75
2.15
<0.051
<0.078
0.188
Removal
wt. %
>99.98
>99.999
>99.999
>99.99
>99.9
>99.9
>99.999
>99.99
89
>92
>99.8
99.998
96.4
>96
>99.4
>97
>99.89
>99.91
>99.8
>99.8
>99.1
>99.2
>99.8
>99.7
>99.8
>99.8
>99.84
>99.93
>99.67
99.03
99.78
>99.89
>99.83
99.88
421
-------
Table 2. (Continued)
Inlet
Outlet
Concentration
Pollutant oorrrw
Plant D (continued)
Tetrachloroethene
Total TO)
Plant E
Nitrobenzene
2-Nitrotoluene
4-Nitrotoluene
Total TOC
Plant F
Benzene
Chlorobenzene
1, 2-Dichlorobenzene
1, 3-Dichlorobenzene
I/ 4-Dichlorobenzene
Ethylbenzene
Toluene
O-xylene
M-xylene
P-xylene
Total TO
Plant G
Benzene
Toluene
Ethyl Benzene
Isophorone
Naphthalene
5-Ethyl-l, 2-Msthylpyridine
1,2,3, 4-Tetrahydronaphthalene
Acetophenone
2-Methyl-l, 3-Cyclopentanedione
Total Organics
162
1994
505
78
51
634
1.84
1.47
3.04
3.51
3.29
2.45
779
1.00
1.53
1.00
798
5.06
1.07
1.46
1.04
12.0
96.4
76.7
9.91
12.1
225
Concentration
pcmw
<0.171
4.90
41.0(<0.8)a
2.4(<0.8)
4.4(<0.8)
47.8(<2.4)
0.0020
0.003
0.005
0.002
0.003
0.002
0.283
0.002
0.002
0.002
0.305
0.012
0.007
0.005
0.026
0.026
9.18
0.789
0.127
0.127
10.3
Removal
wt. %
>99.89
99.75
91.8(>99.8)a
96.9(>98.9)
91.4(>98.4)
92.4(>99.6)
99.89
99.80
99.84
99.94
99.92
99.92
99.96
99.80
99.87
99.80
99.96
99.62b
98.94b
99.44b
96.02b
99.66b
84.54b
98.33b
97.91b
1_
98.29"
92.45b
a For Plant E, values shown in parentheses are effluent concentrations and
removal efficiencies for the treatment of the influent stream by a
combination steam stripping followed by carbon adsorption.
Percent removal based on mass flow rates of organics in stripper influent
and effluent streams. Average flow rates of influent and effluent streams
were 1,831 kg/hr and 2,973 kg/hr, respectively.
422
-------
tion of steam stripping and carbon
adsorption.) The first four plants
were treating v/astewater streams
contaminated with halogenated ali-
phatics, with individual compound
initial concentrations as high as
0.56%. These compounds have Henry's
Law constants, H, ranging from 4.3 x
10-4 to 1.2 x 10-2 atm-m3/mole.
Halogenated aromatics having simi-
larly high H values were also
removed effectively, as shown by the
data presented for chlorobenzene and
dichlorobenzene at Plant F.
Plants E, F and G treat non-
halogenated aromatics, ketones and
other compounds which are more water
soluble and have lower Henry's Law
Constants than the above halogenated
organics. Good removal efficiencies
were also achieved for these com-
pounds. Benzene, toluene, xylene
and ethyl benzene were removed with
greater than 99% efficiency, while
naphthalene and 1,2,3,4-tetrahydro-
naphthalene are removed with better
than 98% efficiency.
Plant E achieved greater than
90% reduction of nitrobenzene and
nitrotoluene using steam stripping,
with overall removal efficiencies of
greater than 98% for these compounds
when carbon adsorption was used for
polishing. The nitro group on the
aromatic ring suppresses the Henry's
Law Constant by over an order of
magnitude for these compounds com-
pared to halogenated organics,
requiring increased energy for their
removal. This led Plant E to decide
to operate a treatment system which
utilized carbon adsorption for final
effluent polishing prior to dis-
charge. Despite the relatively low
H value for these chemicals (2 to 7
x 10-5 atm-m3/mole), steam stripping
is still practical since the low
solubility of nitrobenzene and the
nitrotoluenes, 1,900 mg/1 and
approximately 600 mg/1, respectively,
allow a large percentage of these
compounds to be separated by decant-
ing from the aqueous condensate
produced by the stripping process.
Ninety-four percent of the stripped
nitrobenzene was removed from the
condenser, while over 98% of the
nitrotoluene isomers were removed.
CONCLUSIONS
Steam stripping can achieve
better than 95% removal efficiency
for a range of organic compounds
which are insoluble or slightly
soluble in water. Effluent concen-
trations of less than 10 ppm, and
for most compounds less than 1 ppm,
can be obtained using steam strip-
ping. Polishing of the steam
stripper effluent using carbon
adsorption may be the most cost-
effective means of achieving these
low concentrations for the less
volatile compounds in the group
represented by these seven tests.
The presence of solids in waste-
waters can foul steam strippers and
therefore it is generally advanta-
geous to remove these solids before
stripping. Decanters are typically
used to achieve solids removal and
concurrently remove any insoluble
organics which will also interfere
steam stripper operation.
AO^OWLEDGEMENT
The author wishes to thank the
USEPA Office of Air Quality Planning
and Standards (OAQPS) for providing
data for three of the plants dis-
cussed in this paper.
REFERENCES
1. U.S. Environmental Protection
Agency, National Survey of
Hazardous Waste Generators and
423
-------
Facilities Regulated Under RCRA.
in 1981 r U.S. Environmental
Protection Agency (1984).
2. Breton, M., etal., Technical
Resource Document; Treatment
Technologies for Solvent Oon-
tajnim Wiastesf U.S. Environ-
mental Protection Agency Report
No. EPA/600/2-86/095 (1986).
3. Boegel, J.V., Air Stripping and
Steam Stripping, Standard Hand-
book of Hazardous Waste Treat-
ment and Disposal,. H. Freeman,
Ed., MCGravHHill (1988).
4. Olexsey, R., etal., Technolo-
gies for the Recovery of
Solvents from Hazardous Wastes,
Hazardous Waste and Hazardous
Materials. 5(4) (1988).
Allen, C., et al., Case Studies
of Hazardous Waste Treatment to
Remove Volatile Qrganicsr Vol.
I, U.S. Environmental
Protection Agency Report No.
EPA/600/2-87/094a.
U.S. Environmental Protection
Agency, Field Measurement of
Rail-Scale Hazardous Waste
Treatment Facilities; Organic
Solvent Wastesr U.S. Environ-
mental Protection Agency Report,
NTIS No. PB89-138853.
Disclaimer
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency peer and administrative review policies and approved for
presentation and publication.
424
-------
SITE PROGRAM DEMONSTRATION OF
THE CF SYSTEMS INC. ORGANICS EXTRACTION UNIT
Richard Valentinetti
U.S. Environmental Protection Agency
Office of Research and Development
401 M Street, SW
Washington, DC 20460
ABSTRACT
The Superfund Innovative Technology Evaluation (SITE) Program demon-
stration of the CF Systems organics extraction technology was conducted at the
New Bedford Harbor Superfund site in Massachusetts. The demonstration was
conducted concurrently with pilot dredging studies managed by the U.S. Army
Corps of Engineers, from which samples of contaminated harbor sediments were
obtained for use in the demonstration. Several tests were conducted on a
trailer-mounted, pilot-scale unit to obtain specific operating, analytical,
and cost information that could be used in evaluating the potential
applicability of the technology to New Bedford Harbor and other Superfund
sites. The primary objective of this demonstration was to evaluate the
developer's treatment goals for extracting PCBs from harbor sediments.
Secondary objectives included an evaluation of (1) the unit's performance in
terms of extraction efficiency and a mass balance, (2) system operating con-
ditions, (3) health and safety considerations, and (4) equipment and system
materials handling problems.
CF Systems achieved an overall PCB concentration reduction over 90 per-
cent for sediment samples that contained 350 ppm and 2,575 ppm. The unit
generally operated within specified conditions for flowrates, pressure
temperature, pH and viscosity. Deviations from operating specifications could
not be correlated to changes in extraction efficiency. No significant
releases of pollutants to the atmosphere or surrounding area soils occurred.
Results of the demonstration tests show that the CF Systems technology is
capable of reducing the PCB content of contaminated sediment by greater than
90 percent without a risk to operating personnel or the surrounding community.
This information will assist in the engineering design and costing of a
commercial scale unit at New Bedford Harbor in addition to identifying other
sites where the technology may be economically applied.
425
-------
INTRODUCTION
Through the Superfund Innova-
tive Technology Evaluation (SITE)
program, the U.S. Environmental
Protection Agency (EPA) is assisting
technology developers in the devel-
opment and evaluation of new and
innovative treatment technologies.
The SITE program objective is to
enhance the commercial availability
and use of these technologies at
Superfund sites, as an alternative
to land-based containment systems
that are used most often at Super-
fund sites. Part of the SITE pro-
gram involves field demonstrations
to gather real-world data on a
technology. The developer is
responsible for the cost of oper-
ating the equipment during the
demonstration, while EPA is respon-
sible for the analytical costs and
evaluation associated with the
demonstration. In most cases, the
demonstration is performed at an
actual Superfund site that provides
appropriate site and waste charac-
teristics for the specific technol-
ogy to be tested.
CF Systems Inc., of tfaltham,
Massachusetts, developer of a
liquified propane extraction tech-
nology, was selected to demonstrate
their pilot-scale system. New Bed-
ford Harbor was chosen for the dem-
onstration of CF Systems' technology
because the harbor sediments are
contaminated with polychlorinated
biphenyls (PCBs), a complex organic
substance amenable to extraction
with CF Systems' process.
The developer's pilot-scale
treatment technology is a trailer
mounted unit designed to handle
pumpable soils, sludge, or sedi-
ments. The system was designed to
be operated in a continuous, counter
current mode. The unit operates in
the six basic steps shown in Figure
1, that can cover extraction, phase
separations, and solvent recovery.
A mixture of liquified propane and
butane was used as the extraction
solvent.
The six process steps are: 1)
pumpable (slurried) solid waste is
fed into the top of an extractor; 2)
the solvent, in this case a propane/
butane mix, is condensed by com-
pression and allowed to flow upward
through the same extractor. In the
extractor the solvent makes non-
reactive contact with the waste,
dissolving out the organics it con-
tains. This is a somewhat non-
specific organic extraction process,
though it is based on the solubility
of the organic waste in extracting
liquified gas; 3) the residual mix-
ture of clean water or water/solids
can be removed from the base of the
extractor; 4) the mixture of solvent
gas and organics leaves the top of
the extractor and passes to a
separator through a valve which
partially reduces pressure—the
reduction of pressure causes the
solvent to vaporize out of the top
of the separator; 5) extraction gas
is then collected and recycled
through the compressor as fresh
solvent; and 6) the organics left
behind are drawn off from the
separator.
The demonstration was designed
to evaluate the treatability of New
Bedford Harbor sediments and to
provide operating and scale-up data
to assess potential commercial-scale
applications. The demonstration
included equipment setup; a "shake-
down" stage to set process condi-
tions; and daily start-up, opera-
tion, and shutdown. When tests were
completed, the demonstration con-
cluded with equipment decontamina-
tion and site closure. Thus, all of
the major components of a full-scale
cleanup of New Bedford Harbor were
demonstrated.
426
-------
PURPOSE
Criteria were established to
provide a basis for evaluating the
pilot-scale unit. These criteria
addressed treatment claims and
operational claims made by the
developer. Health and safety issues
were also addressed throughout the
demonstration. Data were collected
during 4 tests for comparison
against treatment and operational
criteria and for assessing health
and safety issues. The PCB concen-
trations contained in harbor sedi-
ments fed to the unit and the number
of passes through the unit were
varied for each of the tests.
Treatment Claims
CF Systems' treatment claim was
to remove PCB contaminants from har-
bor sediments over the course of 4
tests. Test 1 was a shakedown test
only. The feed rate and solvent to
feed ratio were set and operating
conditions were observed. Test 2
was conducted to show that harbor
sediments with 350 ppm of PCB could
be reduced by at least 90 percent
after 10 passes through the unit. A
pass was defined as one cycle,
wherein treated sediment would be
recycled through the unit. In Test
3, a 50 percent reduction was
claimed after 3 passes for a 288 ppm
sediment. Test 4 consisted of
reducing a high concentration
sediment, 2,575 ppm, to the Test 2
feed level, 350 ppm.
Operational Claims
System operating criteria were
set during the shakedown portion of
the demonstration. These criteria
are shown in Table 1.
Extractor pressure was con-
trolled at the unit's main
compressor and at the organics dis-
charge from the extraction segment
of the unit. Solvent flow rate and
the solvent to feed ratio are set
after laboratory bench-scale tests
were run on various mixtures of
solvent and feed.
The feed temperature represents
the temperature of the material
pumped into the feed unit. Feed in
the extractor was maintained above
60°F to avoid the possibility of
hydrate formation, which could have
interfered with the extraction pro-
cess. If the feed is above 120°F,
it must be cooled to prevent vapori-
zation of the solvent.
The feed flowrate represents
the rate at which material is pumped
from the feed kettle into the unit.
Operational flow rates above the
listed maximum can force segments of
the system such as decanters and
control valves, beyond their effec-
tive hydraulic capacity.
The viscosity and solids con-
tent must be such that the feed
material is pumpable. Feeds with a
viscosity above the listed range
were slurried with water to yield a
pumpable viscosity. In order to
prevent damage to the process
equipment, the pilot-scale unit has
a maximum limit for solids size.
Health and Safety Issues
Criteria were not established
to evaluate health and safety
issues; however, the health hazards
associated with project activities
were evaluated. The principal chem-
ical hazards of concern for this
project included polychlorinated
biphenyls (PCBs) and toxic metals,
including cadmium, chromium, copper,
and lead. These chemicals were
known to be present in contaminated
harbor sediments. It was also
suspected that some levels of
gaseous propane, other organic
vapors, and hydrogen sulfide would
be encountered.
427
-------
Volatilization of PCBs to harm-
ful airborne vapor levels and/or
increased airborne particulate con-
centrations, containing PCBs during
initial sampling and feedstream
preparation operations, were not
considered to be likely because of
the low vapor pressures of the PCBs
and the wet characteristics of the
sediment material. However, because
of the toxicity of the PCBs, moni-
toring, repiratory protection, and
complete dermal protection were
required when handling contaminated
sediments.
Air sampling and personnel mon-
itoring were conducted to emulate
chemical releases to the atmosphere.
Soil samples were taken before and
after the demonstration to determine
if spills or atmospheric release had
contaminated soils in the area where
tests were staged. Finally, a car-
bon adsorption cannister mounted on
the unit vent was analyzed to deter-
mine the amount of PCBs contained in
propane vented from the unit during
system shutdown.
APPROACH
Treatment Claims
Sampling and analysis were
conducted for Tests 2, 3, and 4 to
evaluate treatment claims. Test
results are shown in Table 2. Sam-
ples were taken of (1) the sediments
initially fed to the unit for each
test, (2) sediments treated after
each pass, or cycle, through the
unit, and (3) extracted organics
collected at the end of each test.
The critical analytical method
was that used for the measurement of
PCBs, since analytical data gener-
ated by this project would be used
to evaluate quantifiably the devel-
oper's claims of PCB removal by the
demonstration technology. EPA
Method 8080 was used throughout the
demonstration project to analyze
PCBs. Other analyses were process
control observations, and/or ulti-
mate disposal determinations of
project residues.
A detailed sampling and analy-
sis plan and an approved Quality
Assurance Project Plan (QAPP) were
developed in accordance with EPA
Office of Research and Development
guidelines. The following analyses
were conducted:
o PCB (soils, water)
o PCB (Soils)
o Waste Dilution
o Semivolatiles
o Trace Metals
o Particle Size
o Total Recoverable Oil and Grease
o Percent Solids
o EP Toxicity (metals)
o pH in Calcareous and Noncal-
careous Soils
o Total Metals
Operational Claims
Process controls, wastestream
masses, and utilities were measured
at various intervals during each
test. Listed below are critical
operational parameters and measure-
ment frequencies for each test:
o Feed temperature, viscosity, and
pH—measured at each pass
o Feed sediment and treated sedi-
ment mass—measured at each pass
o Feed flow rate—measured every
10 minutes
o Extractor pressure and tempera-
ture—measured every 10 minutes
428
-------
o Solvent flowrate — measured every
10 minutes
o Extracted organics mass — measured
each test
The accuracy of devices used to
measure pressure, temperature, flow,
viscosity, and weight was below 5
percent relative standard deviation.
Health and Safety Issues
Several types of portable moni-
toring equipment were used during
the tests, including:
o Portable Organic Vapor Analyzer
(Century© OVA)
o Portable Photoionization Meter
(HNu©)
o Combustible gas /oxygen/hydrogen
sulfide meters (MSA© and
Enmet-Tritector©)
o Detector tubes (Sensidyne-
Gastec©)
o Personal air sampling
(Dupont-P200®).
pumps
The OVA and HNu© meters were
used to monitor for organic vapors
at all work stations on the unit,
while personnel monitored process
equipment. The OVA also was used as
a survey meter on the process
equipment to search for possible
fugitive emissions from the equip-
ment. Two portable combustible gas
meters were used to check for ele-
vated levels of propane and hydrogen
sulfide during the equipment shake-
down period and for spot testing
during the demonstration. The pilot
unit also contained two integral
combustible gas detectors located on
either end of the unit, which were
observed during the tests.
Personal sampling was conducted
using personal sampling pumps and
150-mg charcoal tubes and Florisil©
tubes to determine personal expo-
sures to organic vapors and PCBs,
respectively.
Soil samples were taken from 10
locations in the test staging area
to determine PCB levels in the soil
prior to the demonstration. Samples
of soil were taken at zero to six
inches of depth. The 10 locations
were resampled after CF Systems
removed their equipment from the
site and debris was removed from the
site. The unit's carbon cannister
was removed and sampled at the end
of the demonstration after low
pressure propane had been vented
from the unit through the cannister.
At the conclusion of the tests,
toluene was run through the unit to
decontaminate unit hardware.
Toluene was introduced to the unit
as a feed material and system efflu-
ents were sampled and analyzed for
PCB.
PROBLEMS ENCOUNTERED
The appropriateness of EPA
Method 8080 for analysis of PCBs was
questioned by several reviewers. It
was believed that the extraction
technology could selectively remove
higher or lower molecular weight
PCBs congeners. This possible
selective extraction would not be
apparent through use of Method 8080.
The use of EPA Method 680 for PCB
analysis was suggested by the
reviewers.
Therefore, both methods were
used during Test 4 and results were
compared before the decision was
made to use Method 8080 for all
samples collected. Method 8080 is
an analysis method for determining
PCBs as Aroclors®. Analytical
results reported for Aroclors©
reflect an aggregation of individual
PCB compounds, or congeners. With
Method 680, each of the 209 PCB con-
429
-------
geners is analyzed and reported as
an individual compound. Analytical
results obtained with Method 680
showed that any selective extraction
that may occur would not be signifi-
cant. The lower molecular weight
PCBs were extracted with a 97 per-
cent efficiency while the higher
weight PCBs were extracted with 93
percent efficiency. Consequently,
Method 8080, a lower cost method,
was selected.
The duration of the test pro-
gram was extended because of opera-
tional difficulties experienced with
the sampling and analysis effort.
Foaming occurred in the treated
sediment collection tank which hin-
dered sample collection. The unit
irregularly accumulated and dis-
charged feed material solids which
prevented calculation of a PCB mass
balance on a pass-by-pass basis.
Finally, the lack of onsite PCB
analytical capabilities did not
allow field personnel to make quick
adjustments to the process opera-
tions or sampling procedures. These
types of problems were anticipated
since the unit was designed for
continuous operation but was oper-
ated in a batch mode during this
demonstration. The developer
believes that each of these sampling
issues can be accommodated in the
design and operation of a full-scale
commercial system.
RESULTS
Treatment Claims
PCB concentration percentage
reductions of PCB and reduction
goals were as follows:
PCB Concentration Reduction
Test Reduction (%) Goal (%)
~5 89 90
3 72 50
4 92 86
Treatment claims were met for
Tests 3 and 4. The treatment claim
for Test 2 was not met after the
tenth pass, however, a 98 percent
concentration reduction was achieved
by the ninth pass. It is believed
that solids retained in the unit
cross-contaminated treated solids
discharged at the tenth pass which
caused the tenth pass concentration
to be higher than the ninth pass.
EPA is using this information to
identify the types of hazardous
waste cleanup situations where the
technology could be economically
applied.
The performance of the treat-
ment unit was evaluated in terms of
extraction efficiency and a mass
balance. Extraction efficiency per
mass is defined as the input PCB
concentration minus the output PCB
concentration divided by the input
PCB concentration (multiplied by 100
percent). Table 2 shows the extrac-
tion efficiency per pass for each of
the three tests. Extraction effi-
ciencies greater than 60 percent
occurred at the first pass of each
test. However, efficiencies ranged
from zero to 84 percent in later
passes. This wide range is probably
due to solids retention, since
changes in extraction efficiency did
not correlate with changes in
operating conditions.
Table 2 also shows solids out-
put from the unit as a percent of
the solids fed to the unit per pass.
The data show that solids were
irregularly retained and discharged
by the unit. This was probably due
to the small volumes that were
batch-fed to the unit during each
pass. Ordinarily, the unit would be
run continuously over an eight hour
working shift. However, schedule
limitations only allowed run times
of several hours per day. The
nominal flow rate is 720 gallons per
430
-------
shift. During this demonstration,
only 50 to 100 gallons were run per
shift. Overall, a total of 794
pounds of solids were processed over
19 passes in Tests 2, 3, and 4. Of
the total, 93 percent of the total
solids were accounted for in
effluent streams.
Even though solids retention
caused cross-contamination of
treated sediments, significant PCB
removals occurred. For example,
Test 2, Pass 9 contained 8 ppm of
PCB. Compared with Test 2 feed (350
ppm), this represents an extraction
efficiency of 98 percent. System
decontamination procedures showed
that PCBs were separated from the
sediment since nearly all, 91
percent, of the PCBs were contained
in extract subsystem hardware. Of
the 81 grams of PCB fed to the unit
during Tests 2, 3, and 4, only 4
grams remained in the final treated
sediments. This indicates an over-
all removal efficiency of 95
percent.
Operational Claims
The unit generally operated
within the specifications listed in
Table 1 with only several excep-
tions. Criteria were met for feed
flowrates, solids content, maximum
possible size, viscosity, and pH as
well as extractor pressures. The
solvent flow rate and solvent to
feed mass ratios fluctuated above
and below criteria throughout the
tests but did not have an observable
affect on pass-by-pass extraction
efficiency. Temperature of the feed
sediments fell below the minimum
temperature criterion during passes
6, 7, 8, 9, and 10 of Test 2.
Commercial-scale designs for
application of the technology should
ensure that operating specifications
are maintained. Feed materials are
likely to be well below 60°F
throughout winter months and this
could affect system performance.
Therefore, heat must be added to
sediments fed to a commercial-scale
unit. Coarse solids removal will be
required to maintain feed sediment
paricle sizes below one-eighth inch.
Wide fluctuations in the feed to
solvent ratio should be minimized.
Extraction efficiency is directly
related to the amount of solvent
available for solubilizing organics
contained in the feed.
Health and Safety Issues
The Health and Safety Plan
established procedures and policies
to protect workers and the public
from potential hazards during the
demonstration. Implementation of
these procedures and health and
safety monitoring showed that OSHA
level B protection is necessary for
personnel that handle system input
and output. Although, only OSHA
level C protection is necessary for
unit operators.
Combustible gas meters indi-
cated that levels at approximately
20 percent of the lower explosive
limit for propane were encountered
while samples were taken. Back-
ground air sampling and personnel
monitoring results indicate that
organic vapors and PCB levels were
present at levels below the detec-
tion limit for the analytical
methods. Site soil samples taken
before and after the demonstration
indicate that demonstration
activities did not result in
increased PCB levels in the staging
area soils. Therefore, no site
contamination occurred due to the
demonstration.
ACKNOWLEDGEMENTS
The technical assistance and
support of John Moses and Christo-
pher Shallice of CF Systems, Inc.,
Frank Ciavattieri of USEPA Region I,
and Richard Hergenroeder of Science
Applications International Corpor-
ation (SAIC) are appreciated.
431
-------
Extn
1.
»— ^
Wuliwiur
or Studflft
FIGUtl 1. SIHrUniD FLOW CUXT
f 1
4. ' * 1
•dor |— fX— «- 1
V - SLj U
L J801"-1' *
K
K
Comprauor
I '
f Water Organic*
3. and/or Solid*
TABLB 1. OPERATING SPECIFICATIONS
Miniauai Hoainal
Extractor Pressure (FSIG) 180 240
Extractor Tup. (*F)
Feed Tap. (*F)
60 100-110
60 70
Solvent Flov (pounds/minute) 8 12
Feed Plovrate (gallons per lainute) 0.2 0.2-0.5
Solvent/Feed Ratio
1 1.5
Feed Solids (percent by veight) 10 30
Solids Size
pa
_
6 7
Viscosity (centlpolse) 0.5 10
MaxiauB
300
120
100
15
1.5
2
60
1/8 inch
12
1,000
TABLE 2. TEST RESULTS
_ Pass-by-Pass Concentration
Test Pass PCB Concentration Reduction Efficiency
Nuaber Husber (pp.) (percent)
2 Feed
2 1
2 2
2 3
2 4
2 5
2 6
2 7
2 a
2 9
2 10
3 Feed
3 1
3 2
3 3
4 Feed
4 1
4 2
4 3
4 4
4 5
4 6
350 Hot Applicable
77 78
52 32
20 . 62
66 Ho Reduction
59 u
41 31
36 12
29 19
8 72
40 Ho Reduction
288 Hot Applicable
47 84
72 Ho Reduction
82 Ho Reduction
2575 Hot Applicable
1000 61
990 i
670 32
325 52
240 26
200 17
Solids Output
As a Percent
Of Input
Hot Applicable
60
90
135
45
125
90
100
150
110
110
Not Applicable
115
90
85
Hot Applicable
80
130
80
90
75
135
432
-------
Disclaimer
Ihis paper has been reviewed in accordance with the U.S. Environmental
Protection Agency peer and administrative review policies and approved for
presentation and publication.
433
-------
ANALYSIS OF VAPOR EXTRACTION DATA FROM APPLICATIONS IN EUROPE
Dieter Killer, Ph.D. and Horst Gudemann, M.S.
HARRESS Geotechnics, Inc.
200 Corporate Center Drive
Coraopolis (Pittsburgh), PA 15108
ABSTRACT
Vapor extraction, an in-situ process to remove volatile organic
compounds (VOC) from soils of the vadose zone, has been applied in
Europe since the early 1980s. With considerably more than 1,000
systems_ operating under virtually all subsoil conditions, vapor
extraction is considered to be a standard procedure in Germany.
In a vapor extraction well a negative differential pressure is
created by a blower or similar device. The differential pressure
generates a steady flow of soil gas towards the extraction well and
thus provides a flushing of the soil with air undersaturated in
respect to the contaminant concentration. Contaminants will
evaporate into the gaseous phase both from the liquid phase and from
the soil. Differential pressures applied range from 15" - 350" of
water. The contaminated discharge air can be treated by activated
carbon or other suitable methods. The effective radius of vapor
extraction systems (VES) ranges typically from 20' to ISO7
underneath non-sealed - and up to 300' underneath sealed surfaces.
Parameters that influence the performance of the VES are the
type_ of soil, depth to the groundwater, length of the screened
section in the well, position of the screen, permeability of the
ground surface and the type of blower used.
Discharge data from several hundred cases reveal typical common
characteristics, regardless of the type of soil, type of blower and
specific performance characteristics. Contaminant discharge is high
during a first phase of about two weeks. A short second phase is
marked by a transition to a stable level at around 10% of the
initial concentration. In a third phase of several months to about
two years the contamination decreases to background levels.
434
-------
INTRODUCTION
Contamination from
volatile organic compounds
(VOC) have turned out to be
widespread due to their almost
ubiquitous presence in
industrial processes.
Specifically, VOC include
halogenated hydrocarbons like
TCE, PCE or TCA, aromatic
hydrocarbons like benzene,
toluene, xylene and volatile
fuels like gasoline.
Especially halogenated
hydrocarbons exhibit physical
properties that enable them to
penetrate even coated concrete
pads and seep into the ground
rapidly, both as liquids and
vapors. Significant
contamination not only occurs
at underground storage tanks,
but also in drum storage,
handling and application areas
and along pipelines or trans-
port paths.
Traditionally, soils that
were contaminated with VOC were
excavated and disposed of at a
proper hazardous waste site.
Due to the enormous amounts of
waste generated and only
limited storage capacity at the
available disposal sites, the
remediation is usually costly
and only moves the problem-
VOC contaminated soils - to
another location.
Recognizing the problem,
regulatory agencies increas-
ingly favor on site and in situ
remediation of soil and ground-
water contamination. In
addition, future regulations
could further limit the
availability of hazardous waste
disposal sites and are likely
to increase the already high
costs of disposal. In the case
of VOC contaminations, vapor
extraction has shown to be an
effective and economically
feasible in situ remediation
(3).
The high specific
retention capacity of the
vadose zone for VOC represents
a continuous source of
groundwater contamination, as
even after a complete
volatilization of the liquid
phase material, vapors will
continue to migrate. Remedi-
ation of the unsaturated zone
by vapor extraction inter-
rupts this migration path and
is cost effective, since it
captures the contaminants
prior to dissolution in the
groundwater (3).
Since vapor extraction is
a widespread and generally
accepted technology in Europe,
particularly in Germany, the
present paper draws on the
experience gained by HARRESS
Geotechnik since the early
80 's through operation of
roughly one thousand vapor
extraction systems in Germany
and throughout Europe and
summarizes the results.
PURPOSE
The comparison of case
histories and evaluation of
vapor extraction serves
several purposes. It
- outlines parameters,
which determine the
effectiveness of vapor
extraction.
- shows the common
discharge character-
istics of vapor
extraction systems in
different geologic
435
-------
settings and varying
concentrations of
contaminants.
allows predictions on
the development and
progress of the
remediation
allows for design of a
treatment system for the
extracted vapors
demonstrates that the
technology has
progressed from an
experimental stage
into a proven and
reliable remedial
process.
APPROACH
Six case histories have
been selected for demonstra-
tion. The specific cases were
chosen on the basis of their
documentation, variety of
settings, difference in
contaminant concentration,
type of contaminant, and
optimum results of vapor
extraction. The justification
for selecting cases on the
basis of the ideal performance
of vapor extraction lies in the
intention of this paper to show
the predictability of the
clean-up process, if all
relevant parameters for the
remediation are adequately
considered. Knowledge of what
to expect in the ideal case
allows for detection of
deviations which can point to a
deficiency in the design of the
system.
For comparison purposes,
all discharge contaminant
concentration data from the
different vapor extraction
systems were compiled in one
graph. Integrations of the
graphs were performed to
estimate the volumetric
balance during the
remediation, i.e. percent of
total contaminant discharge
per unit time.
PROBLEMS ENCOUNTERED
The cited case histories
represent the ideal progres-
sion of a remediation by vapor
extraction. However, numerous
sources for a deviation from
the ideal pathway exist,
leading to either an extended
remediation or in the worst
case to a failure to clean up
the contamination. The two
most common problems for clean
up of the vadose zone are the
placement of the vapor extrac-
tion system outside or at the
periphery of the contamination
center and a continuous
recharge of contaminants into
the subsurface. A continuous
input can result from ongoing
leakage/spills or from evap-
oration of VOCs from
considerably contaminated
groundwater. Misplacement of
vapor extraction wells can be
avoided by carefully defining
the extent and the center of a
source area.
A precise source defini-
tion is most efficiently done
with soil gas investigations,
the results of which also
supply an appropriate basis
for the calculation of
proj ected contaminant concen-
trations in the discharge air.
Given this information, a
vapor extraction system might
still operate inefficiently
if, in the presence of a very
shallow groundwater table,
436
-------
vertical extraction wells with
short screens are used. In
such cases, a horizontal
screen installation provides an
effective solution.
RESULTS
Case Histories
Data from 6 case histories
were compiled into a composite
graph (Fig.l) and summarized on
Table 1. In each of the case
histories, vapor extraction has
performed exceedingly well,
either having already achieved
a clean up or progressing in a
way that a clean up can be
predicted within the near
future.
The curves appear to
reflect three phases merging
into each other. In all cases,
the discharge concentrations
decreased by 80% - 90% within
the first 2O days of operation.
After a steep decline lasting
for 4-7 days (Phase 1) , a
less pronounced transisiton
phase (Phase 2) is observed for
another 10 - 12 days, even-
tually continuing in a gradual
asymptotic decrease to back-
ground concentrations for the
remainder of the operation
(Phase 3) . The initial phase
is more pronounced if the
discharge starts out at high
concentration levels. During
the long lasting third phase,
it is conspiciuous that in five
of the six case histories, the
absolute concentrations vary
within a comparatively narrow
range. This is even more
expressed if mass flow rates
are regarded.
The progression of the
decline of contaminant
concentrations with time seems
to be independent of the
particular type of soil,
initial concentrations or the
specific characteristics of
the vapor extraction system
and well applied in each
particular case.
In most cases, the
measured data can be approx-
imated by two regression
curves. One of the curves
fits the steep branch of the
empirical contaminant concen-
trations curve, the second
regression curve approximates
the asymptotic decline in
contaminant concentrations.
While the regression curves do
not allow a precise prediction
of the time required for a
complete clean up, they can
serve as a guideline to
monitor the progress of the
remediation.
The distinct concentra-
tion decrease during Phases 1
and 2 could feign an equally
rapid removal of contaminants
from vadose zone soils.
However, mass balance calcu-
lations illustrate the
considerable contribution of
the long-term operation at low
concentrations. While 50% of
the total amount of removable
contaminants are discharged
after a time period ranging
from 1.5 weeks to 4 months,
the removal of the remaining
50% is only achieved within 63
days to as many as 570 days of
operation. In the cases
presented, the total clean-up
time ranged from more than 100
days to approximately 2.5
years.
437
-------
FIGURE 1
VAPOR EXTRACTION
Discharge Performance
Discharge Concentration (ppm)
400 -ft
300-
200-
100-
X
A
-X-
30
Bin
60 90 120 150 180 210 240 270 300
Elapsed Time (Days)
Ulm
Stg5
Wue4
0 Wue8
TABLE 1
CASE HISTORY DATA
soil!:
CONTAMINANT:
VOLUME rLOffi
RAKGE OP
INFLUENCE:
BERLIN
(BLN)
aediiua & fine
grained sands
PCE
390 CFH
180'
WURZBURG
(WUE4)
clayey silts
w/limestone
fragments
PCE
68 CFM
40'
WORZBURG
(WUE8)
clayey silts
w/limestone
fragments
PCE
70 CFM
40'
REDWITZ
(RED5)
sandy-
clayey
silts
PCE
50 CFM
60'
STUTTGART
(STG5)
weathered
claystone,
silt
TCE
75 CFM
50'
ULM
(ULM)
coarse
grained fill,
silty sands
PCE
103 CFM
75'
438
-------
Physical Processes in the Soil
Unless a continuous
recharge of contaminants at
high rates is encountered, the
presence of free product in
soils is typically limited to
small fluid particles trapped
in soil pores (4) . At the
expense of these fluid drop-
lets as well as of compounds
dissolved in the soil moisture
or adsorbed to the soil
matrix, a contaminant vapor
phase will develop and spread
over time controlled by the
concentration gradient.
After an initial exchange
of the soil gas volume in the
pores, ambient air will be
drawn continuously from outside
the contaminated area. While
passing through the subsurface
the air will be charged with
evaporating VOC and subse-
quently be discharged through
the vapor extraction system
(1) . The process resembles a
continuous flushing of the
soil with clean air and will
continue until volatilization
and desorption of contaminants
is complete.
The three phases
describing the system perfor-
mance are attributable to
different processes occuring
during the operating time.
During the first phase the
contaminant saturated soil gas,
which is present in the pore
space under equilibrium
conditions, is discharged. A
rapid evaporation of free
product droplets due to
disturbance of the equilibrium
is presumably also in part
occurring during the first
phase. The short transitional
period, when the contaminant
concentration in the
discharged soil gas has
already decreased by more than
80%, is most likely
characterized by a shift in
source of the contaminants.
Rather than from an evapor-
ation of liquid particles, the
contaminants in the discharge
vapors result from a
desorption of contaminants
from soil particles. A second
process, which is believed to
become important in this
phase, is the transition of
contaminants previously
dissolved in the soil moisture
into the gaseous phase (3).
The third phase repre-
sents the comparatively slow
desorption process and a
gradual reduction of the
contaminated soil volume. The
time this period requires
depends on the physical
properties of the compounds
involved and the mass of VOC
retained in the soil. As long
as sufficient supplies of
contaminants are existing, a
nearly constant extraction
rate is accomplished over a
period of several months up to
about two years.
Data from air flow models
indicate a rapid decrease of
pressure differentials to very
low levels with increasing
distances from the vapor
extraction well (2) . The
differential pressures
throughout the majority of the
range of influence are too low
to be directly responsible for
enhanced volatilization of the
contaminants. However, even
very small differential
pressures create a pressure
gradient and thus induce the
flow of air towards the vapor
439
-------
extraction well. The equilib-
rium disturbance caused by this
process, rather than enhanced
volatilization through large
differential pressures is
considered to be responsible
for the extraction of
contaminants.
Design Parameters
Aside from the type of
soil, the range of influence of
a vapor extraction system is
determined by a number of
factors, in particular by the
length and position of the
screened interval in the
extraction well, the thickness
of the vadose zone, and the
permeability of the surface.
Differential pressures decrease
exponentially with increasing
distance from the extraction
well. While higher differ-
ential pressures create
considerably higher volume flow
rates, the effects on the range
of influence are small. For
example, in medium grained
sands an increase of the
differential pressure raises
the volume flow exponentially,
while the effective radius
remains almost unaffected (2) .
Given a constant differ-
ential pressure, variations of
the screen length reveal a
linear relationship between the
length and the volume flow
rate, and also a distinct
increase of the effective
radius. The position of the
screened interval creates
reciprocal effects: shifting
it to deeper portions of the
vadose zone reduces the volume
flow but increases the range of
influence (2).
Under the assumption of a
screened interval of constant
length positioned in the
middle portion of the vadose
zone, an increasing thickness
of the vadose zone results in
a distinct increase of the
range of influence, while it
seems to have only minor
effects on the volume flow
(2).
The same effect can be
observed if variably sized
areas of surface sealing are
regarded. The presence of an
impermeable layer, such as a
concrete pad, gives rise to
significantly increased ranges
of influence at almost
constant volume flow rates.
However, a sealed surface is
not a prerequisite for the
successful application of
vapor extraction systems.
Aside from theoretical
model calculations, the actual
range of influence of single
systems or arrays should be
determined in the field. This
can be done in several ways,
ways, by actually repeatedly
measuring the soil gas
concentrations at varying
distances by measuring the
differential pressure created
in air monitoring points, or
just by qualitatively
determining the existence of a
differential pressure through
the observation of smoke
trails created by air current
tubes at such points.
SUMMARY
In order to achieve a
successful remediation of the
vadose zone by vapor extrac-
tion, a continued input of
440
-------
contaminants into the
subsurface, evaporation of
contaminants from a plume of
contaminated groundwater and
positioning of the vapor
extraction well outside the
contamination center must be
ruled out. A thorough
assessment of the contamin-
ation pattern is, thus, a
prerequisite.
To properly address the
problem, various system design
parameters are to be
considered. They include the
length and position of the
screened well interval(s), as
well as the selection of the
appropriate suction device.
The application of the
results of previous experi-
ences with vapor extraction as
exemplified in the case
histories allow:
- a prediction of the
development of
contaminant concen-
trations in the
discharge air with time.
- design of a treatment
system for the
discharged vapors
- timely recognition of
factors impeding or
preventing a successful
clean up by comparing
the actual curve with
the proj ected ideal
curve.
- optimized design of
vapor extraction wells.
REFERENCES
(1) BRUCKNER, F. , HARRESS,
H.M. & KILLER, D. (1986) : Die
Absaugung der Bodenluft- ein
Verfahren zur Sanierung von
Bodenkontaminationen mit
leichtfluechtigen chlorierten
Kohlenwasserstoffen.-
Brunnenbau, Bau von
Wasserwerken, Rohrleitungsbau
(bbr), 37. pp 3-8.
(2) CROISE, I. et al. (1989):
Computation of Air Flows
Induced in the Zone of
Aeration during In Situ
Remediation of Volatile
Hydrocarbon Spills. - In:
KOBUS & KINZELBACH (Eds.):
Contaminant Transport in
Groundwater. - Balkema,
Rotterdam, pp 437-444.
(3) GUDEMANN, H. & KILLER, D.
(1988): In Situ Remediation of
VOC Contaminated Soil and
Groundwater by Vapor
Extraction and Groundwater
Aeration. - Proceedings
HAZTECH International '88,
Cleveland, Ohio, pp 2A-90-
2A-111.
(4) SCHWILLE,
Leichtfluechtige
wasserstoffe in
klue ft igen
Modellversuche.
F. (1984):
Chlorkohlen-
poroesen und
Medien-
Besond.
Mitt. Deutsch.
Jahrb., 46. 72
Gewaesserkundl.
+ XIIpp
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
441
-------
IN-SITU BIOREMEDIATION OF CYANIDE
Robert C. Weber, P.E., Gregory Smith, Joseph Aiken
ENSR Consulting and Engineering
Westmont, Illinois 60559
Dr. Richard Woodward, Dr. David Ramsden
ENSR Consulting and Engineering
Houston, Texas 77098
ABSTRACT
So that manufacturing operations would not be interrupted, in-situ
bioremediation was chosen for soil remediatioji after a release of cyanide
beneath the main floor of an electronics manufacturing facility. Laboratory
studies were conducted to (1) determine the effectiveness of stimulating
naturally occurring bacteria to effect bioremediation of cyanide and (2)
determine optimum nutrient conditions for the bioremediation system. Seven
equal fractions of a basic medium containing 2,400 ppm cyanide were placed
into capped reactor tanks; carbon and phosphate concentrations were adjusted
differently for each tank. Cyanide was removed at a very rapid rate in all
of the tanks. While the tanks varied in their initial and final rates of
cyanide removal, all achieved similar overall rates. Engineering design for
construction and operation of nutrient mixing and feed systems and for
groundwater extraction are presented.
INTRODUCTION
An active electronics manufacturing
facility experienced a subsurface
release of zinc cyanide plating solu-
tion from an underground storage tank
located beneath the main floor of the
facility. The tank excavation was
limited, to avoid damage to the build-
ing structure. Although removal of
the tank was accomplished, contamina-
ted soil and shallow groundwater
remained. Soil cyanide concentrations
were on the order of 15,000 ppm at the
excavation walls, decreasing to less
than 100 ppm within about 25 feet of
the excavation area. Remediation was
complicated by the need to restore
the floor area so that production
operations could resume. A remedia-
tion approach was needed which would
be as unobtrusive as possible.
