4
Committee on the
Challenges of
Modern Society
EPA/600/R-93/012C
February 1993
      Demonstration of Remedial
       Action Technologies for
Contaminated Land and Groundwater

              Final Report

            Volume 2-Part 2
                Number 190
          North Atlantic Treaty Organization
                1986-1991

-------

-------
NATO/CCMS
          COMMITTEE ON
       THE CHALLENGES OF
        MODERN SOCIETY
           Number 190
                FINAL REPORT
                      Volume 2
                    Appendices
              Part 2: Pages 663 through 1389
       Demonstration of Remedial Action Technologies
          for Contaminated Land and Groundwater
                  Pilot Study Directors

            Donald E. Sannlng, United States - Director
              Vokler Franzius, Germany - Co-Director
            Esther Soczo, The Netherlands - Co-Director


                 Participating Countries
                 Canada
                 Denmark
                 France
  Germany
The Netherlands
 United States
              Observer and Other Countries
                 Austria
                  Italy
                  Japan
   Norway
   Turkey
United Kingdom
                    1986-1991
                                          Printed on Recycled Paper

-------
 Disclaimer

 This publication has been prepared under the auspices of the North Atlantic Treaty Organization's Committee on
 the Challenges of Modem Society (NATO/CCMS).  It is not  a publication of the United States Environmental
 Protection Agency or by an agency or department of any other country.

 The project reports included in Volume 2 are as they were presented.  They have not been revised to bring to a
 consistent level of detail.  For example, some project reports have little baseline data or little discussion of inde-
 pendent evaluations of the project.  They are  included  here to provide general information about that type of
 technology.

 The tables and figures presented in Volume 2 were reproduced without improvement from the author's originals.
 The reproducibility of these tables and figures cannot be guaranteed.


 Printing support of this document has been provided by the United States Environmental Protection Agency.
 Keywords

Aerobic
Anaerobic
Aqueous extractments
Aroctor(s)
Binder screening
Binders, generic
Binders, Portland cement
Bodegradatfon
Biological treatment
Bforeactor
Cement, Portland
Chemical treatment
Chloroform
Composting
Dehalogenation
Dtahlorobenzene
Dfoxln(s)
Electro-osmosis
Electro-phoresls
Electro-reclamation
Emissions, stack gas
Encapsulation
Extracting agents
Extraction
Fixation
Fractured rock
Furan(s)
Ground water, contaminated
Ground water, contaminated, treatment
Hazardous waste
Hazardous waste site
High-pressure jet
Immobilization
In-srtu
Incineration
Indirect heating
Infrared incineration
Kiln, rotary
KPEG
Landfarming
Leach testing
Microbial treatment .
Microorganisms      ,      .,
Oxidation
PCB                         .
Pesticides
Pozzonanes
Pyrolysis
Remedial investigation/Feasibility study
Remediation
Remedy selection
Separation
Soil, contaminated
Soil, contaminated, treatment
Solidification
Stabilization
Stack gas emissions
Thermal desorption
Thermal destruction
Trichloroethene

-------
Foreword
The Council of the North Atlantic Treaty Organization (NATO) established the "Committee on the Challenges of
Modem Society (CCMS) in 1969.  The CCMS was charged with developing meaningful environmental and social
programs that complement other international programs with leadership in solving specific problems of the human
environment within the NATO sphere of influence; as well as transferring these solutions  to other countries with
similar challenges in environmental protection.

Ground water and soil contamination are among the most complex and challenging environmental problems faced
by most countries today, and there is ah ongoing need for more reliable, cost-effective cleanup technologies to
address these problems.   Many governmental and private organizations, in many countries, have committed
resources to the  development, test and evaluation and demonstration  of technologies to meet this need.  The
ongoing challenge to these organizations is how to maximize the value of these technology advancements and
effectively transfer the information to people responsible for making decisions and implementing remedial actions.

Consequently, a NATO Committee on the Challenges of Modern Society  (NATO/CCMS) Pilot  Study on the
Demonstration of Remedial Action Technologies for Contaminated Land and Ground Water was conducted from
1986 through 1991.  It was designed to identify and evaluate innovative,  emerging, and  alternative remediation
technologies, and transfer technical performance and economic information on them to potential users. The Study
was conducted under the joint leadership of the United States, Germany, and The Netherlands. In addition to these
co-pilot countries, Canada, France, and Denmark actively participated  throughout the five year study.  Norway
participated as an "observer" nation, and the United Kingdom, Department of the Environment was represented at
conference and workshop meetings.  Japan was represented at the initial International conference.  Organizations
from Hungary and Austria attended the Fifth International Meeting.

This is the detailed report of the findings, conclusions and recommendations produced by that Study. It is intended
to serve as a reference to the state-of-the-technologies examined by the participants. It is not intended to be a
manual on technology applications but as a guide to the potential application of different technologies to various
types of soil and ground water contamination. The conclusions reached from this Study revealed both the strengths
and weaknesses of current technologies as well as what efforts are  needed to increase the effectiveness of
remediation tools and their application.

There are several volumes to this report. Volume 1 is the report itself. Volume 2 is the Appendices, and  comes in
two parts.  Part 1 contains overviews of national environmental regulations, and papers by NATO/CCMS Guest
Speakers; it consists of pages 1 through 662.  Part 2 contains the final  reports of the NATO/CCMS Fellows, and
reports of the individual projects; it consists of pages 663 through 1389.

A limited number of copies of this report are available at no charge from two sources: NATO Committee on
Challenges to Modern Society, Brussels, Belgium; or U.S. Environmental  Protection Agency, 26 West Martin Luther
King Drive, Cincinnati, Ohio 45268, United States. When there are no more copies from these sources, additional
copies can be purchased from the National Technical Information Service,  Ravensworth Building, Springfield, Vir-
ginia 22161, United States.

                                          Donald  E. Sanning
                                         Pilot Study Director
                                  U.S. Environmental Protection Agency
                                            United States
              Volker Franzius
              Co-Director
              Umveltbundesamt
              Germany
Esther Soczo
Co-Director
Rijksinstituut voor volksgezondheid
en milieuhygiene (RIVM)
The Netherlands

-------
Abstract
This publication reports on the results of the NATO/CCMS Pilot Study "Demonstration of Remedial Action Tech-
nologies for Contaminated Land and Ground Water" which was conducted from 1986 through 1991.  The Pilot
Study was designed to identify and evaluate innovative, emerging and alternative remediation technologies and to
transfer technical performance and economic information on them to potential users.

Twenty-nine remediation technology projects were examined which treat, recycle, separate or concentrate con-
taminants in soil, sludges, and ground water. The emphasis was on in situ and on-site technologies; however, in
some cases, e.g., thermal treatment, fixed facilities off-site were also examined.  Technologies included are:  ther-
mal, stabilization/solidification, soil vapor extraction,  physical/chemical extraction, pump and treat ground water,
chemical treatment of contaminated soils, and microbial treatment.

This report serves as a reference and guide to the potential application of technologies to various types of con-
tamination; it is not a design manual. Unique to this study is the examination and reporting of "failures" as well as
successes.

-------
Acknowledgements

The Pilot and Co-Pilot Study Directors thank all who made significant contributions to the work of the Pilot Study:

Those representing their countries made  a major contribution to the direction of the Study  by recommending
projects within their respective countries which would be of particular interest to this Study, and by discussing the
regulatory and general environmental technology situations in their countries.

The various chapters were written by the respective authors after reviewing reports prepared on the Case Studies
for the meetings of the Study Group.

Good use was made of the NATO/CCMS Fellowship Program to further enhance the value of the Study and a
number of Fellows contributed directly to the preparation of this report. Robert Olf enbuttel of Battelle also served as
the general editor for the report, supported by Virginia R. Hathaway of JACA Corp., editor.

Expert speakers, often supported by NATO/CCMS travel funds, participated in the workshops and conferences of
the Pilot Study and contributed to the work of the Pilot Study Group.

Until his retirement, the NATO/CCMS International Staff was represented by the former CCMS Director, Mr. Ter-
rance Moran. Dr. Deniz Beten replaced Mr. Moran and attended the Fifth International Meeting.

Ms. Naomi Barkley of the Office of  Research and Development, Risk Reduction Environmental Laboratory, Super-
fund Technology Demonstration Division, U.S. Environmental Protection Agency served as Task Project Manager
for this project.

