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
.»§
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

-------