This paper presents the results of
laboratory studies conducted to evalu-
ate the effectiveness of in-situ
bioremediation of the remaining cyan-
ide. The technical approach focused
on stimulation of naturally occurring
bacteria. Optimum nutrient conditions
were determined. The engineering
approaches for construction and opera-
tion of nutrient mixing and feed
systems and for groundwater extraction
are also presented.
442
-------
BIOCHEMISTRY
Although generated by some plants,
many fungi, and a few bacteria, cyan-
ide is toxic to most organisms.
Cyanide is a specific inhibitor of
porphyrin-type biochemical systems.
These include iron-containing cyto-
chromes and hemes associated with
respiration, oxygen transport, and
electron transport. Also sensitive
to cyanide are the plastcyanins (cop-
per-containing proteins associated
with photosynthesis) and cobalt por-
phyrins such as vitamin B-12.
Some microorganisms, however, are
very tolerant of or resistant to
cyanide. Bacillus pumilus is re-
ported to tolerate 0.1M potassium
cyanide.
Although ammonia is readily assimi-
lated and is the preferred nitrogen
source for almost all microbes, some
bacteria can use cyanide compounds as
a nitrogen source. Cyanide metabolism
can serve the dual purpose of sup-
plying nitrogen and detoxifying cya-
nide. The organisms responsible for
cyanide degradation are generally
alkaliphilic.
The simplest systems for microbial
use of cyanide as a nitrogen source
utilize cyanide hydratase and cyanide
oxygenase. The hydratase system
generates formamide and formate with
the release of ammonia. The multi-
component oxygenase system forms
carbon dioxide and ammonia.
Alternative mechanisms for cyanide-
nitrogen assimilation are the syn-
thesis of thiocyanate and its subse-
quent oxidation to carbon dioxide and
ammonia, and of cyanoalanine, which
may be converted to asparagine and
aspartic acid. Generation of thio-
cyano-aminobutyrate is apparently a
dead-end detoxification process.
Not surprisingly, microbes rarely use
cyanide as a carbon source. Carbon
dioxide, the usual carbon product of
cyanide degradation, is energetically
expensive to assimilate. With the
exception of auto trophic bacteria, as
well as scavenging pathways, the vast
majority of microbes prefer or are
obligated to use organic carbon sour-
ces for the vast bulk of their carbon.
APPROACH
Soil borings were advanced from the
surface of the tank excavation and
near the expected boundary of the
contaminated zone to determine the
extent of contamination and to provide
soil samples for characterization of
biological activity and laboratory
treatability studies. Monitor wells
were installed near the expected
boundary of the contaminated zone to
help determine the extent of contami-
nation.
To determine treatability, seven
capped reactor tanks were established
as identically as possible. The basic
medium for all the reactors was mixed
and sampled as one batch before the
additions of phosphate and a carbon
source. Cyanide levels were 2,400 ppm
(about 0.1M). Seven equal fractions
of the basic medium were individually
adjusted for their phosphate and
carbon source concentrations. Phos -
phate concentrations were adjusted
according to the schedule shown in
Table 1.
A 1:1 molasses/Karo syrup (high
fructose) mixture was used as the
carbon source. Aldose sugars are
known to be excellent carbon sources
for cyanide degradation due to a
stabilizing interaction with the
cyanide and an ability to act as an
excellent aerobic energy source. The
concentrations of the mixtures in the
443
-------
Table 1. Concentrations of Compounds in Reactor Tanks (ppra)
Compound
Cyanide
Phosphate
Molasses/
Karo Syrup 1:1
Tank Number
1
2,400
0
100
2
2,400
100
0
3
2,400
100
1,000
2,400
100
10
5
2,400
1,000
100
2,400
100
100
7
2,400
10
100
tanks were adjusted according to the
schedule shown in Table 1. The tanks
were capped gas tight and sampled on
days 0, 4, 7, 11, 14, 21, 28, 32, and
35. The samples were assayed for
ammonia, cyanide, colony-forming units
(CFU; using typticase soy agar (TSA)
plates), and toxicity (Microtox™).
RESULTS
Cyanide concentrations decreased in
every tank (Figure 1). There were
slight differences in time course, but
cyanide in all tanks decreased to
about the same level within 35 days.
In every tank, cyanide removal pla-
teaued between days 11 and 21. This
plateau was associated with ammonia
accumulation (Figure 2), which may
have had a toxic effect on cyanide
removal. An alternative and perhaps
more likely possibility is that both
ammonia and cyanide were responding
to a common effector and not to each
other. Oxygen depletion in the closed
system may have limited the removal
of cyanide and the generation of
ammonia.
Ammonia levels declined in all tanks
after day 21, the day the tanks were
opened to the air. The tanks remained
open for the duration of the study.
Cyanide removal was excellent during
the active periods. The removal rate
for the days 0 to 11, averaged for
all tanks, was 5.03 mg/kg/hr (ppm/hr)
(Table 2).
The marginally best overall treat-
ments were in Tanks 1, 2, and 3.
Cyanide removal rates for treatments
1 and 2, days 21 to 32, were the
highest, although all tanks were
slower during days 21 to 32 than they
were during days 0 to 11. While the
CYANIDE
AMMONIA
2500
2000-
1SOO-
1000
DAY
DAY
Figure 1
Figure 2
444
-------
Table 2. Cyanide Removal Rates (mg/kg/hr)
Tank Number
Time Period (days)
0 - 11
11 - 21 21 - 32 0 - 32
1
2
3
4
5
6
7
Average
4.17
4.55
5.68
5.68
5.30
4.92
4.92
5.03
1.25
0.83
0.83
0.83
0.42
0.42
0.42
0.71
3.03
3.03
1.89
1.52
2.27
2.80
2.80
2.48
2.86
2.86
2.86
2.73
2.73
2.79
2.79
2.80
tanks varied in their initial and
final rates of cyanide removal, all
achieved similar overall rates. Tanks
with faster rates between days 0 and
11 were slower between davs 21 and 32.
Figure 3 illustrates the close cor-
respondence of ammonia generation and
cyanide removal rates in all of the
tanks. The cyanide concentration
appeared Co be controlling the overall
removal rates, suggesting that the
removal of cyanide was enzymatic.
Faster removal earlier gave lower
cyanide concentrations later. Those
lower concentrations then caused a
slower enzymatic removal rate later.
There was little difference between
AMMONIA & CYANIDE
Figure 3
the seven tanks. The stripping of
cyanide observed in other investiga-
tions would not seem to be a mechanism
for cyanide removal in these experi-
ments because the. system was closed
during the initial period of active
removal.
Cyanide disappeared at a very rapid
rate in all of the tanks. Simple mass
balance calculations indicated that
not all of the cyanide removal was
accounted for by the ammonia measured
in the tank liquors. There are sever-
al reasonable explanations for this
effect, none of which diminishes the
excellent removal rates for cyanide.
As Figure 4 indicates, there are
several possible fates for ammonia
generated from cyanide. The most
likely explanation is that, for any
of a multitude of possible reasons,
the CFU data do not reflect actual
cell populations. The counts suggest
that no growth in cell populations has
occurred, which seems unlikely. If
the cell populations actually increas-
ed from 1.5-2.0 x 10* to approximately
1.5-2.0 x 107 cells, a very modest in-
crease for a microbial population,
then the unaccounted-for ammonia could
be explained as cell constituents.
A less likely explanation is that
ammonia was formed, mineralized to ni-
trate/nitrite with the oxygen present,
445
-------
POSSIBLE FATES OF CYANIDE
-»THIOCYANO-AMINOBUTYRATE-W-
crr
>UK AMU ALAN INt
HICON1I , i
— iiinrn'
— kern
ATMOSPHERE
SOLUBLE
1-
ATMOSPHERE
Is
NOi —
SOLUBLE
1
ATMOSPHERE
\I
SOLUBLE
CELL
^CONSTITUENTS
SOLUBLE
Figure 4
and then denitrified to dinitrogen
when the oxygen was exhausted. This
appears unlikely because of low oxygen
availability during the initial days
of the experiment.
Ammonia was generated as cyanide was
removed (Figure 3), suggesting that
the ammonia was generated from the
cyanide. But, as mentioned above,
the total ammonia produced does not
account for the cyanide lost in the
tanks. If the tank having the highest
level of ammonia were kept closed,
the ammonia generated by day 32 would
only account for approximately 1% to
2% of the cyanide removed (Tank 5).
Since the CFU data remained constant,
unaccounted-for nitrogen was not in
cell constituents unless CFU data does
not reflect actual cell populations.
Removal of the caps on the tanks
lowered ammonia, presumably through
loss to the atmosphere. A conclusion
that ammonia was inhibiting cyanide
removal could be drawn since the
decrease in ammonia concentrations was
followed by increased cyanide removal.
An alternative explanation, as men-
tioned above, is that both systems
are dependent on cell metabolism. When
oxygen is depleted, the metabolism
stops. When the metabolism stops , am-
monia and cyanide metabolism stop.
Opening the tanks allowed oxygen to
enter. Oxygen reinitiated the ammonia.
generation and cyanide removal.
Because of the open tanks, ammonia ac-
cumulation was no longer detected at
increasing levels, while cyanide was
removed as before.
Tank 5 accumulated the most ammonia,
and was also the only tank with 1,000
ppm phosphate added. This indicated
an apparent relationship between high
phosphate availability and ammonia
accumulation, which may have been a
function of the increased buffering
capacity of phosphate. There was not
a similarly clear relationship between
cyanide removal and phosphate. The
high ammonia accumulation in Tank 5
was seen at every time point (Figure
5).
Microbial plate counts (demonstrated
by CFU/ml) remained almost constant
in all tanks (Figure 6). The CFU
ranged from 1.4 x 10* to 2.0 x 10*/niL.
Not only was there little growth, but
the numbers of cells detected were
446
-------
CUMULATIVE AMMONIA
MICROBIAL COUNT
DAYO
0 DAY 11
DAY 14
0 DAY 21
D DAY 28
• DAY 32
Q DAY 35
10-
o TANK t
•*• TANK 2
•o- TANK 3
-+• TANK 4
•«- TANKS
-o TANK 6
•*- TANK?
Figure 5
DAY
Figure 6
very low in all the tanks. Soils may
have 106 to 108 cells/g even in deep
subsoil. The results suggest that
either the enumeration medium was not
detecting all of the organisms present
or that cyanide had a sterilizing/in-
hibitory effect. An approximation of
the organism-to-cyanide ratios sug-
gests that over the 35-day period each
microorganism would have to had de-
graded billions of molecules of cyan-
ide per minute. This seems unlikely
and supports the idea that the cell-
counting medium failed to enumerate
all of the viable organisms, or that
a nonenzymatic process was removing
the cyanide.
SYSTEM INSTALLATION
Based on the results of the labora-
tory testing presented herein, a field
bioremediation system was installed
at the subject facility. This in-
stallation coincided with the removal
of the plating solution tank and
utilized the tank excavation as part
of the bioremediation system. The
field bioremediation system consists
of three main elements: (1) nutrient
injection system, (2) nutrient supply
system, and (3) groundwater gradient
control system. A cross-section of
the field bioremediation system is
shown in Figure 9.
The relative toxicities of the tank
liquors were assayed by the Microtox™
technique. The toxicities in all
tanks remained constant and high until
the opening of the tanks at day 21
(Figure 7). Toxicities did not de-
cline, whereas the cyanide concentra-
tions declined from 2,400 ppm to
approximately 900 ppm in each tank.
When the tanks were opened and cyanide
levels fell below 400 ppm, the rela-
tive toxicities decreased dramatical-
ly. The relationship between this de-
cline in cyanide concentration and
decrease in toxicity is approximately
logarithmic (Figure 8).
120
100-
80-
60-
40-
20-
0
•o TANK 1
-*- TANK 2
-o- TANKS
-»• TANK 4
•*• TANKS
•o- TANK S
•*• TANK?
RELATIVE TOXICITY
—T°
10
20
30
DAY
Figure 7
447
-------
CYANIDE vs. RELATIVE TOXICITY
CO
SO-
40
30
20
10
0
n . 09-1
I " AVE. 50-50
AVE. CM (ppm)
Figure 8
The nutrient injection design incor-
porates slotted piping laid on a
gravel bed in the excavation and a
series of injection (well screen)
points located around the perimeter
of the excavation. Both slotted
piping and injection points are con-
nected to a nutrient supply system
that will be used to supply nutrients
to the subsurface. Injection pres-
sures are expected to vary from 2 to
5 psi. Changes in pressure (injection
rate) are due to subsurface soil
heterogeneity and rates of saturation.
Injected fluids will be introduced
into the unsaturated zone. The fluids
will be allowed to percolate to the
water table, where they will be cap-
tured in the radius of influence of
pumping from the extraction/gradient-
control wells. Gradient-control wells
will be used to limit the potential
for migrating fluids from the remedia-
tion area. The gradient-control wells
are designed to have a radius of
influence of 60 ft and a drawdown of
20 ft. Two gradient-control wells,
located east and west of the excava-
tion, should provide sufficient con-
trol to preclude the migration of
water away from the remediation zone.
The nutrient injection piping within
the excavation is laid on a gravel
bed. The area above the piping is
then backfilled with gravel to the
bottom of the floor slab. The surface
is finished with concrete to the
existing floor level.
The injection points are screened
well points (1-ft sections) placed in
sand-backfilled predrilled (4-in.
auger) holes. The depth of the holes
is nominally 5 ft. The holes are
grouted with neat cement above the
screened interval well points. There
are approximately 20 well points
spaced at 10-ft centers arranged near
the perimeter of the remediation area.
All well points are connected by a
header system to the constant head
nutrient supply system.
Bacterial growth is expected to occur
predominantly around the perimeter,
and to gradually move toward the most
contaminated zone at the edge of the
excavation.
CONCLUSION
Cyanide is readily and rapidly re-
moved in the tested system. While the
disappearance of cyanide does not
parallel the formation of ammonia or
the development of biomass, there is
no doubt that cyanide is being drama-
tically removed. The disappearance
of cyanide also is not linearly paral-
leled by decreased toxicity. Cyanide
in all tanks decreased to below 900
ppm before toxicity consistently de-
creased. The decrease is logarithmic
in relation to cyanide concentrations.
This effect is probably explainable
as a threshold value or "window" of
toxicity. The evidence indicates that
bioremediation is a rapid and effec-
tive means of removing cyanide from
contaminated soils and subsoils in
situ, particularly when the proper
nutrient and oxygenation conditions
are maintained.
448
-------
The system is scheduled for startup
of nutrient injection during the
summer of 1989. Ongoing analytical
testing will be performed during the
remediation effort to substantiate the
in-situ biological removal of cyanide,
to supplement data collected during
laboratory testing, and to monitor the
progress of the in-situ bioremedia-
tion.
NUTRIENT FEED SYSTEM
NT TO ROOF
>• WASTE TREATMENT PLANT
f—fWTWEKT LEVEL MONITORING
L STANDPtPE
TRUCTURAL
BACKFILL
INJECTION PIPE
INJECTION
POINT
INJECTION
POINT
BRA.VEL BACKFILLED
t 6RAWEN?
V AAtr*rv*/*t
CONTROL
WELL
PECTED BOUNDARY
OF CONTAMINATION
Revised 3/3/88
Figure 9
Cross-Section of Nutrient Feed System
449
-------
Disclaimer
Ihe work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
450
-------
EVALUATION OF ALTERNATIVES FOR IMPOUNDED HIGH SALT WASTES
CONTAMINATED WITH ORGANIC AND INORGANIC POLLUTANTS
Robert D. Fox and Victor Kalcevic
International Technology Corporation
Knoxville, Tennessee 37923
A waste impoundment with a flexible membrane liner (FML) was suspected of
contributing to subsurface contamination and the Denver area client contracted
with IT for a technical and economic evaluation of alternatives for stopping this
potential source.
The sludge and brine contents of the pond included wastes from the chemical
milling of aluminum, other heavy metals, organics, fluoride, nitrate and high
concentrations of other dissolved salts.
IT first conducted a paper evaluation of alternatives to identify those that were
readily implementable and cost effective. Eight technologies were screened, using
engineering judgement and assumed performance to prepare preliminary estimates
of efficiency and cost. On and off—site disposal of residuals were considered.
Three technologies were selected for additional investigation - direct
solidification, solar evaporation, and a combination of solid/liquid separation and
steam evaporation.
Bench and pilot—scale treatability tests were then performed to experimentally
measure key performance and cost parameters. Solar evaporation was simulated
in an innovative lab system. The data were then used to refine the performance
and cost estimates for the client.
INTRODUCTION
The waste impoundment was used as
an evaporation pond for the myriad of
wastes resulting from chemical milling
of aluminum, surface preparation of
other metals, and various metal
working, plating and finishing
operations. The waste in the pond is
comprised of a predominant sludge
layer and a small overlying aqueous
layer; the total volume was estimated
to be 1 million gallons (5000 tons).
The types and concentration ranges of
the contaminants in this impoundment
are summarized in Table 1.
As expected, an impoundment used for
this purpose contains high levels of
heavy metals, total dissolved solids
(TDS), nitrates, fluorides, and
organics. A large amount of aluminum
is present in the sludge. Identified
volatile organic compounds were <2
ppm, and semi—volatiles (chiefly
phenols and polynuclear aromatic
hydrocarbons) were approximately 10
ppm. Based on TOC content there is
a large amount of unknown organics of
451
-------
low water solubility. This material is
classified as hazardous based on EPA
classifications F001 to F005, pH, and
chemical composition.
The client was located on a pristine
mountain stream that flowed to the
city's drinking water reservoir, so
treatment requirements for discharge
of any treated water were very
stringent.
PURPOSE
IT Corporation contracted to provide
the client with an identification and
evaluation of alternative technologies
for ending this potential source of
subsurface contamination at their site.
The results of this evaluation were a
recommendation of the cost effective
treatment and disposal technology for
closure of the impoundment.
The objective of this project was to
identify and evaluate those options
that are environmentally sound,
minimize long—term liability, minimize
the amount of material to be shipped
off—site and are cost effective.
APPROACH
A two—phase approach was used to
carry out this project. Phase I was a
paper study to screen alternatives on
assumed technical performance and
preliminary cost comparisons. Phase II
was to conduct bench and pilot—scale
testing of the three alternatives
identified in Phase I to generate
experimental data on performance and
to refine the cost estimates of Phase I.
Factors which were considered in
screening technologies were:
— technical performance
— simplicity of operation
— availability of equipment,
preferably mobile and for lease
— minimal technological risk
— ease of scale—up
— air emissions potential
— cost
— feedback from client
Two basic disposal options for each of
the closure alternatives were identified:
— on—site disposal of the treated
pond inventory in either the
existing pond area or a solid waste
landfill after delisting, or
— off—site disposal at either a
hazardous waste landfill or at a
solid waste landfill after delisting.
In other words, the pond inventory
could be handled as a hazardous waste
or treated to delisting criteria and be
handled as a solid waste.
Three basic approaches to dealing with
the evaporation pond contents were
identified, and various treatment
schemes within each were
conceptualized. The three approaches
were:
— Solidification of the entire waste
by addition of chemical additives
to convert sludge and aqueous to a
solid mass that would pass the
paint filter test or the TCLP test.
— Separation of water from
chemicals; because of the high
inorganic content, technologies
common to brine concentration
and processing were identified.
— Separation of chemicals from
water; this is the traditional
approach to processing aqueous
wastes, whereby a series of unit
operations are used to remove or
react the various organic and
inorganic contaminants, producing
a purified water for discharge.
Within these categories, several
technologies were screened based on
IT experience and either accepted for
more detailed consideration or rejected.
452
-------
The three technologies identified for
further evaluation in Phase I were
direct solidification, solar sludge drying
beds, and a treatment combination of
solid—liquid separation, evaporation,
and solidification of solids.
For the Phase II experimental testing,
a composite sample of the pond
contents was collected and shipped to
IT's Technology Development
Laboratory in Knoxville, Tennessee.
Bench and pilot scale test results were
used to prepare technical and economic
comparisons of the three technologies
identified in Phase I.
PROBLEMS ENCOUNTERED
The only significant problem was how
to simulate solar sludge drying beds in
the laboratory. In solar sludge drying,
the incident radiant energy of the sun
is used to supply the heat required to
evaporate water in the sludge.
Preliminary estimates in Phase I had
indicated that it might take six months
to evaporate the sludge sufficient to
pass the paint filter test.
IT designed a bench—scale
experimental system that used a sun
lamp positioned above a covered bed of
sludge. Incident energy from the lamp
at various positions on the sludge
surface was determined using a
pyranometer. The data collected were
used to calculate the size of a covered
drying bed and the length of drying
time under solar conditions in the
Denver area.
RESULTS
Fourteen technologies were screened
on a preliminary basis and either
accepted or rejected for further
evaluation.
Eight alternatives were selected for
more detailed assessment of technical
and cost factors in Phase I.
1. Direct solidification of the entire
pond contents with cementitious
additives.
2. Evaporation with sludge drying
beds under a translucent cover.
3. Evaporation by direct flame
injection of slurry in a sonic pulse jet
dryer.
4. Evaporation in a horizontal, heated
dryer with agitation, additional
heating, and forward movement of the
solids provided by a hollow screw
through which steam is circulated.
5. Sludge separation into solids,
water, and organics with the B.E.S.T.
solvent extraction process.
6. Solid/Liquid Separation by a
mobile filtration or centrifuge system.
7. Evaporation to treat the aqueous
phase from the solid/liquid separation
step to separate organics, dissolved
solids, nitrates, and fluorides from the
water.
8. Treatment of evaporated water
condensate.
The alternatives recommended for
further testing and evaluation in Phase
II were:
1. Direct solidification of evaporation
pond sludge and liquid to either a)
pass the paint filter test for land
disposal as a hazardous waste at an
approved off—site facility or in an on—
site vault or b) to fixate sludge and
hazardous constituents for possible
de—listing and landfill disposal.
453
-------
2. Drying of the sludge in an
evaporative drying bed, covered but
open to the atmosphere; disposal of
the dry solids as is or after
stabilization.
3. Treatment via solid/liquid
separation, evaporation of separated
liquid, condensation and treatment of
evaporator condensate, and
solidification of all solids produced.
These alternatives are shown
schematically in Figures 1, 2, and 3.
1. Direct Solidification Test Results
— Fly ashes and cement kiln dusts
local to the Denver area were used to
solidify the composite sample to two
degrees: to pass the PFT and
stabilized to pass the TCLP test.
Bench scale testing showed that 0.1
tons of fly ash or 0.2 tons of cement
kiln dust per ton of sludge were
sufficient to pass the PFT. For
stabilization the required amounts were
1.0 tons of fly ash or 0.75 tons of
cement kiln dust per ton of sludge, and
volume expansion was 67%, and
unconfined compressive strength
ranged from 3—5 tons per ft2- The
samples passed the TCLP test.
2. Solar Sludge Drying Beds —
Calculations showed that the rate
limiting step for solar sludge drying
was the transfer of the heat required to
evaporate the water. The volume of
air and its relative humidity were not
rate limiting factors. To prevent
rainwater from falling onto the sludge
during drying, an elevated plastic cover
made of materials designed to transmit
most of the sun's energy is placed over
the drying beds. The covered area is
open on the sides to allow free air
movement over the sludge.
The experimental laboratory testing
set—up consisted of a polyethylene
tray to hold the sludge placed in an
insulated enclosure covered with a
piece of Sunglo 17 CL reinforced,
colorless fabric covering furnished by
the Covertec Corporation of
Birmingham, AL. Its solar
transmittance is 87%. The tray was
filled to a depth of approximately 3
1/4" with 5.3 kg of sludge. The drying
chamber was purged with air at a rate
well below the 8.5 mph average wind
velocity in Denver. Humidity,
temperature, and hydrocarbons in the
air purge were monitored.
Solar energy was simulated by
irradiating the surface of the sludge
with a 375 watt infrared lamp with a
silvered internal reflector. Adjustment
of lamp power output was by a
variable power transformer. The lamp
was positioned above the center of the
tray at a distance of 0.5 meters. The
energy imparted to the sludge through
the cover at different lamp power
settings was measured using a direct
measuring pyranometer. Readings
were taken at the center and at the
four corners in the tray. Actual energy
transparency to the wavelength ranges
measured by the pyranometer was
found to be in the range of 60-65%.
These data were used to calculate the
energy input to the surface of the
sludge during drying.
Drying with heat (lamp on) was done
periodically to simulate alternating day
and night periods. The sludge surface
was irradiated for 8 hours per day for
10 days; the total drying time was 218
hours. The air purge was
uninterrupted for the total time.
During that time the water content of
the sludge was reduced from 80% to
approximately 20%.
454
-------
The sludge was mixed by hand using a
rake-type implement twice per day, at
the beginning and end of the light
period. The weight loss of the sludge
in the tray was measured daily.
The air purge humidity ranged from
20—40% and its temperature ranged
from 65 to 96 F. The hydrocarbons in
the purge air were measured using a
GC/FID instrument and ranged from 2
to 14.9 ppm (average 5 ppm) during
the first 100 hrs, then dropped to 1-2
ppm during the next 40 hrs. Also, no
odors were detected.
Using the sun lamp calibration data,
the hours of irradiation, and the daily
water evaporation, the energy which
was necessary to evaporate one pound
of water per day from the sludge was
calculated. The average for days 1
through 8 was 2967 BTU/lb water.
These sludge drying daily energy
results are shown in the graph shown
in Figure 4. The changes in the slope
of the drying curve readily shows the
loss of free water, steady state drying,
and a sharp rise in slope during final
drying.
For a sludge drying period of six
months in the Denver area, the total
solar radiation incident at the surface
were obtained from USGS data. Daily
solar energy ranges from 310 Langleys
(gram—calories/cm2) in October to
525 Langleys in June. Total Langleys
for the May—October period are
81,249. Converted to BTUs/ft2 it is
299,000. Correcting for the 87%
transmittance rating of the cover, only
260,000 BTU/ft2 are available.
To evaporate 1,000,000 gal of sludge @
80% water to 40% water, the energy
required is 1.98 x 1010 BTU. This
translates to a sludge bed drying area
of 1.75 acres. An area of 2 acres
makes an allowance for downtime and
other inefficiencies. This would
produce 1670 tons of dried solids from
5000 tons of sludge.
The dried solids from the sludge drying
tests, which were at 20% moisture,
were re-hydrated to 40% water. This
additional moisture was added to
provide water for hydration of the
cement kiln dust used for stabilization.
Stabilizing with 3.75% kiln dust and
4% lime yielded a product that passed
the TCLP test.
3. Sludge Dewatering/Evapor—
ation/Solidification - Bench and
pilot—scale tests were used to
characterize the performance of this
treatment sequence (see Figure 3) in
separating chemicals from water.
Batch sludge dewatering tests were run
in a pilot belt filter press (BFP) and a
bench—scale vertical solid bowl
centrifuge. The BFP gave the highest
solids content, 43%; anionic flocculant
at 650 ppm was found to produce the
best dewatering characteristics. These
BFP data translated to a production
rate of 17 gal/min on a 2 meter wide
press, a relatively low filter press rate.
Filtrate from the pilot BFP tests and
centrate from the bench centrifuge
tests was combined and batch
evaporated in a glass bench—scale test
unit. Heat was supplied by an electric
heating mantle. Vaporized water was
condensed and analyzed for TOC, then
subjected to bench—scale tests of
carbon adsorption, gas stripping, and
chemical oxidation to further reduce
TOC in the condensate.
Evaporation of 94% of the volume of
the aqueous feed produced a
condensate that ranged from 15—20
ppm TOC. The 6% bottoms product
was a salt slurry that contained 64%
455
-------
solids. No scaling of the glass heat
transfer surfaces was noted.
Treatment of the condensate with
activated carbon in IT's micro—column
adsorption test apparatus produced no
removal of TOC. Purging of the
condensate with nitrogen reduced TOC
by » 10 ppm. Further treatment of the
stripped condensate with hydrogen
peroxide and bleach produced no
reduction of TOC.
The wet cake from the BFP tests and
the salt slurry from the evaporator
were combined in a 10/1 ratio and
solidified with 75% cement kiln dust.
This level of additive is equivalent to
0.48 tons per ton of original pond
sludge. This stabilized sample passed
the TCLP test.
SUMMARY
For the three alternatives evaluated in
Phase II, the summary of the
quantities of treated pond waste for
disposal are presented in Table 2.
For off—site disposal of residuals from
these three alternatives, transportation
and placement in a hazardous waste
landfill in Utah was the lowest cost.
The combined transportation and
disposal cost was $260/ton.
A summary of the estimated treatment
cost for each of the alternatives is
presented in Table 3. It shows the
costs for on-site disposal and off-site
disposal, stabilized to either pass the
PFT or the TCLP.
The recommended approach to closure
of this pond was direct solidification of
the entire pond contents with sufficient
fly ash or cement kiln dust to pass the
TCLP and have sufficient strength to
support installation of a temporary
cap.
Table 1. Evaporation Pond Sludge DatarGeneral Physical/Chemical Parameters
Parameter
H
Water
Total Organic Carbon
Chemical Oxygen Demand
% metals (35 metals analyzed)
Anions
Nitrogen as Nitrate
Fluoride
Cyanide
Phosphorus as Phosphate
Concentration
Range
Mean
10.0 - 10.4 10.2
73 - 85% 79%
6,600-14,000 mg/kg 11,000 mg/kg
2,400-24,000 mg/kg 10,000 mg/kg
29 - 33% 31% (dry weight)
11,000-22,000 mg/kg 14,000 mg/kg
11,000-36,000 mg/kg 21,000 mg/kg
4.8 - 54 mg/kg 30 mg/kg
2,500-4,600 mg/kg 3,900 mg/kg
456
-------
Table 2. Summary of Quantities for Disposal, tons
Alternative
1. Direct Solidification
2. Solar Drying Beds
3. Solid/Liquid Separation, Etc.
Solidified
to Pass PFT
5,500-6,000
1,670
3,190
Stabilized
to Pass TCLP
10,000-8,750
1,800
5,575
Table 3. Summary of Estimated Costs (all dollars in thousands)
Alternate 1
Pass PFT
Pass TCLP
Alternate 2
Pass PFT
Alternate 3
Pass PFT
sFwar1 :
On-Site
Disposal
Treatment
T 118
LP 265
T 1,262
LP 1,300
!
T 742
LP 929
^
«
Treatment
118
303
1,253
1,291
675
862
I
*
Off-Site Disoosal
Disposal
1,430
2,275
434
468
829
1,450
Total
Treatment
Cost
1,548
2,578
1,687
1,759
1,504
2,312
EVITORATKM POfO
-H-H-H---O
PUO MILL
MIXIN3 PUNT
FIGURE I
SCHEMATIC DIAGRAM OF ALTERNATE I
DIRECT SOLIDIFICATION
CUtINO PILE
457
-------
V
EVAPORATION PCUD
*> f—&
s I -^
J
COVERED
DRYING
ueCHANICAl.
LOADER
FIGURE 2
SCHEMATIC DIAGRAM OF ALTERNATE 2
SLUDGE DRYING BEDS
SCHEMATIC FLOW DIAGRAM
ALTERNATE NO. 3
FIGURE 4.
SLUDGE DRYING SUMMARY
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency- The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
459
-------
KPE6 Application From The Laboratory To Guam
Alfred Kornel, Charles J. Rogers, Harold L. Sparks
U. S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Drive
Cincinnati, OH 45268
ABSTRACT
The novel reagent generically named APEG (alkaline polyethylene
glycolate) can be very effective for dehalogenation of a variety of halo
aromatic pollutants. The PCBs, PCDDs, PCDFs even PCTs (polychlorinated
terphenyls) can be effectively dehalogenated by APEG reagents to yield non-
toxic products. The reagent has been shown to be effective on these classes
of pollutants in a variety of matrices from sediment to soils and waste oils.
Early work in 1986 with KPEG performed on PCDD and PCDF contaminated oil
(9000 gal.) in Butte, Montana demonstrated the effectiveness of the KPEG
system on these types of hazardous wastes. Early tests of the in situ
application of KPEG on dioxin contaminated soil in Missouri demonstrated some
limitations of the reagents applicability. The most severe drawback for i_n
situ KPEG application is the reagents extreme hygroscopisity. These early
experiences led to the design of the pilot-scale chemical reactor system for
the effective use of the APEG reagent. This reactor system was first
demonstrated on a heavily PCB-contaminated site in Guam, USA during April-May
1988. This pilot-scale unit was capable of treating from one to two tons of
contaminated soil per batch. Further refinements were made to the reactor
system after careful examination and analysis of the first eight reactor
batches. During September-October of 1988 these refinements were employed in
a second series of KPEG treatments in Guam. Contaminated soil containing from
500 to 2,600 ppm of PCB closely resembling Aroclor 1260 was treated with the
KPEG reagent. The only residual PCB peak detected in the soil treatments was
a tetrachlorobiphenyl, ranging from approximately 1 ppm to nondetectable
levels. The KPEG reagent was also applied to treating all of the contaminated
tyvek clothing, gloves and boots from the combined runs during the 1988 year.
Sampling of the reactor contents revealed the presence of tetra-, penta-, and
hexa- chlorobiphenyls but at below the one part per million range. These
results demonstrate the capability of the KPEG reagent to perform chemical
dehalogenation on haloaromatics in a variety of matricies, ranging from oils
to soils and contaminated clothing.
460
-------
INTRODUCTION
Compounds which have gained
notoriety for their persistence as
well as their toxicity are the
halogenated aromatics. Most notable
among this class or organics are the
chlorinated dibenzodioxins (PCDDs)
dibenzofurans (PCDFs) and the
polychlorinated biphenyls (PCBs).
In the last few years PCDDs and
PCDF's have increasingly been
identified in chemical product waste
streams as well as effluents from
incineration processes. The
accumulation of PCDDs, PCDFs, PCBs
and other toxic halogenated
compounds in the environment and
living systems is a serious problem
that has been well documented.
Although a great amount of work has
been done by many groups in the area
of direct chemical decomposition of
halogenated organics, relatively
little effort has been directed
toward on-site chemical
detoxification.
j
There are currently some chemical
methods available to alter or
decompose PCBs and other
haloaromatics in contaminated oils.
The methods developed by Acurex,
Goodyear and Sun Ohio, involving
dispersion of metalic sodium in oil
or the use of sodium-biphenyl or
naphthalene mixtures are
commercially available. However,
due to the reactivity of these
sodium formulated reagents with
water they present a serious draw
back with their use on haloaromatic
contaminated soils, sediments,
sludges or dredgings. Other
chemical reactions have been
evaluated for dehalogenation of
environmental pollutants but have
not been found to be adaptable to
field conditions. (1, 2, 3)
Up to this time one of the most
reactive chemical dehalogenation
systems for use in the
aforementioned area of
decontaminating soils has been the
alkaline polyethylene glycolate
moiety (APEG). The alkali most
commonly used for this reagents
preparation is potassium hydroxide
(K), in conjunction with a
polyethylene glycol ranging from
molecular weight of 300 to 600
Dal tons. This reagent, labeled
KPEG, is tolerant of wet matrices in
which it is to be used for
dehalogenation of said haloaromatics
eg. the PCBs PCDDs and PCDFs.
Though the chemistry of APEGs can be
traced back to the early 1970s it
was not until the summer of 1978
that this reagent was demonstrated
efficacious at dehalogenation of PCB
in contaminated oils. (4) Since that
time a series of APEG reagents have
been prepared, which with heating
produce rapid dehalogenation of
haloaromatic compounds (5, 6, 7, 8).
The basic reaction scheme for the
dehalogenation of an haloaromatic is
shown:
(1) HOPEG + KOH - KOPEG + H20 (2)
Aryl-Cl+KOPEG - Aryl-0-PEG+KCl (3)
Aryl-0-PEG - Aryl-OH+Vinyl-PEG
In equation (1) the appropriate
polyethylene glycol is reacted with
potassium hydroxide to form the
reactive APEG species. This
preparation step may be performed
directly in the contaminated matrix
which is to be treated. Reaction
(2) takes place over a wide
temperature range from ambient to ca
110'C. The third reaction (3) shows
conversion of the ether linked
PEG/Aryl moiety to a phenolic with
consequent release of a vinyl
terminal polyethylene glycol. This
reaction generally takes place at
temperatures above IIO'C. In some
of these reagent formulations
dimethyl sulfoxide is added as a
cosolvent to enhance reaction rate
kinetics presumably by improving
rates of extraction of the
461
-------
haloaromatic contaminant into the
alkoxide phase. (7)
In 1982, detailed investigations
were initiated to determine the
effects of variable reaction
parameters on the rate and extent of
chemical decontamination of soils
(9). This research focused almost
exclusively on the direct chemical
treatment of PCB contaminated soil.
The first field investigation aimed
at identifying treatment conditions
for chemical destruction of PCDDs
and PCDFs in oil recovered from the
ground of a wood preserving site in
Butte Montana was initiated in
January of 1986. (10) In November
of 1987 plans were made to perform a
Pilot Scale Evaluation of the KPEG
processes' effectiveness in treating
PCB (Aroclor 1260) contaminated soil
at the U.S. Navy's Public Works
Center in Guam. Two separate Pilot
Scale tests were performed, the
first during March/April 1988 and
the second during September/October
1988.
PURPOSE
During January 1986 research and
field investigations were initiated
to determine if a chemical reagent
ie. the KPEG reagent, could be used
to effectively treat PCDD and PCDF
contaminated oil at an industrial
wood preserving site near Butte
Montana. The site contained
approximately 9,000 gallons of a
light petroleum oil collected
previously from groundwater over a
two year period. The oil contained,
3.5% pentachlorophenol and PCDD and
PCDF homologs ranging from 422 part
per billion (ppb) of tetra-isomers
to 83,923 ppb octa-isomers. The
process successfully decontaminated
the petroleum oil during July 1986.
The PCB contamination at the U.S.
Navy Public Works Center in Guam
ranges from a few hundred ppm to
well over 40,000 ppm of Aroclor
1260. This contamination resulting
from the previous practice of
emptying and draining transformer
oil on to the very porous coral soil
surrounding the transformer
rebuilding shop. The contaminated
area is of several acres in size.
Due to the large expense of
containerizing and shipping this
soil from Guam to an approved
incineration facility, the U.S. Navy
decided to examine the potential for
on-site decontamination of this soil
utilizing the KPEG reagent.
APPROACH
Decontamination of PCDD and PCDF
containing oil
In April 1986, U.S. EPA Region 8
agreed, after a review of laboratory
data, that the KPEG process could be
used to decontaminate the PCDD/PCDF
tainted oil on-site. The site being
a former wood treating facility
located near Butte Montana.
The mobile field equipment employed
to implement the previously
mentioned chemical decontamination
process comprised of a 2,700 gallon
batch reactor mounted on a 45 foot
trailer equipped with a boiler,
cooing system and a
laboratory/control room area.
Heating of the raw oily waste/KPEG
reagent mixture was achieved by the
recirculation of the oil and reagent
through a pump, a high shear mixer
and a tube heat exchanger which was
heated by a boiler or cooled through
a series of fin-type air coolers.
The process was successfully
employed in July of 1986 to
decontaminate some 9,000 gallons of
this PCDD/PCDF tainted oil. The
results are shown in Table 1.
462
-------
Table 1. TREATMENT OF CONTAMINATED OIL, BUTTE, MONTANA
Contaminants
CDD/CDF
Concentration in
Untreated Oil (ppb)
Concentration in
Treated residue (ppb)
70°C, 15 min. 100°C, 30 min.
Minimum detectable concentration in parts per billion (ppb)
*MDC
TCDD (2,3,7,8-)
TCDD (total)
PeCDD
HxCDD
TCDF (2,3,7,8-)
TCDF (Total)
PeCDF
HxCDF
HpCDF
OCDF
28.2
422
822
2982
23.1
147
504
3918
5404
6230
—
-
-
-
12.1
33.3
-
4.91
5.84
-
0.65
0.37
0.71
2.13
0.28
0.35
0.36
0.76
1.06
2.62
Decontamination of Arolor 1260
Containing Soil
In November 1987 a 400 gallon
Littleford reactor was purchased for
use in demonstrating a Pilot scale
application of the KPEG process on
the Aroclor 1260 contaminated soil
located in Guam. This reactor and
all ancillary equipment required for
its use were shipped via the Navy
from Port Hueneme California to Guam
in February 1988. During March 1988
the reactor was outfitted and
provisions were made to employ the
analytical instruments located at
the Naval FENA laboratory on Guam
for the PCB treatment analyses. The
goal of this KPEG treatment of the
PCB contaminated soil was to achieve
a maximum residual concentration of
no more than 2 parts per million
(ppm) of any chlorobiphenyl isomer
to remain in the soil.
Processing of PCB Contaminated Soil
in Guam I
Typically from one to one and one
half tons of contaminated soil were
loaded into the reactor. This was
followed by addition of 50% by
weight of Polyethylene glycol
average molecular weight 400 Dal tons
(PEG-400) and the appropriate
quantity of potassium hydroxide (89%
flake) to yield an equi molar ratio
to the PEG-400. After the chemicals
had been added to the reactor, the
unit was sealed and heated to 140-
150°C for 3-5 hours. Any distillate
was captured and condensed, final
venting was through a carbon trap.
The results of this first series of
KPEG treatments is in Table 2.
As these results show, there are
greater than 2 ppm residual PCB
congeners in batches 1, 6, 7 and 8.
These therefore did not comply with
the permit requirement of less than
2ppm per resolvable congener
remaining after treatment as imposed
by the U.S. EPA.
This resulted in a series laboratory
scale KPEG reactions being performed
in the U.S. EPA RREL Research
Laboratory in Cincinnati. Here it
was discovered that the increase of
base, eg KOH over the equi molar
ratio to PEG-400 of from 1:1 to 1.3
- 1.5 : 1 resulted in the
463
-------
Table 2: Comparison of Guam soil prior to and after KPEG treatment.
Batch # Aroclor in PPM PPM of residual PCB
Prior to Uncorrected
Treatment
Corrected *
1
2
3
4
5
6
7
8
1,990
1,740
1,490
1,540
1,010
1,140
1,140
1,860
Tetra-Cl BP
penta-Cl BP
hexa-Cl BP
tetra-Cl BP
tetra-Cl BP
tetra-Cl BP
tetra-Cl BP
tetra-Cl BP
tetra-Cl BP
tetra-Cl BP
= 2.11
- 5.42
= 7.33
= 1.87
= 1.36
= 1.82
= 1.80
= 4.12
= 3.27
= 4.64
= 2.24
= 6.46
= 6.50
= 1.99
= 1.45
= 1.94
= 1.91
= 4.38
= 3.38
= 4.94
* Corrected values calculated by extraction efficiency of mono through deca
chlorobiphenyls from this soil.
desired lowering of all the residual
PCB congeners previously mentioned.
This resulted in a return trip to
Guam during September - October 1988
to demonstrate this improvement on
the KPEG process.