The names and addresses of all participants in the Study Group are given in Volume Two.
                                                 in

-------

-------
Contents:   Volume 2
Foreword  	i
Abstract	ii
Acknowledgements  	 iii

Part I

Appendix 1
1 - A  IMATO/CCMS Tour de Table: National Regulations and Other Overview Topics
      (Given by Countries and Institutions Involved in the NATO/CCMS Pilot Study)

      Participating Countries
        .Canada	 3
         Denmark	  17
         France	  23
         Germany	  27
         The Netherlands	  37
         United States	  45

      Observer and Other Countries
         Austria	  71
         Norway	  83
         Turkey   	  91
         United Kingdom	  99

      United States/German Bilateral Agreement on Abandoned Site Clean-up Projects	113

1 - B  Presentations by NATO/CCMS Guest Speakers	147

      Brett Ibbotson, Canada - AERIS, an Expert Computerized System to Aid in
         the Establishment of Cleanup Guidelines  	149
      Colin Mayfield, Canada - Anaerobic Degradation	161
      A.  Stelzig, Canada - Cleanup Criteria in Canada  	171
      Troels Wenzel, Denmark - Membrane Filtering and Biodegradation	211
      Herve Billard, France - Industrial Waste Management in France	223
      Christian Bocard,  France - New Developments in Remediation of  Oil
         Contaminated  Sites and Ground Water  	255
      Jean Marc Rieger, France - Incineration in Cement Kilns and Sanitary
         Landflliing  	273
      Bruno Verlon, France - Contaminated Sites - Situation in  France	285
      Fritz Holzwarth, Germany - Cleanup of Allied Military Bases in the
         Federal Republic of Germany  	293

-------
       James Berg, Norway - Cold-Climate Bioremediation: Composting and
          Groundwater Treatment Near the Arctic Circle at a Coke Works	295
       Gjis Breedveld, Norway - In Situ Bioremediation of Oil Pollution
          in Unsaturated Zone	307
       Gunnar Randers, Norway - The History of NATO/CCMS	 . .  .	315
       Robert L. Siegrist, Norway - International Review of Approaches
          for Establishing Cleanup Goals for Hazardous Waste Contaminated
          Land; and Sampling Method Effect on Volatile Organic Compound
          Measurements in Solvent Contaminated  Soil	 .321
       Guus Annokkee, The Netherlands - Biological Treatment of
          Contaminated Soil and Groundwater	479
       D.B. Janssen, The Netherlands - Degradation of Halogenated Aliphatic
          Compounds by Specialized Microbial Cultures and their Application
         for Waste Treatment	481
       Rene Kleijntjens, The Netherlands - Microbial Treatment		497
       Karel Luyben,  The Netherlands - Dutch Research on Microbial Soil
         Decontamination in Bioreactors	503
       Yalcin B. Acar, United States - Electrokinetic Soil Processing: A
         Review of the State of the  Art	507
       Douglas Ammon, United States - United States "Clean Sites"	535
       Roy C. Herndon, United States - Environmental Contamination in Eastern
         and Central Europe	 567
       Gregory Ondich, United States - The Use of Innovative Treatment
         Technologies in Remediating Waste	 579
       Ronald Probstein, United States - Electroosmotic Purging for In Situ
         Remediation	603
       Hans-Joachim  Stietzel, European Economic Community	! . . 635

Part 2

1 - C  Final Reports by NATO/CCMS  Fellows	663

      Alain Navarro,  France - New French Procedures for the Control of
         Solidificated Waste	665
      Peter Walter Werner, Germany - Biodegradation of Hydrocarbons	 677
      Alessandro di Domenico,  Italy - Sunlight-Induced Inactivation of
         Halogenated Aromatics in Aqueous Media:  Photodegradation Study
         of a Benzotrifluoride and an Evaluation  of Some Industrial Methods	711
      Sjef Staps, Jhe Netherlands - International Evaluation of In Situ
         Bioremediation of Contaminated Soil and Groundwater	741
      Resat Apak,  Turkey - Heavy Metal and Pesticide Removal from
         Contaminated Ground Water by the Use of Metallurgical
         Solid Waste Solvents	 •.,.. . .757;
      Aysen Turkman, Turkey - Cyanide  Behaviour in Soil and Groundwater
         and its Control	815
                                           VI

-------
      Robert Bell,  United Kingdom - Environmental Legislation in Europe	839
      Michael A. Smith, United Kingdom - In Situ Vitrification	843
      James M. Gossett, United States - Biodegradation of Dichloromethane
         Under Methanogenic Conditions	859
      Merten Hinsenveld, United States - Innovative Technologies for Treatment
         of Contaminated Soils and Sediments; and Alternative Physico-Chemical and
         Thermal Cleaning Technologies for Contaminated Soil	875
      Robert Olfenbuttel, United States - Summary Report: NATO/CCMS Pjlot Study on
         Demonstration of Remedial Action Technologies for  Contaminated Land and
         Groundwater	901
      Wayne A. Pettyjohn, United States - Principles of Ground Water:
         Fact and  Fiction  	903

Appendix 2
Thermal Technology Case Studies

2 - A Rotary Kiln Incineration, The Netherlands	911
2 - B Indirect Heating in a Rotary Kiln, Germany	929
2 - C Off-site Soil Treatment, Japan  	973
2 - D Electric Infrared Incineration, United  States  	987

Appendix 3
Stabilization/Solidification Technology Case Studies

3 - A In Situ Lime Stabilization (EIF Ecology), and Petrifix Process (TREDI), France  	995
3 - B Portland Cement (Hazcon, presently  IM-Tech), United States  	1011

Appendix 4
Soil Vapor Extraction Technology Case Studies

4 - A In Situ Soil Vacuum Extraction, The  Netherlands	1017
4 - B Vacuum Extraction of Soil Vapor,
      Verona Well Field Superfund Site,  United States	1031
4 - C Venting Methods, Hill Air Force Base, United  States  	1053
4 - D Additional case studies, United States	1075

Appendix 5
Physical/Chemical Extraction Technology Case Studies

5 - A High Pressure Soil Washing (Klockner), Germany	1081
5 - B Vibration (Harbauer), Germany	1101
5 - C Jet Cutting Followed by Oxidation (Keller), Germany	1111
5 - D Electro-reclamation (Geokinetics), The Netherlands 	1113
5 - E  In Situ Acid  Extraction (TAUW/Mourik), The Netherlands	1135
5 - F  Debris Washing, United States	,	1157
                                            VII

-------
 Appendix 6
 Ground Water Pump and Treat Technology Case Studies

 6 - A Decontamination of Ville Mercier Aquifer for Toxic Organics,
       Ville Mercier, Quebec, Canada	1167
 6 - B Evaluation of Photo-oxidation Technology (Ultrox International),
       Lorentz Barrel and Drum Site, San Jose, California, United States  .......	1207
 6 - C Zinc Smelting Wastes-and the Lot River, Viviez, Averyon, France  	1211
 6 - D Separation Pumping, Skrydstrup, Denmark	1231

 Appendix 7
 Case Studies on Chemical Treatment of Contaminated Soils: APEG

 7 - A  Supplementary Information on the APEG Process, Wide Beach, United States	1259
 7 - B  The AOSTRA-Taciuk Thermal Pyrolysis/Desorption Process, Canada  	1271
 7 - C  AOSTRA-SoilTech Anaerobic Thermal Processor Wide Beach, United States	1289
 7 - D  Site Demonstration of the SoilTech "Taciuk" Thermal Desorber, Waukegan
       Harbor, United States  	1293

 Appendix 8
 Mfcrobial Treatment Technology Case Studies

 8 - A  Aerobic/Anaerobic In Situ Degradation of Soil and Ground Water,
       Skrydstrup, Denmark	1299
 8 - B   In Situ Biorestoration of Soil, Asten,  The Netherlands  	   1325
 8 - C   In Situ Enhanced Aerobic Restoration of Soil and Ground Water,
       Eglin Air Force Base (AFB), United States	 . .          1327
 8 - D   Biological Pre-treatment of Ground Water, Bunschoten, The Netherlands  . . . '•.	1349
 8 - E   Rotary Composting Reactor  for Oily Soils, Soest, The Netherlands  	1363

Appendix 9
Lfst of NATO/CCMS Pilot Study Participants 	1377
                                           VIII

-------
                     Appendix 1-C




Final Reports by NATO/CCMS Fellows
 663

-------

-------
                                    NATO/CCMS Fellow:




                                    Alain Navarro, France




New French Procedures for the Control of Solidificated Waste
                       665

-------
                 NATO/CCMS  CONFERENCE
                                   ANGERS
                          Wednesday  November  7th
                      BULK AND SOLIDIFIED  WASTE
                        AN ADAPTED  PROCEDURE
 I - INTRODUCTION.

      The admission of hazardous industrial -waste or waste, containing hazardous
 substances is the object of a regulatory procedure. Certain wastes can either be admitted
 directly or absolutely excluded. All other categories of waste must be submitted to an
 admission procedure. In France, this procedure includes a leaching test carried out
 according to the AFNOR standard.
      The principle of this standard is to carry out a leaching test in order to determine the
 immeadiately soluble fraction when brought into contact with water. The results of this
 test therefore integrate both the solubility of the waste and the conditions of access of
 water to the waste. Hence, it is necessary to crush the waste to particle size of no greater
 than 4mm. Under these conditions, the test may primarily be considered as an arbitrary
 means of waste characterization, rather than a tool for predicting the mean or long term
 behaviour of the waste.
      Two large categories of waste should be excluded from the field of application of
 this standard:                                                                ,
      *   Bulk waste : the definition of bulk waste is, any waste that when produced
          consists of fragments greater than 4mm, on condition that this state be durable
          in time when under the influence of physical, mechanical or chemical agents.
     *   Solidified waste  : the definition being waste, either liquid or solid
          originally, but which has undergone a solidification treatment by the addition
          of binders (cement, molten glass, organic binders) according to a specific
          procedure.                                         ,
     In order to more accurately predict the behaviour of these 2 categories before
admission to landfill (class I), the possibility of developing a better adapted leaching test
•was studied, rather than applying the AFNOR standard.
                                    666

-------
     The first phase of this study was to carry out a literature survey followed by a
procedure proposal. The proposed procedure is now being tested in the laboratory, and it
is only when this second phase has been completed that any official decision can be made.
     The first phase of this work will be discussed here.