Guam II
The treatment during this second
phase of the pilot scale
investigation consisted of loading
the reactor with the same 50% weight
of PEG-400 to soil but with an
increase of 30% of KOH to PEG-400
over stoichiometric. As is
demonstrated by the results in Table
3, Retreatment of Original Guam
Soil, the PCB congeners which remain
are reduced to below the 2 ppm
limit. The efficacy of the improved
KPEG process is further demonstrated
by the results as shown in Table 4,
Continuation of Guam Soil PCB
Treatment, where all residual PCB
congeners are either not detected or
below the 2ppm limit.
As condensate was emitted in this
treatment process, it had to meet
emission standards as for the
treated soil. The results of
condensate analysis are shown in
Table 5, Reactor Emission,
Condensate Samples.
464
-------
Table 3: Results of Retreatment of Original Guam Soil
Batch #
1
6
7
8
PPM OF RESIDUAL
1ST Run
TETRA-CL BP =
PENTA-CL BP =
HEXA-CL BP =
TETRA-CL BP =
TETRA-CL BP =
TETRA-CL BP =
2.24
6.46
6.50
4.38
3.48
4.94
PCB (corrected)
RETREATMENT
= 0.48
= ND
- ND
= 0.85
= 0.15
= 0.66
ND = NONE DETECTED
Table 4; Continuation of Guam Soil PCB Treatment
BATCH # AROCLOR IN PPM
PRIOR TO
TREATMENT
PPM OF RESIDUAL PCB
CORRECTED
9
10
11
12
13
ND
*1
*2
919
19 PPM *1
298
529
*2
NONE DETECTED
TETRA-CL BP = 1.01
TETRA-CL BP = 0.22
ND
TETRA-CL BP = 0.59
TETRA, PENTA, AND HEXA-CL BP
ALL LESS THAN 1 PPM
MATERIAL RESIDING IN REACTOR AFTER VENT FAILURE
CONSISTED OF PRIMARILY A PENTA AND HEXA CHLOROBIPHENYL.
THIS LOAD CONSISTED OF SHREDDED CONTAMINATED TYVEK CLOTHING,
GLOVES AND BOOTS. THE ORIGINAL PCB CONCENTRATION WAS NOT
DETERMINED.
465
-------
Table 5: REACTOR EMISSIONS,
CONDENSATE SAMPLES
BATCH #
*i
6*1
7*1
8*1
9
10
11
12
13
PPM RESIDUAL PCB
NO
< 1 PPM
< 1 PPM
< 1 PPM
ND
< 1 PPM
< 1 PPM
< 1 PPM
*2
ND - NONE DETECTED
*1 - THESE ARE THE BATCHES WHICH
WERE RETREATED.
*2 - NO CONDENSATE SUBMITTED FOR
ANALYSIS.
PROBLEMS ENCOUNTERED
The treatment of PCDD/PCDF
contaminated oil with the KPEG
reagent was very efficacious. There
were no problems encountered in
scale-up of the laboratory procedure
used in testing the KPEG reaction on
samples of the contaminated oil.
The application of the KPEG process
to PCB contaminated soil in Guam
revealed a potential problem. This
problem, namely residual PCB
congeners being above 2 ppm in
concentration in some of the treated
batches of Guam contaminated soil
was overcome. The retreatment of
these aforementioned batches and
successful treatment utilizing the
improved KPEG treatment process was
demonstrated.
This is not to say there were no
other problems encountered in either
of these treatment demonstrations.
There were logistical problems, ie
difficulties in having all equipment
at all the required places at the
time of use. However, as is
demonstrated, these problems were
also overcome.
RESULTS
As can be seen in Tables 1 thru 5
the treatment of both oil and soil
contaminated with haloaromatics is
feasible from a laboratory to pilot
scale utilizing KPEG reagents. The
reagent can successfully reduce
PCDD/PCDF levels from thousands of
parts per billion to non detectable
levels in a contaminated oil matrix.
Further the KPEG reagent can be used
to reduce PCB levels resulting from
Aroclor 1260 contamination in soil
to less than 2 PPM, within a
reasonable time frame.
This demonstrates the use of the
KPEG Systems applicability to a
variety of haloaromatic pollutants
in a variety of matrices.
ACKNOWLEDGEMENTS
We would like to thank Dr. D. B.
Chan of the U.S. Navy Civil
Engineering Laboratory Port Huneme
California for his interest in the
application of this chemistry and
his help in coordination of the
pilot scale demonstrations in Guam.
466
-------
References
1. Miller, J. Nucleophilic
Aromatic Substitution.
Elseiver Press, Amsterdam,
1968
2. Yoshikozu, K. and Regen, S.L.
J. Organic
Chemistry 47, 1982, (12) 2493-
2494
3. Andrews, A., Cremonessi, P.,
del Buttero, P.,
Licondra, E. and Malorano, S.
Nucleophylic Aromatic
Substitution of Cr (CO) , -
Complex Dihaloarenes witn
Thiolates J. Organic
Chemistry 48, 1983, 3114 -
3116
4. Pytlewski, L. L. A Study of
the Novel Reaction of Molten
sodium and Soluent with PCBs.
U.S. EPA Grant #R806659010,
1979
5. Kernel, A. and Rogers, C. J.
"PCB Destruction
A Novel Dehalogenation
Reagent" Journal of
Hazardous Materials 12, 1985,
161-176
6. Freeman, H. M. Standard Hand
book, of Hazardous Waste
Treatment and Disposal 1988
Section 7.5 Dehalogenation
7. Peterson, R. L. Method for
Decontaminating Soil
U. S. Patent #4,574,013 1986
8. Bruneile, D. J. and Singleton
D. A. Chemosphere, 12 (2),
1983 183-196
9. Rogers, C. J. Chemical
Treatment of PCB in the
Environment. EPA-6001 9-83-
003, 197-201
10. Peterson, R. Potassium
Polyethylene glycol Treatment
of PCDD/PCDF - Contaminated
Oil in Butte, Montana. It
Corp/Gal son Research Corp.,
Project #86-706, July 1986
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
467
-------
TESTING NATURAL ZEOLITES FOR USE IN
REMEDIATING A SUPERFUND SITE
Robert L. Hoye
PEI Associates, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
and
Jonathan G. Herrmann
and
Walter E. Grube,Jr.
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
ABSTRACT
The U.S. Environmental Protection Agency's (EPA's) Risk
Reduction Engineering Laboratory (RREL) has recently completed a
project to evaluate the effects of adding natural zeolites to a
soil contaminated with metals. The objectives of this project were
to: (1) determine the engineering properties (including hydraulic
conductivity) of several amendment-soil combinations; and (2)
identify those metal ions in the soils that could be sorbed, and
therefore immobilized by the zeolites.
A screening program was performed to determine which of five
commercial zeolite products were most effective in controlling
metal ion mobility. The effectiveness of the five zeolites was
evaluated using a simple batch leaching procedure: The Monofilled
Waste Extraction Procedure (MWEP). The MWEP extracts were analyzed
for several metal ions, including lead, chromium, and zinc, using
ICP. This was done to determine which zeolite product retained the
most metal ions following the MWEP procedure. The two zeolites
which removed the most metals of interest from the MWEP solution
were then used in a more detailed laboratory investigation.
468
-------
The laboratory investigation involved several tests in
addition to the HWEP. These were: (1) permeameter seepage tests
to measure the effect of the zeolites on permeability, and (2)
standard physical and engineering soils tests (e.g., moisture
content, particle size distribution, Atterberg limits, specific
gravity). The total test program also included the Bunker Hill
Superfund site soil amended with agricultural limestone for
comparison purposes. The results of the MWEP and permeameter
seepage tests are discussed in this paper.
INTRODUCTION
The Region X Office of the
U.S. Environmental Protection
Agency (EPA) is responsible for
the remediation of numerous
Superfund sites in the
northwestern United States
where large quantities of soils
have been contaminated with
heavy metals. These metals can
and have caused environmental
and human health impacts.
Environmental pathways for
migration of metal contaminants
include mobilization of metals
present in soils by solubi-
lization in groundwater and/or
surface water, dispersion of
metals with fugitive dusts, and
uptake of metals present in
soil by plants. Each of these
pathways presents potential
exposures to humans and
wildlife. EPA Region X
officials, who were aware of
the ability of natural zeolites
to preferentially sorb metals,
requested that EPA's Risk
Reduction Engineering
Laboratory provide technical
assistance in evaluating
zeolites as soil amendments.
PURPOSE
The purpose of the work
described herein was to: 1)
select and evaluate the
relative effectiveness of
several commercial grades of
the naturally occurring
zeolite, clinoptilolite, in
reducing metal concentrations
in water extracts of amended
Superfund soils; and 2)
evaluate the effect of zeolite
amendments on the leachability,
and physical and engineering
properties of amended Superfund
soils.
LITERATURE SURVEY
More than 150 synthetic
zeolites and 40 natural
zeolites are known (2) .
Synthetic zeolites have enjoyed
a wide range of applications,
but to date they have been too
expensive to be used in large-
scale applications for
remediation of hazardous waste
sites. Until recently it was
believed that natural zeolites
were not abundant enough to be
an economically feasible
replacement; however, large
deposits of zeolites discovered
in the western United States
make these minerals a viable
alternative to synthetic
zeolites. The most common
types of zeolites (clinoptilol-
ite, mordenite, chabazite, and
erionite) originated by the
natural alteration of volcanic
ash in alkaline environments.
Some of the other
applications in which natural
zeolites are used for their ion
469
-------
exchange ability are in
detergents, as water softeners,
ill ammonia/ammonium removal
from fresh-water effluents, in
radio isotope removal from spent
nuclear reactor pile effluent,
and in agriculture as a carrier
of ammonium and potassium.
In addition to ion
exchange, the uniform pore
sizes of zeolites (ranging from
0.3 to 0.8 nm) can selectively
adsorb or reject molecules or
ions based on their size or
shape; this is known as the
molecular sieve effect.
Another type of adsorption
separates components based upon
differences in their selectiv-
ities. Catalysis can also occur
within the intracrystalline
void.
Several researchers have
found that clinoptilolite can
be used to selectively remove
heavy metal ions from water.
Semmens and Seyfarth (8)
conducted studies that showed
that clinoptilolite samples,
one treated with NH4C1 and one
washed with acid, exhibited
very high selectivities for
barium and lead and somewhat
lower selectivities for copper,
cadmium, and zinc in displacing
sodium. The lead and zinc
exchanges were not highly
reversible. Semmens and Martin
(7) looked at the removal of
lead, silver, and cadmium by
clinoptilolite in the presence
of competing concentrations of
calcium, magnesium, and sodium.
The selectivity sequence was
found to be Pb2+>Ag+>Cd2+, with
excellent removal of lead.
Lead and cadmium removal
decreased with the presence of
competing cations in the order
Mg2+>Na+>Ca . Calcium concentra-
tions significantly inhibited
cadmium removal.
Studies have also been
conducted on ammonium removal
by natural zeolites (3).
Blanchard, et.al (1) inves-
tigated the removal of ammonium
and heavy metal ions from
drinking water and found that
the selectivity of the sodium-
exchanged clinoptilolite
decreases in the order
Pb^NH^Cu2*, Cd2*>Zn2*,
Co2">Ni2+>Hg2*. Lo i z idou and
Townsend (5) found that lead
was exchanged to a greater
extent than cadmium, in the
absence of competing cations,
on sodium clinoptilolite. They
calculated a "maximal level of
exchange" (a ratio of mol/kg
exchanged metal to mol/kg of
total aluminum) of 0.795 for
lead and 0.656 for cadmium.
Work has also been
conducted on using zeolites as
soil amendments, which are
defined as substances that aid
plant growth indirectly by
improving the condition of the
soil. Soil amendments should
not be confused with plant
nutrients, such as nitrogen,
that are used directly by the
plants. Natural zeolites have
been used in Japan as soil
amendments for years because of
their ion exchange and water
retention capabilities. Lai
and Eberl (4) added NH4*-
saturated clinoptilolite and
phosphate rock to soil.
Phosphate rock provides
nutrients, but its use as a
fertilizer is limited because
of its low solubility. The
zeolite serves as a sink for
Ca ions that exchange with the
NH4* ions, thereby reducing the
concentration of Ca2+ in
solution and allowing more
phosphate rock to dissolve.
The ion exchange ability of
zeolites may also be able to
470
-------
trap heavy metals in soil and
prevent their uptake by plants.
Nishita and Haug (6) conducted
studies indicating that
addition of clinoptilolite to
soil contaminated with Sr
decreased strontium uptake by
plants.
APPROACH
The experimental approach
involved determination of the
relative ability of several
zeolite treatments to reduce
the concentrations of metals in
water extracts of contaminated
soils. In addition, the
leachabilities of the amended
soils were determined. The
specific natural zeolites used,
clinoptilolites, were obtained
from three mines in eastern
Oregon. Soils from two
Superfund sites were initially
included in a screening test
program; however, only one soil
was used in the full test
program. This soil was
obtained from the Bunker Hill
Superfund site in Kellogg,
Idaho. Lead, up to two percent
by weight, was the major
contaminant in this soil. The
second soil, used in the
screening test program only,
was obtained form the United
Chrome Products, Inc., (UCPI)
Superfund site in Corvallis,
Oregon. Chromium was the major
contaminant in the UCPI soil.
Both the zeolites and the test
soils used in this evaluation
were collected by EPA Region X
personnel prior to commencement
of the study.
A screening test of the
relative efficacy of five
natural zeolites as soil
amendments for the two
Superfund site soils was
conducted. These five grades
are mined near Adrian, Oregon,
and were designated as CH-5,
CH-20, XY-5, XY-20, and SC-35
by the mining company. The
letter symbols differentiate
the zeolites based on the type
and location of the natural
deposit and the quality
characteristics of the mineral;
and the numbers indicated the
screen mesh size through which
the commerical product passes.
The relative efficacy was
determined by comparison of
dissolved metal concentrations
in water extracts of amended
soils and nonamended soils
(i.e., controls). This was
accomplished by mixing the
zeolites, in six mix ratios,
with the test soils and
subjecting the amended soils to
the Monof illed Waste Extraction
Procedure (9). This procedure
(MWEP) consists of four
sequential extractions of
sample using ASTH Type 2
deionized water in a
liquid/solid ratio of 10:1.
Based on these screening tests,
two zeolites and one Superfund
soil were prepared, as were
nomamended control samples and
samples amended with
agricultural limestone. The
limestone amendment was evalua-
ted to allow comparison of test
results with a standard
agricultural practice.
Additionally, the physical and
engineering properties of the
soils were determined.
RESULTS
MWEP Screening Test Program
The screening test program
resulted in selection of CH-20
mesh and XY-5 mesh (from five
clinoptilolites obtained and
tested from three mines) and
three application rates (4:100,
12:100, 20:100 - dry weight
471
-------
compared to the extracts
of the control (Figure
la).
5) Cumulatively, more lead
was extracted from the
Bunker Hill soil amended
with the CH zeolite than
from the control. These
differences were
significant at the 95
percent confidence level.
Extracts of two XY
treated soils, 4:100 and
20:100 mix ratios, and
the extracts of
limestone-amended soils
had significantly lower
cumulative lead values
than extracts of the
controls. Extracts of
the XY treated soil mixed
in a 12:100 mix ratio
were not significantly
different than the
control (Figure la).
6) Cumulative lead values in
extracts of soils amended
with limestone were
significantly (95%
confidence level) lower
than those in extracts of
soils amended with the CH
zeolite and the XY 12:100
mixes. The cumulative
lead values in the XY
4:100 and 20:100 mixes
were not statistically
different from those in
limestone extracts
(Figure la).
7) Both CH and XY zeolites
effected a reduction of
50 percent or more in the
concentration of lead in
the first MWEP extracts
as compared with the
controls. This
observation is consistent
with the MWEP results
obtained during the
screening test program
(Figure la).
8) On a cumulative basis,
zinc and cadmium values
were statistically lower
in all four sequential
extracts of CH, XY, and
limestone-amended soils
as compared with extracts
of control samples
(Figures 2a and 3a). The
extracts of CH and XY
amended soils (except the
XY 4:100 mix) contained
significantly less
cadmium than did the
extracts of the limestone
amended soils; the XY
4:100 mix was not
statistically different
from the limestone.
Permeameter Seepage Program
The seepage from the
permeameters (Fig. 4) used to
determine hydraulic conduc-
tivity was analyzed to
determine concentrations of
target metals released with
time and the cumulative amount
released during the
determinations. The
conclusions drawn from these
tests are summarized as
follows:
1) Concentrations of lead
and zinc in seepage from
the test permeameters
(Figures Ib and 2b)
corresponded with the
results of the water
leach tests (Figures la
and 2a). Cadmium in the
permeameter effluents
(Figure 3b) was not
consistent with its con-
centrations in MWEP
extracts (Figure 3a).
472
-------
't * 111 *
I
(U
•w
rH
tj g
u co
U
(U 0)
60 3
Cfl rH
M 3
(U M-l
>
o
H
Sui 'papeiixa q j jeioi
2
u
cd
X!
(U
P-I
Pk
(U
•U
CO
CO
i—i
w
§
3ui 'popejixa qj JBIOJ.
473
-------
I
Sui 'pope.i|xa uz jciox
M
rH
•H CO
•l-i 8
O 10
4-1
0
-------
4-1
cu
Ol
a.
u a)
N
"H
•H rt
•H O
|,5
O CO
4-1
CD (3
oo a)
cd 3
)-> i-H
Q) 14^
t> " '
•5 a»
.0
en
Siu 'poiaejjx
P
o
O
cd
9
PM
C
•H
•H
4-1
cd
cd
en
g
Sui 'popeajxs p;) JBIOX
475
-------
2) Cumulative amounts of the
target metals in seepage
from the permeameters were
still increasing after
passage of five pore
volumes, which indicated
that all of the available
target metals in the
contaminated soil had not
been leached into solution
(Figures Ib, 2b, and 3b).
CLOSURE
The results of the MWEP
water extractions and other
tests used during this research
project are interpreted as
being indicative of what would
happen in nature (i.e., how
much of a target metal would
leach from amended soils on a
relative basis). Because the
tests were conducted in the
laboratory using deionized
water, however, these tests may
not accurately predict what
would be observed in a natural
system. Therefore, caution
must be used in the application
of these data to the Bunker
Hill Superfund site.
PRESSURE INPUT
INLET
DISTILLED / DEIONIZED
WATER
LEACHATE OUTLET
(INNER)
SOIL
I GEOTEXTTLE FABRIC
DOUBLE-RING
BASE PLATE
-LEACHATE OUTLET
(OUTER)
Figure 4. Schematic of the double-ring permeameter in operation.
476
-------
ACKNOWLEDGEMENTS
The project described in
this paper was part of a
complex undertaking that
required the cooperation and
coordination of a
multidisciplinary team of
research scientists and
engineers. Mr. John Barich of
EPA Region X originated the
request for technical
assistance to RREL for the
evaluation of natural zeolites
as soil amendments and led the
EPA personnel that collected
the soil samples from both the
Bunker Hill Superfund site and
the United Chrome Products,
Inc. Superfund site. Mr.
Daniel Krawczyk of EPA's
Corvallis Environmental Re-
search Laboratory (CERL) in
Corvallis, Oregon, provided
rapid turnaround analytical
support for this program by
conducting the analyses of
metals in numerous samples.
Their contributions to the
successful completion of this
project are recognized and
appreciated.
REFERENCES
1. Blanchard, G., M. Maunaye,
and G. Martin. 1984.
"Removal of Heavy Metals
From Waters by Means of
Natural Zeolites." Water
Research. Vol. 18, No. 12,
pp. 1501-1507.
Kirk-Othmer.
Encvlopedia of
1981.
Chemical
Technology. Volume 15.
Molecular Sieves. Third
Edition. John Wiley and
Sons.
477
3. Klieve, J.R., and M.J.
Semmens. 1980. "An
Evaluation of Pretreated
Natural Zeolites for
Ammonium Removal." Water
Research. Vol. 14, pp.
161-168.
4. Lai, T.M., and D.D. Eberl.
1986. "Controlled and
Renewable Release of
Phosphorus in Soils From
Mixtures of Phosphate Rock
and NH4 Exchanged
Clinoptilolite. "
Zeolites. Vol. 6, pp. 129-
132.
5. Loizidou, M., and R.P.
Townsend. 1987. "Ion-
Exchange Properties of
Natural Clinoptilolite,
Ferrierite, andMordenite:
Part 2. Lead-Sodium and
Lead-Ammonium Equilibria."
Zeolites. Vol. 7, pp. 153-
159.
6. Nishita, H., and
R.M. Haug. 1972.
"Influence of
Clinoptilolite on
Sr90 and CS137 Uptakes
by Plants." Soil
Science. Vol. 114,
No. 2, pp. 149-157.
7. Semmens, M.J., and W.
Martin. 1980. "Studies
on Heavy Metals Removal
From Saline Waters by
Clinoptilolite." The
American Institute of
Chemical Engineers
Symposium Series: Water^
-1979.
8. Semmens, M.J., and M.
Seyfarth. 1978. "The
Selectivity of Clinop-
tilolite for Certain
Heavy Metals." Natural
Zeolites: Occurrence.
Properties. Use. Edited
-------
by L.B. Sand and F.A.
Mumpton. Pergamon Press,
pp. 517-526.
9. U.S. EPA Office of Solid
Waste and Emergency
Response. 1986.
"Procedure for Estimating
Monofilled Solid Waste
Leachate Composition:
Technical Resource
Document." SW-924,
Second Edition.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
478
-------
TREATMENT OF WATER REACTIVE WASTES
\
John Parker
Lanstar Wimpey Waste
Manchester, England
ABSTRACT
In Europe there is a small but growing amount of water reactive waste.
These wastes (e.g. titanium tetrachloride) react violently with water to
produce huge quantities of hydrogen chloride. They therefore pose a
serious waste disposal problem because conventional treatment methods are
not suitable. In addition to their reactivity these wastes pose a number
of other problems, namely; they are often contaminated with halogenated
organic chemicals which cause the waste to form a sludge, the waste is
almost always packed in 200 litre drums, and finally it is potentially
a very corrosive mixture.
Lanstar Wimpey Waste was approached by a number of companies to see
if we could treat this kind of waste so a laboratory project was undertaken
and when this produced encouraging results the process was scaled up to
a pilot plant that was capable of handling full 200 litre drums. The
results of this pilot plant work are discussed in this paper. In essence
the process involves; homogenisation of the waste, hydrolysis, phase separ-
ation, neutralisation of organic and aqueous layers, incineration of organic
solvents, dewatering of the neutralised aqueous slurry with the filter cake
going to a clay containment landfill site and the filtrate going to a
sewerage works.
This process ensures that these types of wastes are disposed of by
the Best Practicable Environmental Option.
INTRODUCTION
The European chemical and
fibre optic industries produce
chemical wastes that are water
reactive. Typical wastes are
titanium tetrachloride, antimony
pentachloride and phosphorous oxy-
chloride. These wastes react
violently with water releasing
huge quantities of hydrogen chlo-
ride. This type of waste is pro-
duced in small but regular quantities
and almost always in 200 litre
drums. In its most difficult
state the waste is produced by
the organic chemical industry as
a spent catalyst. In this form
the waste contains halogenated
hydrocarbons that often form a
sludge in the bottom of the drum.
Conventional disposal methods
such as landfilling, incineration,
479
-------
and chemical neutralisation are
inappropriate for this type of
waste. For instance, landfilling
of liquids in drums is simply not
allowed in the U.K., incineration
of predominantly reactive inorganic
wastes is not recommended, whilst
direct neutralisation of such wastes
is not possible due to the genera-
tion of vast quantities of hydrogen
chloride.
In 1987 detailed laboratory
investigations were under-taken
to determine whether it was possible
to: (a) perform controlled hydrol-
yses on such wastes, (b) treat
the products of the hydrolyses.
Following successful laboratory
treatment of a number of wastes
it was decided to do some pilot
plant work on full drums. This
paper is a report on these trials.
PURPOSE
The purpose of this project
was to develop a commercially viable
process for the hydrolysis and
subsequent treatment of water reac-
tive wastes. Whilst the main
emphasis was placed upon the hydrol-
ysis reaction the whole project
also depended upon there being
disposal routes for the acids and
organic solvents generated. It
was hoped that these disposal routes
would exist 'in house'.
APPROACH
Best Practicable Environmental
Option
The Company is committed to
finding the Best Practicable En-
vironmental Option for any given
waste. This means analysing the
costs and benefits of different
waste disposal options for a given
waste so that the greatest pollution
abatement will be given for the
minimum costs. So the first app-
roach was to see whether this cate-
gory of water reactive wastes could
be fitted into any of the Company's
existing portfolio of treatments.
These were; chemical treatment
(neutralisation, oxidation, re-
duction, etc.), physical treatment
(filtration, centrifugation, dis-
tillation, etc.), biological treat-
ment, incineration or landfill.
When it became apparent that
these wastes required pre-treatment
prior to other disposal methods
coming into play a laboratory in-
vestigation was undertaken.
Laboratory Investigation
A known weight of water was
added to a 2 litre reaction vessel.
The waste in question was added,
under gravity, at a known rate
whilst the temperature was continu-
ally monitored. Periodically
small quantities were removed from
the reaction vessel to determine
acid strength. All hydrogen chlo-
ride fumes were vented from the
reaction vessel and scrubbed.
Initially only inorganic wastes
that were completely liquid, e.g.
titanium tetrachloride, underwent
this evaluation. However, as
a routine process was developed
more difficult wastes were tested,
notably those that were contaminated
with halogenated solvents.
In all cases, when a waste
was successfully hydrolysed in
the laboratory it was then tested
to determine if it could then be
processed through existing pro-
cesses. The key thing was to
evaluate whether the waste could
go through the normal acid neu-
480
-------
tralisation process. Thus, the
following questions required answer-
ing: Could neutralised sludge
be dewatered, was the filter cake
acceptable for landfill and was
the water acceptable for sewer
discharge?
From this work a standard
method of operation was developed
in which the initial charge of
water was fixed, the final concen-
tration of hydrochloric acid select-
ed and a rate of addition recommend-,
ed.
Pilot Plant Evaluation
The recommendations of the
laboratory investigation were taken
to the pilot plant. This plant
consisted of a 2,250 litre glass
lined reactor fitted with cooling
water and connected to a packed
column caustic scrubber (normal
operating strength 10%).
During this evaluation trials
with one or two drums were under-
taken. The process followed the
laboratory trials. Thus a known
volume of towns water was charged
to the reactor, during the addition
of the waste (again under gravity)
the temperature and the weight
of the drum were continually mon-
itored. Furthermore, the strength
of the caustic in the scrubber
was measured at regular intervals.
PROBLEMS ENCOUNTERED
Drums
These wastes are so reactive
that as a protective measure they
are transported in overdrums.
Thus a foolproof system for removing
the drums from their overdrums
and delivering them to the treatment
plant was essential.
The drums, though often new drums,
could not be considered pressure
vessels and therefore the contents
had to be fed to the reactor by
gravity. This meant firmly fixing
the drums in a stillage. In order
to aid this process dry nitrogen
was bubbled through the liquor
but in such a way that no pressure
build-up occurred. N.B. Nitrogen
was introduced through a specially
designed adaptor which kept the
pipe to the reactor free from sludge
build-up.
A major difficulty occurred
with those wastes that had been
catalysts in the organic chemical
industry. In this case there
was always some sludge formation
in the bottom of the drum. Gen-
erally this was overcome by rolling
the drum for 30 minutes prior to
fixing the drum in its stillage.
However, for particularly awkward
drums, when most of the liquid
had been removed from the drums
a dry chlorinated solvent (such
as perchloroethylene) was added
to the drum which was rolled and
then re-fitted to the stillage.
In all cases the drums were
completely emptied (checked by
weighing and dipping) before being
quenched with water, re-emptied
and crushed.
Due to dealing with drums
there was always the problem of
continually coupling and decoupling
lines between the drum and reactor.
This always produced fumes of hydro-
gen chloride and to prevent this
being a problem there was a vent
over the reactor and drum that
fed straight into the scrubber
system.
Reactivity
The wastes in question are
481
-------
so reactive it is essential to have
an adequately sized scrubber to
cope with all the hydrogen chloride
generated.
Also, in those instances where
there are organic chemicals present
the lining of the reactor and holding
tank is of prime importance. In
fact, with one particular waste
there were chlorofluorocarbons
present with the result that on
hydrolysis a small quantity of
hydrofluoric acid was produced.
In order to minimise the affect
of this aggresive acid calcium
chloride was added to the water
prior to the hydrolysis reaction
so that during the reaction calcium
fluoride would precipitate, but
this did not minimise the care
needed about choosing the reactor
linings for these kinds of waste.
Acid Strength
Experience has shown that the
optimum acid strength is around
15% w/w. At this concentration
all the waste goes into solution.
At higher concentrations a slurry
is sometimes produced. Furthermore,
at 15% strength the acid waste can
be directly neutralised with 10%
calcium hydroxide slurry and the
resulting slurry remains pumpable.
At greater than 15% acid strength
it has sometimes been found necessary
to dilute the acid.
When the acid strength is
maintained around 15% the temperature
is always maintained less than 50°C.
A maximum operating temperature
of 70°C is therefore used as a
control guard against 'unusual
reaction1 and to prevent an overload
on the scrubber system.
Phase Separation
This problem only arose with
those wastes containing separable
organics but in this case it was
essential to have a phase separation
tank. As the organics were never
present in large quantities it
was always possible to draw off
'clean' acid from the holding tank
whilst allowing a build-up of
organics on top of the tank.
However, on removing the
organic layer it was always found
to be acidic. This meant neutral-
isation with sodium hydroxide in
a separate reactor followed again
by phase separation. However,
because this was only needed infre-
quently both processes were done
in the same reactor.
The neutralised organic layer
then went for incineration whilst
the neutralised aqueous layer was
processed with other waste acids
being treated on site.
RESULTS
All the results are from
pilot plant studies.
Typical Reactions
TiCl + 2H 0
4 2
2HC1 + Ca(OH) -
The Process
4HC1
CaCl + 2H 0
£t £
Figure 1 shows the block flow
diagram for water reactive wastes.
Experience has shown that the
following are the parameters that
give the best results:-
Hydrolysis
Drum rolling
30 minutes
Water charge to the reactor (2250
litres) - 1400 litres
482
-------
Rate of addition of waste to reactor
- 5 litre/minute
Typical time to empty 200 litre
drum (see Figure 2) - 70 minutes
Maximum number of 200 litre drums
processed per batch - 2
Maximum operating temperature - 70°C
Typical operating temperature - 35°C
(see Figure 3)
Solvent for washing organic sludges
perch.loroethylene
Maximum acid strength - 20% w/w as HC1
Typical acid strength - 15% w/w as HC1
Neutralisation
Neutralisation medium for the acid
layer - calcium hydroxide
Dewatering equipment- Vacuum belt
filters
Total chlorinated hydrocarbons in
filtrate - less than 100 microgram/ltr
Landfill of filter cake - Site has
minimum of 8 metres of clay
Incineration of solvents:
Temperature : 1200°C
Oxygen : Excess
Time in flare: Greater than 1 second
N.B. The solvents are currently
incinerated at sea. It is more
than likely that between 1992 and
1994 the solvents will be transferred
to a landbased incinerator.
Based upon this work done on
the pilot plant and the fact that
there is a need in the U.K. for
this method of processing a decision
has been taken to install a 9,000
litre reactor. This will operate
on the regular work whilst the
pilot plant will continue to operate
on 'spot' work and, of course,
further development work.
The plant has gone through
a full Hazop process and .a full
set of operating instructions have
been issued.
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
483
-------
484
-------
__ 3 , _l _^
HI _J_
' I
-
• i
i,
iy;
+-7
71
I '
-i-
-4—
V>
—! —
—I"
' I
01:
01.
4J!
3;
•rli
o "1
r- o!
Ei
O i
in
t !
1 !_
r~i
i
i
i
I
•nod
i
i
~ i
i
i >
i i
1 _ !
' ' 1
! i '
i ! 1
~ | ' ' 1
J ' ' ,- -
1 | 1 1
1 ! !
1
r-*vO in ** rocM rHC
roro ro ro roro ro o
S^fS
!
1
~*
-
-
J
1
-
3
1
_J —|
1 1
-— |
1 .
1
_| -
' ' J
"i
1 i
-- t "r
I
r _, '_
-+ i „
t !
i j
y\ CD f
N CM 0
Oo •
T
._ _
_!__
—
1
— r
• '
_!
1 _
V
j C
3^nq.e
- - T
'— i
-- i
1
"I
-1
'
i
1
i"
o
M
JtsrJuie
i
- \
i '
. „ i
1
_
-------
T
___ i
TO
r
__— _L
1 I
_L -
_| \_
1
_ _. , i i i
i \ , . 1-'-1 1 —' - -
u
r
i _
• M -
-o
"T
J _
- i
\
<
r
to
,
I H
486
-------
USE OF INNOVATIVE FREEZING TECHNIQUE FOR IN-SlTU TREATMENT
OF CONTAMINATED SOILS
Olufemi A. Ayorinde, Lawrence B. Perry and Iskandar K. Iskandar
U. S. Army Cold Regions Research and Engineering Laboratory
Hanover, New Hampshire 03755-1290
ABSTRACT
In the past few years, CRREL has been investigating the use of artificial freezing as an innovative technique for
soil decontamination. A preliminary laboratory study was conducted specifically to evaluate and analyze the possibil-
ity of mobilizing different types of contaminants by freezing in Lebanon silt. Contaminants investigated were explo-
sive residues most extensively found at the U.S. Army ammunition plants as well as volatile organic compounds
(VOCs), such as chloroform and toluene. Explosives studied were 2,4,6-trinitrotoluene (TNT), hexahydro-1,3,5-
trinitro-1,3,5-triazine (RDX), octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX), 2,6-dinitrotoluene (2,6-DNT),
ortho-nitrotoluene (0-NT), and meta-nitrotoluene (M-NT).
Preliminary data from the laboratory column studies suggested that there was a certain degree of movement of
both explosives and VOCs when soil columns of Lebanon silt saturated with these contaminants were frozen uni-
directionally from the bottom up. Slopes of the control and frozen soil concentration profiles were statistically anal-
yzed and a comparison between them was made. One freeze cycle at an average freezing rate of 0.5 cm/day was
used. Insignificant amounts of movement (< 10% change) were observed for RDX, HMX and TNT. Relatively greater
movements (20-40% change) were observed for 2,6 DNT, 0-NT, M-NT, toluene and chloroform. For given freezing
rate, freeze-thaw cycles, soil and moisture content, it was hypothesized from this and other previous experimental
data that the ability to move any contaminant by freezing strongly depends on the type, initial concentration level and
the soil/chemical interaction of the contaminant.
INTRODUCTION
Several military bases and U. S. Army ammunition plants have soils and sediments contaminated with explosive
residues resulting from the manufacture, use, and disposal of organic-based explosives. Some of the explosive com-
pounds include TNT, RDX, HMX, 2,6-DNT, M-NT and 0-NT. These military facilities and other industrial sites con-
tain VOC contaminants in addition to explosive residues. Cost-effective and safe control, treatment and disposal of
hazardous/toxic wastes remain as some of the major environmental problems and challenges facing the U.S. and
other industrialized countries. In particular, the transport of the contaminants to ground and surface water is perhaps
the main issue of public concern. Iskandar and Houthoofd (1) and Iskandar and Jenkins (2) summarized the review
of available techniques for remedial action at uncontrolled hazardous waste sites, and concluded that current tech-
niques are neither adequate nor cost-effective for protecting ground and surface water.
Artificial freezing is being studied as a means to decontaminate explosives- and VOCs-contaminated soils and
may have the potential to be a cost effective, environmentally safe, and readily implemented technology.
PURPOSE
In the past decade, the growing problem of hazardous wastes has contributed to a recent increase in the public
awareness and preference for a research effort in developing new cleanup technology with a minimum negative
impact on the environment and surrounding aesthetics. Consequently, there is an urgent need for continued re-
487
-------
search to investigate other possible innovative and cost-effective techniques for treating and controlling hazardous/
toxic wastes. Iskandar et al. (3) and Ayorinde et al. (4) have evaluated laboratory methods for the potential use of
artificial freezing to decontaminate soils contaminated with VOCs and their data indicated moderate success. Perry
and Ayorinde (5) and Taylor (6) have experimentally demonstrated that different types of explosive residues can be
effectively moved in water by freezing.
The purpose of this paper is to present the preliminary results of a laboratory investigation of exploring artificial
freezing as an innovative potential technique for in-situ soil decontamination. In particular, the paper describes and
summarizes the experimental approach and analysis developed for a laboratory column study to evaluate possible
freezing-induced transport of volatile and nonvolatile contaminants in soil.
APPROACH
The test apparatus used consisted of 7-cm-OD by 0.325-cm-thick by 12.5-cm-high Lucite cells for containing
Lebanon silt, an environmental chamber for housing the soil columns and for controlling ambient temperature, a
backsaturation test setup, a 2-L carboy for preparing contaminant concentration, thermocouples, sampling tools,
programmable cooling baths (each with rapid response time and a temperature ramp controller), sample storage
facilities, an oven and an analytical laboratory for conducting high-performance liquid chromatography (HPLC). Each
Lucite cell was fabricated and carefully fitted with aluminum end caps and stainless steel inlet and outlet ports at the
bottom and top, respectively. Polyethylene filters, 0.5 cm in thickness, were also placed at the top and bottom of the
cell. The Lucite cell was also instrumented with type T (copper-constantan) thermocouples for monitoring tempera-
ture during freezing.
Sample Preparation:
Because of both the unknown complex chemical behavior of organic compounds, and the undefined potential
interaction between different types of volatile and nonvolatile organic compounds under low temperature conditions,
the explosive compounds were separated from the VOCs for our investigation. At the time of the experiment, there
was no information in the literature about any possible chemical interaction when explosive residues and VOCs are
mixed together to produce a spike solution. To avoid any additional complexity, separate soil sample columns were
prepared for the explosives and the VOCs. Since then, the in-house CRREL organic chemistry group (Jenkins [7])
has noted that no adverse chemical reactions occur when mixing solutions containing explosive compounds and
VOCs. This assessment agrees with the conclusion of a
study conducted by Karickhoff et al. (8).
Eight sample columns were prepared. Two columns were
used as controls for the VOC (column A) and the explosive
contaminants (column E), respectively. Three columns (col-
umns F, G and H) were used as replicates for the explo-
sive-contaminated soil while the remaining three columns
(columns B, C, and D) were replicates for the VOC-contam-
inated soil. Each Lucite column was packed with oven-dried
Lebanon silt to an average density of 1.67 g/cm3 and a por-
osity of about 36%. The packing was done in 2-cm layer
lifts to achieve relatively uniform density throughout the col-
umn. Once the test cell was full, an end cap was placed
with the filter on top of the soil, and the cell was sealed.
Both the bottom inlet and the top outlet were then attached
with three-way control valves fitted with Teflon tubing. A
typical Lucite test cell used for the experiment is shown in
Figure 1 with a soil sample height of 12 cm.
488
7"0,D.
Lucite
Cell
Porous
Filter Disc
<»yp,J
Alum i num
Endcop
(typ )
12cm
Inlet
Figure 1. Typical Lucite soil test cell.
-------
Aout
Soil Cell
A
Buret
3-way Valves
(typ.)
t
AB in
Preparation of Contaminant Spike Solutions:
The VOC spike solution consisting of chloroform and toluene was prepared and their concentrations were
determined in accordance with the previously established procedures (Ayorinde et al. [4]). The expected concentra-
tions for chloroform and toluene were 990 and 230 mg/L, respectively. Explosive residues were added to 200 mL of
Milli-Q water and the resulting solution was poured into a 2-L carboy in which the solution was mechanically mixed
for about 24 hours. The expected concentration for each of the explosive residues was 20 mg/L. Before backsatu-
rating the soil columns with the spike solution, the VOC contaminant concentrations were 604 and 84 mg/L for
chloroform and toluene, respectively, and about 10 mg/L for each of the explosive contaminants.
Backsaturation Method and Breakthrough Curve Procedure:
To obtain a reasonably uniform distribution of the contaminant concentration through each soil column prior to
freezing, each column was backsaturated with 10 pore volumes of spike solution. An increment of 0.5 pore volume
was used until a total of 10 pore volumes was reached. The backsaturation apparatus used is shown in Figure 2.
The burette with the spike solution was connected to the Lucite soil columns using the Teflon tubing to form a closed
loop as shown in Figure 2. The burette
was placed high above the soil columns to
provide enough head to move the spike
solution very slowly through the soil from
the bottom up. Two of the eight prepared
soil columns were simultaneously back-
saturated at each time. During the back-
saturation of a new set of soil columns,
already saturated columns were kept in-
side the environmental chamber. The ef-
fluent solution from each column outlet
was collected in a closed overflow flask to
minimize volatilization of the VOCs.
. . . . , Figure 2. Backsaturation apparatus.
For each incremental pore volume, a
the inlet solution at the bottom of each soil
column and the outlet solution at the top of the column were continuously monitored and collected for chemical
analysis. HPLC was used to measure the concentration levels of the explosive residues and VOCs in the collected
solution based upon the procedure developed for determining nitroaromatics and nitramines (Jenkins et al. [9]) and
VOCs (Jenkins [7]) in water. As shown schematically in Figure 2, the inlet solution from the burette is identified as
ABln, while the effluent solution coming out of samples A and B is identified as Aout and Bout, respectively. Examples
of typical breakthrough curves obtained for all the contaminants studied are shown in Figures 3 and 4 for RDX,
toluene, M-NT and 0-NT.
The soil was assumed to be fully saturated after a flow of 10 pore volumes of the spike solution, and both the
inlet and outlet valves for columns were closed. The soil sample columns were then placed in the environmental
chamber for subsequent freezing. The backsaturation apparatus with a closed loop system depicted in Figure 2 was
developed as a rational method of minimizing the complex problem of the volatilization losses of the spiked VOCs
while saturating the soil samples.
Sample Freezing and Sampling Methods:
The environmental chamber which housed the eight saturated soil columns was set at +1.0°C and maintained
at this temperature throughout the experiment. Prior to freezing, all the soil columns were equilibrated at +1.0°C for
48 hours. 'Soil columns B, C, D, F, G, and H were then placed on a large cooling plate inside the environmental
chamber. Control columns A and E, kept unfrozen throughout the test, were placed on top of an insulation pad and
489
-------
100 200 300
Elapsed Time (min)
400
Rgure 3. Breakthrough curves (cone, vs time) for
RDX and toluene.
4. 2 -
100
200
300
400
Elapsed Time (min)
Figure 4. Breakthrough curves (cone, vs time) for
M-NTandO-NT.
7"0. D.
Frozen
1 cmF
,
-T- Top
12cm
Bottom
Rgure 5. Sampling layout of half soil
column after flash-frozen in liquid nitro-
gen.
Outlet
I cm
T
Lucite
Cell
Thin
Sections
Porous
Filter Disc"
(typ.)
Inlet
I I
10
Aluminum
Endcop
(typ.)
12cm
Figure 6. Test cell soil sampling scheme with thin
section dimensions.
490
-------
isolated from the bottom cooling plate. Then, another large cooling plate maintained at the inside chamber tempera-
ture of +1.0°C was placed on the top of the eight columns to control the top end boundary temperatures of the soil
columns. The temperature data for the cooling baths, the environmental chamber, and the soil column thermo-
couples were collected on a DIG! Ill datalogger, interfaced with an IBM PC.