II - THE AFNOR STANDARD
     (see overhead 1)
III  -  LITERATURE SURVEY.

     The new procedure must achieve 2 objectives.

        *    to be adapted to the specific case of bulk or solidified waste, concerning the
             transfers which take place between the waste and water, and to the
             structural integrity.

     *      to be coherent with the AFNOR standard.

     The literature survey shows that we have a large number of operational tests at our
disposal, on a national and international level to:

           appreciate the physical and mechanical properties which govern the state, in
           the long term, of bulk and solidified waste.

           evaluate the teachability of waste, in particular,  solidified radioactive waste.

1) Principal results of leaching test.

     Globally, leaching tests can be divided into 3 main categories.(see overhead 2)
                                     667

-------
            THE  AFNOR  STANDARD X31 210
                          September  1988

This standard describes a leaching test to obtain, under defined conditions,
a soluble fraction which can be analysed for characterization purposes.

        It consists of :

                 1) general recommendations for waste sampling, either at
                   the  production or disposal stage.
                 2) sample preparation procedures for laboratory tests

                 3) procedures for leaching tests
              1
                             sample preparation
                             continual stirring
                             with  aqueous solution
lOOg waste + lliter water
demineralized water
stirring : 60/min
contact time 16 h
room temperature
                              leachate separation
                             analysis of leachate
       the residual material may be
       submitted to further leaching tests
filtration : (0.45|im)
centrifugation
                 4) an official experimental report

                                668

-------
A - Tests without renewal  of contact solution.

     A known quantity of liquid is brought into contact with a known quantity of waste
for a given period of time. 3 subcategories can be distinguished :

     - tests with stirring
     - tests without stirring
     - tests with sequenced extraction

 B  - Dynamic  tests.

     The solution is renewed,  either continuously or periodically. 3 subcategories can
 also be distinguished:

     BATCH TESTS : the waste is in particle form in a confined medium.

     FLOW AROUND TESTS  : the waste  is  in block form and the solution flows
 round it.

     FLOW  THROUGH  TESTS  :  the  waste is in block form and the  solution
 percolates through it.

 C - Hybrid tests.

      These can be divided into 2 categories:

      - saturation tests
      - extraction tests (SOXHLETfor example)
      All these categories can be distinguished by their different operating conditions

      * Nature of the  solution  :

      - water or dilute aqueous solution (natural, distilled, deionized, acidified)
      - more concentrated aqueous solutions (acids, brine)

                                      669

-------
       It would seem better not to use brine (particular case ofseawater) and strong acids,
  which correspond to unrealistic situations,
       Deionized water would be a better solution, eventually acidified with acetic acid
  (case of mixed waste containing biodegradable waste.

       * Sample preparation :

       All crushing must be excluded but certain wastes are of a dimension that special
  standard samples must be made either by moulding (solidified waste) or by cutting out
  (bulk waste). In the case of bulk waste of dimensions smaller than the chosen standard
  sample, a standard volume was considered.
      * Mode of contact with water :

      Systems with stirring are the best adapted to our approach : mechanical stirring
 (backwards and forwards) using a bar magnet, or a rotor, or by bubbling gas through the
 solution (nitrogen or carbon dioxide).

      * Liquid/solid ratio  :

      This ratio can easily be expressed for dry waste ; it must be more precisely defined
for wet waste. In the literature, the ratio solution I waste varies from 1: 1 to 100:1.

      *  Contact time :

     It can vary from 16 hours to 1 year. The AFNOR standard fixes this time at 16
hours.
     * Final separation of leachate ;

     In the case of tests without stirring, separation is carried out by filtration. In other
cases, the separation is  often  described precisely  (choice of filtration  mode,
centrifitgation, filter size. etc...
                                    670

-------
2) Tests used  to  evaluate the  structural integrity of the bulk materials.

        - visual appreciation : a rigourous nomenclature must be defined concerning
          fragmentation, softening, fissuring, colour changes, reduction to dust, etc...

        - mechanical strength.

     The mechanical behaviour of the waste has to be known in order to qualify it as
 "bulk waste" and to predict its behaviour at the time of testing.
     The literature shows that, to our knowledge, no specific procedure exists for waste,
for this aspect, and that we have to use procedures applied to construction materials. For
 example we can quote tests such as :

         - tensile strength

         - compressive strength ,for which a number of devices exist together with very
            rigourous procedures.
 3) Freeze/thaw  tests

      A number of procedures have been listed. The parameter "material" (porosity and
 tensile strength) and the operational parameters (water content at the time of freezing,
 temperature gradient, number of cycles) must be taken into account.

 4) Wetting I dry ing  tests

      The water content and water absorption capacity are the two essential parameters
 which must be known in order to carry out this test. The  various tests in the literature
 differ in their operating modes and as well as in the  criteria  chosen to evaluate the
 structural integrity.

 S) Resistance to  biological agents  test.

      The freeze!thaw tests  and the wetting/drying tests have given  rise to diverse
 protocols adapted to waste. This is not the case for biological tests which have been very
  little used. It is necessary to refer to tests concerning the biodeterioration of materials.

                                      671

-------
      - bacterial corrosion of metals,
      - biodegradation of plastic, paint, fuel, tarmac, rubber,
      - composting of municipal and industrial waste,
      - bioleaching of ores,
      - biodegradation of diverse materials (piping systems, joints, surfaces) in contact
        •with drinking -water,

6) Diverse ageing tests.


     Procedures exist related to the nature of the concerned materials and to their
mechanisms of physico-chemical degradation.
                                    672

-------
IV - SPECIFICATIONS FOR THE DEVELOPMENT OF AN ADAPTED
   LEACHING  TEST
     In France, there is a standardized test for waste prior to landfill. As bulk and
solidified waste have been excluded from the field of application of this test, it is
important to develop a specific procedure for these two categories.

     To define the new test it is important to :

     a)   to define the conditions which differentiate classical waste from bulk or
           solidified waste.

     b)    to determinie the shape of the standard sample to be adopted.

     c)    to adopt a leaching procedure which can be common to both bulk and
           solidified waste  and also  as close as possible to the existing standard
           procedure.

     d)   to  define the list of complementary tests which allow us to predict the
           behaviour of these wastes in their ultimate disposal site.

     e)   once this list of tests has been determined, to choose the techniques and tools
           adapted to each of these tests.
 V - EXPERIMENTAL PROTOCOL

      At the end of this study, and after discussion with the laboratory representatives
 concerned, a provisory protocol was elaborated. This protocol which we will describe
 here is presently being tested in the framework of an experimental programme, and it is
 only after this second phase that a definitive protocol will be elaborated.

      It is important to remind you here that :

      1)    this protocol is common to both bulk and solidified wastes.

      2)    It has tried to remain coherent with the existing french standard procedure.
                                     673

-------
       3)   It does not have the ambition of being a trustworthy simulation of the
            behaviour of waste, but is to be considered as a tool, easy to implement,
            reliable and applicable to these types of waste of any origin.

       4)   it must not be considered as a test destined to "validate" solidification
            technologies  in general. However such validation work could include the
            principles of these tests.

       5)   carrying out this test does not mean that the analytical operations for waste
            characterization are no longer necessary.

  VI -  CONCLUSION.

       The procedure presented here cannot be considered as defintive. It is in the process
  of being validated and may undergo modification at the end of the experimental phase.
  Taking into account the probable evolution of conditioning techniques and management
  strategies of landfills which accept industrial waste (class I), it is necessary to have such
,  control tools at our disposal
      ft must be noted that this procedure, which has been voluntarily simplified, cannot,
 by itself, be used as a technique for the validation of solidification processes in general.
      Finally, by confronting this procedure with procedures used abroad, the ultimate
 objective may be achieved, which is to elaborate a common international procedure.
                                 A. NAVARRO 1990
                                     674

-------
             TEST PROTOCOL
    Waste
                                       standard leaching test
    Preliminary test
    bulk or solidified?
       NO
                                    AFNOR X31
                                    210
            YES
   sample size 4 x 4 x scm
   possible?
YES
                  I
NO
 | Cutting  |      | Crushing]    10 < 0 < 20mm
Structural integrity test: mechanical resistance
and other tests (freeze/thaw,
humidification/drying,... ) according to needs
             YES
  further leaching tests
       675

-------

-------
            NATO/CCMS Fellow:



     Peter Walter Werner, Germany



   Biodegradation of Hydrocarbons
677

-------
         Report on activities in the frame of the
              NATO/CCMS Fellowship Programme
                         Title:
       Demonstration of Remedial Action Technologies
          for Contaminated Land and Groundwater
Aspects  of In  Situ Removal  of Hydrocarbons  from
   Contaminated  Sites-by  Biodegradation

                by  Peter Werner
                          678

-------
Dr. Peter Werner
Report on activities in the frame of the NATO/CCKS Fellowship Prograrane
Title:
Demonstration of Remedial  Action Technologies  for  Contaminated
Land and Groundwater
Activities:  1987: Participation  in   the
                   (Florida)
                   November, 9, 1987
                                SETAC-Conference   in   Pensicola
                    Visit  of the  Eglin Airbase  site
                    November,  10,  1987

                    Participation in  the  NATO/CCMS Meeting  in Washington, DC
                    November,  11-13,  1987