The top and bottom boundary temperatures of the soil columns, as well as the environmental chamber tempera-
ture were controlled by programmable cooling baths. These programmable cooling baths are capable of maintaining
temperatures within 0.1 °C. The freezing of the soil columns was gradual and .was from bottom up in order to elim-
inate the possible effect of gravity on the mobility of contaminants. By controlling the cooling rate of the bottom plate
at a constant temperature decrement rate of about 0.4°C per day to freeze the soil columns, an approximate freeze
rate of 0.5 cm/day was achieved. Because of the good temperature control of the environmental chamber, a one-
dimensional vertical freezing of the soil columns was also achieved. Columns B, C, D, F, G and H were frozen ap-
proximately halfway (6 cm) from the bottom.
At the end of the freezing process, all samples were taken out of the environmental chamber and flash frozen in
liquid nitrogen. Thin sectioning was used to provide estimation of the sample soil solute concentration profiles. Fro-
zen soil columns were transferred to a coldroom and cut vertically in half as shown in Figure 5. One half was kept in
the freezer, while the other was cut into 1 -cm-thick sections on a band saw as depicted in Figure 6.
Two sample duplicates of each thin section were obtained, representing about 2/3 of each thin section made
from 1/2 of the original sample. The remaining 1/3 of the thin section was used for estimating the sample moisture
content profile. The thin-section sample duplicates were placed in sealed glass scintillation vials to be extracted for
chemical analysis. About one milliliter of the extract was then transferred into mini-autosampler vials for direct con-
centration determination using HPLC. By this procedure, the concentration profile of the contaminants along each
soil column was obtained.
PROBLEMS ENCOUNTERED
One of the problems concerned the difficulty in achieving initial uniform distribution of the contaminants in soil.
There were several contributing factors. These included the inherent soil inhomogeneity, flow tortuosity, local soil/
contaminant interactions due to sorption and other chemical and physical forces, and local losses by absorption, bio-
degradation and volatilization. Since these influencing parameters could not be completely eliminated, the experi-
ment was designed to include controls. To minimize the abovementioned contributing factors, triplicate soil columns
were used for freezing treatment with duplicate samples taken at different locations along each column height. Other
problems centered on the limitation in obtaining and chemically analyzing a large number of soil samples in a timely
and efficient way that would be representative of the soil columns for the experiment. It was only possible to analyze
about 1/3 of each column for contaminant concentration determination. The underlying assumption was that the soil
sample was relatively homogeneous without (a) any local preferential sorption of the contaminants to the soil, (b)
any significant tortuosity, and (c) appreciable accumulation of any of the contaminants in the soil portion used for the
moisture content estimation.
RESULTS
Data Analysis
Theoretical analysis of freezing-induced organic contaminant transport in soils requires, in general, the use of
advective-dispersive equations with sorption and volatility effects coupled with heat transfer with phase change. The
set of equations involved here is very complex and difficult to analyze. However, in order to model the transport of
organic compounds in soils due to freezing, this type of complex coupled equations would have to be developed and
solved. In this paper, no attempt is made to develop an analytical model for the data presented. Such an effort is in-
cluded in our research investigation in the future. Our current research goal is to develop reliable experimental meth-
ods for obtaining data that can be used for developing and testing models that describe freezing-induced contami-
nant transport in soils. 491
-------
As noted above, there were two duplicate concentration measurements at each location along each soil column.
Duplicate analysis using Youden's Method (Bauer [9]) was used for all measured data. Thus, at each location along
each soil column,
C, = meanCE [CF,(i) + CG,(i) + CH,(i)] ) ; i = 1 to 2 (1)
; i = 1to2 (2)
where i represents the number of duplicates per location along the soil column, and f and t subscripts denote frozen
and thawed (unfrozen) soil columns, respectively.
The standard deviation (STD) and the corresponding coefficient of variation (CV) for the measured concentra-
tion at each location along each soil column were then calculated using Equations 1 and 2. Thus:
STD = (Z [(c,^) - C, „,)*] /(N-1))« ; j = 1 to 6 (3)
for subscript f and j = 1 to 2 for subscript t
CV(%) = 100xSTD/Ctor, (4)
where N is the number of data measurement locations along the soil column.
In an attempt to minimize inherent variability in the measured data arising from unavoidable variations in meas-
urement, sampling, sample preparation and soil properties (e. g., heterogeneity), the concentration value per loca-
tion along each column was normalized with respect to the mean concentration value (Ctmean) for the whole height of
the unfrozen control column E. Hence,
k);k=1...N (5)
where N is the number of data measurement locations along the unfrozen control column E. And the normalized
concentration for the soil column profile is given by
Normalized Cone. = Ctor/Ctmean (6)
As indicated above, the purpose of this preliminary experimental investigation was based on the hypothesis that
freezing can move organic contaminants in soils just as it has been demonstrated for water (Perry and Ayorinde
[5],Tayfor [6]). To test this hypothesis, two simple analytical approaches were adopted to analyze the measured
concentration data.
The first approach compared the average of the concentration profile for the frozen bottom half (0-6 cm) of the
freezing-treated columns (columns F, G and H) with the corresponding average concentration for the bottom half
(0-6 cm) of the unfrozen control column (column E). As it may be recalled, it was pointed out earlier that the col-
umns F, G and H which were subjected to freezing only froze to about 6 cm from the cold bottom ends of the col-
umns. The bottom-half average calculated concentration was each normalized with respect to Ctmea . The difference
between the frozen column normalized concentration and that of the unfrozen control column was expressed as the
percent normalized reduction in concentration. Thus
Normalized Reduction in Cone. (%) = 1 00(CtBH - C(BH)/Clmean (7)
where C1BH is the average bottom-half (0-6 cm) control column concentration, CfBH is the average bottom-half (0-6
cm) frozen columns concentration, and Ctaean is given by Equation 5. The calculated values for the normalized re-
duction in concentration for the explosive contaminants are given in Table 1.
Analogous analysis performed for the explosive contaminants with Equations 1-7 was done for the VOCs. The
unfrozen control column A data was used to replace column E in the above equations. The frozen columns B, C and
D replace columns F, G and H. Calculated values for soil profile concentrations for the VOCs are shown in Table 2.
The second analytical method used to test experimental hypothesis compares, using the Student t statistical
significance test (Bauer [10]; Draper and Smith [11)], the regression slope of the concentration profile for the
492
-------
Table 1. Approximate changes in explosive contaminant concentration in Lebanon silt presumably induced
during freezing.* (Approximate freezing front location = 6 cm from bottom end.)
DISTANCE
TYPE FROM SAMPLE
OF BOTTOM
CONTAMINANT (cm)
UNFROZEN NORM. AVG.
CONTROL CONTROL
SAMPLE SAMPLE
CONC. CONC.
(ct)
HMX
RDX
TNT
2 , 6-DNT
O-NT
M-NT
Bottom-Half
(0-6 cm.)
Bottom-Half
(0-6 cm.)
Bottom-Half
(0-6 cm.)
Bottom-Half
(0-6 cm.)
Bottom-Half
(0-6 cm.)
Bottom-Half
.(0-6 cm.)
2.30
0.77
4.71
5.86
3.21
4.04
0.95
1.02
1.07
1.05
1.07
1.08
2.51
0.71
4.34
4.96
2.00
2.65
1.04
0.94
0.99
0.89
0.67
0.71
-8.81
8.28
8.40
16.10
40.36
37.19
* Calculated concentration changes were obtained by dividing each
column into two equal segments, and by averaging concentration
profile over the bottom frozen half segment.
Table 2. Approximate changes in volatile organic concentration in Lebanon silt presumably induced during
freezing.* (Approximate freezing front location = 6 cm from bottom end.)
UNFROZEN NORM. AVG.
TYPE
OF
CONTAMINANT
CHLOROFORM
TOLUENE
DISTANCE
FROM SAMPLE
BOTTOM
(cm)
Bottom-Half
(0-6 cm.)
Bottom-Half
(0-6 cm.)
CONTROL
SAMPLE
CONTROL
SAMPLE
CONC. CONC.
(ct) (ct/ctmean)
64.05
9.10
0.96
1.12
AVG. NORM. AVG.
FROZEN
SAMPLE
CONC.
(Cf) (Cj
48.83
7.36
FROZEN
SAMPLE
CONC.
0.73
0.91
NORM.*
REDUCTION
IN CONC.
(ct-c^)/ctmean)
22.83
21.40
* Calculated concentration changes were obtained by dividing each
column into two equal segments, and by averaging concentration
profile over the bottom frozen half segment.
unfrozen control column with that of the average concentration profile for the three frozen columns for each analyte.
AH the data for the three replicate frozen columns were combined and used to least-squares fit the combined 78
data points for the frozen slope estimate. Duplicate values at each location were used for the control column.
For each analyte, a value of t was calculated based on the slope comparison analysis, and the significance level
of the difference between the slopes was obtained from the Student t table. A significance level less than 95% was
considered insignificant for any difference between the slopes to be considered substantial for the control and frozen
sets of data. And a probability level of equal to or more than 95% was considered significant. Also, an F-test was
Used to check if the intercept of the regression line equation is different from zero before the Student t statistical
comparison analysis was performed. The results of the comparison between the regression slopes for the frozen
and control concentration profiles are summarized in Table 3.
493
-------
Table 3. Comparison between the regression slopes of contaminant concentration
profiles for control and combined frozen silty soil columns using t-test significance
analysis.
TXPE
OF
CONTAMINANT
HMX
RDX
TNT
2,6-DNT
MNT
ONT
CHLOROFORM
TOLUENE
CONTROL
SLOPE
/*cf/9
3.25E-03
-3.58E-03
-1.02E-01
-8.92E-02
-1.07E-01
-7.85E-02
4.77E-01
-1.82E-01
COMBINED
FROZEN
SLOPE
per cm
-9.55E-02
-6.89E-03
-5.60E-02
-2.99E-02
-6.70E-03
-1.17E-03
2.38E+00
-9.58E-02
t-TEST
SIGNIF.
LEVEL
(%)
86.5
41.0
82.0
76.0
97.5
96.5
74.0
35.0
OCTANOL-WATER
PARTITION GENERA
COEFFICIENT COMMEN1
*
1.36
7.59
67.6
97.0
263.0
199.5
93.3
490.0
NS?
NS;
NS?
NS?
S?
S?
NS?
NS?
< 95%
< 95%
< 95%
< 95%
> 95%
> 95%
< 95%
< 95%
S ^ Statistically Significant Difference Between Slopes
NS = No Statistically Significant Difference Between Slopes
* Values obtained from Hansch and Leo (12), Leggett (13) and
Jenkins (14)
Discussion
Rgure 7 shows a composite of the soil moisture content profiles for unfrozen control sample and the three
replicate frozen samples used as triplicates involving the explosive contaminants. The main purpose of using
triplicate samples is to diminish the effect of several variabilities inherent and usually encountered in a complex
nonhomogeneous system such as soil. The moisture profiles for these four samples appeared to agree very well
within the measurement accuracy. Similarly good agreement between the unfrozen and frozen sample moisture
profiles can be observed for the VOC contaminated soil columns (Fig. 8).
The comparison between the normalized HMX contaminant concentration profile for the unfrozen control soil
condition and that for the partially frozen soil condition is shown in Figure 9. The observed anomaly in the control
30
2 20
I
10
o Control Column E
• Frozen Column F
» Frozen Column G
A H
30
2 20
4 8
Distonce from Sample Bottom Cold End (cm)
12
10
o Control Column A
• Frozen Column B
A Frozen Column C
* D
I I
8
Distance from Sample Bottom Cold End (cm)
12
Figure 7. Moisture content profiles of soil columns with explo- Figure 8. Moisture content profiles of soil columns with VOCs.
494
-------
1.6
o Average Control
• Average Frozen
4 8
Distance from Sample Bottom Cold End (cm)
Figure 9. Normalized HMX concentration profiles in Leb-
anon silt.
1.6
1.2
- 0.8
0.4
ROX
o Average Control
• Average Frozen
4 8
Distance from Sample Bottom Cold End (cm)
12
Figure 10. Normalized RDX concentration profiles in Leb-
anon silt.
concentration profile around the top mid-portion of the sample height was supported by the high values in the sam-
ple standard deviation (STD) and coefficients of variation (CV) at these locations.
Very little data scatter was observed for the RDX concentration profiles both for the control and frozen samples,
as shown in Figure 10. Moreover, the extreme low STD and CV values for the control sample and the corresponding
moderate values for the frozen sample supported this observation. The slight or absent freezing-induced mobility
may be due, in part, to the relatively low initial concentration level used. Hence, it was inferred from the data that one
freeze cycle at a rate of 0.5 cm/day would not significantly move RDX analyte with an initial average concentration of
about 0.76 jxg/g.
Concentration profiles for TNT and 2,6-DNT are depicted in Figures 11 and 12, respectively. Very little data
scatter along the sample height for control and frozen sample profiles could be observed for both TNT and 2,6-DNT
as shown in their respective low STD and CV values. There appeared to be no significant difference in the control
and frozen sample concentration profiles for TNT, since its normalized concentration reduction within the frozen por-
tion was about 8%. The concentration reduction for 2,6-DNT was about 16%, indicating no appreciable movement
caused by freezing. Comparison of the regression slopes between control and frozen samples for both TNT and 2,6-
DNT showed that there was no statistically significant difference between the slopes.
1.6
E
O
1.2
t: 0.8
0.4
TNT
o Average Control
• Average Frozen
_J
1.6
1.2
~ 0.8
O.4
2,6-DNT
o Average Control
• Average Frozen
_J
4 8
Distance from Sample Bottom Cold End (cm)
12
4 8
Distance from Sample Bottom Cold End (cm)
12
Figure 11. Normalized TNT concentration profiles in Leb- Figure 12. Normalized 2,6-DNT concentration profiles in
anon silt. Lebanon silt,
495
-------
1.6
1.8
"
S 0,4
1
0-NT
o Average Control
• Average Frozen
1.6
1.2
t o.a
_ M-NT
0 4 3 12
Distance (ram Sample Bottom Cold End (cm)
Rgure 13. Normalized 0-NT concentration profiles in
Lebanon silt.
o Average Control
• Average Frozen
I I
4 8
Distance from Sample Bottom Cold End (cm)
12
Figure 14. Normalized M-NT concentration profiles in
Lebanon silt.
1,6
1,2
0.8
0.4
1.6
1.2
$ 0.8
S 0.4
Toluen
o Average Control
• Average Frozen
4 8
Distance from Sample Bottom (cm)
Rgure 1 5. Normalized chloroform concentration profiles in
Lebanon silt.
048
Distance from Sample Bottom Cold End (cm)
Figure 16. Normalized toluene concentration profiles in
Lebanon silt.
12
Rgures 13 and 14 compare the concentration profiles between the control and frozen samples for 0-NT and (VI-
NT, respectively. The normalized concentration reductions within the frozen portion behind the freeze front between
the control and frozen samples were about 40% for 0-NT and about 37% for M-NT.
Normalized concentration profiles for chloroform and toluene are depicted in Rgures 15 and 16. Sharp peaks at
some locations indicated the large data scatter along the sample height. Large values for CV and STD represented
the large differences in the duplicate measurements at different locations. At some locations indicated by. the CV
value of 100% for the control soil column, one of the duplicate samples was either missing or was below detection
limits. In particular, the use of HPLC for the chemical analysis of chloroform in soil was found to be very difficult.
Such a technique is still under development at CRREL (Jenkins [7]).
Conclusions
The following conclusions were drawn from this experimental study:
1 . For given freeze rate, freeze-thaw cycles, soil and soil moisture, it was postulated that the ability to move a
contaminant by freezing strongly depends on the type, initial concentration level and the soil/chemical interaction of
the contaminant. Also the ability to easily detect each contaminant in the soil affects how to assess whether or not
freezing moves a given type of contaminant.
496
-------
2. Among the explosives, 0-NT and M-NT analytes were significantly reduced by freezing.
3. By comparing frozen and control columns, movement of HMX, RDX, TNT and 2,6-DNT attributed to freezing
was statistically insignificant for the relative low concentration used under the freezing condition of one freeze cycle
with 0.5-cm/day freeze rate.
4. The inherent volatility of VOCs, the soil spatial variability and the complex chemical interaction (i.e. absorp-
tion) between the organic compounds and the soil particles represent some of the sampling problems encountered
in the use of artificial freezing as a potential soil decontamination method.
5. Even though there was a distinct difference in the slopes of the concentration profiles for control (unfrozen)
and frozen soil columns contaminated with chloroform and toluene, the data in this experiment showed no statisti-
cally significant movement induced by freezing at a freeze rate of 0.5 cm/day and one cycle of freezing.
6. An insignificant, freezing-induced reduction in concentration (< 10%) was observed for RDX, HMX and TNT.
Relatively greater reduction (20-40% change) was observed for 2,6 DNT, 0-NT, M-NT, toluene and chloroform.
ACKNOWLEDGMENTS
This work was financially supported, in part, by the U.S. Environmental Protection Agency (EPA) under the
Interagency Agreement Project No. DW96931421-01-2 with the U.S. Army Cold Regions Research and Engineering
Laboratory (CRREL), and, in part, by the U.S. Army Corps of Engineers RDTE Project No. 4A161102AT24, Work
Unit SS/020, Prediction of Chemical Species Transport in Snow and Frozen Ground. The authors wish to thank T.F.
Jenkins, P. Schumacher, S. Taylor, D. Pidgeon and P. Miyares for their cooperation and assistance in the chemical -
and computer data analyses. We also thank E. Wright for editorial review and Dr. C.M. Reynolds of CRREL, Prof.
John Sullivan of the Worcester Polytechnic Institute and Dr. Richard Dobbs of the USEPA for their technical review
of the manuscript. The EPA Project Officers for this project were Janet Houthoofd and Doug Keller, Risk Reduction
Engineering Laboratory, Cincinnati, Ohio.
REFERENCES
1. Iskandar, I.K. and J.M. Houthoofd, 1985, Effect of freezing on the level of contaminants in uncontrolled hazardous
waste sites. Part I. Literature review and concepts. Proceedings, Eleventh Annual Research Symposium,
Cincinnati, Ohio, 29 April-1 May, 1985.
2. Iskandar, I.K. and T.F. Jenkins, 1985, Potential use of artificial ground freezing for contaminant immobilization.
Proceedings, International Conference on New Frontiers for Hazardous Waste Management, 15-18 Septem-
ber, 1985, Pittsburgh, Pennsylvania, p. 128-137.
3. Iskandar, I.K., L.B. Perry and T.F. Jenkins, 1986, Artificial freezing for treatment of contaminated soils - A pilot
study. Land Disposal, Remedial Action, Incineration and Treatment of Hazardous Waste, Proceedings of the
EPA Twelfth Annual Research Symposium, Cincinnati, Ohio, 21 -23 April, 1986.
4. Ayorinde, O.A., L.B. Perry, D.E. Pidgeon and I.K. lskandar,1988, Experimental methods for decontaminating soils
by freezing. Proceedings, Test Technology Symposium, 26-28 January, 1988, John Hopkins University,
Laurel, Maryland.
5. Perry, L.B. and O.A. Ayorinde, 1988, Exclusion of nonvolatile organic compounds in water during freezing-
Experimental Data. U. S. A.Cold Regions Research and Engineering Laboratory Internal Report 1023,
Hanover, New Hampshire.
6. Taylor, S.,1988, Ice/water partition coefficients for RDX and TNT. U. S. A. Cold Regions Research and Engineer-
ing Laboratory CRREL Report 89-8, Hanover, New Hampshire.
7. Jenkins, T.F.,1988, Personal Communication. U.S.A. Cold Regions Research and Engineering Laboratory,
Hanover, New Hampshire.
8. Karickhoff, S.W., D.S. Brown and T.A. Scott, 1978, Sorption of hydrophobia pollutants on natural sediments.
Water Research, vol. 13, pp 241-248.
497
-------
9. Jenkins, T.F., P.H. Miyares and M.E. Walsh,1988, An improved RP-HPLC method for determining nitroaromatics
and nitramines in water. U. S. A. Cold Regions Research and Engineering Laboratory Special Report 88-23,
Hanover, New Hampshire.
10. Bauer E.L.,1971, A statistical manual for chemists, second edition, Academic Press, New York.
11. Draper, N. and H. Smith,1976, Applied regression analysis, Wiley, New York.
12. Hansch, C. and A. Leo, 1979, Substituent constants for correlation analysis in chemistry and biology, Wiley, New
York.
13. Leggett, D.C.,1985, Sorption of military explosive contaminants on bentonite drilling muds. U. S. A. Cold Regions
Research and Engineering Laboratory CRREL Report 85-18, Hanover, New Hampshire.
14. Jenkins, T.F.,1989, Development of an analytical method for the determination of extractable nitroaromatics and
nitramines in soils. Ph.D. dissertation, University of New Hampshire, Durham.
Disclaimer
The information contained in this paper represents the authors' opinions and not necessarily those of EPA or
CRREL Hence, no official endorsement should be inferred.
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
498
-------
ENHANCING LIQUID-LIQUID AND SOLID-LIQUID
PHASE SEPARATION BY INTEGRATING ALTERNATING
CURRENT ELECTROCOAGULATORS WITH PROCESSING
AND WASTEWATER CONTROL SYSTEMS
Patrick E. Ryan
Electro-Pure Systems, Inc.
Thomas F. Stanczyk
Recra Environmental, Inc.
ABSTRACT
Coal, pigments, pharmaceutical
solids, ceramics, carbon, clays,
metallic powders and ores are among
the categorical groups of products
which are wasted as suspended
solids in aqueous-based wash solu-
tions. Phase separation and reco-
very of these solids by
conventional dewatering systems is
costly and relatively inefficient.
Alternating current can be used to
neutralize the electrical charge on
fine and ultra-fine particles in
aqueous suspensions, and facilitate
agglomeration and settling of these
particulates without using chemi-
cals while improving recovery effi-
ciencies. The AC electrocoagulator
has demonstrated the phase separa-
tion of emulsified oils and sub
micron particles, thereby purifying
water while improving the perfor-
mance of conventional wastewater
control systems. Applications for
the removal of soluble pollutants
are being investigated.
This paper provides an over-
view of the technology and
discusses applications and benefits
in the areas of in-plant pro-
cessing, industrial wastewater
treatment, site remediation and
water purification.
INTRODUCTION
In our industrial society,
there is an ever increasing aware-
ness of the adverse impacts of
inorganic and organic chemicals on
the quality of water. Human health
and environmental concerns dictate
process improvements, substitutes
and control systems effective in
preventing and minimizing related
risks. Wastewaters are of prime
concern and warrant reassessment in
terms of both volume reduction and
pollutant removal. Research must
continue to search for effective
and efficient solutions to con-
tamination resulting from specific
industrial sources as well as con-
tamination related to groundwater
and surface run-off.
Advancements
chemistry permit
detect and
micals at
centrations
capability
new and stricter
dards which take
in analytical
scientists to
quantify hazardous che-
extremely low con-
in aqueous media. This
is expected to support
regulatory stan-
into account the
cumulative impacts of chemical
exposure, as well as the potential
for chemical transport and trans-
formation.
499
-------
Wastewater treatment tech-
nologies are required to provide
optimum removal of suspended and
soluble pollutants. Technical
strategies emphasizing the reduc-
tion of pollution before wastewater
generation have been adopted in an
attempt to achieve these higher
performance levels. These strate-
gies include the separation of
fine, and ultra-fine solid products
which were previously wasted in
waterwash operations. These solid
products may be suspended,
emulsified and/or partially solubi-
lized in aqueous media to an extent
that could be deemed significant in
terms of potential hazard and toxic
impact. Most of these washwaters
require chemical addition to
enhance solid agglomeration and
settling before conventional mecha-
nical dewatering systems can be
employed. Unfortunately, the addi-
tion of these chemicals add to the
volume of waste generated.
As an alternative to chemical
conditioning and flocculation,
recent developments indicate that
liquid-liquid and solid-liquid
phase separation can be achieved
using alternating current electro-
coagulation (AC/EC). The AC
electrocoagulator has been used to
flocculate and settle fine solids
without the use of chemical aids
(1,2,3,4). Recent pilot-scale data
(5) demonstrated phase separation
of wastewater containing suspended
and emulsified oils, thus mini-
mizing potentially toxic pollu-
tants. AC/EC may be easily inte-
grated with conventional process
and control systems to enhance
solid product recovery and water
purification. Waste reduction
goals may be accomplished by
integrating this technology with a
variety of operations which
generate contaminated water.
500
This paper discusses the
theory of electrocoagulation with
alternating current. Operating
variables are reviewed and poten-
tial advantages and benefits are
highlighted. The current stage of
development and plans for future
research are discussed in light of
applications dealing with water
purification, wastewater treatment
and site remediation.
STATEMENT OF PROBLEM
Contaminants influence the
physical, chemical and electrical
properties of water. These proper-
ties, in turn, are used to identify
environmental concerns requiring
control to ensure regulatory
compliance. Water is used as a
universal "solvent" and its proper-
ties vary as a function of use.
Solid products subjected to water
wash operations create suspensions
of finely divided colloidal matter
in aqueous media. These wastewa-
ters are generally difficult to
phase separate and the suspended
solids contribute to the loadings
of inorganic and organic pollutants
soluble in the aqueous matrix.
Regardless of origin, waste
water may also contain the various
forms of colloidal matter sum-
marized below:
o solid particles in the form of
colloidais with a mal-
distribution of electrons which
are in magnetic suspension in
the water media.
o solution components present as
water soluble fractions and
suspensions due to magnetic for-
ces.
o chemically stable and soluble
"salts" displaying an inter-
mediate stable existence in the
form of colloidal suspensions of
unstable matter.
o suspended inert matter that is
colloidal as well as susceptible
to precipitation.
-------
Molecular hydrogen bonding is
also a consideration. It has a
major impact on the "bridging"
effect between the water and solid
molecules of aqueous sludge. The
mechanisms influencing the water
associated with solid particles can
be summarized as follows:
o interior adsorption
o surface adsorbtion
o capillary absorption
o interparticle absorption, and
o adhesion water.
Interior and surface adsorp-
tion are referred to as "free"
water which is usually removed by
mechanical techniques. The other
three mechanisms require energy
intensive techniques such as ther-
mal drying for phase separation.
The presence of an electrical
charge on the surface of particles
is often a prerequisite to their
existence as stable colloids. This
surface charge also depends on the
properties of the aqueous phase
because adsorption or binding of
solutes to the surface of the
colloids may increase, decrease, or
reverse the effective charge on the
particle. The adsorption may occur
as a result of a variety of binding
mechanisms: electrostatic attrac-
tion or repulsion, covalent
bonding, hydrogen-bond formation,
van der Waals' interaction or
hydrophobic interaction.
Flocculation and filtration
destabilizes suspended colloids by
enhancing aggregation or the
attachment tendency of these
cotloids.
BACKGROUND AND RELATED THEORY
Several studies (6,7,8,9)
suggest that most solid particles
suspended in aqueous media carry
electrical charges on their sur-
face. When the particles are
larger than atomic or molecular
dimensions, they will tend to
separate from the aqueous media
under gravitational force unless
they are stabilized by electrical
repulsion or other forces. Such
forces can prevent aggregation into
larger particle masses or floes
which are more prone to settling.
These surface charges, may exist as
an ionic double layer or a neutra-
lized electric depole, as concep-
tually depicted in Figure 1.
Generally, the gravitational
force on small particles is weaker
than the other forces which can act
on the particles. Collisions bet-
ween particles due to Brownian
motion often result in aggregates
held together by Van der Waals for-
ces and coagulation may occur in
the following ways:
o The particle crystal lattice may
contain a net charge resulting
from lattice imperfections or
substitutions. The net charge
is balanced by compensating ions
at the surface such as zeolites,
monmorillonite and" other clay
minerals.
o The particle solids may contain
ionizable groups.
o Specific soluble ions may be
absorbed by surface complexes or
compounds formed on the particle
surface.
501
-------
PRINCIPLES OF AC ELECTROCOAGULATION
The electrocoagulator process,
invented by Moeglich, et.al.
(10,11,12) is based on colloidal
chemistry principles using AC power
and electrophoretic metal hydroxide
coagulation. The process employs
two main principles:
o electrostriction, whereby the
suspended particles are stripped
of their charges by subjection
to alternating current electri-
cal field conditions in a tur-
bulent stream, and
o electroflocculation, whereby
minute quantities of metal
hydroxides are emitted from the
electrodes to assist in floc-
culation of the suspended par-
ticles.
The theory of electrofloc-
culation, or metal ion floc-
culation, is well established.
Iron and aluminum ions have been
widely used to clarify water.
Recently, Parekh, et.al. (3,13),
developed a coagulation system
involving the use of metal
hydroxide and fine particles. They
reported that the optimum coagula-
tion of a metal ion/particle system
takes place at the iso-electric
point of the metal hydroxide preci-
pitate. Jensen(14) suggests that
optimum coagulation may not
necessarily occur at the exact
point of zero charge, since other
mechanisms such as bridging are
also important.
A better understanding of the
mechanisms which underly the opera-
tion of AC/EC is expected to result
from research initiated by the
State University of New York at
Buffalo in June, 1989. The current
hypothesis for AC/EC operation is
summarized as follows:
o Polar molecules adsorbed on the
surface of small particles are
neutralized by an equivalently
charged diffuse layer of ions
around the particle. A zero net
change results.
o Non-spherical particles have
non-uniformly distributed
charges (dipoles) and elongated
neutralizing charge clouds
surrounding them.
o These dipoles come into play
when the charge clouds are
distorted by external forces or
close proximity of other charged
particles.
o External forces such as electric
fields can: (a) cause dipolar
particles to form chains; and
(b) unbalance electrostatic for-
ces resulting in dramatic phase
changes (coagulation).
o AC electric fields do not cause
electrophoretic transport of
charged particles, but do induce
dipolar chain-linking and may
also tend to disrupt the stabi-
lity of balanced dipolar struc-
tures.
PROCESS DESCRIPTION
INTRODUCTION
AC/EC system designs will vary
depending on the characteristics
and quantity of waste or process
steams being treated, treatment
objectives and location.
Characteristics, such as particle
size, conductivity, pH and chemical
constituent concentrations, dictate
operating parameters of the coagu-
lator. The quantity and flow rate
of the raw solution will effect
total system sizing, coagulator
502
-------
retention time, and mode of opera-
tion (recycle, batch or
continuous). Treatment objectives
will establish the type of gravity
separation system to use, establish
recovery criteria, identify the
utility of side stream treatment,
define effluent standards to be met
and determine the advantages of
recycle or multiple staging.
Treatment objectives may include
product recovery or simply precon-
ditioning prior to using an
existing process or as a polishing
step after treatment. Location
will impact design by imposing phy-
sical size constraints and pumping
requirements. In-plant industrial
applications, for example, may be
configured differently than a
mobile pn-site system used for
remediation or treatment of ponded
water.
BASIC PROCESS
A basic process flow diagram
for AC/EC is presented in Figure 2.
Coagulation and flocculation occur
simultaneously within the coagula-
tor and in the product separation
step. The redistribution of
charges and onset of coagulation
occur within the coagulator as a
result of exposure to the electric
field and catalytic precipitation
of aluminum from the plate electro-
des. This reaction is usually
completed within 30 seconds for
most aqueous suspensions. The
solution may be transferred by gra-
vity flow to the product separation
step.
Product separation may be
accomplished in conventional gra-
vity separation and decant vessels.
Coagulation and flocculation con-
tinue in this step until the
desired degree of phase separation
is achieved. Generally, the rate
of separation is faster than
methods which employ chemical floc-
culants or polyelectrolytes, and
for some applications the solid
phase is denser than the solids
resulting from chemical treatment.
A recent feasability study (15)
demonstrated 95 to 99.5 percent
recovery of submicron fines from a
0.6 percent stable suspension after
1.5 hours settling time. Alter-
native treatment achieved only 80
percent removal after 1.5 hours.
In many applications, the
electrocoagulator retention time
may be reduced and performance
improved by agitating the solution
as it passes through the electric
field. This turbulence can be
induced by using a static aerator
concept or simply diffusing small
bubbles of air or nitrogen through
the solution in the space between
the plates. Air has been used in
full scale applications treating
pond waters and removing fines from
coal washwaters. Bottled nitrogen
and bottled air have been used in
the laboratory to conduct treatabi-
lity tests. Since the gas used to
create turbulence may also strip
volatile organics, it is necessary
to analyze the vent gas stream,
especially when treating hazardous
wastes. When appropriate, the vent
gases may be collected and treated
using available conventional tech-
nologies and thus control air
emissions within acceptable limits.
After the product separation
step, each phase (oil, water,
solid) is removed for reuse,
recycle, further treatment or
disposal. A typical hazardous
waste decontamination application,
503
-------
for example, would result in a
water phase which could be
discharged directly to a stream or
to a local wastewater treatment
plant for further treatment. The
solid phase, after dewatering,
would be shipped off-site for
disposal, the dewatering filtrate
being recycled. Any floatable
material would be reclaimed, re-
refined, or otherwise recycled or
disposed.
OPERATING REQUIREMENTS
The AC/EC operates on low
voltage, generally below 110 VAC.
It is designed to work at
atmospheric pressure, and is vented
to alleviate any problems with gas
accumulation. As previously men-
tioned, air abatement apparatus may
be added, if necessary.
The internal geometry allows
for free passage of particles less
than 1/4 inch. While normal opera-
tion is relatively maintenance
free, some problems can be encoun-
tered if process upsets allow heavy
particulates to inadvertantly enter
the lines. In this case, material
build-up could restrict passage and
thus retard flow. No permanent
damage has been experienced in
these cases and the problem has
always been alleviated by reverse
flushing or minor disassembly and
cleanout.
While there has been some
question regarding electrode
deterioration, in practice none of
consequence has been noted. Minor
etching occurs on the electrode
skins. As nearly as can be
theorized, the alternating current
cyclic energization retards the
normal mechanisms of electrode
attack that are experienced in DC
systems and reasonable electrode
life has been proven. Electrodes
were replaced after four months ol
continuous operation (20 hours pe»
day) in a 250 gpm commercial unit.
Electrical energy costs v
based on the solution being treatec
and the specific application..
Commercial units have treated coa
wash waters for $0.40 per l.OOol
gallons at power costs of $0.05 perl
kwH. This cost is more than offset!
by savings in the chemical costs!
associated with alternative methods
which require the use of poly-
electrolytes and chemicals tol
adjust pH.
RESIDUAL EFFECTIVENESS
Bench-scale tests (15) and I
full scale field applications (2)
have demonstrated a phenomenon
referred to as residual effec-
tiveness. Once the solution has
passed through the coagulator and
settling is complete in the product
separation stage, the separated
products can be remixed and sub-
sequent phase separation will recur
without further treatment through
the electrocoagulator. It appears
as if the charge redistribution and
coagulating forces remain effective
for extended periods of time. This
phenomenon is important in that
mixing and pumping can be accom-
modated after coagulation, if so
dictated by other system design
conditions, without losing the
phase separation effectiveness.
This also indicates that in some
applications only a portion of the
total contaminated solution would
504
-------
need to be treated. For example,
in removing constituents from a
pond, a portion of the solution may
be treated and returned to the pond
until the desired phase separation
results. Phase separations have
been accomplished by passing as
little as 25 percent of the total
volume through the electrocoagula-
tor.
effluent that
tamination.
resists con-
EFFECTS AND APPLICATIONS
Studies suggest that alter-
nating current coagulation causes
the following effects on the
resulting by-products:
o the magnetic forces associated
with liquid suspensions are
destroyed.
o sludges tend to dewater and den-
si fy, suggesting a disruption
and/or destruction of the hydro-
gen bonding of water molecules.
o electronic or ion exchange
creates an electro-chemical
environment:, causing various
reactions dependent on con-
taminating constituents.
o oils, soap, detergents and
cellulosic material can be phase
separated from water.
o inert clay colloidais can be
removed from aqueous media.
o the generation of OH~ as well as
the potential for 03, H£, Oj and
\\2$2 may influence soluble
pollutants by chemical oxida-
tion.
o soluble oils phase separated
from aqueous media can extract
other toxic constituents which
are preferentially soluble in
the oil fraction.
o water characteristics created by
electrocoagulation and sub-
sequent clarification result in
a long lasting demagnetized
505
The following advantages were
identified as a result of using
AC/EC in the coal industry:
o improved fine coal recovery.
o improved dewatering rates.
o reduced filtration time.
o reduced recirculation of coal
and clay fines in closed loop
water.
o reduced buildup of fines and
clays on dewatering screens.
o neutralization of plant water
pH.
o removal of heavy metal, and orga-
nic carbon from water.
o reduced plant maintenance.
o increased plant availability.
o increased coal yields without
sacrificing quality.
o increased quality at the same or
increased yieldi
o reduced freezing of treated
coal.
Test data such as presented in
Table 1 and 2 support the cited
advantages, as well as many of the
pertinent principles of the tech-
nology.
SUMMARY
The use of alternating current
electrocoagulation to break
emulsions and phase separate
aqueous solutions has been success-
fully demonstrated without using
chemical aids. Based on pilot-
scale results and an assessment of
potential physico-chemical reac-
tions, the applicability of this
technology to various industrial
and hazardous waste management
applications has been identified.
Research activities are underway
and/or planned to further investi-
gate this technology and to better
define applications and benefits.
On-going field demonstrations and
treatability studies of emulsions,
-------
slurries and suspensions from
various industries as well as
hazardous waste sites contribute to
understanding effective operating
parameters. Cost effectiveness is
derived from the reduction or eli-
mination of chemical aids, perfor-
mance improvement of conventional
mechanical separation
an overall reduction
tity of wastes
Applications can be
almost every industrial sector for
a wide range of wastes and in-plant
processes.
systems, and
in the quan-
generated.
found within
Disclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
506
-------
TABLE 1 APPLICATIONS OF ELECTROCOAGULATION
APPLICATION
RESULTS
Reference
1. Participate removal
a. Water from contaminated
soil wash
b. Clay colloids in ponded
water
c. Removal of coal fines
d. Micron size participates
2. Removal of soluble orgam'cs
a. Ponded water
b. Creosote suspension
c. Emulsion of rolling coolant
and waste oils
d. Lake water (DC electro-
coagulation)
3. Metals
a. Ponded water
b. Acid mine wastes
4. Enhanced dewatering
a. Coal fines
>washwater clean enough to
recycle to the ground
>99% removal of suspended
solids
improved performance without
chemicals
99 - 99.5% recovery
>99S removal of TOC*
>98S removal of TOC
>90S removal of TOC
>95% removal ot TOC
>99% removal of Fe, Mn.Al
removal of Fe.Cu, Al
17
1
1 - 18
15
>dewatering rate increased
by 30 - 50%
1
19
1 - 2
* TOC = total organic carbon
TABLE 2 FIELD TEST RESULTS
ELECTROCOAGULATION OF PONDED WATER
Parameter
PH
Suspended Solids
Dissolved Solids
Soluble Iron
Total Iron
Manganese
Aluminum
Alkalinity
TOC
POND A POND
Raw Water After Treatment Raw Mater
6.4 7.7 7.3
197 ppm 1 ppm 195,000 ppm
7,212 ppm
88 ppm 0.13 ppm __
285 ppm 263 ppm 3,500 ppm
(in sludge)
3 ppm 1.9 ppm 104 ppm
304 ppm
48,500 ppm
11,000 ppm
B
After Treatment
8.3
15 ppm
3,344 ppm
_..
0.18 ppm
0.02 ppm
0.08 ppm
400 ppm
30 ppm
Adapted from PI antes, Reference [1]
507
-------
FIGURE 1
SURFACE CHARGE DISTRIBUTIONS
IONIC DOUBLE LAYER
NEUTRALIZED ELECTRIC DIPOLE
FIGURE 2
AC/EC PROCESS FLOW DIAGRAM
RAW
SOLUTION
OR SLURRY
I
VENT OR
TREAT GAS
PRODUCT SEPARATION
AC/EC
COAGULATOR
t
OIL
LIQUID
SOLID
AIR FOR
TURBULENCE
CONTROL
FEED
RATE
508
-------
Machine Coolant Maintenance Leading to Waste Reduction
Barb Loida
Donna Peterson
Terry Foecke
Minnesota Technical Assistance Program
University of Minnesota
Minneapolis, MN
ABSTRACT
The volume of machine coolant waste can be reduced through
maintenance. Coolants are generally considered a waste due to
rancidity, not due to a loss of their cooling properties.
Through a grant study and intern projects administered by
MnTAP the steps for maintaining coolants and the means to implement
them were uncovered. Maintenance involves the removal of tramp
oils from the machine's coolant sump, control of bacterial growth
in the coolant and maintenance of the proper coolant to water
ratio.
INTRODUCTION
There are approximately
14,000 companies performing
machining operations in the
United States, generating sales
of $20 billion per year.
Machine shops make parts from
metal stock for a wide variety
of industrial products, using
processes such as drilling,
turning, lapping, grinding, and
broaching. Machine tools used
in these operations are cooled
by fluids, generally referred
to as "coolants", in order to
extend tool life and enhance
product quality. When these
coolants reach the end of their
useful life, their disposal may
be regulated as a hazardous
waste in some states.
Spent coolants may contain
contaminants which can pollute
groundwater, surface waters,
and upset wastewater treatment
plant (WWTP) operations. These
contaminants may include tramp
oils, high levels of biocides,
heavy metals and nitrates. In
addition, they are generally a
high strength waste as measured
in biochemical oxygen demand
(BOD), chemical oxygen demand
(COD) and total suspended
solids (TSS). Preliminary
studies on the treatability of
coolants indicate that the
material is biodegradable if
sufficient acclimation time is
provided (acclimation time is
the length of time between the
addition of a substance and a
noticeable change in the oxygen
uptake by the organisms). It
should be noted that these
results are believed to be
plant specific and the
acclimation time will vary
depending on the exposure to
high strength industrial waste
the organisms have received in
the past. In addition, the
acclimation time may exceed the
509
-------
detention time of the WWTP and
result in the material passing
through the WWTP into the
receiving water.
The Minnesota Technical
Assistance Program (MnTAP)
became involved with this
wastestream when a number of
generators called regarding
proper coolant management and
disposal. A company approached
MnTAP in 1986 for an intern to
help them establish an
acceptable management strategy
for their coolant waste.
The early part of the
intern project focused on
treatment methods (chemical
splitting technologies) which
would break the emulsion for
each of the two brands of
coolants the company used.
Treatment would generally
result in the water portion
being sewered and the
coolant/oil portion recycled
through used oil haulers.
After repeated efforts at
treatment it was determined
that the ability to split the
coolant emulsion by chemical
treatment was coolant specific
and therefore not broadly
applicable.
During
information
maintenance
reduce the
generated
this project,
about coolant
as a means to
volume of waste
was also emerging.
Since the factors affecting
coolant life were not coolant
specific, the steps required
for coolant maintenance and
implementation were identified
as areas which needed to be
further explored.