              1988:  Participation in  the  TNO/BMFT Congress  in Hamburg  (Germany)
                    April, 14, 1988

              1989:  Visit  of  the  institute  of  Erik  Arvin   (Department   of
                    Environmental Engineering in Copenhagen,  Denmark)
                    May, 11,  1989

                    Participation in an  International Symposium  on  Processes
                    and Governing the Movement and Fate of Contaminants in   the
                    Subsurface Environment, Standford University
                    July, 24-26,  1989
                                       679

-------
McCarty,  Stanford   Univarsity*   Stanford,

          Stanford   University,,   Stanford,

          Stanford   University,,   Stanford,
Contacted persons: Prof.  Perry
                   California
                   Dr.  Lewis   Semprini
                   California
                   Prof.  Paul    Roberts
                   California
                   Prof. Herb Ward,  Rice University,  Houston,  Texas
                   Prof.  Erik   Arvin,   Technical   University   of  Copenhagen,
                   Denmark
                   Prof. John Wilson,   Robert  S.   Kerr  Environmental   Research
                   Laboratory,  EPA,  Ada,  Oklahoma
                   Dr.  S.  Hutchins, Robert  S.  Kerr   Environmental   Research
                   Laboratory,  EPA,  Ada,  Oklahoma
                   Dr.  Doug Downey,  U.  S. Air  Force, Tyndall, AFB,  Florida
                   Prof.   Dr.   Muntzer,   Institut   de   Mecanique  des  Fluids,
                   Universite Louis  Pasteur, Strasbourg  (France)
  680

-------
Introduction

In the late 70's, the public interest in contaminated sites increased rapidly.
First aid measures consisted in replacing the polluted soil with  non-polluted
material.  The  more  people  investigated  soil  and  groundwater,  the  more
contaminated.areas could be found. Gradually, the public became more and  more
aware of the problems. Together with an increasing feeling of  responsibility,
a large industry of remediation measures developed during the 80 s.

Besides landfill, encapsulation, extraction, and incineration, which are  just
mentioned here, there is a strong interest in biological methods to  remediate
contaminated sites. The target of these methods using microbial activities  is
a complete mineralisation of the pollutants [1, 2].

The technical aspects  of  biological remediation  using on-site  and  in-situ
methods are described in a  great number of publications of  which only a  few
are refered to here [3, 4, 5, 6, 7].

The task of my fellowship  programme is not to  repeat  things and facts  which
have  already been described in  different journals or  presented at  different
congresses. Nor  is  it good  to believe  everything  announced  in  advertising
brochures of firms which offer biological remediation already commercially.

As  a  microbiologist,  I want to focus especially on problems still existing  in
the field of biological processes used for remediation of  contaminated  sites.
This  will help us to  get  a  better understanding for  the ongoing   processes.
Moreover, it is  necessary  to regard biological  processes on a realistic  base.
It  is not worth exaggerating  these methods   not knowing  enough   about  the
background. The  knowledge  on  biodegradation itself  must   be increased  before
the methods can  be  applied realistically. We must be aware of the problems and
must  admit that  we  do not  know  enough  about them. Otherwise we might be blamed
later on for having   dealt amateurishly  with  difficult question  such  as  the
remediation  of   contaminated sites.  Certainly,  it  is  very useful  to  apply
biological methods,  where   they  really   seem useful.  But  it is   impossible  to
solve all problems  with microbiology.  The  aim  must always  be  a combination  of
                                       681

-------
 different methods.

 The reputation of  microbial  processes  in the  frame.of  remediation  methods
 still  is very good and we should take care that it remains like that.

 In  order to increase  the possibilities   to use microbial   activities for   the
 elimination of  contaminants,   I  put the main  interest   in   the   fellowship
 programme on two problems which prohibit the biodegradation processes.

     1}  Bioavailability of the  contaminants
     2)  Question of oxygen sources  (air,  oxygen,  nitrate,  hydrogenperoxide)

 For  a  better understanding of the problem,  I  have to restrict my  explanations
 to the biodegradation  of  hydrocarbons.

 1) Bioavailability of  the contaminants

 Contaminations of  subsurface by  oil products  in general  result from  accidental
 spills or leaking  underground storages. Even  when mobile oil has been  removed
 by pumping,   residual  trapped  oil  can  be  a   long-lasting  source  of  water
 contamination by   soluble  hydrocarbons which  therefore   should  be  removed.
 Enhanced biodegradation   should  be  used  with the  support  "of  other  means,
 described elsewhere  [7].

For the application of in-situ  measures the following  preconditions must   be
guaranteed:
                                      682

-------
Table "I: Requirements for biological in-situ treatment
         MICROBIOLOGICAL POINTS OF VIEW:
         Biodegradability of the contaminants
         Concentration of the contaminants
         Absence of toxic substances (e. g. heavy metals)
         Solubility of the contaminants

         HYDROGEOLOGICAL POINTS OF VIEW:
                                      -4
         Hydraulic conductivity   5x10   m/s
         Flushing circuit
         Water treatment before infiltration
         Homogenious distribution of the contaminants
         Prevention of spreading of the contaminants
         Homogenious flushing through the contaminated soil
A detailed description of the procedure is given in [2].

Among the different factors, which limit biodegradation, listed in Table 2,  I
first want  to focus  on  the  bioavailability of  the  contaminants  for  the
bacteria.

Table 2: Limiting factors for the biodegradation in contaminated sites

         1) Physiological conditions  (composition of inorganic nutrients, con
            centration of the nutrients)
         2) toxic substances  (e. g. heavy metals)
         3) biodegradability of the pollutants themselves  (kinetic data)
         4) bioavailability (solubility, spatial separation)
                                      683

-------
The  problem  of  bioavai lability  is  characterized   in  just  four  words,,  given   in
Figure  1.
               bacteria  J
                 *>£
                                      soil
( pollutants j
Figure 1; Distribution-of pollutants  and microorganisms  in  polluted  soil
To enhance  the biodegradation  of  contaminants,  it   is   necessary  to  bring
microorganisms and pollutants into close  contact. Normally, the  contaminants
are very well attached to the soil or even  integrated  into the soil matrix  -
as it is known in the case of clay. We have to consider that the pollution  of
the soil normally  takes a  very long time  (in the  case  of  coalgasification
plants about 50 years) and  so diffusion plays a predominant  role. We  cannot
expect that the organics are released within a short period of time. So it  is
an urgent task to increase the concentration of the pollutants in the  aqueous-
phase to enable  the bacteria to  biodegrade them.  In  my  opinion,  it is  not
worth increasing the number of bacteria artificially because they are not able
to penetrate the soil, at least in the case of clay.

The first step of remediation  is to establish and  maintain conditions  which
allow biodegradation. The second step is  done by the bacteria themselves:  if
they find acceptable physiological conditions they will settle and proliferate
voluntarily.

In order to illustrate the problem, an  example of a laboratory experiment  is
described here. A column has been filled with sand that had been  contaminated
artificially with a consortium of  10 different PAH's.  The initial  concentra-
                                      684

-------
tions of the soil used  in the experiment are given  in the first  row of  the
next table.
Table 3: Concentrations of PAH's during biodegradation  in a laboratory  pilot
         plant [8]


Substance

Naphthalene (NAP)
Acenaphthylene (ACY)
Acenaphthene (ACE)
Fluorene (FLO) ,
Phenanthrene (PHE)
Fluoranthene (FLA)
Pyrene (PYR)
Benzanthracene (B2A)
Chrysene (GHR)
Benzo(a)pyrene (BZP)
Sum
Concentration (mg/kg)
Operation Time (d)

1 35 56 87
265.11 1.21 0.64 nn
318.66 13.52 9.93 1.37
356.45 43.42 16.78 3.03
378.19 146.76 88.79 13.10
379.42 226.13 171.80 23.52
400.13 357.27 398.88 55.62
' 397.68 385.36 339.42 58.27
40.53. 44.65 35.93 11.60
41.74 36.15 37.82 T9.66
40.99 38.52 36.65 24.39
2619 1293 1137 211



137
nn
0.94
2.30
7.53
12.55
29.56
30.47
11.26
17.28
22.95
135
Some  results  of the experiment   are  shown  in Figure 2  on which the number  of
hydrocarbondegraders  is  plotted  with respect to the operation time. As a proof
for the  bacteria, the incubation was done  with naphthalene  as a single  carbon
source.  In  the beginning,  there  is an increase  in the population of 2  orders
of magnitude, followed  by a  plateauphase which is  almost constant  over   a
period of 5 weeks. The concentration of  the PAH's during this time is about 23
mg/1  in  water and 2.6 g/kg in soil.  After  the plateauphase, there is a decline
of the number of bacteria   combined  with  a decrease in  the concentration  of
PAH's in water to a   minimum of  0.03 mg/1. The  concentration of PAH's in  the
soil, however, still  is  very high  (1.0 mg/kg soil).
                                       685

-------
 In the experiment presented here,   we added 0.5  % acetone  to   the  system  and
 were thus able to increase the concentration of hydrocarbons in the water. The
 result was another  increase in the number  of bacteria  in the system.  The
 microorganisms where  neither prohibited   by  acetone  nor   were they   able  to
 biodegrade it. At  the end  of the  experiment,   a concentration of about  140
 mg/kg soil  could be detected.  The value is  given  separately because there  was
 another step done at the   end  of the experiment which   should be discussed  in
 another scope. In order to increase the biodegradability we used UV-radiation.
                       10 15 20  25 30 35 40 45 50 55  60 65 70 75 80 B5  9C
                               OPERATION TIME  (d)

Figure 2; Hydrocarbondegraders  with  respect  to  the  operation  time  in  a
          laboratory experiment, showing the mineralisation of PAH's
          (In a parallel  experiment, in  which
                         i  sligl
          stipping effects [8]).
                                       HgCl   was   used  to   kill  the
bacteria, only a  slight decrease  of PAH's  could   be found due  to
From these data we can  conclude that the  solubility of PAH's   is  the   growth
limiting factor.
                                      686

-------
The elution  of pollutants  from  different  soils is  very  unlike  from  the
geological point of view and  can be  shown in an  experiment demonstrated  in
Table 4.   .      ...