PURPOSE
MnTAP wanted shop based
data which would indicate the
steps required to accomplish
maintenance and show how they
could be implemented. The
literature indicated that
coolant life could be prolonged
by removing the tramp oil which
accumulates in the machine's
coolant sump, controlling the
bacterial growth in the coolant
and maintaining the proper
coolant to water ratio. Two
opportunities were presented to
MnTAP to test these principles
in a shop setting. Washington
Scientific Inc. conducted a
year long grant study which was
administered by MnTAP, and
Midwest Electric Products was
selected as a company for a
1988 MnTAP intern project. The
details of these projects and
their results are outlined in
the next section.
APPROACH AND RESULTS
It is important to note
that both companies listed
above approached the projects
with the following elements in
mind:
1) Starting with the lowest
cost techniques and
expanding them as needed.
2) Examining the site
specific problems of
implementation and
modifying the operations
if feasible where needed.
3) Starting with simplest (or
easiest) equipment to
operate.
510
-------
Washington
S c ientific
Industries, Inc.
Background
This machine shop has
approximately 150 employees
doing precision machining and
assembly of motors and drives
for the computer industry and
other specification
manufacturers. A wide variety
of machine tools are used which
have in the past generated as
much as 120 55-gallon drums per
year of waste water-soluble
coolant.
The research project was
designed to study the effects
of maintenance practices and
coolant type on coolant life.
Baselines were established for
coolant maintenance practice,
coolant life, and tool
performance. Coolant life was
found to be as short as two
weeks in some machine sumps,
due to the formation of
hydrogen sulfide by anaerobic
bacteria and, in some cases,
reduced tool performance.
Maintenance involved coolant
removal and sump cleaning as
required because of excessive
hydrogen sulfide odor. A
specific water-soluble coolant
which does not require the
labor and expense of biocide
additions for bacterial control
was substituted for the
coolants used in all operations
except grinding and magnesium
machining. The performance of
this coolant was evaluated
throughout the project for tool
performance, resistance to
bacterial contamination, and
health effects on operators.
Development and Implementation
It took several months to
characterize current practice
and coolant performance. Oil
skimmers, a centrifuge, and
coolant changing/coolant sump
cleaning practices were
evaluated for oil removal
efficiency and effect on
coolant life. Use of the
mobile centrifuge was tested by
cleaning the coolant in the
sumps of a group of 25
machines. It was found that
this is the maximum size group
for one person to maintain,
i.e., the first sump required
cleaning by the time the 25th
was done. Even though the
coolant was cleaned adequately,
this was judged to be an
inefficient use of labor.
At the same time that the
centrifuge was being evaluated,
disk and belt oil skimmers (see
Figure 1) were installed on a
Disc Skimmer
Figure 1
511
-------
study group of five machines
which represented a spectrum of
machining operations and raw
materials. All machines are a
common size, one to four years
old, and used on two-four week
production runs. Operators,
material and processes were
held constant as much as
possible. The project also
evaluated different coolant
change procedures for labor
requirements and effectiveness.
When coolant is changed or
cleaned (oil removal, solids
removal), it is important to
clean the coolant sump to
minimize carry over of bacteria
and other contamination. The
ideal method is to vacuum and
clean the sump until all
contaminants are removed.
However, this can be difficult
in practice because of limited
sump access. Therefore, the
objective of the project was to
determine the best combination
of effectiveness and
efficiency.
Results
The selected coolant, even
after being recycled for over
seven months, met or exceeded
all performance requirements.
There was no incidence of
dermatitis among the operators.
It is important to note that no
biocide additions were required
to maintain this particular
coolant. The combination of
normal pumping agitation and
oil skimming was sufficient to
control bacterial growth.
A coolant change and sump
cleaning practice was developed
which standardized a procedure
to be used with any machine.
It was found that some sumps
are easier to clean than
others, especially when access
covers can be fully removed and
512
corners (along bottom and
sidewalls) are rounded. It is
believed that this change
routine, which requires
approximately five hours per
sump, extends coolant life.
However, it was not evaluated
in isolation from other
changes.
Both disk and belt
skimmers effectively reduced
coolant oil and grease
concentrations. It proved
necessary to install timers and
pumps to reduce the amount of
coolant carried out with the
oil, but these were simple
changes. It was also necessary
to modify several sumps to
provide access for the skimming
equipment. In some cases, a
skimmer was moved from sump to
sump; on other machines, the
installation was permanent.
Both modes of operation were
equally effective in extending
coolant life.
The results of tests
conducted during this project
show that it is possible to
extend the life of machine oil
coolant. Simple skimming of
tramp oil using any of several
different methods was cost-
effective and controlled oil
and grease concentrations
well.
Midwest Electric Products
Project Background
This company, with
approximately 250 employees
manufactures weatherproof
electrical boxes. Many of the
machine operations use a water-
soluble coolant. The company
had asked for assistance in
reducing this waste because of
concerns over increasing
-------
charges for BOD/COD levels, and
the uncertainty about the
continued acceptance of the
wastewater by the city.
The intern project started
with a waste survey, in which
the sites of waste generation
were identified and the waste
quantities from each site
were measured. There were
seven machine sumps generating
batch dumps ranging from 10-40
gallons. In most cases machine
operators dumped the coolant
when it became rancid. The
total annual waste coolant was
estimated to be about 2000
galIons.
Development and Implementation
At one site in the plant,
several machines already were
situated close to each other,
which made the idea of plumbing
their sumps together attractive
as that would reduce the time
required for maintenance. At
the same time that plans were
being made to do this, the
student conducted tests to
evaluate -the effectiveness of
the methods identified in the
literature on prolonging
coolant life. While results
were not always conclusive,
preliminary findings supported
the validity of these methods
(removal of tramp oil, bacteria
control, maintaining proper
coolant to water ratio).
Therefore, plans were
developed for a central sump,
shown in Figure Two, to service
the eleven drilling and tapping
machines. The system was
designed to include periodic
oil skimming, removal of metal
fines and agitation of coolant
once oil .skimming was
completed. The central sump
was fabricated at the company
site, starting with a purchased
cattle trough for the holding
tank. Baffles were welded into
the center of the tank so that
an oil skimmer could remove the
oil at one end of the tank
before the coolant flowed over
the weir to the other end of
I&lit ud
Finn Serin
Central Sump
Figure 2
513
-------
the tank where it is agitated
by a pump prior to
recirculation to the machines.
Agitation was incorporated into
the design so as to discourage
the growth of anaerobic
bacteria. A programmable timer
was purchased so that the disc
oil skimmer would operate only
periodically, thus reducing
quantities of coolant removed
with the oil. Stainless steel
screens were fabricated to
collect fines at each
worktable. In addition a bag
filter placed over the inlet to
the central sump also removes
metal fines. The cost for the
components purchased and the
labor needed to install the
tank was approximately $800.
While agitation, removal
of fines and oil happen
automatically in the central
sump, pH control and bacteria
level require manual
monitoring. Strips are used to
read pH and also bacterial
count weekly. When these
parameters register outside the
expected operating range,
caustic and/or biocide are
added. In addition, coolant
concentration is monitored with
a refractometer, and adjusted
as needed.
Results
In late August, the new
system became operational.
Four months later, the same
coolant was still being used in
the sump. Weekly, the operator
monitors pH, bacterial count
and coolant concentration and
makes the necessary changes.
While this involves some effort
on the part of the operator,
the bonus is that employees
have a more pleasant machining
environment as dermatitis is
not a problem nor is there a
rotten egg smell associated
with hydrogen sulfide
formation. In addition the
wastewater strength has been
reduced by discharging less
coolant waste. Preliminary
results indicate that both the
BOD and COD of the total
wastewater from the company
have been reduced by
approximately 50 percent since
the centralized sump became
operational.
SUMMARY AND CONCLUSIONS
The Department of Energy
has estimated that proper
maintenance has the potential
to reduce the volume of
emulsified oil waste by 80
percent. The steps for
maintenance for one type of
emulsified oil, coolants, are
briefly outlined below:
Properly designed sumps to
give access for
skimming/cleaning
equipment. Sumps should
also be constructed out of
material easily cleaned
and should allow the
coolant to circulate
freely.
Routine sump cleaning
(either chemically or with
steam) needs to be
performed when coolant is
replaced to remove
residual sump bacteria.
Oil skimming should be
performed routinely. The
two most practical
skimming devices for
coolant maintenance are
disc and belt skimmers.
514
-------
Some additional factors to
keep in mind when using
skimming devices include:
- access to the sump will
limit the type of
skimmer which can be
used
- use of timers to
intermittently remove
oil and reduce the
amount of coolant
carried out of the sump.
of
skimmers
>,
- placement
near the sump's pump
since the pumping action
will draw the tramp oil
towards it.
- use of a low speed/
volume pump to reduce
the possibility of
drawing the tramp oil
into the pump.
o Metal chip removal should
be performed routinely to
minimize the machine
fouling and minimizing
sites for bacterial
growth. Screens can also
be used to prevent the
chips from entering the
system.
o Adjusting the pH of the
coolant or addition of a
biocide to control
bacterial growth.
o Refractometers or coolant
proportioners should be
used to maintain the
proper coolant to water
ratio. Both are fairly
inexpensive devices.
Although the maintenance
steps are not coolant
dependent, higher quality
coolant did not require the
addition of biocides or pH
adjustments. In some cases
where tooling machines did not
have access for maintenance
equipment, the higher grade
coolant did have a longer life.
The maintenance steps were also
found to be more streamlined by
using a single coolant in the
shop.
At some point waste
coolant will still need to be
treated for disposal. A number
of treatment technologies exist
including: high speed
centrifuges, chemical splitting
processes, ultrafiltration,
coalescing and reverse osmosis.
These technologies require
further research to determine
their performance.
Field work and other
research directed by our office
has shown that the volume of
machine coolant generated as
waste can be reduced by
implementation of relatively
simple technologies and
techniques. Factors important
to successful implementation
include a limited number of
coolants, tolerance of selected
coolant to a wide variety of
conditions, and careful
solution and equipment
maintenance.
Barriers to implementation
include lack of a willingness
to change established
processes, limited availability
of shop-based testing data, and
initial capital costs,
especially machine tool
modifications. As treatment
and disposal costs continue to
rise, more attention will be
given towards waste
minimization.
515
-------
ACKNOWLEDGEMENTS
MnTAP wishes to acknowledge and
thank the following people:
Laura Newcombe - the 1986
intern student who uncovered
the possibility of maintenance
in prolonging coolant life for
MnTAP.
Joe Pallansch of Washington
Scientific Industries, Inc. -
for his diligent efforts at
exploring the maintenance
options and means to implement
then.
Don Neu - the 1988 intern
student at Midwest Electric
Products for the design of the
central sump and for the
maintenance information he
uncovered.
Ed Grazulis of Midwest Electric
Products - for his support in
showing that maintenance can
work and his support of MnTAP.
Lee Anne Johnson of the
Metropolitan Waste Commission -
for the treatability studies
she conducted on coolants.
MnTAP would also like to thank
the Minnesota Pollution Control
Agency for the program's
funding and the University of
Minnesota's School of Public
Health for its support.
REFERENCES
Cutting
and
Grinding
Fluids;
Selection and
Appli cat i on , American
Society of Tool and
Manufacturing Engineers,
Dearborn, MI, 1967.
7.
"Emulsified Industrial Oils
Recycling", U.S. Department
of Energy - Division of
Energy Conservation, 1982.
Joseph, J.J., Coolant
Filtration, Joseph
Marketing, East Syracuse,
NY, 1985.
Newcombe, L. "Reduction art
Treatment Options for Water
Soluble Coolants,"
Minnesota Technical
Assistance Program, Intern
Report, 1986.
Neu, D. "Maintaining
Coolant Quality to Reduce
Waste," Minnesota Technical
Assistance Program, Intern
Report, 1988.
Pallansch, J.
"Machine
Coolant Waste Reduction By
Optimizing Coolant Life,"
Grant Funded by U.S. EPA -
Administered by Minnesota
Technical Assistance
Program.
"U.S. Industrial Outlook
1988," U.S. Department of
Commerce, 1988.
MnTAP was established at
the University of Minnesota in
1984 as a non-profit, non-
regulatory program
grant from the
Pollution Control
MnTAP's goals are
Minnesota businesses
management and waste
Assistance
telephone
through a
Minnesota
Agency.
to assist
with waste
reduction.
is provided through
consultations, on-
site visits, student
internships, the gathering of
technical information and
acting as a clearinghouse.
516
-------
Disclaimer
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
517
-------
WASTE GYPSUM - ITS UTILIZATION
AND ENVIRONMENTAL IMPACTS
Ryszard Szpadt
Technical University of" Wroclaw
5O-370 Wroclaw, Wybrze±e Wyspiariskiego 27. Poland
Zdzlslaw Augustyn
Wroclaw Geological Enterprise
Koila^taja 24, 50-007 Wroclaw, Poland
W±adysiaw Grysiewicz
Research Center "Hydro-Mech" Kowary
58-S30 Kowary., Poland
ABSTRACT
Waste gypsum , also referred to as phosphogypsum - the
main by-product of the wet process of phosphoric acid manufac-
ture - is amongst the most significant wastes generated by the
chemical industry of Poland. Up till now most of these waste
materials have been disposed of to landfills without talcing any
conventional safety measures. It resulted in a serious pollu-
tion of soil and groundwater in the vicinity of WG landfills.
WG from apatite processing contains
significant amounts of valuable and recoverable rare elements
as yttrium and lanthanide series CY+LSX The total amount of
these elements in one of the WG landfills approaches 9200 tons.
A Joint concept of Y+LS recovery and production of building gy-
psum has been developed. Presented are also the results of pi-
lot studies.
INTRODUCTION
Existing and abandoned in-
dustrial landfills create se-
rious hazards to the environ-
ment. In Poland, most of them
fail to be properly located;
they meet neither legislative
regulations nor environmental
requirements. Many kinds of
wastes are actually suitable
for reuse in the -industry, di-
rectly or after appropriate
treatment.
Among them is also WG ge-
nerated during the manufacture
of wet phosphoric acid.
WG landfills in Poland cover a
total area of dozen hectares
518
-------
and are noxious to their vici-
nity because of dust, emis-
sions, acidic leachates and
runoff.
Systematic stripping of
the landfills. processing and
reuse of the deposited waste
material are very important
from both economical and eco-
logical viewpoints. As a re-
sult. Polish industry obtains
fresh starting materials, and
the landfills receive conside-
rably lower quantities of ha-
zardous wastes. Needless to
say that these lead to a sig-
nificant abatement of environ-
mental pollution in general.
and in the immediate vicinity
of the landfill in particular.
PURPOSE
The purpose of this study
can be itemized as follows:
1. Identify main sources of"
environmental pollution caused
by the WG landfill.
2. Evaluate actual level of
environmental pollution in the
landfill vicinity,
3. Propose proper remedial me-
asures in order to minimize
environmental damage.
4. Evaluate total resources of
Y+LS in the WG landfill.
5. Propose a concept and con-
duct a pilot-scale study on Y+
LS recovery and on the pro-
duction of building gypsum.
APPROACH
Cooperation
This study was conducted
under cooperation of the fol-
lowing institutions:
Institute of Environment
Protection Engineering of the
Wrociaw Technical University,
Institute of Inorganic Che-
mistry and Metallurgy of Rare
Elements of the Wroclaw Tech-
nical University,
Wrociaw Geological Enter-
prise, and
Research Center "Hydro-
Mech" of Kowary.
Wastes
Chemical analyses as well
as pilot-scale studies on Y+LS
recovery were conducted on the
waste material from current
production of phosphoric acid
Ccalcium sulfate hemihydrate
or mixture of hemihydrate and
dihydrate> and on the material
from the landfill Cdihydrate>.
Deposited WG was sampled in
13 boreholes at every 0.5 m of
landfill depth.
Landfill site
In the time spam of ±97O
to 1983 the landfill received
only WG in the form of dihyd-
rate.
Since 1983. disposal has been
carried out in the following
manner: hemihydrate or mixtu-
res of hemihydrate and dihyd-
rate have been placed in the
higher part of landfill, whe-
reas mixtures of partly dewa-
tered wastewater sludge and
sodium fluosilicate sludge re-
moved from a lagoon have been
disposed of to the lower part
of the landfill.
The WG landfill under study
covers an area of ca. 11 hec-
tares. Total volume and weight
of deposited WG have been de-
termined as amounting to 1.6
million cubic meters and 2.O
million tons. respectively.
519
-------
Fig. 1 shows the landfill site
and its vicinity.
The landfill subsoil is a
glacial outwash aquifer- with a
thickness of ca. 7 m. It con-
sists of beds and lenses of
fine to coarse sand and gravel
with thin lenses and beds of
fine to medium sand and silt
inter-bedded with the coarser
material. The glacial outwash
aquifer is underlain by a gla-
cial till.
Groundwater was sampled in
2i monitoring wells located on
the landfill foreland. Soil
samples have been collected
near the monitoring wells.
Procedures
Standard methods were used
for examination of groundwa-
ter, surface water, runoff,
and leachates. Moisture con-
tent in WG was determined by
drying at temperatures of 5O°C
or below in order to avoid de-
hydration and removal of cry-
stallization water from dihy-
drate. Total Y+LS content was
determined by the titration
method with NazEDTA or by co-
lorimetric analysis, using xy-
lenol orange as indicator.
Atomic absorption was used for
the examination of particular
elements.
PROBLEMS ENCOUNTERED
Environmental pollution
in the vicinity of the WG Ian-
fill accounts only for a part
of the overall pollution pro-
duced by the chemical plant
itself. It was practically
impossible to exactly attri-
bute the environmental damage
to respective source having at
hand the existing system of
environmental monitoring.
According to generalized es-
timations, release of noxious
substances with runoff and ie-
achates as well as with dust
blowing from the landfill does
not exceed 2% of the total
emission from the whole plant.
RESULTS
The results of the che-
mical examination of WG. as
well as of the mixture of slu-
dges are presented in Table 1.
The main component of WG -
calcium sulfate dihydrate or a
mixture of hemihydrate and di-
hydrate - approaches for ca.
90 % of dry wt.. The wastes
also contain CaFz. CaaCPO-Oz,
SiOz. NazSiFo. HaPO*, and
HzSO-i.
Rare elements are the most
valuable compounds of the WG.
Yttrium - Y and lanthanide se-
ries - LS CLa, Ce, Pr. Nd, Sm.
Eu, Gd. Tb, Dy, Ho, Er. Tm,
Yb, Lu> defined as the sum of
oxides
Y, 151; La. 1063; Ce. 2142;
Pr, 144; Nd, 887; Sm. 92; Eu.
25; Gd, 86; Tb, 21; Dy, 38;
Ho, 12; Er, 15; Tin, 0.8; Yb,
5.4; Lu, not determined. EIL
A special issue is the
estimation of the radioactivi-
ty level in WG. It is interes-
ting to note that - unlike the
gypsum coming from phosphorite
processing
-------
521
-------
TABLE 1. Average chemical composition or wastes and or their
water extracts
Parameter
Unit Waste gypsum
rrom current rrom
production landTill
Mixture
or sludges
Water content
at 50°G
Water content
at 105°C
Loss on ignition
at 600°C
Calcium
Iron
SulTates
Water-soluble
phosphates,PO*a
Water-soluble
riuorides, P~
YaOa+LSzOa
Total water hol-
ding capacity
CField capacity)
Water retention
capacity
wet wt. %
wet wt. %
dry wt. %
dry wt. %
dry wt. %
dry wt. %
dry wt. %
dry wt. %
dry wt. %
wet wt. %
wet wt. %
20.2
25.3
6.6
22.5
O.O4
58.8
1.15
O.O8
0.62
54.O
28.7
11.2
23.3
6.1
21.2
0.10
53.5
1.39
0.21
O.62
40.O
13.8
61.2
12.8
9.4
1.8
3.73
O.19
Water extracts
pH
Conductivity
SulTates
Fluorides
Phosphates
Calcium
mS/cm
g SCU2/m3
g F'xm3
g PO*Vm
g CaXm
2.6
3.8
2328
57
86O
800
1.7-4.7
2.8
1649
39
974
577
4.1-5.1
2.3
824
67
V fl
146O
88
or dry wt. are related to the
REMARKS: All data expressed in
drying residue at 1O5 C.
Water extracts were prepared by shaking 1OO g or raw wastes with
1 liter or distilled water.
522
-------
TABLE 2. Radioactivity of WQ C33
we
from apatite pro-
cessing:
- raw
- after recovery of
Y+LS
from phosphorite
processing
Radioactivity,
Bq/kg
40 K 21<$ Ra
110
98
113
31
16
575
dry wt.
228 Th
11
16
19
gypsum. This enables a safe
utilization of the recovered
gypsum for building purposes
(Table 2> C31.
The pH of the waste and
of its water extract is stro-
ngly acidic. Both acids - pho-^
sphoric and sulfuric - are
first leached from the wastes
deposited on the landfill.
This process also occured in
the column tests.
The investigations car-
ried out in the vicinity of
the landfill have shown that
the highest degree of soil
contamination is measured im-
mediately at the landfill foot
(primarily at the monitoring
wells P2, P1O, P14 and, also
at P12, P16 and P17>. Those
soils are characterized by in-
creased contents of sulfates,
phosphates, fluorides, cal-
cium, and potassium which a-
re the main components of was-
te gypsum CTable 3>. Slope
runoff and top dust blowing
and deposition have been iden-
tified as the main sources of
soil contamination in the nea-
rest vicinity (in a radius ca.
80 m > of the landfill.
Groundwater in the land-
fill vicinity is highly mine-
ralized Cin terms of calcium
sulfate concentration which,
in some samples, approaches
the saturation level>.
Leachates and excess water
from the sludges deposited in
the lower part of the landfill
(located predominantly in a
sand excavation} have been
considered as the principal
sources of groundwater pollu-
tion.
The remaining are as follows:
runoff from landfill slopes.
roads and neighbouring area as
well as leaching of dust depo-
sited on the soil in the land-
fill vicinity (mainly from the
area of the chemical plant>.
The highest concentra-
tions of pollutants are found
to occur in water samples from
the monitoring W11JL P17
(1903-2131 _ « SO* Smf O.26-
2.79 g P~/m , conductivity of
2.63-2.87 mS/cm, pH of 3,92-
4.31>. Water sampled at the
monitoring wells located on
-------
the foreland of the higher-
part of the landfill Is chara-
cterized by considerably lower
concentrations of main pollu-
tants CO-0.3 g F~/m3, 518-154O
g SO* /m , cond. of 1.O1-2.32
mS/cm>.
Natural, relatively high con-
centration of calcium sulfate
is typical for the groundwater
of this area C224-42S g Sol2/
m , 9S-180 g Ga/ma>,thus only
a part of calcium sulfate can
be attributed to the landfill
impact. Oroundwater also car-
ried Y+LS at total concen-
trations below 1.0 g/m3, exp-
ressed as YzOa + LSzOa CTabie
3>. Because of groundwater
pollution, it. was necessary to
discontinue any use of the
farm wells in the radius of
ca. 1 km from the landfill
either for household or for
farming purposes.
The results show that in-
filtration of rainwater inside
the WO Cin the higher- part of
the landfill) is considerably
decreased by
TABLE 3. Average composition of groundwater
collected in the landfill vicinity [1,41
and soil samples
Monito-
ring pH
Oroundwater
Sulpha- Fluori-
Soil
YzOa + Sulpha- Pluo- YaOa
well
P 1
P 2
P 3
P 4
P 5
P 7
P 8
P 9
P 10
P 12
P 13
P 14
P 15
P 16
P 17
P 18
P 19
P 20
P 21
P 22
P 23
6.5-7.3
5.6-6.8
6.5-6.8
6.4-7.2
6.1-7.0
6.0-6.3
6.3-7.1
5.8-6.7
5.4-5.9
5.8-7.0
5.9-7.1
6.2-6.6
5.5-6.5
5.7-6.3
3.9-4.3
6.0-7.1
4.9-6.5
6.6-7.6
7.3
6.7
6.6
tes
gXm
727
1271
1031
767
216
926
845
1213
1400
883
976
855
1186
1174
2022
1162
949
525
915
425
224
des
g/m
O.09
O.13
0.11
O.13
O.15
0.02
O.18
O.1O
O.O6
0.14
0.17
0.07
0.10
0.42
2.O7
0.10
O.O7
0.12
n.d.
n.d.
n.d.
LSzOa
g/m3
1.10
O.48
O.51
O.64
1.O1
O.59
0.40
0.46
0.91
O.47
0.68
0.38
0.67
0.91
0.38
O.46
0.42
O.30
_
_
—
— ~»
tes
2800
30OO
—
—
_
trace
_
3200
42OO
3300
_
4400
_
2900
_
_
_
_
_
_
_
rides
9
12.4
_
w
10
M
15
28
7.4
9.7
11.2
12.6
*•»
3.6
^
w
LSzOa
800
1200
400
500
500
Rn
_
600
^
— B
—
—
900
600
_
^
n.d. - not detected
-------
Cl> hydx-ation of liemihydrate
to dihydrate which makes a
crust, form on the top and on
the slopes of the landfill,
thus promoting more intensive
runoff,
C2> advantageous properties of
the compacted material . It
is very difficult to determine
the changes in the macrostruc-
ture of the landfill, because
infiltration and leaching oc-
cur predominantly in the mic-
rostructure of the WG. Low
concentrations of phosphates
and fluorides in groundwater
(compared to that of suifates)
should be attributed to two
major factors:
Cl> small infiltration rate,
and C2> chemical precipitation
-------
BOREHOLES
4 —5 —7—8-9
d.w.% = % of dry weight
at 105°C
0.1 0.2
F-, d.w. %
- JWS^S //
0 55 101 139 212 244 311 371 431 486
647
830 980 1038
Distance , m
FIG. 2 DISTRIBUTION OF WATER-SOLUBLE FLUORIDES AND
PHOSPHATES IN LANDFILL PROFILE
526
-------
Component
H20
Solids
with :
Y203+LS203
P205
others
Total
tons
11.50
30.30
0.182
0.52
29.60
41.80
I
0
1
WASTE GYPSUM
^
p41.8 tons
CRUSHING and
MILLING
^ i
r
2| DIGESTION
,154.14 tons
H2SOA.96%
|3| EXTRACTION |
13.06 tons
H20
167.2 tons
41.8 tons
Filtrate and
Ammonia water.25% Washings, 165,1 t
17.63 tons
H20
6.23 tons
FILTRATION and
WASHING
Cake CaSO/, • 2 H2 0, 43.9 tons
Y+ LS
PRECIPITATION
Washings
182.73 tons
Component
Solids
H20
Total
tons
29.4
14.5
43.9
FILTRATION and
WASHING
112.34 tons
Filtrate
70.39 tons
Y*LS concentrate,6,231
Cake
CONCENTRATION,
CRYSTALLIZATION
53.39 tons
Condensate
17.0 tons
(NH/,)2 S04
Component
H20
Solids
with :
LS203
Ca
others
Total
tons
4.98
1.25
0,10
0.10
1.05
6.23
18 DEHYDRATION |
14.5 tons
Calciner gas
Calcined gypsum . 29.4 tons
Component
Y203+LS203
K
Na
CaS04 •
1/2 H20
Total
tons
0.08
0.14
0.03
0.09
29.06
29.40
FIG.3 FLOW CHART AND MASS BALANCE OF WASTE-GYPSUM PROCESSING
-------
made on the basis of pilot-
scale results is given in Fi-
gure 3. Extraction of Y+LS
from WG was carried out using
10 to 12 9£ sulfuric acid Cpro-
duced by the same plant where
this waste is generated). Re-
covered concentrate of Y+LS is
subject to further processing
and refining in order to ob-
tain a useful product for fi-
nal utilization. Overall effi-
ciency of Y+LS recovery which
may be achieved with the tech-
nology proposed varies between
50 and 60 %. Further studies
are necessary to make it rise
to a reasonable level Cabout
80 %> either by improving this
technology or by replacing
sulfuric acid with nitric acid
to extract Y+LS from WG. In
this last process. calcium
nitrate is a major by-product
which can be utilized as a
fertilizer. Recovered gypsum
meets the requirements of the
Polish Standard for binding
material of this kind.
Actually, the chemical plant
in which WG is generated is
not able to implement a tech-
nological line for Y+LS and
gypsum recovery. Implementa-
tion requires an investement
effort which, under today's
economic climate, is not pos-
sible to sustain by the plant
alone. Thus the WO landfill
is temporarily considered as a
prospective source of Y+LS and
gypsum for the Polish indus-
try. Recently, new possibili-
ties of a Joint-venture for WG
processing have been examined.
Short-term operational and
remedial measures have been
proposed in order to elimina-
te Cor, at least, minimize)
further environmental pollu-
tion in the immediate vicinity
of the landfill prior to the
start-up of material stripping
and processing.
These are as follows:
strong compacting of the
WG on the landfill surface in
order to minimize top dust
blowing and to maximize rain-
water retention in the surface
layer to promote more intensi-
ve evaporation,
- current covering of sludges
in the lower part of the land-
fill with waste hemihydrate to
reduce leaching of easily so-
luble pollutants,
- covering of landfill slopes
with earth and stabilized was-
te material from a neighbou-
ring municipal landfill which
is obsolent,
- biological cultivation of
covered slopes,
- construction of a ditch sur-
rounding the landfill slopes
to collect runoff water, and
send it to the industrial was-
tewater treatment plant for
purification before discharge
into a watercourse.
ACKNOWGLEMENTS
This study was part of the
Central Research Programs No.
O3.O8 and No. 03.11 sponsored
by the Polish Government. The
authors are greatly indebted
to their co-workers affiliated
with the cooperating institu-
tions for valuable assistance.
REFERENCES
1. Augustyn,Z., 1987, Reso-
urces of Rare Elements in the
Phosphogypsum Landfill. Wroc-
528
-------
taw Geological Enterprise, CIn
Pollsh>.
2. Grysiewicz.W., A.Ostrowski,
and A.Szczytowski, 1987, Pilot.
Studies on the Recovery of
Rare Elements from Phospho-
gypsum. Report, of the Research
Center "Hydro-Mech" Kowary,
Tech. Univ. of Wroclaw, CIn
Polish>.
3. Kijkowska R., J.Kowalczyk,
C.Mazanek and D.Pawi o v/sJca—JLo—
zlnska, 1988, Apatite Phospho-
gypsum - Material for Obtai-
ning Rare Elements and Gypsum.
Wydawnictwo Geologiczne, War-
sssawa CIn PoiishX
4. Szpadt, R., 1988, Environ-
mental Impacts of the Phospho-
gypsum Landfill. Report of the
Inst. of Env. Prot.
Tech. Univ. of Wroclaw.
CIn Polish>.
Disclaimer
The work described in this paper was not funded by the U.S.
Environmental Protection Agency. The contents do not necessa-
rily reflect the views of the Agency and no official endorsement
should be inferred.
529
-------
OBSTACLES AND ISSUES IN SOURCE REDUCTION OF CHLORINATED
SOLVENTS-SOLVENT CLEANING APPLICATIONS
Azita Yazdani
Project Engineer
Source Reduction Research Partnership
Los Angeles, CA 90054
ABSTRACT
The Source Reduction Research Partnership (SRRP) is a
unique joint venture of The Metropolitan Water District of
Southern California and the Environmental Defense Fund. The
partnership is sponsoring a study to estimate the potential for
source reduction of chlorinated solvents, contaminants commonly
found in ground and surface water system across the United
States.
The project focuses on a particular chemical class, the
chlorinated solvents, and targets principle solvent using
industries including metal cleaning, dry cleaning and textile
processing, electronics, paint stripping and coating, aerosols,
chemical manufacturing and intermediates, and adhesives. The
chlorinated solvents that are addressed in the course of study
are: trichloroethylene (TCE), 1,1,1-trichloroethane (TCA),
methylene chloride (METH), perchloroethylene (PERC), and
trichlorotrifluorocarbon (CFC-113).
The project team has developed detailed industry profiles
that will include a discussion of source reduction measures,
which will reduce solvent use and toxic discharges in a
multimedia fashion. Source reduction options that will be
addressed in the course of the project include chemical and
production substitution, process control and process changes,
and housekeeping measures. Field interviews will also be
conducted to verify information. Finally, the source reduction
potential will be estimated taking into account institutional
and economic considerations.
This paper will present a discussion of the obstacles that
may^be encountered in implementation of the source reduction
options in solvent cleaning applications.
530
-------
INTRODUCTION
Cleaning with
chlorinated solvents is a
common practice in many
diverse sectors of
industry. As much as 311
thousand metric tons (mt)
of chlorinated solvents is
used annually for this
purpose (Yazdani, 1988).
All five major chlorinated
solvents—PERC, TCE, METH,
TCA and CFC-113—are used
in cleaning applications.
Solvent cleaning is
normally required before
assembly, fabricating,
painting or welding. It
can also be used to reduce
contamination in downstream
production processes. It
should be noted that the
solvent cleaning process is
a significant source of
chlorinated solvent
emission and hazardous
waste generation.
Solvent cleaning can
be classified into two
major categories: cold
cleaning and vapor
degreasing.
Cold cleaning is used
to remove drawing
compounds, cutting and
grinding fluids, polishing
and buffing compounds and
miscellaneous contaminants
such as metal chips. Cold
cleaning techniques include
wiping, dipping, spraying,
soaking, swabbing,
immersing, brushing and
ultrasonic agitation. Cold
cleaners are the simplest
type of cleaner. In some
cases, blends of solvents
such as chlorinated
solvents and alcohols may
be considered for cold
cleaning applications.
About one-third of
chlorinated solvents is
devoted to cold cleaning
(Yazdani, 1988).
The vapor degreasing
differs from cold cleaning
in that it is normally
conducted at the boiling
temperature of the solvent.
In general, a vapor
degreaser is a steel tank
with a stream or electrical
heating coil below the
liquid level and a water
jacketed vapor cooling and
condensing zone above the
vapor level. The work piece
is placed in the vapor
zone. Boiling solvent
vapors rise inside the tank
to the level of the primary
condensing coils. Solvent
vapors condense on the
cooler workload as it enters
the vapor zone and dissolves
and removes the
contaminants. When the
surface temperature of the
object reaches that of the
vapor, condensation ceases
and cleaning is completed.
Impurities accumulate in the
sump.
PURPOSE
The chlorinated
solvents, a class of
interrelated chemicals is
currently under regulatory
scrutiny for a variety of
reasons (Wolf et al, 1987).
Chlorinated solvent
cleaning is a source of
release of hazardous
substances into the
environment. There are two
major loss mechanisms in
solvent cleaning
operations. The^
first—emissions5 to the
atmosphere—are process
531
-------
emissions which are usually
significant fractions of
total emission up to 95
percent. The second loss
mechanism is the waste
generation which can
constitute between 5 to 15
percent of solvent loss
(SRRP, 1988). This implies
that virtually all the
solvent used is emitted to
the atmosphere.
Solvent emissions can
occur both directly or
indirectly from all types
of solvent cleaning
equipment. Major causes of
emissions include loss of
solvent from the cleaning
tank due to diffusion,
carry out of solvent on
cleaned parts, and leaks
from tanks and associated
equipment.
The hazardous waste
generated in the cleaning
processes include
contaminated liquid
solvent, bottoms from
stills when on-site
recycling is practiced, and
bottom sludge that arises
in cleaning degreaser
tanks. Nowadays,
generators commonly employ
the services of an off-site
recycler or do recycling
on-site.
In the course of the
study, a variety of source
reduction options for
chlorinated solvents has
been examined in the
solvent cleaning industry.
These fall into the general
categories of chemical
substitution, process
substitution, recovery and
reuse of vapors, recycling
of waste solvent, and
operating practices. The
two most attractive options,
that is: chemical
substitution and process
substitution are presented
in Table 1 and are discussed
in more detail in the
following sections. The
result of the analysis
suggest that such measures
are complex and that
attention need to be paid in
order to avoid unexpected
effects.
APPROACH
In what follows, a
substitution analysis of
chlorinated solvents in
cleaning operations is
presented. This is limited
to the technical suitability
dimension. Other
dimensions, such as cost and
health effects are beyond
the scope of this analysis.
These sections will discuss
the available substitutes,
will specify the
contaminants that can be
removed and the
characteristics that the
alternative solvents must
possess to meet reasonable
standards in cleaning.
Problems Encountered
As Table 1 summarizes,
there are a variety of
source reduction options to
replace chlorinated solvents
in the solvent cleaning
industries. In some cases,
the various options
themselves may have adverse
features which would not
allow for complete
replacement of the
chlorinated solvents in all
sectors.
532
-------
Table 1. Source Reduction Options for Chlorinated Solvents in
Cleaning Applications
Chemical Substitution
0 Low Molecular Weight
Organic Solvents
0 High Molecular Weight
Organic Solvents
0 Chlorofluorocarbons (CFCs)
and Their Blends
Process Substitution
0 Aqueous Cleaning
0 Emulsion Cleaning
A. Chemical Substitution
In principle, an ideal
substitution would consider
all possible substitutes
and compare them along
three dimensions. The
first dimension is
technical suitability. All
candidates should be
compared in terms of their
capability to.accomplish a
specific task. The second
is economic. The complete
cost analysis of a
substitute includes the
cost of raw material,
associated equipment,
process changes, regulatory
requirements and disposal.
The third dimension is
health implications. The
possible consequences of
producing, using, and
disposing of each
candidate. Human health
impacts based on
toxicological data for each
substance would also be
necessary. In this paper,
we focus on the first
dimension, the technical
suitability.
The effects of
substitution are complex
and can be unexpected since
possible alternatives to a
hazardous chemical may
themselves be hazardous but
in a different way, and
there may be no valid way to
compare them.
In cleaning
applications, there are a
number of chemicals that are
technically feasible
substitutes for both cold
cleaning and vapor
degreasing applications
(SRRP, 1988; Yazdani, 1988}.
Low molecular weight
solvents such as aliphatic,
aromatic and oxygenated
solvents are possible
chemical substitutes for
chlorinated solvents. Most
of these solvents show good
solvency characteristics,
however, they have several
limitations such as low
vapor pressure, flammability
and are considered
precursors to photochemical
smog (Table 2 >.
They may be used to
clean surfaces with typical
contaminants susceptible to
solvent removal including
most greases, oils, waxes,
resins, and polymers. These
chemicals are presently used
as simple substitutes for
halogenated solvents in cold
cleaning applications. No
major equipment modification
is required, the solvents
533
-------
«4H t-
•
CO ^
s
re 4->
rH C
(i, .rt
s.
MH
O > -H
> 0)
o in
8 8
r- oo
o o o
VD
o o
VD,
rH rH rH rH
o in co
o oo
r- in o tM
O CN 00 03
CM in n n
O O O O
4-1 4J 4-> 4->
rH ^ rr, 03
O in rH rH
V rH i-H rH
o CM r-
CM 03
o o o o
O O
r~ n
03
rH in rH r-l
in •q- o> co
-
fu
in rH
(Tl O
(0
4J to a>
W M E
O 0)
a) C J-l rH
c a) (d td
(0 M T3 H
4J O T3 tt)
ft ^ o c
a) a) 4-1 -H
3: W co s
-
•H M
4J 10
id u
•r'O
o c c a>
m « « e
rH J3 ft tt)
rH STrH
O O 3 >1
CO EH fi X
to
.1
•H (fl
4J O
(0 O
oo r~ in r» \D
o in vo 03 CM
in 4-1 4-1
M1 O in 03 C—
o in 10 co
nCMPICM rHCMrHrHi-H
tt) tt) tt) tt) tt)
C C C C C
o o o o o
C C C C
a> a> a) a)
T3
rH rH i-H rH O
a> a> 0) 0) o
u o o o cp
w w w w
03
cn
03
cn
33 CJ
EH OS
(0 to
C 4-1
•H C
tj a>
o >
rH rH
•C. O
CJ W
U
U
"
C C
a) (0 -H a>
0) -H
e to
a) c
13 X> >
C .H
0) 0) C O
0) U
0) CO
D (0 tt) >
rH ft CO -H tt)
nj to a) ty ,c
> ,c 4->
T3 4-> tO
0> C -rH tt)
,C -H W rH
4-> rH tt) 4-> -H
X> X C 4J
n to a> rd
tt) tt> -H > rH
4J rH rH O
O (fl
O >
4-> O W
tt) 0) O (0 10
43 C O 0)
E-< a) <0 "H rH
ft O
4-1 Ct)
• O C >i,C
M 4-> 0) 4-> 4-1
tt) > -rl
S 4-1 rH rH »
aC O -rl 4J
0) 10 4-> C
4J rH (1) rH
c o .c o
tt) W 4-> >
rH a) ^ tt) C
O .C tt) > -H
tfl 4-1 4-1 -H rH
4-> 4-1 -H
tt) MH a> (0 o
J3 O Xt rH X)
a
.
O -H 4-> tt) 4->
o xi x a>
•H co a> a> x
ro a> C D -H
o
ra
c w j= xi x;
•H 4-1 +J 4J
O 4J
C 0) h C •>
10 rH tt) (0 •
IM 43 4-J 0)
w tt) a>
13 .c c
Sj C EH W -H
a
g c w c o>
(3 -H -H tt) C
JJ 4J H > "H
(0 4-> XI rH rH
ft tt) a) O '"H
S T) W O
4J 0) tt) X
•H J3
O 8)
rH EH 4-> 10
a) 4->
rH£ O O O
10 tt) 4J 4-) >4
> <0 C tt)
tt) rH O "H 'O
X O rH O t-l
EH to tH Q, 0
534
-------
show high solvency for
common contaminants and
chemical compatibility with
all ferrous and nonferrous
metals (SRRP, 1988).
Organic solvents have
several limitations. They
cannot be used in enclosed
systems because of solvent
vapor build up. These
solvents can not be used in
vapor degreasing
applications because of
their flammability.
Another disadvantage of
these solvents is that they
have short atmospheric
lifetimes and they form
precursors that contribute
to photochemical smog.
High molecular weight
organic solvents such as
terpenes, N-methyl
pyrrolidene (NMP) and
dibasic esters (DBE) are
very new to the solvent
cleaning market and are not
widely used. These
chemicals are considered
biodegradable substances by
their manufacturers and are
rated combustible due to
their flash point by the
National Fire Protection
Association. The flash
point of Bioact DG-1 (a
terpene product marketed by
Petroferm, Inc.) is 47°C
(117°F). These substances
because of their
combustible nature should
be used at room
temperature. Furthermore,
a terpene supplier contends
that the parts need to be
washed with water
afterwards to remove the
terpene residue, which are
too heavy to volatilize.
This in itself may pose
some problems since the
stream may have to be
handled properly (i.e.
pretreated) prior to sewer
discharge. To overcome the
flammability problem,
nitrogen inerting systems
are proposed which will add
to the expense of the unit.
It should be noted that at
this time, not much is known
about the toxicity of these
chemicals and it is not
clear if they can be
recycled for reuse.
The new CFC that is
under investigation as a
replacement for CFC-113 and
other chlorinated solvents
is CFC-123, which has shown
good solvent power,
stability, low surface
tension, and low ozone
depletion potential, because
it contains hydrogen.
CFC-123 has a low boiling
point (27.7°C) which is
advantageous for some
applications such as cold
cleaning but disadvantageous
for others. A consortium of
CFC producers has organized
a toxicity testing program
to test the chemical for
chronic toxicity effects.
New CFCs would not be
commercially available for
the next five years.
Presently there is no
production process for
production of CFC-123 and
there is active research to
identify a process which
gives reasonable yield (CMR,
1988).
Some other solvent
blends such as Freon SMT and
ISC-108 are available as
partial substitutes for
chlorinated solvents since
they contain a small
percentage of chlorinated
535
-------
solvents. The Freon SMT
which has been marketed by
DuPont has 68 percent
CFC-113, 25 percent 1,2
Dichloroethylene and 6
percent alcohol. ISC-108
which has 0.4 percent TCA
by weight in an alkaline
solution with a surfactant,
is considered a
biodegradable compound and
requires neutralization
prior to discharge to the
sewer.
B. Process Substitution
A number of processes
are available and are
potential for chlorinated
solvent degreasing.
Aqueous cleaning is popular
for removing contaminants
from various media.
The water based
cleaning methods usually
employ simple hot water
with some additives in
combination with
mechanical, electrical or
ultrasonic energy. Aqueous
cold cleaning is popular
for removing contaminants
from metal furniture,
fabricated product, and
transportation equipment.
Water, although it may
be used in a variety of
different industries, can
not be substituted for
cleaning of products where
its use may induce material
or substrate corrosion. At
the same time, because of
high surface tension of
water {Table 2), parts with
small interstices may not
be effectively cleaned.
Aqueous cleaning can remove
potentially damaging
chloride residues in
industrial soils that can
not be removed by vapor
degreasing (SRRP, 1988).
The extent to which aqueous
cleaning can replace solvent
applications is not known;
complete substitution is
probably not possible and
applicability needs to be
determined on a case by case
basis.
Water has several
disadvantages. It has a low
solubility for organic soils
such as greases. It
evaporates slowly, it
conducts electricity, it has
high surface tension and it
causes rusting of ferrous
metals and staining of
nonferrous metals. Aqueous
cleaning is likely to leave
residual water on cleaned
parts which can promote
corrosion. Manufacturers of
aqueous systems sometimes
recommend the use of a final
rinse with rinse inhibitor
to resolve this problem.
Drying of parts can be
achieved using a hot air
knife, soak tank or rotary
drum washers (SRRP, 1988).
Also, water which is a
nonhazardous material
becomes hazardous when used
and requires appropriate
handling such as
pretreatment prior to
discharge to the sewer. In
spite of the pretreatment
costs, present equipments
need to also be modified for
use of the aqueous cleaning
process.
Emulsion cleaning is
another possible substitute
for chlorinated solvent
cleaning applications.
Emulsion cleaning is
primarily used in the
process of cleaning metal
536
-------
parts to remove pigments or
impediment drawing
compounds, lubricants,
cutting fluids and metal
chips. Emulsion is not a
substitute for vapor
degreasing since it leaves
a light film residue or oil
onto the parts. This film
may be good for rust
protection or bad for
maintenance and repair
operations that require
parts without oil or
residue. Sometimes
subsequent cleaning with
alkaline cleaners is
necessary.
RESULTS
The chlorinated
solvents used today for
cleaning purposes amounts
to 311 thousand mt. Some
source reduction
opportunities have been
examined in this article.
Chemical substitution
on the average has "medium"
(5-10 years) potential.
Alternative chemical
substitutes probably cannot
replace chlorinated
solvents in all
applications, and such
substitutes may have
drawbacks making one
hundred percent replacement
of chlorinated solvents
unlikely. In many cases,
testing of alternatives,
some of which are new to
the market may extend the
time frame for
substitution. In most
cases, R&D may be required
to thoroughly investigate
conversion potential. We
believe such testing will
result in a medium time
frame for implementation.
Process substitution,
especially conversion to
aqueous and emulsion
cleaning techniques—is
possible where such
conversions are technically
satisfactory. Mechanical
cleaning methods and cold
cleaning with other solvents
are other examples of
process substitutions that
can have limited
applications. Overall, we
believe process substitution
has "medium" potential.
Because equipment conversion
is required to accomplish
process substitution, a 5 to
10 year implementation will
be necessary.
REFERENCES
1. Chemical Marketing
Reporter (CMR), "CFC
Alternate Aimed at Use
in Electronics" Jan.
1988.
2. Morrison, Paul and
Wolf, Katy,
11 Substitution
Analysis: A Case Study
of Solvents," The RAND
Corporation, Journal of
Hazardous Materials, 10
(1985).
3. Source Reduction
Research Partnership
(SRRP), "Solvent
Cleaning Industry,
Profile - Draft," June
1988.
4. Wolf, Katy, et al.
"Chlorinated Solvents:
Market Interactions and
Regulation," The RAND
Corporation, Journal of
Hazardous Materials, 15
(1987).
537
-------
5. Yazdani, Azita,
"Reducing Hazardous
Waste at the Source:
A Unique Approach,"
World Conference on
Industrial Risk
Management and Clean
Technologies, UNIDO,
Vienna, Nov. 1988.
Disclaimer
Ihe work described in this paper was
not funded by the U.S. ESivironmental
Protection Agency. The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
538
-------
Use of EPA's Synthetic Soil Matrix (SSM) in the Evaluation and
Development of Innovative Soil Treatment Technologies
Richard P. Traver, P.E.
Releases Control Branch
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Edison, NJ 08837
M. Pat Esposito
Bruck, Hartman & Esposito, Inc.
Cincinnati, OH 45241
ABSTRACT
This paper reviews the development and use of EPA's synthetic soil matrix
or SSM. This surrogate soil was developed to be broadly representative of
contaminated soils occurring at CERCLA sites. It is composed of natural
soil fractions such as clays, topsoil, sand, gravel, and silt, and is spiked
with a variety of organic and inorganic chemicals known to occur
frequently at CERCLA sites. SSM has been used to test and compare the
treatment efficiencies of various types of soil treatment technologies,
including emerging and innovative systems. Currently, it is being spiked
with gasoline, diesel, and waste oil and used to evaluate the application of
soil washing to the remediation of contaminated soils and fill material at
petroleum underground tank sites where spills or leaks have occurred.
539
-------
BACKGROUND
The USEPA's Risk Reduction
Engineering Laboratory is
currently evaluating several
innovative treatment technologies
for contaminated soil from
Superfund and RCRA sites. Soil
washing, dechlorination, stabili-
zation, vitrification, thermal
desorption, incineration,
biodegradation, vacuum
extraction, and steam stripping
are examples of technologies that
have been recently evaluated.
In 1987, a program was initiated
to develop and produce a standard
soil for treatability testing. The
goal was to develop a consistent
and standard test material that
could be used to test and compare
the treatment efficiency of several
different types of technologies,
including emerging and
innovative treatment systems.
After considering several
alternatives, EPA decided to
develop a synthetic soil matrix
(SSM) that could be easily
reproduced from readily available
stocks of clay, sand, gravel, and
other soil materials. This would
provide a long term source of the
soil for EPA's research purposes
as well as allow other scientists
the opportunity to obtain it or
reproduce it for their own
research programs.
SSM DEVELOPMENT
Development of the SSM began in
the spring of 1987 with a review of
the soil types that typically are
encountered at Superfund sites.
Information on soil type was
gleaned from EPA Records of
Decision, as well as from a study
of eastern US soil composition.
From this review it was
determined that the soil should
have a well-stratified grain size
distribution containing at least
25% fine clay-sized particles, a
moderate cation exchange content
of 30-50 meq/100 g, a total organic
carbon content of 3-6%, a pH of 5-8,
and a 10-20% moisture content.
Following several weeks of bench
scale formulation tests which
sought to optimize the soil for
these parameters, a final formula
for the basic (unspiked) SSM
emerged, as shown in Table 1.
Thirty thousand pounds of the soil
were prepared from these dry
ingredients in the early summer
of 1987 at a sand and gravel quarry
in southwestern Ohio using
standard quarry materials
handling equipment. The soil
(which had a 5% moisture
content) was packaged in 55-
gallon steel drums, 500 pounds to
the drum, and transported to
EPA's Center Hill Research
Facility in Cincinnati. There it
was spiked with additional
moisture and seventeen selected
organic and inorganic chemicals
known to frequently occur in
contaminated soils at Superfund
sites. The chemicals were added
to the SSM at different levels to
produce a series of four spiked
formulations that have become
known as SSM-I, SSM-II, SSM-
III, and SSM-IV. The formula for
each of these SSMs is given in
Table 2.
Each of the SSMs so prepared was
used in the summer of 1987 to test
and compare the treatment
efficiency of five technologies,
namely soil washing,
540
-------
dechlorination, stabilization,
thermal desorption, and
incineration. These technologies
were selected because they were
believed to have the greatest
potential for application to soils
that would be banned from
landfilling as a result of the
Hazardous and Solid Waste Act
Amendments (HSWA).Results of
these tests have been recently
published. 1"6
EDISON
FACILITY
PRODUCTION
In the fall of 1988, EPA moved the
SSM production facility from its
original location in Cincinnati to
EPA's research facility in Edison
NJ. 7 New mixing equipment was
purchased and the system was re-
engineered to optimize the
soil/contaminant blending process
and reduce potential workplace
hazards. This facility was
completed in October of 1988 and is
currently operational. A new
30,000 pound supply of unspiked
SSM is stockpiled at the NJ facility
for use in future SSM-related
studies. The new facility in
Edison consists of three functional
areas or zones:
1. The mixing area (exclusion
zone): This is the area where
chenicals are blended with the
soil. It includes a 10 cubic foot
capacity Marian tilt-tub mixer
with explosion proof motor, access
platform, ventilation system,
bermed floor, work benches,
weighing and measuring
equipment, umbilical air-line
support for four operators in level
B, and other facilities and
equipment ancillary to the batch
production of SSMs.
2. The decontamination area
(contamination reduction zone):
This area acts as a buffer between
the exclusion and clean zones to
prevent the clean area from
getting contaminated.
3. The clean/support area (clean
zone): This area is designated for
all support services. It includes
an office trailer, storage cabinets
for supplies, a storage area for the
clean SSM supply, emergency
equipment, and phones.
EPA has prepared a video,
brochure, and user's manual
covering the operations and
availabililty of SSM from this
facility. 8'10 These resource
materials are available from the
Risk Reduction Engineering
Laboratory in Edison (contact
Richard Traver at 201/321-6677).
MODIFIED SSM FOR THE SITE
PROGRAM
In late 1988, a modified version of
SSM formula IV containing only
chrome, zinc, tetrachloroethylene,
bis(2-ethylhexyl)phthalate, and
anthracene at relatively high
dosage levels was prepared and
used to pretest the efficiency of an
innovative centrifugal plasma
torch thermal treatment system
for contaminated soils. This
technology, under development by
Retech of California, provides both
destruction of organics and
vitrification of residual metals
within the molten soil mass.
For this test, the SSM was
required to have a minimum
analyzable content for each of the
five selected analytes, in order to
541
-------
insure the potential
demonstration of RCRA's 99.99%
destruction and removal efficiency
(DEE) performance standard for
high temprature thermal
treatment devices. The formula
for the SSM used in this program
is given in Table 3. The table
shows the specified minimum
composition of this batch, the
expected concentrations (based on
dosages applied), and the average
actual concentrations achieved as
defined by chemical analysis of 16
samples taken from the batch.
Although the tetrachloroethylene
content of the batch was
considerablly higher than
expected, the rest of the analytes
were close to the expected
concentrations and the batch was
accepted. The high contaminant
levels in the SSM provided wide
windows of opportunity for
evaluating the system's ability to
deliver 4-nines DRE on the
organics and adequate metals
retention or capture in either the
vitrified soil residue or the air
pollution control system-
Testing of the Retech unit with the
SSM was carried out under the
SITE program in the spring of
1989. Results of this test program
have not yet been made available,
but are expected to be released
later this year.
UST SOIL TREATMENT R&D
EPA's Risk Reduction
Engineering Laboratory has been
actively involved at its Releases
Control Branch in Edison, New
Jersey with research and
development efforts to address the
problem of leaking underground
storage tanks (USTs). Under this
effort, EPA is currently evaluating
several corrective action
technologies for cleaning up soil
contaminated by the release of
petroleum products from USTs.
Several key concerns for applying
these technologies to petroleum
contaminated soils are the
effectiveness of the technology to
remove the contaminants from the
soil, the management of the
residuals, and the total system
costs.
As part of the research program,
soil washing technology is being
evaluated to determine the
feasibility of applying this
technology to the cleanup of soils
contaminated with petroleum
products, this process has been
shown to be effective for the
removal of organics and
inorganics common to petroleum
products, namely xylene,
ethylbenzene, toluene, benzene,
and lead.
At the present time, about 6000
pounds of the clean SSM are being
used to prepare a series of new
SSM test soils spiked with
petroleum products for use in the
soil washing feasibility study. The
SSM will be spiked with various
amounts of gasoline, diesel, and
waste oil and used as the test
matrix for bench and pilot scale
demonstration tests. EPA's pilot-
scale mini-soil washer will be
utilized in the study. This unit is
capable of processing 600 pounds
of soil per hour; it was designed
and constructed in 1988 from
state-of-the-art specifications
developed by RREL. The test
program will be carried out at the
Edison NJ facility, and in the field
at UST sites of opportunity.
542
-------
Results of this testing program
should be available in late 1989.
REFERENCES
1. Esposito, M. P. et al. Results of
Treatment Evaluations of a
Contam-inated Synthetic Soil.
JAPCA 39(3):294-304, March 1989.
2. Esposito, M. P., B. B. Locke, J.
S. Greber, and R. Traver.
CERCLA BDAT Standard
Analytical Reference Matrix
(SARM) Preparation and Results
of Physical Soils Washing
Experiments. Presented at the
14th Annual EPA Research
Symposium: Land Disposal,
Remedial Action, Incineration
and Treatment of Hazardous
Waste, Cincinati, Ohio, May 1988.
3. Szabo, M.F., R.D. Fox, R.
Thurnau. Application of Low
Temperature Thermal Treatment
Technology to Hazardous Waste.
Presented at the 14th Annual EPA
Research Symposium: Land
Disposal, Remedial Action,
Incineration and Treatment of
Hazardous Waste, Cincinati,
Ohio, May 1988.
4. Weitzman, L., L.G. Hamel, and
E. F. Earth. BOAT for
Solidification/ Stabilization
Technology for Superfund Soils.
Presented at the 14th Annual EPA
Research Symposium: Land
Disposal, Remedial Action,
Incineration and Treatment of
Hazardous Waste, Cincinati,
Ohio, May 1988.
5. Esposito, M. P. et al. BDAT
Incineration of a Surrogate
Superfund Soil Using a Piolot-
Scale Rotary Kiln System.
Presented at the 3rd Chemical
Congress of North America/195th
American Chemical Society
National Meeting, Toronto,
Canada, June 5-10,1988.
6. Thurnau, R. and P. Esposito.
TCLP as a Measure of Treatment
Effectiveness. In Press, 1989.
7. Rosenthal, S., P. Esposito, and
R. P Traver. Followup Report on
the Continued Use of EPA's
Synthetic Soil Matrix (SSM) as a
Test Medium for New Treatment
Technology Demonstrations.
Presented at the Annual Meeting
of the Air and Waste Management
Association, Anaheim, CA, June
1989.
8. Traver, R. P., S. Rosenthal, and
P. Esposito. Development and
Application of the USEPA's
Synthetic Soils Matrix for
Hazardous Waste Evaluations.
Video produced and released by
USEPA's Risk Reduction
Engineering Laboratory, Edison,
NJ, 1989.
9. USEPA. Synthetic Soil Matrix
Blending System--SSM.
Standardized Test Material for
Innovative Technology Evaluation.
Brochure produced by USEPA's
Risk Reduction Engineering
Laboratory, Edison, NJ, 1989.
10. USEPA. Synthetic Soil Matrix
(SSM) User's Manual. Risk
Reduction Engineering
Laboratory, Edison, NJ, October,
1988.
543
-------
TABLE 1. BASIC SSM COMPOSITION
Soil Component Weight %
Clays
montmorillinite
kaolinite
Sand
Gravel (No.9)
Silt
Topsoil
5
10
31
6
28
20
544
-------
a
g" _
> co £5
S 0 E
co^-S
OT O.F23
IE"*1
— E "cc
— CO •*£
~ E? c
2: o ^
CO 5 ^
.0" ^_
- rt Is
4 B®
^™ C^ t"j
CO °
co 5 5
'E ^
— jo g
^ O) Ct
CO ° E
CO .C g
.S> j|
0
t£
"o
c
o o o o o o
ooooooo oo
cooocMoocM inm
CO CO f- CO CO CM
gooogoo oo
CO CO T™" 00 CO CVI
°00000° gg
CO CO *^~ CO CO CM
o o o o o o
ooooooo oo
COOOCMOOCM inm
co" •* «> „- ^ , co ><
0 "CD 0 £ ® 5
0) NO® 2 ^c-c"
flJo^QrScDcOC: 'Zi-CM
•S|i~t&«4 if?
> ^ O t— LJJ CO I — >I CO — t°~l
f~J r^ t~l ^"^ ^^ c 3 ^^
05 co «« i in
^^ C\l CO ^\l
^- ~ C^J ^^r
o
CM
O a* O
CM . ^^ CM
CO 00^ CO
• co i ^ .
—^-O § ^ .a 2i>
O T§ g'g °-. o
M "^ co «8Q CO Z O^
o w ^5- co -0 c
1 [?li.E ocBw.l
£
-------
TABLE 3. SPECIFICATIONS FOR MODIFIED SSM-IV FOR THE
RETECH DEMONSTRATION
Analvte
Tetrachloroethylene
Anthracene
Bis (2-ethylhexyl) phthalate
Zinc
Chromium
* Based on dosage applied to
Specified
minimum
nnm
1000
6500
2500
22500
1500
soil
Expected
ppm*
2075
7420
3785
26475
1775
Actual
t>r)m**
3277
7361
3702
28306
1898
Disclaimer
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency peer and administrative review policies and approved for
presentation and publication.
546
-------
PILOT-SCALE INCINERATION TESTING OF AN OXYGEN-ENHANCED
COMBUSTION SYSTEM
Larry R. Waterland and Johannes W. Lee
Acurex Corporation
Environmental Systems Division
Mountain View, California 94039
Laurel J. Staley
U.S. Environmental Protection Agency
Risk Reduction Engineering Research Laboratory
Cincinnati, Ohio 45268
ABSTRACT
A series of demonstration tests of the American Combustion, Inc., Thermal Destruction
System was performed under the Superfund innovative technology evaluation (SITE) program.
This oxygen-enhanced combustion system was retrofit to the pilot-scale rotary kiln incinerator
at EPA's Combustion Research Facility. This system's performance was tested firing contami-
nated soil from the Stringfellow Superfund Site, both alone and mixed with a hazardous coal
tar waste (decanter tank tar sludge from coking operations — K087). Comparative performance
with conventional incinerator operation was tested.
Test results show that compliance with the hazardous waste incinerator performance stan-
dards bf 99.99 percent principal organic hazardous constituent (POHC) destruction and
removal efficiency (DRE) and paniculate emissions of less than 180 mg/dscm at 7 percent O2
was achieved for all tests. The Pyretron oxygen-enhanced combustion system allowed in-com-
pliance operation at double the mixed waste feedrate possible with conventional incineration,
and with a 60 percent increase in charge weight than possible with conventional incineration.
INTRODUCTION
Under EPA's Superfund innovative
technologies evaluation (SITE) program,
several innovative waste treatment
technologies are being evaluated to
determine their applicability to Superfund
site waste cleanup efforts. Under this
program, EPA supports the evaluation
while process vendors are responsible for
the treatment process construction and
operation.
A demonstration of the American
Combustion, Inc., (ACI) oxygen-enhanced
burner system (referred to as the
Pyretron Thermal Destruction System)
interfaced with the Combustion Research
Facility's (CRF) rotary kiln incinerator
system (RKS) was performed under the
SITE program. This program was
547
-------
conducted using waste material
excavated from the Stringfellow
Superfund site near Riverside, California.
In most tests the Stringfellow soil was
combined with a high heating value listed
hazardous waste, K087 (decanter tank
tar sludge from coking operations).
PURPOSE
The objective of the demonstration
tests was to provide data to evaluate
three ACI claims regarding the Pyretron
system:
• The Pyretron system with dynamic
oxygen enhancement reduces the
magnitude of the transient high
levels of organic emissions, CO, and
soot ("puffs") that occur with
repeated batch charging of waste
fed to a rotary kiln
• The Pyretron system is capable of
achieving the RCRA mandated
99.99 percent destruction and
removal efficiency (DRE) of principal
organic hazardous constituents
(POHCs) in wastes incinerated at a
higher waste feedrate than
conventional, air-only, incineration
• The Pyretron system is more
economical than conventional
incineration
This paper addresses only the first
two objectives
APPROACH
As noted above, the demonstration
tests were performed using a prototype
Pyretron system retrofitted to the rotary
kiln incineration system (RKS) at the
CRF. A simplified schematic of this
system is given in Figure 1.
Venturi
inlet duct
filte
Cyclone Packed
separator tower
scrubber
Reclrculatlon
pump
Reclrc illation
tank
Figure 1. CRF rotary kiln system.
548
-------
The prototype Pyretron system
retrofitted to the CRF RKS consisted of
the following: a propane-fired burner
installed at the waste feed end of the
RKS kiln; a similar burner in the RKS
afterburner; gas metering and control
assembly (valve trains) for controlling
propane, air, and oxygen flows to both
burners; an oxygen supply consisting of a
trailer-mounted liquid oxygen tank with
evaporator; and a system for injecting
water into the kiln to afford additional kiln
temperature control.
The waste incinerated in these tests
consisted of contaminated soil from the
Stringfellow Superfund site mixed with
the listed hazardous waste K087. The
mixed waste contained 60 percent K087
waste and 40 percent Stringfellow soil.
The K087 waste was included in the test
mixture to give it heat content and
thereby present a challenge to the
incineration process. In addition, the
K087 contributed to the POHCs in the
waste mixture. The K087 waste
contained several polynuclear aromatic
hydrocarbon compounds in percent
quantities. Six of these, naphthalene,
acenaphthylene, fluorene, phenanthrene,
anthracene, and fluoranthene were
selected as the POHCs for the test
program.
For these tests a fiber pack drum
ram feeder system was used to feed
waste to the kiln. This system feeds 5.7-L
(1.5-gal) fiber pack drums in a cyclical
batch charge operation. Drums contained
between 4.1 and 7.9 kg (9 to 17 Sb) of
waste depending on the specific test
underway.
Six tests were performed to supply
data to evaluate the ACI claims. Since
the ACI claims state that the Pyretron
system offers superior performance when
compared to conventional incineration,
one set of operating conditions reflecting
the limit of the capabilities of conventional
incineration in terms of waste batch
charge mass and total waste mass
feedrate was tested.
The capability limits for conventional
incineration were defined via several
scoping tests. These tests confirmed that
a waste feed schedule of 10.9 kg (24 Ib)
every 10 min resulted in unacceptable
transients in kiln exit flue gas CO levels.
These transient CO puffs survived
passage through the afterburner and
gave unacceptable CO spikes at the
stack. A waste feed schedule of 9.5 kg
(21 Ib) every 12 min resulted in
acceptable incinerator operation. This
feed schedule was defined to be the
capability limit of conventional
incineration and was denoted the
optimum conventional operating
condition. Two emissions tests
(replicates) were performed at this
condition.
The other four tests were performed
with the Pyretron system in operation.
The optimum conventional operating
condition was repeated with the Pyretron
system. Then a waste feed schedule of
15.5 kg (34 Ib) every 19.5 min was tested
with the Pyretron system to evaluate the
ACI claim that the Pyretron system can
reduce the magnitude of transient puffs.
Finally, a waste feed schedule of 9.5 kg
(21 Ib) every 6 min was tested with the
Pyretron system to evaluate the ACI
claim that higher waste feedrates would
be possible with the Pyretron system.
Two tests (replicates) were performed at
this last test condition as well.
Table 1 summarizes the incinerator
operating conditions tested.
RESULTS
Figure 2 plots the variation in
incinerator operating parameters for the
conventional incineration attempt to feed
549
-------
TABLE 1. AVERAGE INCINERATOR OPERATING CONDITIONS FOR THE TESTS
PERFORMED
Waste
Operation
Charge Charge Feedrate,
weight, Interval, kg/hr
kg (Ib) minutes (Ib/hr)
Kiln exit Afterburner exit
Flue Flue
gas gas
O2 Temperature O2
Temperature
°C °F
Conventional, scoping
Conventional, optimum
Conventional, optimum
replicate
Pyretron, at conventional
optimum
Pyretron, increase charge
mass
10.9
9.5
9.5
9.5
15.5
(24)
(21)
(21)
(21)
(34)
10
12
12
12
19.5
65.5
47.7
47.7
47.7
47.7
(144)
(105)
(105)
(105)
(105)
1,027
954
921
1,035
963
(1,880)
(1,750)
(1.090)
(1.895)
(1,765)
9.9
13.3
12.8
17.6
14.5
1,121
1,121
1,121
1,121
1,121
(2,050)
(2,050)
(2,050)
(2,050)
(2,050)
6.4
7.7
7.4
15.2
15.0
Pyretron, optimum
Pyretron, optimum
replicate
9.5 (21)
9.5 (21)
6
6
95.5 (210)
95.5 (210)
979 (1,795)
979 (1,795)
13.9 1,121 (2,050) 14.0
14.6 1,121 (2,050) 15.3
'Replicate tests.
1.200
960-
800
800-
0
t
70-
,-. 270-
30
9.
f^/^m^M^^
j—LJ
11
,0
—r~
12
,
16
J«wJ
tf '
TW« of Day (hr)
Test
Stop
18
Figure 2. Klin data for the conventional incineration scoping tests: 65.6 kg/hr
(10.9 kg every 10 mln).
550
-------
10.Q kg (24 Ib) of mixed waste every 10
min. The figure shows that, early in the
test period, kiln exit temperature varied
from about 870 to 980°C (1,600 to
1,800°F) over a charge cycle. Kiln exit O2
ranged from about 7 to 16 percent O2
over a cycle, and kiln exit CO levels were
generally low. However, intermittent CO
spikes up to 2,200 ppm occurred. As this
test proceeded, kiln temperature
increased such that, after about 3 hours
of operation, kiln exit temperature was
ranging from 980 to over 1,150°C (1,800
to over 2,100°F) over a charge cycle. Kiln
exit flue gas O2 peaked at about 15
percent just prior to initiating a batch
charge, but decreased to 0 as the puff of
volatilized waste from a charge filled the
kiln. Kiln exit CO levels peaked at about
3,000 ppm under these depleted O2
conditions. These CO puffs survived
through the afterburner and resulted in
CO peaks of above 100 ppm at the
stack.
In contrast, operating conditions for
conventional operation were much more
controlled with a waste feed schedule of
9.5 kg (21 Ib) every 12 min, as shown in
Figure 3. At stabilized operation, kiln exit
temperature ranged from about 900 to
1,080°C (1,650 to 1,970°F) over a charge
cycle. Kiln exit CO peaks were less than
about 50 ppm with the one exception, a
spike early in the test. These were
reduced to less than 10 ppm at the stack
after passage through the afterburner.
Figure 4 shows the variation in
operating parameters for the Pyretron
system test at an increased charge mass
of 15.5 kg fed every 19 min. For this test,
average kiln exit temperature was
comparable to the conventional
incineration optimum condition test at
about 960°C (1,765°F), though
„ l.*<™ _
w|_
^fc? 900 -
Ul a.-—"
1 800
8~ 32°° "
5~ 800 -
0
£ .vm 15 '
UJ ^-'
O M
0
_ 270 -
-*C1 90-
" 30
l«^fe 9-
S.2-6
^AA/VVV/\A/W\AMAA^^
i .
WVVV^YVVVVYV^^
Uv.i_r r „ r,,-^-
-^^
•~^....._ •». >.._.^..,^^»..^.^^^-^~.<— ^^_ >^^^^>-p^-..^».<-
^J
m
f
J
i
Test
Start
Tint* of Day (hr)
Stop
Figure 3. Kiln data for the optimum conventional incineration tests: 47.7 kg/hr
(9.5 kg every 12 min).
551
-------
1.200
960-
800
o aoa
"I
a*-* 200
0
70-
_. 270-
wVA/vvAr/vW'^^^
Instrument
Calibration
13 Test
start
—I—
15
T
17
Test
stop
Ttm« of Day (hr)
Figure 4. Kiln data for the Pyretron system test at Increased charge
mass: 47.7 kg/hr (15.5 kg every 19.5 mln).
temperatures as low as 870°C (1,600°F)
and as high as 1,065°C (1,950°F) were
routinely experienced. Kiln exit flue gas
O2 generally ranged from about 13 to
about 19 percent over a charge cycle.
However, kiln exit flue gas CO was
generally below 10 ppm. This test clearly
established that a 60 percent increase in
waste batch charge mass (9.5 to 15.5 kg)
over the limit of conventional incineration
was possible with acceptable emissions
transients with the Pyretron system.
Figure 5 shows the variation in
operating parameters for the Pyretron
system test with a feed schedule of 9.5
kg (21 Ib) every 6 min. This represents
double the feedrate achievable under
conventional operation. As shown in the
figure, average kiln exit temperature was
about 980°C (1,795°F) with routine
variations from about 925°C (1,700°F) to
about 1,035°C (1,900°F). Kiln exit flue
gas O2 generally ranged from 11 to 17
percent. Kiln exit flue gas CO peaks of
about 100 to 300 ppm occurred when kiln
exit O2 fell below about 10 percent.
However, for other than these periods,
CO levels in the kiln exit flue gas were
usually about 30 ppm. This test clearly
shows that a waste feedrate double that
possible with conventional incineration
can be achieved with acceptable
emission transients with the Pyretron
system.
Table 2 summarizes the DREs
achieved for the POHCs in the mixed
Stringfellow soil/K087 waste at a location
in the flue gas that would correspond to
the stack discharge from a typical
industrial rotary kiln incinerator. This
location is at the packed tower scrubber
discharge at the CRF. None of the
POHCs designated for these tests were
detected in the flue gas at this location.
The DREs noted in Table 2 reflect
method detection limits.
552
-------
j, 1,200
3_
£? 960-
J O.—*
I? BOO
800.
SI
fi-3 200-
70-
0
270
30
11
^^/>V/\^Vv'^^^v^^
~^-w?v»p^^
iiufc»Yyrv^^w'w--wr*
U.>^J.Jl>N/HJV*_A99.99934
>99.99936
>99.99924
>99.9964
>99.99896
>99.99978
>99.99989
>99.99990
>99.99991
>99.99990
>99.9975
>99.9976
>99.99958
>99.99961
>99.99965
>99.99961
>99.9945
>99.9947
>99.9985
>99.9986
>99.9951
>99.9953
>99.99914
>99.99919
>99.99930
>99.99923
>99.99976
>99.99977
>99.99980
>99.99978
>99.99922
>99.99927
>99.99935
>99.99929
>99.9979
>99.9980
>99.9967
>99.984
>99.9955
>99.99910
>99.9933
>99.968
>99.9937
>99.9981
>99.9984
>99.9920
>99.9979
>99.99948
>99.9945
>99.974
>99.9933
>99.9983
>99.9974
>99.988
>99.9960
>99.9989
>99.99949
>99.99952
>99.99953
>99.99948
553
-------
As shown in Table 2, DREs at the
scrubber discharge were greater than
99.99 percent for all POHCs. In many
instances, detection limits allowed
establishing DREs greater than 99.9999
percent for POHCs at higher waste feed
concentrations. Since all POHC DREs for
all tests were >99.99 percent, no relative
statement concerning conventional
incineration performance compared to
Pyretron system performance is possible.
The good DRE performance in all tests is
understandable since all tests were
performed at relatively high kiln and
afterburner temperatures.
Particulate levels in the scrubber
discharge flue gas were in the 20 to 40
mg/dscm at 7 percent O2 range
regardless of test conditions. These
levels were below the incinerator
performance standard of 180 mg/dscm at
7 percent O2.
The composite scrubber blowdown
liquor and kiln ash samples from each
test were analyzed for the test POHCs
and other Method 8270 semivolatile
organic hazardous constituents and none
were detected. Since semivolatile
organics were not detected in any
residual sample, firing mode
(conventional versus three conditions of
Pyretron operation) had no measurable
effect on residue composition.
CONCLUSIONS
The objective of the demonstration
tests was to provide the data to evaluate
the three ACI claims regarding the
Pyretron system discussed in the
Introduction, although only two objectives
are addressed in this paper.
With respect to the first ACI claim,
test results are inclusive. Initial scoping
tests confirmed that a waste feed
schedule of 10.9 kg (24 Ib) every 10 min
(65.5 kg/hr (140 Ib/hr) total feedrate)
gave unacceptable operation under
conventional incinerator operation. The
Pyretron system was capable of
acceptable operation at increased
charged mass of 15.5 kg (34 Ib) but
charge frequency was decreased to
every 19.5 min. Thus, total feedrate was
decreased to 47.7 kg/hr (105 Ib/hr). Kiln
exit flue gas CO levels under this
condition were comparable to
conventional incineration operation at the
same overall waste feedrate.
With respect to the second ACI
claim, test results clearly indicate that
99.99 percent POHC DRE was achieved
with the Pyretron system with waste
feedrate doubled over the limit
established under conventional operation.
Acceptable operation with the Pyretron
system was achieved at a feed schedule
of 9.6 kg (21 Ib) every 6 min, or 95.5 kg/
hr (210 Ib/hr). Greater than 99.99 percent
DRE for all POHCs, and particulate
emissions of significantly less than 180
mg/dscm at 7 percent O2 were
measured.
Disclaimer
The work described in this paper was funded by the U.S. Environmental Protection
Agency. The contents do not necessarily reflect the views of the Agency and no official
endorsement should be inferred.
554
-------
THE PARTITIONING OF METALS IN ROTARY KILN INCINERATION
Gregory J. Carroll, Robert C. Thurnau and Robert E. Mournighan
U.S. EPA, Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
Larry R. Waterland, Johannes W. Lee and Donald J. Fournier, Jr.
Acurex Corporation
Mountain View, California 94039
ABSTRACT
fate of trace metals in rotary
tower-scrubber particulate- and
using a factorial experimental
among kiln ash, scrubber water,
Synthetic waste formulations
This research project investigated the
kiln incineration with venturi- and packed
acid gas-control. A test plan was developed,
design, to study the partitioning of metals
flue gas particulate, and flue gas vapor. .
included the hazardous trace elements arsenic, lead, cadmium, chromium, and
barium, as well as the non-hazardous trace elements copper, magnesium, bismuth
and strontium, spiked into a clay absorbent material. The independent
variables chosen for evaluation were chlorine content of the feed, kiln
temperature, and afterburner temperature.
Cadmium, lead and bismuth appeared volatile over the range of kiln
temperatures tested, while the other six metals displayed refractory
properties. Of the three independent variables tested, feed chlorine content
had the strongest effect on changes in metal partitioning across the tests; as
chlorine content increased, metal volatilization appeared to increase, while
scrubber efficiency for metals decreased.
INTRODUCTION
Metals, such as arsenic, barium,
beryllium, cadmium, chromium, lead,
mercury, nickel, and zinc are of
concern in waste incineration because
of their presence in many hazardous
wastes and because of possible
adverse health effects from human
exposure to emissions. Incineration
will change the form of metal
fractions in waste streams, but it
will not destroy the metals. As a
result, metals are expected to emerge
from the combustion zone essentially
in the same total quantity as the
input. The principal environmental
concern centers around where and in
what physical or chemical form the
metals end up in the combustion
system, i.e., bottom ash, air pol-
lution control device (APCD) resi-
dues, or stack emissions.
Most interest has traditionally
focused on stack emissions of metals.
However, increasing attention is
being given to the quality of resi-
duals from incineration of metal-
bearing wastes since disposal of
these materials may be subject to
restrictions on land disposal under
555
-------
the Hazardous and Solid Waste Act of
1984 (HSWA). (1)
A major role of the U.S. EPA
Combustion Research Facility (CRF) in
Jefferson, Arkansas is to perform
research in support of regulatory
development by the U.S. EPA Office of
Solid Waste (OSW). Accordingly, the
CRF has conducted research to study
the partitioning of trace metals
during the incineration of a metal -
bearing surrogate waste. (2)
The testing described herein
took place in the Rotary Kiln System
(RKS) of the CRF, utilizing venturi-
and packed tower-scrubbing for par-
ticulate- and acid gas-control. This
research is part of a larger test
program designed to evaluate the fate
of metals in incineration using a
variety of APCDs.
PURPOSE
Field studies to date indicate
that emissions of metals from incin-
erators burning wastes of relatively
low metal content probably do not
pose an unacceptable level of risk.
However, as the level of metals in
wastes which are incinerated rises,
there is concern that some inciner-
ators could create unacceptable risk.
Current regulatory thinking suggests
implementation of metal emission
limits and metal feed-rate limits,
calculated using health risk levels
and dispersion modeling.
Information gathered from the
subject research will be used to
further examine the impact of metals
upon environmental emissions and to
assist in developing and refining
regulatory strategies for dealing
with such an impact.
The research was designed to
identify:
the distribution of metal emis-
sions among ash, scrubber blow-
down water, flue gas parti cul ate
and flue gas vapor
- the sensitivity of metal fate to
RKS operating conditions
the dependence of metal emissions
on chlorine content of the incin-
erated organic waste.
This testing was conducted in
conjunction with the development of a
metal partitioning model under a U.S
EPA contract with EERC Corporation;
data from these tests will be used in
part to evaluate the predictive capa-
bility of the model.
APPROACH
As indicated in Figure 1, the
RKS consists of a rotary kiln primary
combustion chamber followed by an
afterburner. Combustion gases exit-
ing the afterburner are quenched
after which they enter a primary air
pollution control system. This
system consists of a venturi scrubber
and packed column scrubber, and is
followed by secondary air pollution
control consisting of a carbon bed
adsorber and a high efficiency
particulate filter.
The feed material consisted of
synthetic waste formulations mixed
into a clay absorbent material. In a
four-week series of eight parametric
tests, aqueous mixtures were spiked
into the solid material, which was
then screw-fed to the RKS.
The tests examined the fate of
the hazardous constituent trace ele-
ments arsenic (As), lead (Pb), cad-
mium (Cd), chromium (Cr), and barium
(Ba). In addition, the non-hazardous
elements, copper (Cu), magnesium (Mg),
bismuth (Bi), and strontium (Sr) were
spiked into the test feed to
establish whether their discharge
distributions were comparable to
those for the hazardous elements.
The feed concentrations of
metals are presented in Table 1.
556
-------
ATMOSPHERE
AFTERBURNER
PROPANE
TRANSFER
DUCT
CARBON BED HEPA
ADSORBER FLTER
PACKED TOWER
SCRUBBER
CYCLONE
SEPARATOR
FRESH
PROCESS
WATER
10 FAN
TO DRAM
SLOWDOWN
COLLECTION OR
DISPOSAL
RECIRCULATION RECIRCULATION
PUMP TANK
Figure 1. Schematic of the CRF Rotary Kiln System
These values represent the sum of
metals in the spike solution and
background levels of metals in the
feed clay. In addition to the
metals, toluene, chlorobenzene, and
tetrachloroethylene were added to the
solid material in predetermined
amounts to achieve the necessary
chlorine levels and heat content in
the feed.
Kiln temperature, afterburner
temperature and the chlorine content
of the organic liquid were varied
based on a factorial experimental
design for three variables over three
levels.
The target conditions for the
three variables covered the following
ranges: feed liquid matrix chlorine
content (0 to 8 percent); kiln temp-
erature (816° to 927°C [1500° to
1700°F]); and afterburner temperature
(982° to 1204°C [1800° to 2200°F]).
Table 2 gives the values for each of
the test variables for seven test
points specified by the factorial
Table 1. Nominal feed metal concen-
trations.3
Metal Concentration (ppm)
Arsenic
Barium
Bismuth
Cadmium
Chromium
Copper
Lead
Magnesium
Strontium
50
50
180
10
90
500
50
17,000
300
aBased on average clay matrix metals
concentrations of: Bi (12 ppm);Cr (53
ppm); Pb (3 ppm); Mg (2.2 percent);
Sr (34 ppm)
557
-------
design algorithm. The eighth test
point represents a duplicate of test
point'4. This duplicate test condi-
tion was added to provide information
on test measurement precision.
Table 2. Parametric test design con-
ditions.
Test
1
2
3
4
5
6
7
8a
Feed Cl
Content
(wt %)
0
4
4
4
4
4
8
4
Kiln Exit
nTemfi>
°c m
871
(1,600)
816
(1,500)
927
(1,700)
871
(1,600)
871
(1,600)
871
(1,600)
871
(1,600)
871
(1,600)
Afterburner
Exit Temp,
°C (°F)
1,093
(2,000)
1,093
(2,000)
1,093
(2,000)
1,093
(2,000)
1,204
(2,200)
982
(1,800)
1,093
(2,000)
1,093
(2,000)
aTest 8 is a duplicate of test 4 .
Actual test conditions achieved
were very close to target test
conditions for all test points except
point 2. The average afterburner
exit temperature was approximately
22 C lower than the target temp-
erature of 1093°C for this test. For
all other parametric tests, average
actual operating temperatures were
within 10°C of target conditions.
Actual feed chlorine content was
0 percent for test 1; 8.3 percent for
test 7; and ranged from 3.4 to 4.6
percent for tests 2 through 6, and
test 8.
All tests were conducted at the
same nominal kiln exit flue gas
oxygen (12%), afterburner exit flue
gas oxygen (7.5%), and synthetic
waste feedrate (63 kg/hr L140 lb/hr]),
of which 18 kg/hr (40 Ib/hr) was the
organic liquid matrix.
For all
tional speed
0.2 rpm, and
operated at
tests, the kiln rota-
was held constant at
the primary APCDs were
the following nomin-
al settings: scrubber blowdown rate
(2 L/min); venturi liquid flowrate
(75 L/min); venturi pressure drop
(6 kPa); packed tower liquid flowrate
(110 L/min).
Sampling for each test included:
- a composite sample of all feed
material (clay/organic liquid mix-
ture just before kiln introduc-
tion, and aqueous metal spike
solution)
- samples of scrubber blowdown water
over time
- composite samples of kiln ash from
each test
- samples of the flue gas at the
afterburner and scrubber exits for
metal capture in the particulate
and backup impinger train
- samples of the flue gas at the
afterburner and scrubber exits for
volatile organic hazardous consti-
tuents. (2)
RESULTS
Table 3 represents a summary of
metal discharge distribution data
collected during the eight parametric
tests.
Metal Volatility
Metals exiting the kiln followed
558
-------
Table 3. Normalized metal discharge distributions (% of total measured).
Test Number:
Primary
Variable:
Target Value:
Test Average:
Constants 1;2
Const. 1 Avg.
Const. 2 Avg.