In both soils  (sand and  clay) from different  contaminated  coalgasification
plants in Germany there is almost the same concentration of the consortium  of
PAH's (about 2,I/kg soil dry weight). An aqueous elution of 100 g soil in 1  1
of water in the result is almost undetectable in the case of clay while it  is
about 10 mg/1 in the  case of sand. For  that reason,  we have to  look for  a
method to increase the release of pollutants from a clay material.

Table 4: Aqueous elution of PAH's out of sand and clay from contaminated sites





Indene
Indan
Naphthalene
1 -Methyl naphtha! ene
2-Methyl naphthalene
Acenaphthene
Acenaphthylene
1,1-Biphenyl
Fluorene
Anthracene
Penanthrene
Pyrene
Fluoranthene
Crysene
Benz{ a) anthracene
Sum
CLAY
concentration
in the
soil
(mg/kg)
3,0
6,2
70
82
no
210
20
25
265
130
240
285
400
no
165
~2100
in the
supernatant
(mg/1)
>0,01
>0,01
0,04
0,04
0,07
>0,01
>0,01
0,02
0,02
>0,01
>0,01
>0,01
>0,01
>0,01
>0,01
~ 0,2
SAND
concentration
in the
soil
(mg/kg)
2,5
4,0
105
65
210
140
60
60
400
250
310
220
160
90
no
~2200
in the
supernatant
(mg/1)
0,10
0,20
6,00
0,80
1,00
0,90
0,10
0,20
0,15
0,08
0,15
0,09
0,05
0,01
0,02
~10
                                      687

-------
Prof. Paul Roberts and Dr. Lewis  Semprini from Stanford University deal  with
similar problems concerning bioavailability of contaminants in the subsurface.
They measure the effect of  sorption on the rate of  biodegradation [3].  This
group puts the main focus  on the biodegradation of chlorinated  hydrocarbons.
Another research  group at  Rice  University in  Houston carries  out  special
experiments with the biodegradation of  hydrocarbons in clay materials  (Prof.
Dr. Herb Ward). Thanks to the NATO/CCMS fellowship programme granted to me,  I
was able to  contact both  groups which deal  with these  problems. The  close
exchange  of  information  still  continues  and  another  visit  to  Stanford
University is planned with the support of the fellowships'^ financial remains.

The problem we have to tackle is the release of organics bound to the soil  in
order to increase the solubility  and therefore to enable biodegradability  in
the first place.

The disadvantages of the use of organic surfactants are summarized in [1]  and
[9] and are not explained in detail here. I just want to mention the  increase
of biomass and  gasproduction which  decreases the hydraulic  permeability  to
zero. Moreover, toxic surfactants can principally not be used.

The requirements a surfactant has to meet are:

- neither to prohibit nor to enhance biomass production
- no prohibiting effect on the biodegradation of the pollutants themselves
- no influence on the environment.

We were able to  find out  that pyrophosphates fullfil  the preconditions,  at
least on a laboratory scale.

The example of  the experiment  shows the effect  of Na-Pyrophosphate  on  the
release of PAH's attached to  sand. The  efficiency for the  two consortia, ,of
PAH*s is given  in the  figure of  Table 3.  The experimental  conditions  are
described at the top of the figure.  The best efficiency of pyrophosphates  to
increase the PAH content in the supernatant  is in the concentration range  of
0.1 %. Using 0.5 %, a decrease can already be observed which can be  explained

                                      688

-------
by sedimentation processes.  The sedimentation  time  in  the case presented  was
10 minutes.
         •-  30

         <
         Q_

          5  20
          C
          0)
          o  10
          C
          o
          o
                                                 I ACY.ACE,FLO,PHEN
                                                 I FLA,PYR.BZA.CHRY.BZP
Sand:
ACY
ACE
FLO
PHEN
FLA
PYR
BZA
CHRY
BZP
48.96 mg/kg
47.86 "mg/kg
48,12 mg/kg
51,74 mg/kg
49,86 mg/kg
47,62 mg/kg
 5,68 mg/kg
 4,86 mg/kg
 5,60 mg/kg
                     0.1   0.2  0.3   0.4  0.5   O.S   0.7   0.8
                     concentration Na-Pyrophosphate in  %
          0.9   1.0
 Figure 3:  Effect of pyrophosphates  on  the release of PAH's  attached  to sand
 The effect of pyrophoshates  on  the release of 'contaminants  is  even higher  but
 on the other hand it results in a  complete destruction of  the soil   structure
 itself. So far, we were  not  able to find any toxic or enhancing effect on  the
 biodegradation  of  the   pollutants  themselves.  In  the   frame  of  research
 programmes we follow the target to increase the concentration  of pollutants in
 the aqueous phase by releasing  them  from the soil. Besides pyrophosphates,  a
 consortium of other surfactants with similar effects is applied. An  intensive
 exchange of information  on this important  question with the purpose to  solve
 environmental problems  by biodegradation will be continued  in  future.
                                        689

-------
  2?  Question  of oxygen  sources

  Normally, the  biodegradation  of  hydrocarbons optimally  occurs under  aerobic
  conditions.  The  solubility  of   oxygen  in  water,  however,  is  limited  to
  approximately  9-10  mg/1  water   at  10-15°C,, which  is  the  temperature  of
  groundwater  in Central Europe. Because of the high amounts and  concentrations
  of pollutants  normally found in the spill,  the growth limiting factor is  the
  available oxygen. As a rule,  approximately 3 mg oxygen are  consumed for  the
  oxydation of   1 mg  hydrocarbons.  There are  some indications  that  aromatic
  hydrocarbons can be biodegraded under methanogenic conditions but the kinetics
 of these processes are too slow to be used as a remediation method [10].

 One way to increase the oxygen concentration is to use technical  oxygen  which
 will help to raise the values by the factor of 4 to 5. Another way is to   work
 alternatively with other additional oxygen  sources such as nitrates. For  the
 complete mineralization  of 1   mg  hydrocarbons about  3,5-4 mg  nitrates  are
 consumed. Theoretically,  the concentration  of nitrates  can  be raised  almost
 without limit.   From  this  point  of view,   nitrate   would  be  an   excellent
 electronacceptor for the   biodegradation  of contaminants   because  the  optimum
 final  products  are  carbondioxide,  water,  and free  nitrogen.

 But  unfortunately not  all  contaminants  can   be  biodegraded  in the  presence   of
 nitrates. It  is  evident  that  free oxygen   must be  available for the   first
 oxidation step  [5].  This  is  the  case  at least with  aliphatic  hydrocarbons.   On
 the  other hand,  only very  little  is known  about the mineralization of aromatic
 hydrocarbons  under  denitrification conditions.  It  seems as  if some of them  are
 inserted   directly   in  denitrification  processes  [11,  12].  The    problems
 concerning the  biochemical  pathways in  these biodegradation   processes  have
 been discussed  with M.  Hutchins   and  J.  Wilson  from  the   Robert  S.  Kerr
 Environmental Research Laboratory,  Ada, Oklahoma. The exchange  of  information
will be continued because this research group is still focussing, like we  do,
on these  interesting phenomena. The knowledge on the beneficial use of nitrate
for the remediation of contaminated sites is still   very poor. On the one hand,
a lot of fundamental work in the laboratory has to  be done, as it is described
in [12], On  the other hand,  it is  already necessary to  do practical  field
                                     690

-------
work. Our institute is involved  in two field projects in  which nitrates  are
used as additional oxygen sources. The results  will be published in 1990  and
are to be presented at the next TNO/BMFT Congress in Karlsruhe in 1990.

Problems with the usage of nitrates  mainly occur, besides in the  selectivity
of the substrate spectrum, in  the production of the  unwanted nitrite,  which
develops under so far undefined conditions.

Another  possibility  to  bring  additional  oxygen  into  biological  systems
consists in using hydrogenperoxide. There  are no problems with the  disinfec-
ting properties of this agent, as has been expected by microbiologists.  In our
own experiments concentrations up to 2000 mg/1  of HO  could be used  without
killing the bacterial population. These results are confirmed by the experien-
ces of Doug Downey [13].  Intensive discussions on the problems  and an  exchange
of experiences have  taken place  and are  also planned  for the  future.  Own
experiences were made in  a field experiment  comparing the use  of nitrate  and
HO  in two similar sites. The results will  be presented at the Conference  on
  o o
Subsurface Contamination  by   Immiscible Fluids  in  Calgary, Canada,  in  April
1990  [143.

The   advantage  of  the   usage  of   hydrogenperoxyde   is  that  it   works   as
electronacceptor  for  oxygen  and  that it  can  be  diluted  in  water  in  high
concentrations. The disadvantage  lies in the fact that the  agent disintegrates
very rapidly  to water  and free oxygen which  forms bubbles in  the water.

Therefore,  it is   difficult  to  transport  hydrogenperoxyde  in   the   subsurface
into areas  where  it  is  needed. The  question  of  stabilizing  this substances is
not  yet  answered.  On  the  other  side, due to   the  production  of  gas  bubbles in
the  aquifer,  the  hydraulic  permeability  decreases  rapidly.
                                       691

-------
 Conclusions and future plans for the fellowship programme

 The objective of the  fellowship programme  was to elucidate  problems in  the
 frame of remediation measures of contaminated sites using microbial activites.
 Due to  the financial  support  by NATO/CCMS  I had  the  chance  to  exchange
 informations with other research groups. The focus of my work lay on two  main
 topics concerning microbial remediation: the , bioavailability of  contaminants
 on the one side and the question of additional  oxygen sources on the other.

 The first step for a  successful biological degradation is  to bring  bacteria
 and  contaminants  into   close  contact.   Some  methods  to   increase   the
 concentration  of the pollutants in the  aqueous phase are described above.   It
 is only in solution that  optimum biodegradation conditions  can be  installed
 and above all  maintained.  Only when this problem has been solved, the question
 of additional  and/or  alternative oxygendonors   can be  taken into  view.   The
 application of nitrates  and hydrogenperoxide is discussed in the report.

 The NATO/CCMS   fellowship   programme  allowed  me  to  get  into   contact  with
 different research  groups  treating  similar   topics as  I  deal   with,,  All  these
 groups  are still  working in order to find solutions for  the problems,  to get  a
 better   understanding  of    the   limiting   factors,   and  to   overcome    the
 difficulties.  This  will  enable  a  more  successful   application  in  future  of
 microbial  methods for the  remediation  of contaminated  sites.       :

 In  order to keep  in  good touch with  these people,  I  plan  to  participate in  the
 above mentioned conference  in Canada  where  I will   have  the   oportunity  to
 present  results of  our own experiences.

The expenses of these travel activities  can certainly  not be  covered with   the
second part  of the  NATO/CCMS   grant  alone. But  nevertheless  "l  intend  to
participate in the conference and to visit  at least one university in order to
stay in  close contact and  to be  able to  exchange experiences personally, which
seems to be the best  way.  For that reason,   I do  not spare the  rest of  the
travel expenses to be covered privately.
                                      692

-------
References

[ 1]  Werner,  P.,  Brauch,  H.-J.:  Der  Abbau  von  Kohlenwasserstoffen   in
            kontaminierten Standorten.durch in-situ und on-site MaBnahmen. In:
            Altlastensanierung '88, K.  Wolf, J.  von den Brink,  F. C.  Colon
             (Hrsg.), 707-720 (1988)

[ 2]  Werner, P.: Experiences in the Use of Microorganisms in Soil  and Aquifer
             Decontamination. In: Contaminant Transport  in Groundwater,  Kobus
             and Kinzelbach  (eds.),  (1989), 59-63. Balkema. Rotterdam

 13]  Semprini, L.,  P.  V.  Roberts,  G.  D.  Hopkins   and  D. M.  Mackay.  A  field
             evaluation  of   in-situ  biodegradation  for  aquifer   restoration.
             EPA/600/S2-87/096,   U.    S.:   Environmental   Protection   Agency,
             Washington, D.  C.  (1989)   .  .  .   ;

 [  4]   Kobus and   Kinzelbach  (eels.),   Contaminant   Transport  in    Groundwater
             (1989),  Balkema,  Rotterdam

 [5]   Battermann, G.,  Werner,  P.:   Beseitigung einer  Grundwasserkontamination
             mit  Kohlenwasserstoffen   durch  mikrobiellen  Abbau.  gwf-wasser/
             abwasser US (1984), 366-372  .

 [ 6]  Nagel,  G.  Sontheimer,  H.,  KUhn,   W.  Werner,  P.:   Das   "Karlsruher
             Verfahren" zur aktivierten aeroben  Grundwassersanierung, Heft  29
             der  Veroffentlichungen, des  Bereichs  und  des  Lehrstuhls   fur
             Wasserchemie  am  Engler-Bunte-Insitut der  Universitat  Karlsruhe,
             1986

  [  7]  Franzius,  V.  Stegmann, R. und Wolf, K.:  Handbuch der  Altlastensanierung,
             Grundwerke, R. v. Deckers  Verlag,  G. Schenk  (1988)

  [  8]  Stieber,   M.,  Bockle,  K.,  Werner,   P.:  Abbauverhalten   von  PAK   in
             Untergrund (in preparation)
                                        693

-------
  [ 9]  Battermann,   6.   Werner,   P.:   Feldexperimente   zur    mikrobiellen
             Dekontamination. FGU-Korigress  ira  Rahmen  der  BIG-TECH,  Berlin,
             November 1987

 [10]  Grbic-Galic, D.: Microbial  Degradation of  Homocyclic and  Heterocyclic
             Aromatic Hydrocarbons under Anaerobic Conditions. To be  published
             in "Developments in Industrial  Microbiology", Vol.  30 (1989)

 [IT]  Hutchins, S. R.   and Wilson,  J. T.:  Evaluation  of Denitrification  for
             Biorestauration of an  Aquifer  Contaminated  with JP-4  Jet  Fuel.
             Presented  at  the  International  Symposium  on Processes   Governing
             the  Movement  and    Fate  of  Contaminants   in  the   Subsurface
             Environment,  Stanford University,  July 23-26,  1989

 [12]   RIB, A.,  Gerber, J.,   KeBler-Schmidt, M., Maisch,   H. U.,   SchweiBfurth,
             R.: Altlastensanierung mittels Nitratdosierung: Laborversuche  zum
                                  von   Heizol   (EL),   gwf-wasser/abwasser   129
mikrobiellen  Abbau
(1988), 32-40
[13]  Downey:,D. C.: Enhanced Bioreclamation. of a JP 4 Contaminated  Aquifer.
            Presentation at the SETAC-Conference in  Pensicola (Fl), .November
            9-12, 1987

[14]  Werner,  P., Battermann, G.:  In  situ-remediation of hydrocarbon  spills:
            microbiological  field   tests  with nitrate  and  peroxide.   To  be
            presented  at  the  Conference  on  Subsurface  Contamination   by
            Immiscible Fluids- in Calgary,  Canada,  in  April  1990
                                      694

-------
          Report on activities in the frame of the
               .NATO/CCMS Fellowship Programme
                           Title:
        Demonstration  of  Remedial Action Technologies
            for  Contaminated  Land and Groundwater
To be presented at the 4th International  NATO/CCMS Conference
           at Angers (France) November, 5-9, 1990
                          695

-------
 1. Introduction

 The paper presented here is the continuation of my report already printed  and
 documented in the Summary of the "Third International NATO/CCMS Conference  on
 Demonstration of  Remedial  Action  Technologies  for  Contaminated  Land  and
 Groundwater" in Montreal, Canada, November, 6 - 9, 1989.

 The task of my fellowship  programme is not to repeat  things and facts  which
 were already  described  in  different  journals  or  presented  at  different
 congresses.  Nor  is it  good  to believe  anything  announced  in  advertising
 brochures of firms which already offer biological  remediation commercially.

 As a microbiologist,  I want to focus  especially on  problems that still   exist
 in the field  of biological   processes used  for remediation  of  contaminated
 sites. This   will  help  us   to get   a  better  understanding   for  the   ongoing
 processes. Moreover,   it is   necessary  to  regard  biological  processes   on   a
 realistic base.  It is  not worth  exaggerating these methods not knowing   enough
 about  the background.  The knowledge  on biodegradation  itself must  be increased
 before the methods can  be   applied  realistically.  We must   be  aware  of the
 problems  and  must  admit  that   we  do  not  know enough   about them.  Otherwise  we
 might  be  blamed  later  for having dealt  amateurishly  with difficult  question
 such as the remediation  of contaminated  sites. Certainly, it  is  very useful to
 apply  biological   methods  where  they  really seem  appropriate.  But   it  is
 impossible to solve all  problems with  microbiology.  The aim must  always  be  a
 combination of different methods.

 In my  report  mentioned  above, I  put  the  main interest on  the two  problems
which  inhibit biodegradation processes:

                        1) bioavailability of the contaminants
                        2) question of oxygen source.
                                  696

-------
The conclusions of the former  presentation are not to be  repeated here,  but.
the aim of this paper should be  the experiences made when biological  methods
are applied for the remediation  of contaminated sites. Therefore,  I want  to
compare different methods already used worldwide and to describe their  appli-
cability with all  the advantages  and disadvantages.  Moreover,  contaminants
other than hydrocarbons are to be discussed here.