Arsenic
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Barium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Bismuth
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Cadmium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Chromium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Copper
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Lead
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Magnesium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Strontium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
2
Kiln
816
825
4
Exit
871
875
Afterburner
1071
3.7
94.4
2.1-2.9
2.7
74.3
3.8
21.9
25.8
41.5
32.6
<15.2
43-49
42-51
94.7
3.0
2.3
84.2
12.9
3.0
12.6
50.4
37.0
99.4
0.2
0.4
82.9
1.1
16.0
1088
3.8
86.1
3.8-5
8.2
79.6
2.2
18.2
22.2
41.1
36.7
<10.3
56-61
34-39
94.1
2.0
3.9
75.8
15.1
9.1
15.0
48.9
36.1
99.3
0.1
0.6
93.0
1.7
5.3
8
Temperature
871
876
=1093°C; Cl
1093
4.6
92.3
.8 2.3-4.1
3.6
69.9
5.5
24.7
30.0
35.2
34.7
<12.9
42-45
45-55
85.9
1.9
12.2
76.2
17.8
5.9
13.7
50.2
36.0
99.3
0.2
0.5
89.8
3.5
6.6
3
6
4
(°C) Afterburner
927
927
982
984
1093
1088
8
5
Temperature (°C)
1093
1093
1204
1196
=4S Kiln= 871°C; C1=4S
1092
4.2
84.0
6.8-8.4
7.6
69.6
1.6
28.8
22.9
50.7
26.3
<10.7
62-6S
27-31
95.3
2.1
2.6
82.3
14.1
3.6
10.4
67.2
22.4
99.5
0.1
0.4
90.1
1.6
8.3
875
3.4
93.6
2.6-3.
2.6
85.2
2.2
12.5
20.9
47.4
31.6
<13.9
61-69
25-31
95.5
1.1
3.4
79.2
15.2
5.6
5.8
60.6
33.6
99.3
0.1
0.6
94.3
1.3
4.4
878
3.8
86.1
8 3.8-5
8.2
79.6
2.2
18.2
22.2
41.1
36.7
<10.3
56-61
34-39
94.1
2.0
3.9
75.8
15.1
9.1
15.0
48.9
36.1
99.3
0.1
0.6
93.0
1.7
5.3
873
4.6
92.3
.8 2.3-4.1
3.6
69.9
5.5
24.7
30.0
35.2
34.7
<12.9
42-45
45-55
85.9
1.9
12.2
76.2
17.8
5.9
13.7
50.2
36.0
99.3
0.2
0.5
89.8
3.5
6.6
871
3.6
91.2
3.0-4.3
4.6
86.9
1.6
11.5
30.1
37.1
32.8
<14.5
55-62
31-38
89.3
4.2
6.6
75.1
16.0
8.9
13.8
45.0
41.1
99.2
0.1
0.7
81.9
3.7
14.4
559
-------
Table 3. (Cont'd).
Test Number:
Primary
Variable:
Target Value:
Test Average:
Constants 1;2:
Const. 1 Avg.:
Const. 2 Avg.:
1
Feed
0
0
Kiln
872
1094
4
Chlorine
4
3.8
* 871°C
878
1088
8
Content
4
4.6
7
(wt %)
8
8.3
; AB = 1093°C
873
1093
871
1092
Arsenic
Kiln Ash 93.9 86.1 92.3 92.4
Scrub.Ex. Gas 1.7-2.2 3.8-5.8 2.3-4.1 4.0-4.8
Scrub. Water 3.9 8.2 3.6 2.7
Barium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Bismuth
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Cadmium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Chromium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Copper
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Lead
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Magnesium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Strontium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
68.8
2.0
28.8
64.8
15.7
19.5
<29.3
42-54
29-46
95.7
1.4
2.8
97.6
0.8
1.6
83.7
11.6
4.7
99.6
0.03
0.36
91.8
2.5
5.7
79.6
2.2
18.2
22.2
41.1
36.7
<10.3
56-61
34-39
94.1
2.0
3.9
75.8
15.1
9.1
15.0
48.9
36.1
99.3
0.1
0.6
93.0
1.7
5.3
69.9
5.5
24.7
30.0
35.2
34.7
<12.9
42-45
45-55
85.9
1.9
12.2
76.2
17.8
5.9
13.7
50.2
36.0
99.3
0.2
0.5
89.8
3.5
6.6
78.6
2.4
19.0
36.3
38.4
25.4
<9.3
68-74
68-74
92.1
2.8
5.1
58.0
33.2
8.8
6.0
73.6
20.3
99.4
0.1
0.5
90.7
1.6
7.7
one of three pathways; they either
left 1n the kiln ash, or traveled to
the afterburner as fly ash or as
volatilized compounds. In this dis-
cussion, those metals with a
propensity towards partitioning in
the kiln ash will be referred to as
"refractory" metals, while those
which tend to exit from the kiln to
the afterburner can be grouped into a
category of "volatile" metals. These
designations are based on the amount
of each metal found in the kiln ash,
as a percentage of the total metal
measured among the process streams
(ash, scrubber blowdown, and flue
gas). An assumption in making these
distinctions is that metals traveling
from the kiln to the afterburner are
predominantly the result of volatil-
ization, rather than entrainment in
fly ash.
Cadmium, lead, and bismuth
appeared to be the most volatile of
the nine metals over the range of
kiln temperatures tested, with
average partitioning to the kiln ash
of <15, 20, and 32 percent of the
measured amounts for each of the
three metals, respectively.
Barium, copper, strontium,
arsenic, chromium, and magnesium
appeared to be more refractory, with
average partitioning of 77, 79, 89,
91, 93, and 99 percent to the kiln
ash for each of the metals,
respectively.
With the exception of arsenic,
the observed order in the degree to
which the metals volatilize correl-
ates strongly with that which would
be predicted by 'volatility tempera-
tures' (the temperatures at which the
vapor pressure of the most volatile
principal species of each metaJ under
oxidizing conditions is 10 atm).
The fact that arsenic is significant-
ly less volatile than expected sug-
gests that either a refractory com-
pound is preferred over the predom-
inant arsenic species (As203), or
that some chemical interaction, such
560
-------
as strong adsorption to the clay,
occurred. (4)
appeared relatively stable with
increasing feed chlorine.
Kiln Ash Partitioning
The only metals for which kiln
ash partitioning appeared to be
affected by changes in kiln
temperature were the most volatile
metals (cadmium and lead) and
arsenic. For these three metals,
slight decreases in kiln ash
fractions were seen with increasing
kiln temperature. Kiln temperature
effects for the other individual
metal5^ as well as for total metals,
were not significant within data
variability.
The effect of increasing feed
chlorine content on the partitioning
of metals appeared to be much more
significant than that of kiln temp-
erature. As kiln and afterburner
temperatures were held constant, and
feed chlorine content increased from
0.0 to 8.3 percent, the overall
fraction of metals partitioning to
the kiln ash dropped from 81 to 63
percent.
Chlorine content of the feed
affected the volatility of individual
metals to different degrees. Kiln
ash partitioning of the three vola-
tile metals decreased measurably as
chlorine content rose. In the case
of bismuth, the fraction of the
measured metal found in the kiln ash
dropped from 65 to 36 percent as feed
chlorine concentration increased from
0.0 to 8.3 percent. Kiln ash frac-
tions of cadmium and lead dropped
from 29 to 9 percent, and from 84 to
6 percent, respectively, over the
same chlorine range.
Only one of the refractory
metals showed an increase in
volatility with increasing feed
chlorine content. The fraction of
measured copper found in the kiln ash
dropped from 98 to 58 percent as feed
chlorine increased. The volatility
of the other refractory metals
Scrubber Efficiency
The split of metals between
scrubber exit gas and scrubber water
determines the apparent scrubber
efficiency, defined as the ratio of
metals in the water to the sum of the
metals in the two splits.
Scrubber efficiency for total
metals ranged from 33 to 56 percent,
and appeared to be negatively impact-
ed by increases in kiln temperature
and feed chlorine content.
Average scrubber efficiencies
for the individual metals ranged from
a low of 32 percent for copper to a
high of 88 percent for barium. Aver-
age scrubber efficiency for the three
volatile metals was lower than that
for five of the six refractory metals.
Kiln temperature effects on
scrubber efficiency were measurable
in the cases of two of the volatile
metals. Decreases in scrubber effi-
ciency, from 51 to 31 percent for
cadmium and from 42 to 25 percent for
lead, were seen with increases in
kiln temperature from 825°C to 927°C.
This is consistent with the changes
in volatilization of these metals
experienced with changes in kiln
temperature.
The effect of chlorine on
scrubber efficiency for individual
metals was significant for copper,
arsenic, bismuth, and cadmium. In
each case, the effect of increasing
feed chlorine was to decrease
scrubber efficiency. Decreases were
as follows: copper (from 67 to 21
percent); arsenic (67 to 38 percent);
cadmium (46 to 26 percent); and
bismuth (55 to 40 percent).
The effect of afterburner
temperature on scrubber efficiency
for individual metals was less
substantial than that of kiln temper-
561
-------
ature and feed chlorine content.
Modest increases in scrubber effici-
ency .were seen along the increasing
afterburner temperature range for
bismuth and lead.
Flue Gas Phase Distribution
Standard Method 5 trains were
used to collect samples of flue gas
particulate- and vapor-phase metals.
In this discussion, two assumptions
regarding this sampling technique are
made: (1) all particulate-phase met-
als are collected on the filters, and
(2) all vapor-phase metals are
collected in the impingers; vapor-
phase metals do not condense on the
filters, nor do they pass through the
impingers. (In actuality, the 'vapor-
phase' likely includes some water-
soluble metal that has wept through
the filters and collected in the
impingers.)
The distribution of metals in
the flue gas favored the particulate
phase over vapor. Average particulate
phase metals as a fraction of total
flue gas metals was highest for the
three volatile metals, with lead at
96 percent, cadmium at 90 percent and
bismuth at 84 percent. For the
refractory metals, the fraction of
flue gas metals exiting as particu-
late ranged from 82 percent for
copper to 31 percent for barium.
None of the test variables had a
clear effect on the split of total
flue gas metals between the two
phases. Likewise, effects of test
variables on the flue gas phase dis-
tribution for the individual metals
were not significant within data
variability. (£)
CONCLUSIONS
In the subject test series, the
trace metals cadmium, lead and bis-
muth were found to be relatively
volatile, while barium, copper,
strontium, arsenic, chromium and mag-
nesium were relatively non-volatile.
Average kiln ash partitioning
ranged from <15 percent for copper to
99 percent for magnesium. Average
scrubber efficiencies for the indivi-
dual metals ranged from 32 percent
for copper to 88 percent for barium.
Both kiln ash partitioning and
scrubber efficiency appeared to be
impacted negatively by increases in
feed chlorine content and, to a
lesser extent, increases in kiln
temperature.
Average particulate-phase metals
as a fraction of total flue gas
metals ranged from 31 percent for
barium to 96 percent for lead. No
clear effects of the three test
variables on flue gas phase distribu-
tions were apparent.
REFERENCES
1. Cppelt, E.T., "Incineration of
Hazardous Waste: A Critical
Review", JAPCA 37: 558, 1987.
2. Acurex Corp., "Test Plan for
Evaluating the Fate of Trace
Metals in Rotary Kiln Incinera-
tion with Yenturi Scrubber/
Packed Tower Scrubber Particu-
1 ate/Acid Gas Control"; U.S. EPA
Contract No. 68-03-3267; Cincin-
nati; July, 1988.
3. Skinner, J.H. and G.J. Carroll,
"Hazardous Waste Incineration:
Status and Direction", Int'l.
Conference on Incineration of
Hazardous, Radioactive and Mixed
Wastes; San Francisco; May, 1988.
4. Fournier, D.J. Jr. and L.R.
Waterland, "Pilot-scale Evalua-
tion of the Fate of Trace Metals
in a Rotary Kiln Incinerator
with a Venturi Scrubber/ Packed
Column Scrubber - Draft", U.S.
EPA Contract 68-03-3267; Cincin-
nati; April, 1989.
562
-------
Disclaimer
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency peer and administrative review policies and approved for
presentation and publication.
563
-------
TREATMENT OF RCRA HAZARDOUS/RADIOACTIVE MIXED WASTE
M.E. Redmon, M.J. Williams, and S.D. Liedle
Bechtel National, Inc.
P.O. Box 350
Oak Ridge, TN 37831
ABSTRACT
The following paper describes a treatment process for a
radioactive/chemical mixed waste sludge. The waste, which was generated
during remedial action under the Department of Energy's (DOE) Formerly
Utilized Sites Remedial Action Program (FUSRAP), was designated as mixed
because of its uranium content (up to 14,000 picoCuries per gram) and the
presence of chemical constituents which caused the waste material to fail
the Resource Conservation and Recovery Act (RCRA) characteristic test for
ignitability.
Because the sludge was classified as a mixed waste, no commercial or
federal facilities were able to dispose of the material. As a result,
various alternatives were proposed to either eliminate the chemical
constituents which contributed to the ignitability characteristic or to
separate the waste into a RCRA hazardous waste and a low-level radioactive
waste. Based on an evaluation of technical feasibility, regulatory
requirements, and cost, a thermal treatment process was selected to
eliminate the ignitability characteristic and allow for disposal as a low-
level radioactive waste. The treatment process was accomplished in two
phases, the lengths of which were determined during bench scale testing.
Phase I consisted of heating the waste slowly to a maximum temperature of
180°F. This phase was implemented to drive off the highly volatile
compounds which contributed to the ignitability characteristic and to
ensure that ignition of the material did not occur. During Phase II of
the process, the temperature was substantially increased to further reduce
the concentration of the contained organic compounds. Each phase utilized
a ventilation system designed to contain organic chemicals and radioactive
particulates. The system contained sample ports which were used for both
health and safety and process efficiency monitoring.
After the treatment was completed, the moisture content was typically
reduced from approximately 70 percent to about 20 percent. The moisture
content was further reduced for ultimate disposal by mixing with
diatomaceous earth in a 1:1 ratio by volume. Laboratory results from the
analysis of the treated sludge indicated substantial reductions in the
concentrations of volatile and semi-volatile organic compounds as well as
the elimination of the RCRA characteristic of ignitability. The waste has
now been transported and disposed of as a low-level radioactive waste at
a DOE facility.
564
-------
INTRODUCTION
Under the Department of
Energy's (DOE) Formerly Utilized
Sites Remedial Action Program
(FUSRAP), Bechtel National, Inc.
performed remedial actions to
clean up radioactive materials at
the National Guard Armory (NGA) in
Chicago, Illinois. These
radioactive materials resulted
from uranium processing operations
during the early years of the
nation's atomic energy program.
The remedial actions at the
NGA generated sixteen 55-gallon
drums of sludge. This sludge came
from catch basins and drain lines
in the motor pool area of the
armory. The sludge contained
uranium concentrations up to
14,000 picoCuries/gram (pCi/g) and
elevated levels of lead. In
addition, the sludge contained a
variety of organic volatile and
semi-volatile compounds such as
benzene, toluene, xylene,
dichloroethane, trichloroethane,
and phenol. Samples were
collected from each drum by using
a one inch diameter, 36 inch long
vertical sampling device.
Multiple vertical samples from a
drum were blended together to
create a single sample from each
respective drum. Samples were
analyzed by the appropriate
Environmental Protection Agency
(EPA) approved methods. Volatile
organic and semi-volatile organic
analyses were performed by Gas
Chromatography/Mass Spectroscopy
(GC/MS) by EPA methods 8420 and
8250, respectively. Metal
analyses were conducted by
Inductively Coupled Plasma Atomic
Emission Spectrophotometry
(ICPAES), EPA Method 6010. The
organic and lead constituents
present in the sludge were
probably the result of NGA
vehicle maintenance operations.
Laboratory analysis of the
sludge indicated a flashpoint of
<70°F and consequently it met the
Resource Conservation and Recovery
Act (RCRA) definition for the
characteristic of ignitability
(Ref. 1). All RCRA
characteristeric analyses were
perfromed by methods as defined by
EPA. This RCRA characteristic and
the concentrations of uranium
present in the sludge resulted in
a mixed waste classification.
PURPOSE
Because of a lack of
treatment, storage, or disposal
facilities for mixed waste, a
treatment scheme had to be
developed to eliminate either the
ignitability or the radioactive
characteristics from the sludge or
to separate the material into two
distinct wastes, a RCRA hazardous
waste and a low-level radioactive
waste.
APPROACH
Several treatment and
disposal options for the sludge
were explored. These included
disposal at a government low-level
radioactive waste disposal site,
incineration at both a government-
owned and at a commercial
hazardous waste incinerator,
treatment by new technologies such
as supercritical water oxidation
and a microwave technique,
chemical treatment to separate the
uranium from the sludge, and a
thermal treatment process to
reduce the concentrations of
organics and thus, raise the
flashpoint to eliminate the
ignitability characteristic.
The thermal treatment option
was chosen as the method for field
565
-------
operations. Bench scale tests
using the thermal treatment
method proved very successful.
These tests resulted in the
flashpoint of the sludge being
raised from the initial <70°F to
greater than 800°F. In addition
to eliminating the RCRA
ignitability characteristic, a
volume reduction in the total
amount of waste was observed. The
primary reason for the volume
reduction was the moisture content
of the sludge being decreased from
about 70% moisture to less than
20% moisture. The other
alternatives were eliminated due
to cost, technical feasibility, or
regulatory restrictions.
Before preparing for field
operations, several safety
concerns were addressed. The
primary concern was the ignition
of the sludge during processing.
The bench tests had shown if heat
was not evenly distributed, the
sludge would form red embers, and
the potential for fire was
increased. Other health concerns
were worker protection from dust
particles that contained
radioactive or lead particulates
and from benzene gases. The
formation of a crust on the top of
the sludge was also of concern
since this would cause gases to
accumulate in the drum, therefore,
resulting in a potentially
explosive condition.
The system which was designed
for field operations was kept as
simple as possible while
addressing all safety and
operating concerns. Ventilation
systems were designed to meet
capture velocity and flow rate
requirements of the American
Conference of Governmental
Industrial Hygienists (AC6IH) and
the National Fire Protection
Association (NFPA), respectively
(Refs. 2, 3). Off-the-shelf items
were used where feasible.
Field processing of the waste
sludge was divided into two
phases. Phase I of the operation
included heating the sludge in the
original 55-gallon drums in a
controlled manner to drive off the
majority of the organic volatile
compounds and to reduce the
potential for fire and explosion.
Bench tests had shown that the
temperature should not be
increased above 180°F in Phase I
because of the potential for fire.
Commercially available drum
heaters that were thermostatically
controlled with a 35°F/hour
temperature increase capacity were
utilized. The ventilation system
used in Phase I was a low volume,
high negative pressure system
powered by a 200-cubic
feet/minute (cfm), 105-inches of
water suction, high efficiency
particulate air (HEPA) filter
vacuum. This ventilation system
used two activated carbon filter
beds in series between the drums
of sludge being treated and the
vacuum. The carbon filters
adsorbed organic vapors that were
generated during the processing
and the HEPA filter captured
radioactive and metal particulates
that were emitted. To prevent the
sludge from forming a crust and
therefore accumulating organic
gases, a commercial concrete
vibrator was used to continuously
agitate the waste. This also
aided in the liberation of organic
gases. Sample ports were provided
in the system design to allow for
monitoring of the process
efficiency and for health and
safety concerns. Phase I
continued until field flammability
tests were negative and organic
vapor analyzer (OVA) readings had
decreased to 10% of initial
readings. The total time for the
566
-------
completion of Phase I of the
process was approximately 8 hours
per drum. A schematic of the
Phase I system is shown in
Figure 1.
Phase II of the thermal
treatment operation involved
heating the sludge to the highest
temperature attainable with the
drum heaters to further reduce the
moisture content and to continue
to reduce the organic volatile
and semi-volatile concentrations.
The Phase II ventilation system
was also designed with sampling
ports for monitoring activities.
This system included the use of
fume hoods suspended from a mobile
gantry crane that were connected
to a 2000-cfm HEPA blower unit
with an activated carbon filter in
the ventilation line. The gantry
crane was used to eliminate the
need for personnel to handle hot
drums. Once Phase II treatment
had progressed to the point that
the moisture content had been
reduced to approximately 20%, the
treated sludge was further
stabilized with diatomaceous earth
in a 1:1 ratio by volume. This
was done to ensure the sludge met
moisture content requirements for
disposal at a low-level
radioactive disposal site. Phase
II of the operation required an
. .average of 4 hours per drum. The
layout of the Phase II system is
shown in Figure 2.
RESULTS
Once all processing was
complete, including stabilization
with diatomaceous earth, a total
of fourteen 55-gallon drums of
treated waste remained. The
treatment process yielded more
than a 50% reduction in volume
before the sludge was stabilized
with the diatomaceous earth.
Laboratory analysis of the
treated sludge has confirmed that
it no longer demonstrates the RCRA
characteristic of ignitability.
Monitoring during process
operations confirmed the
efficiency of the treatment system
in removing organics from the
sludge while protecting the work
environment from the release of
organic vapors and airborne
particulate hazards. Table 1
shows typical reductions in
volatile and semi-volati le
concentrations after treatment.
This thermal treatment process
proved to be a success as the
waste can now be treated as a
low-level radioactive waste for
the purposes of disposal.
The total cost of this
operation was about $190/gallon of
treated waste. This cost includes
all capital equipment, labor, and
other incidental costs. Because
the capital equipment could be
reused in additional processing of
waste, the total cost would
decrease with increasing volumes
of waste. The cost would be
reduced to about $65/gallon for
6000 gallons of waste and about
$55/gallon for 20,000 gallons of
waste. The breakdown of the final
cost of the 880 gallons of treated
waste was approximately $95/gallon
for capital equipment, $65/gallon
for labor, and $30/gallon other
costs. Labor included a site
health and safety officer,
operations superintendent, 3
support laborers, and engineering
support in the home office.
CONCLUSIONS
Based on the experience from
this operation, a system of this
type would be applicable in many
similar situations, especially to
small quantity generators.
567
-------
O
I
LL
Z
O
O
z
UJ
s
UJ
UJ
£
UJ
O
O
UJ
X
a.
LLI
DC
O
568
-------
I
LTu
O
O
l-
LU
Q.
O
LU
LU
£
cc
LU
D
CO
LU
CO
0.
CM
LU
DC
O
O
569
-------
vo in co
CM CM r-- o
in in in in
CM CM CM CM
V V V
in
vo
H
CM in co
** rH CO
m
CM
CO
CM
•
O CM VO
CM CM OO
in •* co
H
00
CD
rH
•H
-P
U
H ft
CO O
CM O
H O
X
i
Cn
•H
O
vo o
cn co vo H o
H CO CM H O
CD 0) CD 0)
-p -p -p -p
fd fd (d fd
Cn Cn Cn Cn
CD O CD Q)
0)
•H
-P
fd
Cn
CD
C
CD
•rH
-P
Cn
•H
•P
Cn
0)
o
o
r-
V
-P
C
•H
o
ft
X!
03
—
p-
|
55
r-
^
<
£-
J2
c
^— >.
CD
rH
•rl
-P
fd
rH
O
,£.
0)
c
0)
|3
rH
0
EH
*-*>
^ OJ
0) rH
rH -H
•H 4J
•P
r S"" "
CD
0) C
C CD
(D N
rH C
>t
rH ^^
O
> CD
^™^ C
fd
Q) XI
C -P
CD CD
N O
C JH
CD O
X! rH
1** r*^
>i O
r^ -rH
W EH
«^s.
^ Q)
Q) rH
H -H
•H 4J
-P rd
fd rH
H O
° £,
N— •»
O
0) C
C CD
x! -P
-P CD
CD 0
O t-{
rH 0
O H
rH X!
X! 0
O -H
•H M
Q EH
-— s
£ 1
•H 1
•P -H
rH Q)
O 01
1
•rH CD
g C
CD CD
01 rH
— • fd
CD 4->
C X!
Q) ft
rH fd
fd c
XI H
•P >i
X! X!
ft -P
g |
,_»
rH
fd
-P
^ CD
rH g
fd
CD >
g fd
CD
t> ^^
0) g
•^ -H
g
"D O
fd M
rH 6
^^
CD
ft
O
O ••
01 co
•H U
O H
•H EH
•O CO
fd H
v- - pq
EH
CO U
CO rtj
CM K
g ffi
3 u
•rH
C <
fd K
rH U
..
4J
•H
>
•rH
4->
O
fd
(D
,t
t>i
-P
•H
>
•H
01
O
^|
O
o
Cn
•H
..
• • S»j
>i -P
•P -H
•H rH
O -H
•H xt
X fd
O -P
EH -rH
c
A Cn
W H
570
-------
Caution must be exercised,
however, in utilizing this type of
system with regard to the
concentration of organics. The
waste should be proven to be only
flammable and not explosive, the
filter media should have a high
affinity for the contaminants, and
the effectiveness of the system
should be demonstrated prior to
field use.
In addition, the appropriate
regulatory agencies must be
involved in plans for treatment
and disposal of the waste from the
outset to ensure that the
treatment will result in the
required reclassification of the
waste. The success of the thermal
treatment in converting the mixed
waste at the NGA to a low-level
radioactive waste offers one
reliable solution to a problem for
which few alternative solutions
currently exist.
REFERENCES
Office of Federal
Regulation National Archive
and Records
Administration, "Code of
Federal Regulations",
Chapter 40, Part 261.
American Conference of
Governmental Industrial
Hygienists, "Hood Design
Data," Industrial
Ventilation, a Manual of
Recommended Practice. 17th
ed., Edwards Brothers,
Inc., Ann Arbor, MI, 1982.
McKinnon, G.P., ed.,
"Industrial and Commercial
Heat Utilization
Equipment," Fire Protection
Handbook. 15th ed.,
National Fire Protection
Association, Quincy, MA,
1981.
Disclaimer
Ihe work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
571
-------
OPERATING EXPERIENCES OF THE EPA MOBILE INCINERATION SYSTEM
WITH VARIOUS FEED MATERIALS
James P. Stumbar
Robert H. Sawyer
Gopal D. Gupta
Foster Wheeler Enviresponse, Inc.
Edison, NJ 08837
Joyce M. Perdek
Frank J. Freestone
Releases Control Branch, USEPA
Edison, NJ 08837
ABSTRACT
During the past four years, the USEPA's Mobile Incineration System (MIS)
has processed a wide variety of feeds. Besides incinerating the hazardous
materials for which the MIS was designed, the unit has incinerated contam-
inated debris including wood pallets, steel and fiber drums, and plastics.
This paper identifies significant physical and chemical characteristics of
various feed materials and their relationship to MIS performance. The
effects of these feed characteristics on specific MIS components are corre-
lated. Several problem characteristics have been mitigated by corrective
actions. The operating experience with the MIS has provided valuable data
on the limits of its incineration capacity as well as reliability of the
unit in relation to various feedstocks. The information contained in this
paper is directly applicable to field use of mobile and transportable incin-
erators at Superfund and other industrial cleanup sites.
INTRODUCTION
Under the sponsorship of the Office
of Research and Development of the
U.S. Environmental Protection Agency
(EPA), the Mobile Incineration Sys-
tem (MIS) was designed and construc-
ted to demonstrate high-temperature
incineration of hazardous wastes
(1). The system essentially consists
of a refractory-lined rotary kiln, a
secondary combustion chamber (SCC),
and an air pollution control system.
These three components are mounted
on three separate heavy-duty semi-
trailers. Monitoring equipment is
carried by a fourth trailer.
Over the lifetime of the MIS, a
wide variety of feed materials have
been processed (2,3). These mate-
rials exhibited differences in char-
acteristics that affected the MIS in
various ways. Often a particular
characteristic or a combination of
characteristics would affect the MIS
performance adversely. The exper-
iences gained from field operations
of the MIS during the past four
years have increased the understand-
ing of the interplay of feed charac-
teristics with hardware.
This paper describes the effects
of feed characteristics on the MIS
performance; correlates various feed
572
-------
characteristics with affected parts
of the system; describes actions
taken to mitigate the resulting prob-
lems; and discusses the limits im-
posed on capacity and reliability by
the various feed characteristics.
FEED CHARACTERISTICS
Both the physical and the chemical
properties of the feed determine
incineration system performance (4).
Important physical properties in-
clude: heating value, morphology,
density, rheology, ash particle-size
distribution, and fusion characteris-
tics. Important chemical properties
include the composition of the feed
as shown by: organic content,organic
hazardous constituents, acid forming
elements such as sulfur and the halo-
gens, moisture content, and inorgan-
ic ash components. These properties
can affect the operating parameters,
the capacity, and the reliability of
the incineration system. Many of
these properties are interdependent
as far as their effect on the incin-
erator performance. The manner in
which these properties affect the
performance of incineration systems,
based on the experience of the MIS,
is summarized in Table 1.
EFFECT OF HEATING VALUE
Heating value of the feed material
affects both feed capacity and fuel
usage of the incinerator. An in-
creased heating value of feed mate-
rial can raise kiln temperature,
which can become uncontrollable. The
kiln then requires greater amounts
of oxidant to complete combustion
and greater quantities of inert mate-
rial to control kiln temperature.
The temperature increase can limit
feed capacity. The MIS reached its
capacity limit at 1.33 to 1.61 mega-
watts (MW) heat input to the kiln.
Feed materials such as plastics,
trash, wooden pallets, and bromin-
ated sludge had capacity constraints
caused by high calorific values.
Maximum feed capacities for these
materials ranged from 90 kg/hr for
pure plastics to 859 kg/hr for
brominated sludge as shown in Table
2.
Solid materials with high calor-
ific values cause transient be-
haviors that sometimes further limit
feed capacity. Plastics, trash, and
wood ignite almost immediately after
they are fed to the kiln. Gases
evolved from these materials burn
rapidly producing a sharp increase
in kiln temperature and a sharp
decrease in excess oxygen.
Prior to system modifications
implemented in 1987, the MIS was
extremely sensitive to these tran-
sient behaviors, which caused many
feed stoppages. After the addition
of the LINDEK Oxygen Combustion
System (OCS),the MIS response to the
transient behaviors was improved,
and feed stoppages due to low oxygen
were virtually eliminated (5). How-
ever, there were still many feed cut-
offs caused by excessive kiln temper-
atures. These were minimized by
operating the kiln at the lower end
(790°C) of the temperature range
allowed by the RCRA permit, and by
using water injection to control
kiln temperatures.
For brominated sludge, the behavior
of the MIS was somewhat different.
Large oscillations of the kiln tem-
perature and excess oxygen levels
occurred even when the kiln was oper-
ated at 790°C. The resulting over-
temperatures (greater than 1040°C)
caused feed cutoffs and loss of the
kiln burners. Loss of the burners
increased the length of the feed cut-
off period. The operating changes
required to alleviate this phenom-
enon are described below.
573
-------
Ul
S
1
1
CO
1
i
o
a:
13
Z
5
Ul
£
CO
Ul
H>
E
1
IT—
a
E
E
g
»*-
o
•§
U.
S
c-
•t
s.
g
t>
**-
H-
LU
1
1
19
*
, 4-*
g
1
g
t,
O*
•g
4v
a
1
CD
as
X
1
GL
t. a) •»- O
3 CO JS <0 (A CO
9* 5P "v p £ .^ § -n §
C/3 'tl S > CO H-CD 3E! CO
g,
1 s
«
o y?
•^- 3
§3 3
CJ u-
5 ** *j*
Q, X >• X X
3 4-» 4J 4-» 4-*
t 'o '5 *o 'o
.^ .— o
= | _yl 8
^ j= 2 S .£
•o < c -2 a> o
— ' O D t t-
<1> •. O O *r- D
tS. t! 0) 2
>» rt c (^
C- CD CA C ^ fe
fl> H- O *• "O C — O ••- 4-»
ro **- o 4-» *-• co
*J> O -M '> C CA 15
r- O (0 3t CO
4-» 3> t. 0 5 & X
CO L. •«- 3 — * — » O
3 U- ^ Q. CD UU f^
CA
co — » C
E* C *Cfl
*. •* •» cD o _cy »
i JBi^^0"" i
£ ^ 4J^ XC4-»C 4-<
to ^y; to .^ CA O C O co
QJ 4^ flj 4-t 5J ^^ 3 CU O
d) o ct) o 13 o ^r Oj co
u. oi u_ a; oQ-cDou-
s
cu
L.
3
"5
^ g
O X 4>
O 4-i O -^ "-
_T
'o
i2
c
L.
Ul
«r
O
"o
fc
CO
4j
•s
O)
.£
1
U.
0
in
cu
5
u
'^
c.
(D
a_
"S
c_
•a
2
to
I
8
^*
g
f^
D
?5
£_
4->
CA
M
CA
^1
(O
*J
f_
£
aT
"§
5
(A
?
C
1
L.
ca
i
Q
CA
£_
CU
+-•
ffl
irocess w;
u.
8"
(/)
(D
4->
CO
1
•^
•5
4->
g
°
g
§
CO
UJ
*jj
lectros
i
,t^
0
i
§.
1
0>
ca
"o
5
„
1 g
C_ -M
1 1
j;
g
L.
h-
-
•-
CO
CD
a.
g
'5
5
a.
c"
o
'5>
o>
§^
1
g
4~*
ca
4-»
C
~io
ca
O)
cu
«f
o
4->
CA
t_
O
CO
0
574
-------
UJ
u
UJ
o_
C/)
£
g 1
V)
s
UJ
u_
v>
3
Ct*
«t
>
u.
O
UJ
u_
u.
UJ
1
to
u
s
2
G
0
a
UJ
u_
CM
UJ
1
s «
4-* C.
CO 01
L- U
o u
$
H-
s
c
o
4-*
CO
a>
6 *^
E c. C
1 | ^
£ g- G
••— Q) O
g c
rn fli
11"
O 4-1
i s S
o 8
0)
L.
4-f
'5
*«
I s §
CO CO CO
o > o
7; e
— o
4-> •—
L. £
a- o _c
3 to
**
_
0)
ra
1
"8
0)
u.
1 1
* ^3 CO
i o .2
C£ CM u
»*- X c
o L. ±;
CD .-
f? g •*
s "a IT •§
co i o5 r
i " " 1 1
O£ C fcj »*-
oT 15
J 3 ^
4-» ' O CO CO
« 4-M 3 •£ -^
I_ U H^ O O
CD f_ 4J 4_*
y s III
>* "O
£— " "5
"O — • E
o in JD o o
O N. L. *J- CO
0 ex, 0 ^ 3
N- in
O CO
^1_ 1 = =
CO S
m
5 = =
in
oa = r =
o o
^ =
.. C3
f^- = S O
*•* in
CO
• s = s
X ? to
JS B "S
O E=
— ' I- 4-» L.
•^ o •— a>
O H- O) 4-* (0
W QJ C H- U
o Q) CD ••—
J= ;O ,. CQ — • w
CD CO O CO «^ CD
u. O* O Ch O — *
^ *~ *** *- w a.
X i O) i
§ 1 ^ § II
a v^ »- ^ a: E
o .3
>. S CO
CD - o)
> ••- O X
CJ ^5
4-> n ej
g _C «o in
ro w 1 x x
•2 CD" •? 7 S
f g
4-» CK C
<— fc
i «T 5
S .1 *-
. 4J *" °
2> 3 • X 3
_c S. ca 4-> 5
CO* 3 CJ S. O
— * H- O CO >
w co o <5
^^
^^
o
o in
fc "•
CM «—
o co
tr ?5
CO l^»
=
o
CM =
£
^i
, ~
10
*2 ***
*f~ £
CO CD
if— tii
— ' O 0)
*O CD 4^
CO _Q CD
CD O» O^
U. T- *-
_ 1 1
Cm ro-
se o* o*
ill ^^ ^^
i
s
,
s s
If
to 0)
? c &
E tr- 4-»
•8 -S S
CO S
-Q J2 ro 3
C Q.
^ o o 'ra "-
•S £ tJ 5? =
o o o — * i-
U O O W CD
^2
E
|§
= =
= =
S =
A
= O
in
•. _
CO
u
— • CO
•^- CD
0 —
CA Q-
1 1
1
1*1
g
E
-ft
a
IQ
b
4-*
I
^_
S
in
CO
i
N.
in
o
o
o
i
o
in
o
i
o
o
o
in
o
o
s
10
o
CO
2
4-*
CO
U
a
s
o
X —
M "m
> E
C1J CO
4J O
"o •§ ej '5.
_c
CO
'S 2 X~
CD H- Q.
CO CO £5
— ' C)
CO » 4-*
c 8- c S
= f = 1
in
in
OJ
o
o
o
o
o
o
CO
0
CM
o
in
1*0
o
CO
O CD (J
O — » U
•^* — • *•-.
en co o>
ru co
O o in
CD S CD
J^
"8
g
u
575
-------
s?
eg
1 g JJ
4)
0}
i
**
S? « CT
I 1 8
1C •**
- 1
£ S i
"S V S
X
_j
a
fe
*j
«
S
•g
u.
01
n> c » ^
9J D5 flj JO 1 C O
1^= .2 1? +. £ ? £
3w °"USwi!i; T— 01 o Of t-
.CC 4-» **- 0> <*- £ 4-»
4J C C !>C.OOW
— • — ' £ "D*-»"3L-T!""<0
U OT (TJ I2OOCH-C-
Q J3 •-» ro o s- 3: w o "u
4-» «
1 X * 1
ro *j •* •M
" 1 S- g ^
fg ^ -g .5
E -^ 'w Z3
O
K) 3
*p %S o
S! M fo
in
in in ro
*~ «^ Cj
co co ^
i t i
o o «-
s
O O I
in
(M
1 s^
O 1 I
in
ro
8
0 0 S
^
0 S
01
f
« 3
— • w
O 0)
t— 4-* *13
m __j- ^j
U C
1 1 1
4-* 01 C.
(O > CO
B
5?
.c
or
J
I
M
1
O
Cu
1
|Q
^
0
5^
i
o
S
•o
rj
in
•4-
CO
c>
§
1
1
£
3
,5
E
£ %
_^ ~o
m
t—
at
1
X
z e
Ife
Is
_0
'E
(D
ra E?
o
•- J=
•M 5*
'> •*
S P
0
^O
«
=
<0
ol
I
o
s
5
i
to
2
(A
fr
•J
|
•g
£_
0
1
576
-------
To reduce the amplitude of the
temperature excursions, an automatic
feed cutoff was introduced into the
kiln control system. This stopped
feed whenever the kiln temperature
exceeded 945°C or the oxygen level
dropped below 4% (wet). This action
minimized over-temperature inci-
dents, but the oscillation frequency
was still large. About four oscilla-
tions occurred per hour. Feed was
cut off for approximately eight min-
utes during each oscillation. Feed
rate was limited to 450 kg/hr under
these operating conditions.
Observations of the kiln during the
oscillations showed that the sludge
was not igniting rapidly. Several
batches of sludge would be fed by
the ram before ignition occurred.
After ignition, flame would fill the
kiln,and oxygen flow and temperature
would increase rapidly. After the
sludge had burned several minutes,
the flames would extinguish, and the
oxygen flow and temperature would de-
crease rapidly to complete the cycle
of oscillation. It became apparent
that steady ignition of the material
was required to prevent the oscilla-
tions.
The kiln temperature was increased
from 790°C to900°C to provide
the necessary energy to evaporate
the water and volatile organics so
that ignition could be sustained.
This operating change was success-
ful, and maximum feed rate was
increased to 900 kg/hr by the above
changes.
For feeds with high heat content
such as brominated sludge, the capa-
city of the MIS is increased by
water injection. Due to its high
heat capacity, water provides a very
effective heat sink. Consequently,
when it is used to control kiln tem-
perature, moisture increases the SCC
residence time as compared to the
use of excess air. Fig. 1 shows that
the use of water injection can in-
crease capacity by about 20% over
the use of excess air at a given SCC
residence time. The use of oxygen in
the kiln enhances the effect of
water injection by allowing further
capacity increases. At an enrich-
ment to 40% 02 in the combustion
air, capacity can increase by 60%.
EFFECT OF MORPHOLOGY
The morphology of the feed mate-
rial affects the feed system in that
periodic jams usually are caused by
materials that are poorly prepared
by the shredder due to their morpho-
logical characteristics. Problem
materials include wooden pallets,
metal drum closure rings, thick
metal pieces (pipe), plastics,
trash, clothing, and mud (6,7).
The feeding of these materials
restricts the MIS capacity. As
shown in Table 2, relatively dry
soils can be fed at rates up to 2275
kg/hr, but the presence of plastics
and mud reduces the feed rate to 680
kg/hr.
The shredder is used to prepare
the solid feed materials for inciner-
ation. For most materials, the
shredder works extremely well. How-
ever, wooden pallets and metal drum
closure rings often cause feed block-
ages when shredded in the present
equipment. While the shredder breaks
most of the wooden pallet into 5-cm
wood chips, an occasional board will
position itself to go through the
shredder as a 5-cm-wide by 1.3-m-
long sliver. The same is true for
the drum lid rings. The shredder
sometimes drags a ring through,
straightening it but not cutting it.
In each case, plugging of the convey-
or, weigh scale, or ram follows. The
577
-------
UJ
2
ui
o
z
HI
s
CO
Ul
cc
8
(0
I I I I
0 200 400
I i i i i i i r i r
600 800 1000 1200 1400 1600 1800 2000
FEED RATE (kg/hr)
NOTES:
Solid Feed Heating Value 1.50 Kcal/g
Kiln Temperature 925°C
SCC Temperature 1200°C
A - Water injection using 40% oxygen-enriched air for
combustion
B - Water injection using air for combustion
C - Air cooled using either air or 40% oxygen-enriched
air for combustion
Figure 1. Effect of cooling media on SCC residence time.
578
-------
best solution has been to manually
separate and prepare these feed mate-
rials by cutting them into small
pieces (about 24 cm in length) prior
to shredding.
The shredder also has performed
poorly on materials such as plastic,
clothing, trash, and mud. These
poorly shredded materials often have
jammed at the doctor blade or re-
stricting dam that was originally
used to level the granular material
on the conveyor belt as the belt
exits the shredder hopper. The
doctor blade worked quite well for
granular material, but it created
large blockages when materials such
as shredded plastic,clothing, trash,
metal, or mud were fed. A roller
has been installed to replace the
doctor blade, and this has reduced
the number of jamming incidents.
The addition of a television cam-
era at the top of the main feed con-
veyor has permitted the quick detec-
tion of feed material jams in the
weigh scale chute. This allows the
operator to take prompt corrective
action and has prevented many hours
of feed interruption since it has
been installed.
Finely granulated material affected
the operation of the ram feeder. It
would bypass the ram head and col-
lect on the backside. The material
on the backside would periodically
prevent the ram from fully retract-
ing. A small chain-plug conveyor
was installed and timed to convey
the bypassed material to the front
side of the ram. This solution has
worked quite well.
EFFECT OF DENSITY
The density of the feed
determines capacity for many feed
materials. For feeds of typical den-
sities (1.5 g/cc), such as soils,
the maximum feed rate is 2275 kg/hr,
obtained at a kiln revolution rate
of 1.6 revolutions per minute (rpm),
which gives a typical solids resi-
dence time of 30 minutes. For low-
density materials such as vermicu-
lite (0.096, g/cc), feed rates up to
364 kg/hr are feasible.