In addition to  the institutions  already  mentioned in  the first  report,   I
contacted the following firms:

                        TNO, Delft  (NL)                 '
                        Umweltschutz Nord, Ganderkesee  (D)

which have many  experiences  in the  use of microbial  biodegradation  processes.
                                   697

-------
  2. Contaminants                          :


  A lot of concern is concentrated on  the biodegradation of the following   sub-

  stances, that can be found in contaminated sites.  The table below contains the
  biodegradability itself and-the applicability in  practice.
  Table  1:  Contaminants  of  Special  Interest
 Contaminants
Biodegradabi-
bility Proved
in Laboratory
Biodegradabi
lity Proved
at the Site
Applicability
  Proved in
 Remediation
  Measures
 Aliphatic Contaminants

 Aromatic Contaminants

 Polyaromatic Hydrocarbons

 Volatile Chlorinated
 Hydrocarbons

 Phenols

 Non-volatile Chlorinated
 Hydrocarbons

 PCBs, Dioxins,  Furans

 Cyanides

 Heavy Metals
 With  respect  to  aliphatic  hydrocarbons  it  can  be  concluded  that  alcanas  (C
          fairly well  biodegradable.  Isoalcanes  and  alkenes  are  less  biodegra-

dable.  Cycloalcanes are  almost 'resistant.  From this   point  of  view  it  is  of
 special  interest to analyze  in advance the combination of pollutants at  a con-

 taminated  site in order  to decide about the most suitable remediation ; method

 Due to  the great amount  of contaminants normally found on spills with  alipha-

                                  698

-------
tic hydrocarbons, the question of  the additional oxygen source is  important.
Figure 1 shows  the  biodegradation pathway  of aliphatic  hydrocarbons  which
leads to the conclusion that nitro'gen can only be applied in combination  with
oxygen as primary oxidant.         : •
     CH3 - (CH2)n - CH3 + 0
      Aliphatic Hydrocarbon
                                 Oxygenase
•>  CH3  -
          - OH
Alcohol
                                                                   - 2H
                                                    CH3 -  (CH2)n
                                                         Aldhyd
        - C
            \,
                                                                   - 2H
                                                    CH3 -  (CH2)n  -
             .0
             •OH
                                                        Fatty Acid
                                                  Further  biodegradation  by
                                                  oxygen or  nitrates  possible!

 Figure'!:  Biodegradation  Pathways  of   Aliphatic  Hydrocarbons 'with  Respect  to
           the  Use  of  Nitrates
 Phenols  and  most  of  the  aromatic  hydrocarbons  can  be  biodegraded  fairly   well,
 reason why bioremediation  of  these  contaminants  is already used worldwide.  The
 concentration  of  these substances is  crucial for the  success.
                                   699

-------
 Polyaromatic  hydrocarbons  are  especially  important  in  Germany  since  its reuni-
 fication  because  of  the  large  number  of abandoned coke-oven  and coal-gasifica-
 tion  plants,  where these pollutants are predominant.  In  principle, biodegrada-
 tion  is possible. The  problems   of its applicability  in  practice are  already
 described  in  detail  in my  first  report and  should not  be repeated here.

 The biodegradation of  cyanides is also of   special  interest  in the areas  men-
 tioned above. Non-complexed cyanides  are well biodegradable  in  concentrations
 up to about 15 mg/1, although  it is one of  the most toxic inorganic  substances
 known. The experiences  show that  cyanide  in  abandoned industrial  areas  is
 complexed  and therefore not or only less toxic and  less soluble. On  the  other
 side, it is almost not biodegradable  in this form. Table 2 shows the occurence
 of cyanides and their toxic properties. As  a rule, Prussian  blue can be  found
 predominantly in abandoned coal-gasification plants.
Table 2: Occurence and Behaviour of Cyanides
Form
Solubility
Toxicity    Biodegra-
            dability
"Non-complexed"
KCN, NaCN, NH CN Cyanides
4
KOCN, NaOCN, NH OCN Cyanates
4
KSCN, NaSCN, NH SCN Thiocyanates
Zn(CN)
2
"Complexed"
K (Fe(CN) ) red
3 6
K (Fe(CN) ) yellow
KFe(Fe(CN) ) soluble Prussian blue
6 • *
Fe (Fe(CN) ) non-soluble Prussian blue
^ D O

high
high
high
low

high
high
low
not soluble

high +
not toxic +
low +
fairly toxic ?

low
not toxic
not toxic
not toxic
 predominant  in  abandoned  coal-gasification  plants
                                  700

-------
Volatile chlorinated hydrocarbons are fairly  well biodegradable under  diffe-
rent environmental conditions. A lot  of work was  done in this  field in  the
last few years.  The experiences  concerning this subject  were summarized  in
1989 by Werner and Ritter in a literature review called "Abbau- und  Biotrans-
formation von leichtfluchtigen Halogenkohlenwasserstoffen in der Umwelt  unter
besonderer Berticksichtigung der  Vorgange  im Untergrund".  The study  can  be
ordered at the Landesanstalt flir  Umweltschutz im Karlsruhe (Germany). Due  to
the complexity of  biodegradation pathways,  no case of  remediation based  on
microbial processes is known so far.

Only some of the  non-volatile chlorinated hydrocarbons  could be  biodegraded
under optimum conditions on  laboratory scale. An  applicability of  microbial
mineralization of these compounds is not known. PCBs, dioxins, and furans  are
tested for their biodegradability in laboratory set-ups. The results are  only
of scientific interest.and cannot be transferred to bioremediation measures.

Heavy metals cannot be eliminated  microbially. Only indirect mobilization  by
leaching processes are known.
3. Procedure in the Application of Bioremediation

The procedure to check the applicability of biological methods for the remedi-
ation of contaminated  sites using  in-situ or on-site  measures is  shown  in
Figure 2. The complexity  of the  scheme indicates that  the decision  whether
these methods can be applied is time-consuming. Although some of the tests can
be carried  out  parallely,  the  procedure takes  a  long  time  due  to  the
complexity of the contaminations.  The experience  shows that in  the case  of
hydrocarbons at least 4-6 months are necessary.
                                  701

-------
                        Sample of the site
                         (water  and soil)
                                I
        test of viable microorganisms with  the capability
          to biodegrade  the contaminants to be  treated
                  o    .   • x*
                                1
      determination  of
     the biodegradatiori
          potential
addition of
 bacteria
negative,  due to
 non—eliminable
toxic substances
      determination  of
      limiting factors
          alternative remediation
        methods, e.g.  incineration
     tests to select measures
        and to increase  the
     environmental conditions
       tests to prove the applicability
      of the measures in contaminated
        soil and water  samples of the
           site in laboratory scale
      (fermenter and percolation  setups)
        mass balance: biodegradation
                       abiotic processes
                      I
               if  abiotic processes
               are  predominant
          tests in pilot  scale  under
             practical conditions
                      I
Figure 2:  Procedure in the Application of Bioremediation

                              702

-------
4. Mixed Contaminations

Contaminations are generally named after the main substance polluting the site
(e. g. contamination with chlorinated hydrocarbons etc.). Most contaminations,
however, are caused by a mixture of pollutants, i.e. soil and groundwater are
contaminated to a different degree with a variety of chemical substances.  The
occurrence of a single substance is an exception which is usually confined  to
contaminations after transportation accidents. Single substances, however, can
be dealt .     more easily than a mixture of substances.

Typical examples of a mixed contamination can be found in abandoned gas works.
Analyses of soil and seepage water samples  often show high amounts of  pollu-
tants, as the maximum concentrations given in Table 3 prove. The high contami-
nation with aromatic hydrocarbons, such  as benzene, toluene, xylene, as  well
as polycyclic aromatic hydrocarbons (PAH)  is well demonstrated.  Furthermore,
high values of cyanide and  ammonium are to be found.  The water partly  shows
high concentrations of heavy metals and sulphates.
Table 3: Concentrations of Pollutants in Soil  and Seepage Water of  Abandoned
         Coke-oven and Gas Work Plants (Maximum Concentrations found so far)
Soil






Water




















, Benzol
Toluol
Xylol,
Naphthalene
Phenanthrene
PAH (Sum)
Cyanides (complexed)
Phenols
Aromats According to Maximum Solubility
Sulphate
Nitrate
Oxygen
Ammonium
ion
Manganese
ca.
ca.
ca.
ca.
ca.
ca.
ca.
ca.

ca.


ca.
ca.
ca.
5000 mg/kg
5000 mg/kg
5000 mg/kg
5000 mg/kg
5000 mg/kg
1000 mg/kg
1000 mg/kg
1000 mg/kg
• • 1
i
3000 mg/1
0 mg/1
0 mg/1
20 mg/1
20 mg/1
10 mg/1
This example illustrates the mutual influences of the pollutants on  microbial
                                   703

-------
 degradation.

 Substances such as ammonium, which  in the  first place is  no pollutant,  can
 impede the mineralization of hydrocarbons. Nitrification processes, i. e.  the
 oxidation of ammonium to nitrite or nitrate, consume oxygen, which is then  no
 more available for the oxidation  of the pollutants themselves.  This is  also
 true for readily degradable organic substances which compete with the contami-
 nants for the oxygen.

 Oxidation of  1 mg  ammonium  or degradable  DOC requires  3  - 4  mg  oxygen.
 Compared to this,  the oxygen consumption for oxidation of iron or manganese can
 be neglected.

 Hydrocarbons often show a competitive  degradation.  BTX-aromats, mineral   oils,
 and PAH probably do not mineralize at  the same degradation rate.  Better   solu-
 ble components  are in general   more readily degraded.  According to  laboratory
 and field experiences m-  and  p-xylene  are usually biodegraded  a lot  slower
 than other BTX-aromats.  In  some cases, biodegradation  of m-  and p-xylene  does
 not start until  the concentration   of  the  other contaminants  have  sunk to   a
 minimum.