EFFECT OF RHEOLOGY
The rheology of the material
affects either the feed system or
the decontamination behavior. Muddy
soils fed to the MIS would form
clumps of material,which were caught
by the doctor blade; and also would
stick to the conveyor belt, weigh
scale, and ram trough. The resulting
buildup would periodically plug var-
ious parts of the feed system thus
reducing overall feed rates to about
1140 kg/hr. This is approximately
50% of the maximum rates achievable
with dry soils. Addition of vermicu-
lite has eliminated the sticking of
the muddy material while adding only
a small amount to the throughput
weight.
Brominated sludge feed had a ten-
dency to form balls up to 8 cm in
diameter. Since the time required
for burnout of a sphere is propor-
tional to the square of its dia-
meter, the large balls require a
much longer residence time in the
kiln for decontamination. At the
normal 1.6 rpm kiln speed for soil,
small smoking particles would exit
the kiln with the kiln ash. This
required limiting feed rate to about
450 kg/hr. However, when the resi-
dence time of the sludge was in-
creased by reducing the rpm to 0.8,
feed rates up to 900 kg/hr were
achievable without smoking.
579
-------
EFFECT OF HALOGEN AND SULFUR CONTENT
Incineration of brominated and
chlorinated wastes generate the acid
gases hydrogen bromide (HBr) and
hydrogen chloride (HC1). These acid
gases affect the capacity, the blow-
down rates that control total dis-
solved solids (TDS) in the process
water system, and the particulate
emissions.
A capacity limit of 115 kg/hr of
acid-forming organic chlorides was
encountered during the 1987 trial
burn (3). The capacity limit was
caused by pump cavitation in the
quench system, which cools the gases
exiting from the SCC to about
95°C. The cavitation reduced the
quench water flow rate, which acti-
vates the protective instrumentation
resulting in cutoff of the feed and
shutdown of the burners. This cavita-
tion was produced by excessive chlor-
inated waste feed rate as follows:
The quench water was treated with
caustic solution to neutralize acid
gases. Reaction between HC1 in the
combustion gases and sodium hydrox-
ide (NaOH) in the quench sump pro-
duced effervescence. The amount of
effervescence increased to violent
levels as HC1 flow rate increased.
The violent effervescence reduced
the available net positive suction
head (NPSH) of the pump, which
causes cavitation at very high HC1
loads.
High organic chloride loads also
affect particulate emissions through
the phenomenon of mist carry-over
into the stack. The amount of carry-
over is determined by both the HC1
loading of the flue gas and the TDS
of the process water (3).
In tests performed prior to the
1987 trial burn, particulate emis-
sions were found to exceed the allow-
able emissions (180 mg/dscm) by as
much as a factor of three. Analysis
of the Method 5 particulate filter
cakes showed that over 90% of the
particulate was sodium chloride
(NaCl) and sodium hydroxide (NaOH).
The emissions were brought into com-
pliance after a mist eliminator was
installed.
However, data taken during the
1987 trial burn showed that TDS of
the process water also affected par-
ticulate emissions. As shown in
Fig. 2, particulate emissions were
proportional to TDS during the trial
burn tests. The data shows that oper-
ation with TDS at 20,000ppm provides
an adequate safety margin to insure
that particulate emissions are in
compliance at high organic chloride
loadings.
EFFECT OF MOISTURE
Moisture content affects inciner-
ator performance and can adversely
affect rheological behavior as de-
scribed above. Depending upon the
heat content of the waste, moisture
can either improve or impede inciner-
ator performance.
When using feeds with high heat
content, moisture acts as a heat
sink to control kiln temperatures.
Conversely, when using feeds with
low heat content, moisture increases
auxiliary fuel requirements and de-
creases SCC residence time. Feed
rate must be decreased to maintain
SCC residence time. The effect of
moisture on SCC residence time is
shown in Fig. 3.
EFFECT OF PARTICLE SIZE DISTRIBUTION
The particle size distribution of
the ash generated from the waste
determines the amount of particulate
580
-------
carry-over from the rotary kiln to
the rest of the system. The impor-
tance of this characteristic can be
demonstrated by the MIS experience
with Denney Farm soil and Erwin Farm
soil. As shown in Table 2, Denney
Farm soil is much coarser than the
Erwin Farm soil. Up to 25% of the
Erwin Farm soil would carry over
from the kiln to the SCC. This
caused a rapid buildup of solids in
the SCC. The solids buildup necessi-
tated a 70-hr shutdown for removal
of the slag after each 96 to 120
hours of operation (45,000 kg of
soil processed). The behavior of the
silt caused the unit to be unavail-
able for operation an average of 40%
of the time due to the need to clean
out the SCC. On the other hand, the
unit could process the coarser
Denney Farm soil for about 600 hours
(270,000 kg of soil processed)before
a shutdown for slag removal from the
SCC was required. The unit was un-
available for 10% of the time due to
SCC cleanduts with the coarser
Denney Farm soil. In both cases, the
buildup of solids in the SCC signifi-
cantly reduced the availability of
the incinerator.
The problem was mitigated in 1987
when a cyclone was added between the
kiln and the SCC to remove the fines
carried over from the kiln. The sys-
tem operated over a three-month per-
iod and processed over 500,000 kg of
solid material without requiring a
shutdown for slag removal from the
SCC.
Although the cyclone has alleviated
the solids buildup in the SCC, fine
particulates still have caused prob-
lems with the operating instruments.
The large number of fine particu-
lates associated with brominated
sludge have fouled the kiln oxygen
meter and the SCC thermocouple about
once every eight hours. This in-
creases the number of over-tempera-
ture incidents in the rotary kiln,
induces incinerator feed cutoff due
to a false SCC low temperature meas-
urement, and increases the fuel flow
to SCC burners. The thermocouple
problems have been eased by changing
the thermocouple location and using
a thermowell rather than an aspirat-
ing thermocouple. No satisfactory
solution has been found for the kiln
oxygen measurement. An oxygen meter
with a sampling system resistant to
fouling was tried. This produced an
inadequate response time of the oxy-
gen meter, and it showed only a
slight improvement in fouling
resistance.
EFFECT OF ASH FUSION CHARACTERISTICS
The ash fusion characteristics of
some feed materials caused the forma-
tion of hard slag deposits in the
kiln; consolidated deposits in the
ductwork between the kiln and cy-
clone, between the cyclone and SCC,
and in the cyclone exit tube; and
consolidated deposits in the SCC
exit venturi. The ash fusion charac-
teristics are determined by the chem-
ical composition of the ash.
Ashes containing elements such as
sodium (Na) and potassium (K) have
low slagging temperatures. Ashes
from plastics, glasses, wood, and
other components of trash are rich
in these compounds. The increased
slagging tendency of trash was exper-
ienced in the rotary kiln, which re-
quired a system shutdown about every
ten days to remove the slag buildup
caused by incineration of trash.
Ashes containing significant quan-
tities of calcium (Ca), iron (Fe),
sulfur (S), or phosphorous (P) have
moderate slagging temperatures.
Although brominated sludge, contain-
ing both Ca and S, did not slag the
kiln,the ash produced a consolidated
deposit, which fouled the ductwork
581
-------
240
z
o
1
UJ
LU
O
cc
20000 40000
TOTAL DISSOLVED SOLIDS - TDS (ppm)
NOTES:
Solids Feed Rate 1800 kg/hr
Liquids Feed Rate 60-160 kg/hr
Organic Chloride Feed Rate 50-74 kg/hr
Figure 2. Effect of TDS on particulate emissions.
60000
582
-------
UJ
iu
O
i
55
O
2.9 .
2.8 .
2.7 .
2.6 -
2.5 .
2.4 .
2.3 -
2.2 _
2.1 -
2
1.9 _
1.8 .
1.7 .
1.6 -
1.5
10
Permit Limit
30
MOISTURE (%
I
50
NOTES:
Solid Feed Rate 1820 kg/hr
Kiln Temperature 925°C
SCC Temperature 1200°C
30% Oxygen-Enriched Air
Figure 3. Effect of feed moisture content
on SCC residence time.
583
-------
between the kiln and the SCC. A con-
solidated deposit also occasionally
formed in the quench elbow upstream
of the quench nozzles. This fouling
necessitated a system shut down
about every twelve days to remove
the deposits.
Samples of the deposits were anal-
yzed to determine the mechanism of
deposition. More details on this
topic are provided in Reference 8.
The deposit mechanism was found to
be similar to those operative in
boilers fired with sub-bituminous
coal. The deposits were formed by
sintering of calcium sulfate
(CaSOA) in the temperature range
of 870° and 980°C in an ash
containing 14% CaS04, 23% calcium
oxide (CaO), and about 2.5% sodium
oxide (Na20). The formation of
fused calcium silicates as a result
of the decomposition of CaS04 in
the presence of quartz and alumino-
silicates between 870°-980°C was
also an important factor in the
mechanism of deposition.
Most ashes consist mainly of alum-
inum (Al) and silicon (Si), which
generally have good fusion character-
istics (fusion temperatures above
1650°C). Both the Denney Farm and
Erwin soils were composed mainly of
Si02 and AloOo. The lack of
slagging and of troublesome deposits
experienced with the MIS when proces-
sing these materials demonstrates
these good fusion characteristics of
Al and Si.
REFERENCES
1. Yezzi, J.J., Jr. et al. Results of
the Initial Trial Burn of the EPA-
ORD Mobile Incineration Systems.
In: Proceedings of the 1984 Na-
tional Waste Processing Confer-
ence, ASME, pp. 514-534.
2. Mortensen, H. et al. Destruction
of Dioxin-Contaminated Solids and
Liquids by Mobile Incineration.
EPA Contract 68-03-3255, Hazardous
Waste Engineering Research Labor-
atory, Cincinnati, Ohio, 1987.
3. King, 6., and J. Stumbar. Demon-
stration Test Report for Rotary
Kiln Mobile Incinerator System at
the James Denney Farm Site,
McDowell, Missouri. EPA Hazardous
Waste Engineering Research Labor-
atory, Cincinnati, Ohio, 1988.
4. Brunner, C. Incineration Systems
Selection and Design. Van Nostrand
Reinhold Company, New York, 1984.
5. Ho, M., and M.G. Ding. Field Test-
ing and Computer Modeling of an
Oxygen Combustion System at the
EPA Mobile Incinerator. JAPCA,
Vol. 38, No. 9, September 1988.
6. Gupta, G.D.,et al. Operating Exper-
iences with EPA's Mobile Incinera-
tion System, In: Proceedings of
the International Symposium on
Incineration of Hazardous, Muni-
cipal, and Other Wastes. American
Flame Research Committee, Palm
Springs, CA, 1987.
7. Freestone, F.J., et al. Evaluation
of On-site Incineration for Clean-
up of Dioxin-contaminated Mate-
rials. Nuclear and Chemical Waste
Management, Vol 7, pp 3-20, 1987.
8. Bryers, R.W. Deposit Analysis:
Cyclone Riser/Quench Elbow -Denney
Farm Site, Foster Wheeler Develop-
ment Corp., Livingston,New Jersey,
1988.
Disclaimer
This paper has been reviewed in
accordance with the U.S. Envi-
ronmental Protection Agency peer
and administrative review poli-
cies and approved for presenta-
tion and publication.
584
-------
STATE OF THE ART ASSESSMENT AND ENGINEERING
EVALUATION OF MEDICAL WASTE THERMAL TREATMENT
R. G. Barton
G. R. Hassel
W. S. Lanier
W. R. Seeker
Energy and Environmental Research Corporation
18 Mason
Irvine, CA 92718
ABSTRACT
Incineration is a method of disposing of medical waste that is increasingly
utilized to reduce the wastes volume and hazard. Recent field tests indicate
that medical waste incinerators may be prone to emitting high concentrations of
acid gases, toxic organic compounds and other hazardous substances. This study
examined current practices in the design and operation of medical waste
incinerators to identify the parameters which govern toxic emissions. A variety
of design and operating parameters including chamber temperatures, gas phase
mixing and waste feed rate were found to have an important impact on emissions.
INTRODUCTION
Medical waste treatment and
disposal is proving to be an
increasingly imposing crisis for public
health. EPA estimates as much as
14,000 ton/day of medical waste are
produced in hospitals in the United
States (Infectious Waste News, 1988).
Additional waste is produced in
research and diagnostic laboratories,
nursing homes, doctors offices and
veterinary clinics. Not only is this
waste extremely voluminous, but its
infectious and sometimes radioactive
nature of waste requires careful
management. The technologies most
commonly used to treat medical waste in
the United States include steam
steri1i zati on, shreddi ng/chemi cal
disinfection, and incineration.
Approximately 67 percent of all
hospital waste is incinerated on-site.
Incineration can provide up to 95
percent waste volume reduction, as well
as effectively destroying infectious
organisms in the waste. With landfill
capacity rapidly diminishing and
medical waste generation increasing,
incineration is being increasingly
viewed as a viable medical waste
management technique.
While effective for volume/hazard
reduction, incineration of medical
wastes presents a unique set of
problems. Destruction of hospital
wastes by combustion results in
585
-------
formation of air pollutants in solid,
liquid, condensible, and gaseous forms.
This is one of the most sensitive of
air pollution problems, because not
only "is the general public exposed to
these emissions, but the greatest
exposure is potentially sustained by
that segment of the population which
are the least capable of withstanding
any further stress to their health:
hospital patients.
OBJECTIVES
The ultimate goal of this
research is to develop methods for
controlling medical waste incinerator
emissions. However, before such
practices can be established, research
and experimentation is required to
understand the process variables that
govern pollutant formation and emission
from medical waste incinerators. Thus
the objective of this project was to
define specifically what the most
critical emissions problems are and
what methods are currently being used
to deal with the problems.
INCINERATION SYSTEMS
There is a wide variety of
medical waste incineration systems
including rotary kilns, modular
controlled air, and retort
incinerators. Both batch fed and semi-
continuous versions of all of these all
these systems are used. Modular
controlled air systems are by far the
most commonly manufactured system in
the United States at this time. In
Figure 1 is shown a typical unit as
designed by U.S. manufacturers. The
primary zone of these systems are run
fuel rich (oxygen deficient) and
secondary air is usually added at the
entrance to the secondary chamber to
complete combustion. The waste is fed
onto a small grate with pneumatic
pushers that direct the burning bed
through the primary chamber. The
secondary chamber is normally equipped
with an auxiliary fuel burner.
The principal control variables
for modern controlled air incinerators
are temperatures in the primary and
secondary zones. Combustion air and
auxiliary fuel flow are used to control
the temperatures. Waste feed rate is
not easily used to control primary
temperature due to fact that the feed
is non-continuous.
POLLUTANT FORMATION
The pollutants of concern are:
Radioactive materials
Pathogens
Cytotoxic compounds
Toxic metals
Trace organics (e.g.
PCDD/PCDF)
• Criteria pollutants (CO,
NOX,
S0£, HC1, PM, PM10)
Radioactive Materials
Radioactive materials are only a
problem when present in the waste. In
general, radioactive materials are
segregated and disposed of separately.
Pathogens
The destruction of pathogens is
one of the primary goals of medical
waste incineration. There exist some
data that indicate that a properly
designed and operated incinerator is
capable of completely destroying
pathogens (Barba, 1987, Allen, 1988).
These data were obtained by spiking the
feed with a particular bacteria spore
and then testing the residuals and
exhaust gases for spore activity. Only
in an incinerator operated at very low
temperatures, 1100'F, was there found
any residual activity.
While pathogen destruction is
generally complete, an incinerator can
be made to operate with poor
destruction under abnormal situations.
Poor carbon burnout will generally
accompany pathogen emissions, i.e. the
586
-------
pathogens can not survive if organic
carbon is destroyed. On the other
hand, pathogens are likely more fragile
than total organic carbon so that
pathogen destruction could be
accomplished without good carbon
burnout. Nonetheless the factors that
affect carbon burnout and pathogen
destruction are related, and the
parameters that control carbon burnout
can be used as useful indicators of
complete pathogen destruction.
The phenomena impacting carbon
burnout and by analogy pathogen
destruction include the following:
• waste moisture content
• uniformity of combustion zone
conditions
• combustion zone temperature
• residence times of solids in the
combustion zone including both
solids on the bed and particles
entrained into the combustion
gases
• excess air levels
Cvtotoxic Compounds
Dr. Nelson Slavik of the
Envi ronmental Heal th Management Systems
representing the American Health
Association recently was involved in a
Workshop on Medical Waste Management.
He estimates approximately 2 percent of
solid waste generated by hospitals is
contaminated with cytotoxic chemicals.
Cytotoxic chemicals are used in
chemotherapy and can themselves be both
carcinogenic and mutagenic; many are
listed as hazardous compounds in RCRA
regulations. There are currently no
standards available for the disposal of
cytotoxic materials and Dr. Slavik
indicated that the current disposal
practices are based largely on
experience.
There are no data available on
the destruction efficiency of cytotoxic
compounds in waste incinerators.
However, since cytotoxic compounds are
generally organic, organic destruction
efficiency data can be used as a guide
to the behavior of these materials.
Toxic Metals
Metals are present in large
amounts in hospital wastes.
Contaminated needles are the most
obvious source of metals. However,
there are a number of less obvious
sources, ranging from the dyes and inks
used in printed matter to some
pharmaceutical preparations. It is
these trace sources of metals which
represent the greatest potential danger
to human health, for they generally
contain the most toxic metals.
In Table 1 are data from a
hospital waste incinerator in
California (Jenkins et al., 1987).
This study indicated that the fly ash
was highly enriched relative to the
bottom ash with lead, cadmium, chromium
and arsenic. Even when equipped with
a baghouse, emission of cadmium and
lead were found to be relatively high.
A number of mechanisms control the
behavior of metals during the
incineration of hospital wastes. These
mechanisms can be identified by
examining data available from a variety
of incineration systems and are
discussed in detail by Barton et al.
(1988). The key phenomena, illustrated
in Figure 2, are:
• Vaporization and subsequent
condensation of volatile
metals.
Entrainment of
bearing solids
metals
Reactions between metals and
other elements (esp. Cl)
Some medical waste incinerators
are equipped with air pollution control
devices (APCDs). These units are
designed to remove both gaseous
compounds and solid particles from the
587
-------
incinerator's exhaust gases. The APCD
can have a significant impact on metals
emissions. The temperature at which an
APCD is operated has a strong impact on
the units' ability to control metals
emissions. Due to their relatively low
concentrations and low gas phase
mobilities, metal vapors cannot be
effectively captured by most APCDs.
Thus the APCD must be operated at a
temperature at which the metals of
concern will condense.
APCD efficiency is also
determined by the size distribution of
the entrained particles. In general,
APCDs can effectively capture large
particles - those with diameters
greater than 5 microns - and are less
efficient at capturing small particles.
Unfortunately, as discussed previously
many toxic metals are concentrated on
the small particles and the fume. Thus
the ability of a device to capture
small particles plays a key role in
determining metals emissions.
The important hospital waste
incinerator operating parameters that
influence metals emissions can be
identified using field data and an
understanding of the controlling
mechanisms. The most important
parameter is, of course, the amount and
type of metals in the waste. As
previously discussed, the metals that
are volatile at combustion conditions
or are converted to volatile form such
as As, Cd, Pb and Hg are particularly
difficult to control. Other system
parameters expected to influence metal
emissions include the following:
• The maximum temperatures of
solids in the primary zone
• Chlorine content of waste
since chlorides of metals
such as lead tetrachloride
can form which are
extremely volatile
• Primary zone gas velocity
which dictates particle
entrainment
• The temperature of the
particle control device
since this determines how
much of the volatile metals
have recondensed and can
therefore be captured
• The fine particle control
of the APCD and parameters
such as pressure drop which
influence the control
levels.
Chlorinated Dioxin/Furans
One of the greatest challenges
that remains for the disposal of
medical waste by incineration is the
control of emissions of trace organics
such as po1ych1 orinated
dibenzo(p)dioxin and furans
(PCDD/PCDF). There have now been
several studies which have indicated
that medical waste incinerators can
generate relatively high levels of
PCDD/PCDF. In Table 2 are provided
data from a wide variety of combustion
sources, including three medical waste
incinerators that were selected to be
state-of-the-art and were evaluated as
part of ARB's medical waste incinerator
assessment program. These data have an
average total PCDD/PCDF level of
approximately 700 ng/Nm . This is in
contrast to large modern mass burn
municipal waste incinerators such as
Marion, Tulsa, and Wurzburg which have
average total .-PCDD/PCDF levels of less
than 30 ng/Nm.
The U.S. EPA is currently
developing standards for the Municipal
Solid Waste Incinerators to control
PCDD/PCDF. Several proposed standards
are being considered based upon a
combination of what is achievable and
the risk impacts. The current proposal
would restrict new units to as low as
10 ng/Nm'' total PCDD/PCDF (tetra-
through octa.-) and existing units to
125 ng/Nm . Hospital waste
incinerators would have some difficulty
588
-------
in attaining these levels, based on the
limited data base available -.to date.
Dioxins were first detected in
incinerator emissions in 1977 by Ollie
et al. Since that time a large number
of researchers have attempted to
identify the mechanisms which lead to
PCDD/PCDF emissions. However no
consensus has been reached.
Figure 3 illustrates the various
potential mechanisms that have been
proposed. One of the simplest
mechanisms is that the emitted dioxins
were originally present in the waste
and were not destroyed in the
incinerator. While this mechanism may
responsible for a very small fraction
of the emitted PCDD/PCDF most wastes do
not contain sufficient quantities of
PCDD/PCDF to account for the observed
emission levels.
In the second potential
mechanism, dioxins are formed as
intermediates in the oxidation of more
Complex hydrocarbons. The hydrocarbons
may be chlorinated (PVC for example) or
not (cellulose). If the dioxins are
originally unchlorinated the
chlorination must take place as a
second step.
The third potential dioxin
formation mechanism involves reactions
between relatively simple gas phase
precursors such as phenols and
chlorobenzenes. Shaub and Tsang (1983)
developed a kinetic model to study the
characteristics of the proposed
reactions. They compared the models
predictions with emissions data but
were unable to determine if the
proposed reactions were responsible for
the dioxin emissions. Ballschmitter et
al. (1983) and Benenfinati et al.
(1983) examined the emissions from full
scale incinerators and found a close
relationship between the dioxin
emissions and the quantity of
polychlorobiphenol and polychlorophenol
in the exhaust. They suggested that
this indicates that dioxins are formed
by reactions involving PCBs and PCPs.
However, it is. also possible that the
compounds are all formed by similar
sets of reactions.
The final mechanism that has been
proposed calls for fly ash catalyzed
formation of dioxins in relatively cool
regions of the incinerator. This last
mechanism was originally proposed by
Vogg and Stieglitz (1983)and is
supported by laboratory experiments.
In the experiments, fly ash was placed
in an oven and heated under controlled
conditions. They found that the
quantity of PCDD/PCDF increased
significantly during the heating. All
of the chlorine and organic precursors
needed to form the dioxins were present
on the particle, since none were added
during the test. Extensive experiments
are under way in Germany, Canada and
the U.S. to extend Vogg and Stieglitz's
work to full scale systems.
PCDD/PCDF emissions were found to
be closely related to the quantity of
entrained particles and the amount of
gas phase hydrocarbons (Barton et al.,
1988). Based on this analysis and
other data, a mechanism for the
emission of dioxins during the
incineration of medical wastes can be
hypothesized. The mechanism involves
a series of steps. First ash particles
are entrained by the gas flow. At the
same time, some of the hydrocarbons
present in the waste vaporize. A very
small but relatively constant fraction
of the hydrocarbons escape destruction
in the incinerator chamber. Upon
reaching a reaction zone, the
hydrocarbons react on the surfaces of
the particles to form dioxins. This
mechanism is illustrated in Figure 4
for a hospital waste incinerator. The
proposed mechanism is consistent with
recent studies which indicate that the
reactions occur at temperatures between
48CTF and 660°F.
Based on the mechanism described
above, it can be hypothesized that
PCDD/PCDF formation can be minimized by
589
-------
controlling particle emission levels
within the incinerator, minimizing the
time particles are held at key
temperatures (between 480*F and 666"F)
and by maximizing the destruction of
precursors both vapor and particle
bound within the incinerator. Also
dioxins can ultimately be removed from
the flue gas through the use of fine
particle control since PCDD/PCDF will
condense on particles at low
temperatures. This last approach,
however, merely transfers the dioxins
from one media (air) to another (ash).
Criteria Pollutants
The incineration of medical
wastes can generate a variety of acid
gases such as S02, SO,, NO , HC1 and
HF. NO can be formed i>y oxidation of
the nitrogen in air and in the wastes.
The other gases are typically formed by
the chemical reaction of sulfur,
chlorine and other elements in the
waste. The most common occurrence is
the formation of HC1, and thus in most
hospital waste incinerators, HC1 is the
principal acid gas of interest.
During combustion, the organic
chloride in the waste reacts with
hydrogen from the waste, auxiliary
fuel, or water in the combustion
chamber to form hydrogen chloride, HC1.
The organic chloride can be found in
many substances commonly burned in
hospital waste incinerators including
plastic bags, disposable syringes and
plastic tubing. It is possible for
free chlorine from when the combustion
chamber contains an insufficient
quantity of hydrogen convert all of the
organic chlorine to HC1, but this is
very uncommon.
Hydrogen chloride gas is highly
water soluble. Thus liquid scrubbers
are commonly used to control HC1
emissions. The actual performance of
an acid gas removal device is a complex
interaction of design and operating
parameters that is unique to the type
of device. For these devices, three
key parameters can be identified:
• Liquid to gas flow rate
ratio
• pH of the feed liquid
t pH of the outlet liquid
In addition, if a venturi scrubber is
used to remove HC1, the pressure drop
across the scrubber is important.
These parameters combined with the
input levels of chlorine in hospital
waste dictate the HC1 emissions.
SUMMARY
The pollutants of principal
concern from medical waste incinerators
include:
• Volatile metals such as Ar,
Pb, Cd, Hg
• Highly toxic materials such
as: PCDD/PCDF and
hexavalent chromium
t HC1
• Fine particles (PM10)
In certain situations radioactive
materials, destruction of cytotoxic
compounds and destruction of pathogens
may be important issues. If
radioactive materials are excluded,
then the principal area of concern for
both pathogen and cytotoxic compound
destruction is maintaining sufficient
temperature uniformly to the solids for
sufficient time. The data available
indicates that unusually low
temperatures are required for emissions
of either of these substances to be of
concern.
In the case of HC1, the waste
chlorine level and the performance of
any air pollution control device
scrubbers are dominant. Waste chlorine
levels are highly variable and not
easily quantified. Thus the focus
comes back to the proving that the
scrubber efficiency is sufficient to
control any levels of HC1 emissions.
Since even short-term exposures to HC1
590
-------
are a measurable risk, the short-term
control efficiency must be determined.
For toxic metals, the volatile
species such as arsenic, lead and
cadmium are of principal concern
because they escape the incinerator as
a vapor and condense into an ultrafine
fume that is not easily captured and is
highly respirable. However, hexavalent
chromium is so toxic in small amounts
that it can drive risk assessments, and
therefore even small emission levels of
these species are important to
quantify. Again, the amount of these
metals present in the hospital waste
is, of course, the dominant variable
for control of their emission.
However, it is virtually impossible to
either quantify the metals in the input
waste stream or to control the levels
due to the trace quantities of concern
and the ubiquitous nature of many of
the metals such as chromium and lead.
Nonetheless this is the primary method
of control and must be quantified as
much as possible. For known levels of
input metals, the key operating
parameters are those that control the
volatilization of metals (such as
temperature, excess air and chlorine
levels) and those that control fine
particulate control.
Finally, the emissions of
PCDD/PCDF are expected to be the result
of a complicated interrelationship
between waste properties, combustion
conditions, and scrubber/fine
parti cul ate control. By analogy to MSW
combustion, the primary control
parameters were identified as:
0 Combustion uniformity
t Combustion zone mean
temperature
• Fine particle control
efficiency
• APCD temperature
t Particle loading exiting
furnace (determined by
incinerator load,
velocities and waste
properties)
REFERENCES
Allen, R.J. et al., "Bacterial
Emissions From Incineration of Hospital
Waste, Final Report," Illinois
Department of Energy and Natural
Resources, ILENR/RE-AQ-88/17, July,
1988.
Ballschmiter, K., W. Zoller, C.
Scholtz, A. Nottrodt, Chemosphere,
12(4/5), 1983, p. 585.
Barba, P.,"Test Results From Bacterial
Sample Burns From Nine Infectious Waste
Incinerators, "APCAMid-Atlantic States
Section, Nov. 1987.
Barton, R.G., et al., "Prediction of
the Fate of Toxic Metals in Hazardous
Waste Incinerators," Final Report for
EPA Contract 68-03-3365, 1987.
Barton, R.G., et al., "Draft Topical
Report: Analysis of Quebec City
Incineration Tests," EPA Contract 68-
03-3365, April 1988.
Benefenanti, E., F. Gizzi, R. Reginato,
R. Fanelli, M. Lodi, and R.
Tagliaferri, Chemosphere. 11(9/10),
1983, p. 1151.
Infectious Waste News, "EPA Releases
Estimates on Infectious Wastes
Generation for This Week's Meeting,"
November 17, 1988.
Jenkins, A., et al., "Evaluation Test
on a Hospital Refuse Incinerator at
Cedars-Sinai Medical Center, Los
Angeles California," California Air
Resources Board, ARB/SS-87-11, April,
1987.
01 lie, et al., Chemosphere. 6, 1977.
Shaub, W. M. and W. Tsang, Environ.
Sci. and Tech.. 17(12), 1983, p. 721.
Vogg, H. and L. Stieglitz, Chemosphere,
15, 1986, p. 1373.
591
-------
METAL
LEAD
IRON
HANCAHESE
NICKEL
CADMIUM
CHROMIUM
ARSENIC
INLET TO
BAGHOUSE
(MQ/
18051
2401
175
30.7
1792
78
86
BAGHOUSE
CATCH
OJO/OB)
89
3100
240
33
1700
1
4.7
BOTTOM ASH
cw
385-700
700-715
52 -M
400-37500
323
100-160
15.5
27.9
679
1>5-4S
154
70.1
507
2107
TABLE 1. METAL EMISSIONS FROM
CEDARS-SINAI MEDICAL CENTER
INCINERATOR.
TABLE 2.
SUMMARY.
PCDD/PCDF EMISSIONS
JL
/ Secondary Chamber
L.
Waste ^' Primary Chamber
Feedr
Auxiliary
Burner
Figure 1. Typical modular
starved incinerator.
592
-------
•N BURNING SPRAY
OF LIQUID
WASTE
»•*
REDUCING
ENVIRONMENT
INDIVIDUAL /A&\V
PARTICLE 'L,T**\|
OR DROPLET \j ~ *
VAPOR
HOMOGENEOUS
CONDENSATION
COAGULATION
CHLORIDES
SULFIDES
HIDES, ETC.
FLY ASH
RESIDUALS
Figure 2. Phenomena affecting
metals behavior during
incineration.
Formation From
Precursors
Incomplete Destruction1
of Long Chain Organics \
PCDD/PCDF in Wnsle
Low Temperature
Catalyzed Reactions
Particulat*
leaud Ructions
Figure 3. Potential PCDO/PCDF
formation mechanisms.
Figure 4. Hypothesized PCDD/PCDF
formation mechanism.
593
-------
Disclaimer
Ihe work described in this paper was not funded by the U.S. Environmental
Protection Agency. She contents do not necessarily reflect the views of
the Agency and no official endorsement should be inferred.
594
-------
INDEX
(Numbers refer to individual papers as listed in the Table of Contents)
acclimation time, 42 .:.:••.>:. , ;,. K!
acid mine water (AMD), 3 r ' :;"'"' ' -; !!st
activated carbon and high efficiency particulate air (HEPA) filters, 59
adsorption/desorption, 41
aeration, 3
aerobic microbiological degradation, 42
air emissions control, 43
air pollution control device (APCD)
efficiency, 58
air stripping, 3
alkaline polyethylene glycolates (APEG), 48
AMOCO CADIZ, 17
aqueous cleaning, 55
aquifer
evaluation, 8
Aroclor 1260, 48
aromatic hydrocarbons
polynuclear, 40
chlorinated, 40
arsenic, 20
arsenic volatilization during roasting, 20
bacteria
control, 53
bedrock neutralization, 14
bentonite
backfill, 10
cracking of backfill material, 10
best practicable environmental option, 50
biodegradation, 24
aerobic, 40
PCE, 36
TCE, 36
DCE, 36
VC, 36
biodegradation/bioremediation, 41
biofilm, 42
biological oxidation, 41
bioreactors
shake flask, 40
bioreclamation, 24
bioremediation, 22, 24, 42
laboratory treatability studies, 46
treatment system design, 46
full-scale treatment performance, 46
595
-------
blotrap, 23
brine evaporation, 47
Bruin lagoon, 14
Bunker Hill Superfund Site, 49
calcium sulfate dihydrate, 54
California hazardous waste management plan, 30
capillary barrier, 12
cement, 20
casting technology, 20
kiln, 2
CF Systems Inc., 44
CFCS, 55
chemical treatment, 50
chemical waste, 39
storage for non-treatable, 13
C2-wastes, 13
metal sludges, 13
leathertanning sludges, 13
programme for treating in future, 13
chlorinated hydrocarbons, 34
chlorinated solvents, 55
chloroform, 51
chlorophenolic compounds, 42
chromium removal, 4
clay pelletizing/sintering technology, 20
clays, 6
coal-tar contaminants, 41
combustion efficiency, 31
community right-to-know, 26
composting, 24
concentration profiles and reduction, 51
cone penetration testing, 8
contaminated bedrock, 14
contaminated oils and clothing eg. TYVEK, 48
continuous emission monitor, 31
coolant
to water ratio, 53
machine, 53
proportioner, 53
covering systems, 12
crack size and distribution, 10
cracking, 11
cracks
effect on permeability, 10
effect of soil composition, 10
"cradle-to-grave11 management, 26
creosote, 41
criteria pollutants, 61
cyanide
hydratase system, 46
oxygenase system, 46
remediation, 46
596
-------
dehalogenation, 48
depth effect, 42
destruction and removal efficiency (DRE), 31
detection, 8
dewatering, 50
diffusion, 41
diffusion and reaction, 15
dioxins, 61
dissipation testing, 8
distribution coefficient, 23
drums, 50
electroregulators, 52
electrostatic precipitator, 33
emergency planning, 26
emissions control
transient, 32
emulsion cleaning, 55
encapsulation, 19
evaluation, 57
explosive residue contaminants, 51
explosives, 24
exposure pathway analysis, 36
extract analysis, 21
extraksol, 5
fate mechanisms, 41
Federal Republic of Germany, 45
fermentation, 40
field test, 12
flammable solvents, 55
flue gas phase distribution, 58
freezing technique
laboratory study, 51
statistical analysis, 51
future paint removal strategies, 28
geopolymers, 19
Geo-Con deep soil mixing equipment, 18
groundwater
treatment, 25
decontamination, 4
bio-remediation, 46
contamination, 54
cleanup, 4
Hanford site, 4
harbor sediments, 44
hazard communication, 26
hazardous waste
determination, 26
process emissions control, 43
health and safety monitoring, 59
heat release rate, 31
heavy metals, 46
copper, 23
597
-------
recovery, 25
cadmium, 25
copper, 25
mercury, 25
lead treatment, 49
zinc treatment, 49
cadmium treatment, 49
hemihydrate, 5 4
hollow mixing auger, 18
homogenized waste, 2
Hong Kong, 39
hydraulic
conductivity, 8,12
loading rate, 42
pressure tests, 14
hydrological performance of landfill caps, 12
hydrolysis, 50
immobilization, 15
immobilized algae, 25
impounded high salt wastes, 47
incineration, 16,33,34,50,58,61
oxygen-enriched, 32
oxygen Enrichment, 57
oxygen enhanced burners, 57
pyretrun, 57
transient puffs, 57
incinerator
design, 61
operation, 61
ash, 29
residuals, 58
inorganic
cements, 19
chemical wastes, 10
intermediate treatment method, 38
international waste technologies stabilization process, 18
in-situ treatment, 51
iron coprecipitation, 4
iterative alternating direction implicit, 8
kaolinite, 10
Kemiavafall, 2
Kiln puff control, 32
kinetics, 34
KPEG, 1
lagoons, 17
landfill, 13,50,54
multilayered caps, 12
lanthanide series, 54
leachability tests, 18
leachate analysis, 21
leaching, 15
leakage
598
-------
calculations, 8
lime Slurrying, 3
liners
clay, 11,12
flexible membrane, 12
shrinkage potential, li
deterioration, 11
design of, 11
dessication of, 11
effects of gradation on shrinkage potential> 11
effects of clay type on shrinkage potential, 11
degradation under swell/shrink cycles, 11
concrete construction with inner and bottom, 13
filling by portal crane, 13
rainwater protection, 13
movable roofs, 13
clean water basins, 13
permanent cover, 13
liquefied gas,44
liquid membranes
supported, 4
low-temperature thermal desorption, 1
medical waste, 61
mercury, 33
metal
chip, 53
partitioning, 58
plating, 46
removal from soils, 1
volatility, 58
metals, 6i
recovery, 29
sorptioh, 49
microcosm reactors, 41
micro-encapsulation, 15
mixing, 3
mobile incineration systems
slagging, 60
slag formation, 60
ash deposition, 60
feed characteristics, 60
heating value, 60
feed preparation, 60
feed handling, 60
Rheology, 60
Halogen content, 60
moisture, 60
partide size, 60
particulate emissions, 60
modeling, 8,15
monofilled waste extraction, 49
multi-media models
599
-------
transport, 36
transform, 36
municipal solid waste, 16
incineration, 16
combustion, 29
mycelia, 23
National Priority List Sites, 22
neutralization, 50
New Bedford (MA) Harbor, 44
nitrate removal, 4
nitroaromatics
2,6-DNT, 51
D-NT, 51
M-NT, 51
RDX, 51
HMX, 51
TNT, 51
nitrogen oxides reduction, 32
Norway, 37
oil
tramp, 53
skimmers, 53
open pits
abandoned, 9
waste disposal, 9
regulations, 9
characteristics, 9
investigation, 9
techniques, 9
protection, 9
monitoring, 9
organic
chemical contamination, 40
chemical wastes, 10
extraction, 44
sorbates, 40
waste, 6
organics removal, 43
oxidation, 3
oxygen combustion, 32
oxygen-fuel burners, 32
paint
trade-off considerations, 28
parametric metals testing, 58
pathogens, 61
PCS extraction, 44
PCE, 3
penicillium, 23
permeability, 18
petrographic examination, 21
petroleum sludge, 14
phase separation
600
-------
liquid-liquid, 52
solid-liquid, 52
phosphoric acid (wet), 54
PIC control strategies, 34
PIC formation
reaction rates, 34
chain mechanisms, 34
Piezocone penetrometer, 8
pollution prevention, 26,30
polychorinated biphenyls (PCBs) in soil, 18
polynuclear aromatic hydrocarbons, 41
pozzolanic, 21
precipitation/dissolution, 15
pretreatment, 2
propane as extracting solvent, 44
prope1lants, 2 4
radiation sites, 7
radioactive waste, 7,59
radioactivity level, 54
radium, 7
radon, 3
Rare earth elements
yttrium, 54
lanthanide series, 54
rare elements and gypsum
processing, 54
recovery, 54
RCRA, 31
Record of Decision (ROD), 35
recycling, 30
recycling and reuse, 27,38
refinery wastes, 14
refractometer, 53
remediation
chemical extraction, 7
physical separation, 7
soil washing, 7
on-site treatment, 7
off-site disposal, 7
on-site disposal, 7
reverse osmosis, 4
risk assessment, 36
rotary kiln, 2
trial burns, 31
inc inerator, 58
Rotterdam, 13
San Diego County waste reduction program, 30
sand, 42
screening, 22
screens, 53
secondary assimilation, 42
seepage tests
601
-------
permeameter, 49
segregating wastes, 27
semivolatile organics removal, 1
semi-solids, 44
SITE demonstration, 21
SITE program, 44,56,57
site remediation, 19
sludge treatment, 47
sludges, 44
slurry wall, 10
sodium bentonite, 10
sodium hydroxide treatment, 14
soil
contaminated, 40
fractions, 40
extraction of organic species,
treatment for contaminated, 1
washing, l
CERCLA, 1
chlorinated organics removal, 1
superfund, 1
bio-remediation, 46
contamination, 54
treatment, 56
washing, 56
decontamination, 5
organic contaminants, 5
soil systems, 41
soils
metals, 49
vadose zone, 45
contaminated, 48,51
with haloaromatics, 48
solar evaporation
simulation, 47
solid wastes
batch feed, 31
solidification, 15,19,21,47
solidification/stabilization, 17
soliditech
long-term testing, 21
solid/liquid separation, 47
solvent cleaning, 55
source reduction, 30,55
specific fuel, 2
spent catalysts, 50
SSM, 56
stabilizing ashes
field and lab test results, 16
leaching, 16
physical behavior, 16
cementitious binders, 16
40
602
-------
stabilization, 15,19,21
stabilization/solidification, 18,59
static mixer, 3
steam stripping, 43
strippers
chemical/typical use, 28
influence of new paints, 28
pollutants caused by, 28
dry systems, 28
non-halogenated, 28
performance, 28
source segregating to reduce pollutant loadings, 28
air emission standards, 28
substituting less hazardous materials, 27
sulfur dioxide, 33
Superfund sites, 14,26
surface cleaning, 55
synthetic soil matrix, 56
Taiwan, R.O.C.
environmental committee, 38
waste exchange, 38
standards of hazardous waste, 38
short-term program for industrial waste control, 38
waste treatment facilities, 38
control program, 38
terpenes, 55
test fields, 12
thermal decomposition, 34
thermal destruction, 34
thermal treatment, 57,59,61
toluene, 51
trace contaminants, 42
treatability study, 14
treated soil column, 18
unconfined compressive strength, 18
uranium removal from groundwater, 4
UST R & D, 56
Value Engineering (VE), 35
cost effective, 35
vapor extraction, 45
vapor extraction systems
performance characteristics, 45
design parameters, 45
effective radius, 45
predictability of performance, 45
ventilation systems, 59
Ventura Country waste reduction plan, 30
volatile organic compounds (VOCs), 51,59
in situ removal, 45
desorption from soils, 45
volatile organics removal, 1
waste
603
-------
mixed, 59
characteristics, 61
waste fixation
Thiourea enhancement, 20
waste gypsum - phosphogypsum, 54
waste minimization, 30,55
Material Safety Datasheet (MSDS), 26
waste monoliths, 21
waste neutralization, 14
waste reduction
attitudes for, 27
management initiatives, 27
audits, 27
hous ekeep i ng, 27
case studies, 27
technology modi fications, 2 7
local government, 30
waste stabilization
soil and sludge, 20
roasting enhancement, 20
treatability study, 20
wastewater, 23,52
treatment, 25
Predetermined Standards, Guidelines and Criteria (PSCGs), 37
criteria, 37
treatment, 43
water holding and retention capacity, 54
water reactive wastes, 50
wet scrubber, 33
WRITE program, 30
yttrium, 54
zeolites, 49
zinc cyanide plating solution, 46
She editor of these Proceedings wishes to tender special thanks to Danny Yao
for his efforts in the production of this volune.
604
•&U. S. GOVERNMENT PRINTING OFFICE 1989/648-163/00319
------- |