 High concentrations of heavy metals  can  have  a  negative  effect  on   bacterial
 growth. Contaminations with  lead and/or   mercury, which  often   occur  in   aban-
 doned gas works, are  of   special importance. Here,   a microbiological  remedi-
 ation is  out  of  the question.  Even  if  bacterial  strains that  resist  to   heavy
 metals  and mineralize  hydrocarbons  are increased, soil and groundwater   cannot
 be  considered as   remediated due  to the  remaining  heavy  metals.  Presently,
 there is  no biological procedure for the  elimination of  heavy metals.

 Another problem are the cyanides which are found  as complexed cyanides in  gas
works, mostly in the shape of the non-toxic "Prussian blue". Although they  dp
 not  impede the degradation of  pollutants at all, they   cannot be  mineralized
themselves. In contrast, the degradation  of free cyanides is known and applied
in sewage plants of coke  oven works  for the elimination  of this  substance.
Chemical  analysis for  the detection  of  complexed cyanides  is only  possible
                                 ,704

-------
after special pre-treatment, so that often  no difference can be noticed  bet-
ween free and complexed cyanide.

Mixed contaminations, however, also show a mutual influence in a positive way.
Aliphatic chlorinated  hydrocarbons (e.  g.  trichlorethene)  are  aerobically
degradable in the presence of BTX-aromats. this mechanism is based on co-meta-
bolism which certainly plays an important role in the degradation of  contami-
nants. The knowledge about these processes, however, is still very little,  so
that the corresponding procedures can not yet be  applied on a large scale.  A
mixed contamination with pollutants of high and low solubility in water (e. g.
BTX-aromats and PAH) can sometimes bring about  a process in which the  compo-
nent with low  solubility is  mobilized by the  well soluble  one. This  might
increase the availability of the components with low solubility to the  micro-
organisms and thus fasten degradation.

Co-metabolic processes might also be significant in this case. The  importance
of the mentioned processes for  natural degradation and  the question  whether
these processes can be used for  remediation measures cannot be assessed  with
the present state of knowledge. As a conclusion, the following can be stated:

From the analytical data of a water  and soil contamination a possible  micro-
bial remediation procedure  can  be concluded.  Degradability of  a  substance
alone does not make sure that it is in  fact eliminated in a remediation  pro-
cess. In an in-situ as well as an  on- and off-site procedure all  biological,
chemical, and geohydrological factors have to be considered, most of which are
not even known. When bioremediation measures are planned, small scale  experi-
ments in the shape of controllable on-site and in-situ fields are necessary in
the first place. Their size  is determined by the distribution  of the  pollu-
tants and the geohydrological conditions.

All contaminants are  individually composed and  therefore require  individual
treatment. The type and size of contaminations in connection with  geohydrolo-
gical conditions can only be hints to the different possibilities of  remedia-
tion.
                                   705

-------
5. Toxicity and Mutagenicity

Not only toxicity and mutagenicity of the initial substances are decisive  for
the risk assessment with relation to the environment, but al^s'o the development
of these parameters during a remediation measure. Experiments in our own labo-
ratory as well as of other research groups have shown that the toxicity of me-
tabolic products is often higher than of the initial pollutants.

An example for this are the results of an own degradation experiment, given in
Figure 3 a and 3 b.
                                 706

-------
            O
            C7>

            £•
            O

            O

            O
35





30





25





20





15





10
                                      A	A DOC


                                      *	* Kohlenwasserstoffe
                                      "-»..*t.

                                        ^""""—•-».
                                              -<
                                                 V-._^._._.
                                                                 -15
                                                    -10
                                                                 -5
                                                        en

                                                        E
                                                        o
                                                       -4-J
                                                        W

                                                        0)
                                                        w


                                                        o


                                                        c
                                                                 -0
                0  5  10 15 20 25 30 35 40 45 50 55 60 65 70 75 80


                               Versuchsdauer  (d)
Figure 3  a:  Degradation of  a Mixture of Gaswork-specific Contaminants


             (Decane, Hexadecane, Pristane,  Naphthalene)


             - Development of DOC
          LoJ
          n:
           x
           <
100



 90



 SO



 70



 60



 50



 40



 30



 20



 10



  0
                                      Hemmung nach


                                            15 min A	A


                                            30 min *	*


                                                nach


                                             15 min a	a


                                             30 min
                                                  -25
                                                                 -20
 CD
r-

fn
 o
to
o
                                                                 -15   2
                                                                 -10
                                                                 -5
                0  5  10  15  20  25  30  35  40  45  50  55  60  65  70  75  80

                              Versuchsdauer (d)




Figure 3 b: Degradation  of a Mixture  of  Gaswork-specific  Contaminants


             (Decane,  Hexadecane, Pristane,  Naphthalene)


             - Development  of Toxicity (Microtox)
                                     707

-------
 The problem arising in this context is the fact that the metabolites in  gene-
 ral are  soluble  in  water  and  can  therefore  -  without  adequate  safety
 measures - drain off into the groundwater.

 If these substances are more toxic or  mutagenic than the less mobile  initial
 products, the r-isk of a negative impact on the environment through remediation
 is rather^high. So far, there are only few data about the behaviour of metabo-
 ^kr-pfoducts,  which are generally difficult to detect.

 In the procedure to test the applicability of biological measures described in
 chapter 3 the  registration of  the metabolites and their toxicological   impor-
 tance should especially be taken into consideration.
 6. General Conclusions

 The NATO/CCMS fellowship programme allowed me to get into contact with   diffe-
 rent research groups  dealing with similar topics  than I  do.  All  these   groups
 are still  working  on  solutions  for the problems  in  order to  get  a better   un-
 derstanding of the limiting  factors and to overcome  the difficulties  mentioned
 above.  This will,  in  future,  allow a more successful  application  of  microbial
 methods in the remediation of contaminated sites.

 Although biological treatment of  contaminated  soil and groundwater is  already
 in  use  worldwide even  on  a large  scale,  there  are  still a lot of  questions  to
 be  answered about  the  success and   further optimization of the processes.  The
 facts and  data of  the  different measures   applied in  special  cases  are  pub-
 lished  in  detail esewhere. In  order  to  improve  the systems, it   is  of  great
 interest and  importance to know more  about  the problems we have to  tackle  and
which occur during remediation  measures  based on biodegradation.

One of the main problems  we  have to  overcome is the  bioavailability of  the
contaminants for the bacteria.  The  limiting factors are both solubility of the
pollutants  (biodegradation only occurs in the aqueous  phase) and spatial  sepa-
ration mainly due to geological  conditions.
                                  708

-------
Furthermore, the question of additional  and/or alternative •electronacceptors
has to be taken into consideration. A lot of research has to be done to decide
which contaminants can be biodegraded with different oxygen sources.

One of the  main problems  in the  frame  of biodegradation  is based  on  the
mixture of several different pollutants  contaminating a site. A  lot of  work
has to be done to be able to mineralize  all pollutants in an acceptable  time
and with  acceptable efforts.  It  is not  reasonable to  eliminate  just  one
substance out of a whole  consortium. Therefore, the application of  different
methods, of which microbial degradation is one, is advisable.

Last but not  least, there  is the  problem  of metabolites  which has  to  be
solved. If metabolites are not avoidable the risk assessment of them has to be
determined  and taken into consideration when bioremediation is applied.

From this point  of view,  future  research  should  mainly  be  focussed on  the
questions mentioned above, which   have to   be  considered with  respect to  the
application  of microbial methods.
                                    709

-------

-------
                                     NATO/CCMS Fellow:

                             Alessandro di Domen/co, Italy

    Sunlight-Induced Inactivation of Halogenated Aromatics in
Aqueous Media:  Photodegradation Study of a Benzotrifluoride
              and an Evaluation of Some Industrial Methods
                        711

-------
         SUNLIGHT-INDUCED INACTIVATION OF HALOGENATED AROMATICS IN
        AQUEOUS MEDIA:  PHOTODEGRADATION STUDY OF A BENZOTRIFLUORIDE
               AND AN EVALUATION OF SOME INDUSTRIAL METHODS
                Alessandro di Domenico  and Elena De Felip
         Laboratory of Comparative Toxicology and Ecotoxicology,
             Istituto Superiore di Sanita, 00161 Rome, Italy
 ABSTRACT

      Aqueous solutions of 4-chloro-3-nitrobenzotrifluoride (NCTT)  were
 irradiated with a high intensity solar reactor.  No NCTT  disappearance
 was observed in pure water, or  in the presence of 2 %  cetylpyridinium
 chloride alone or  mixed with 2  % cetylpyridinium chloroiodide.   How-
 ever, photodegradation was  observed by adding  the semiconductor  Ti02
 (anatase powder).   NCTT  concentration diminished  by 50  % within  the
 first 4560 min of irradiation,  but the disappearance rate  decreased
 with increasing  length of  irradiation.    The experimental  data  were
 fitted with mathematical  models;  the  linear combination  of a  simple
 exponential with a linear function provided the best fit.
      In addition  to  the  experimental activity  carried  out  in  our
 laboratory,  this paper briefly describes  photochemistry and  photolysis
 equipment and  application for  industrial purposes.    Two examples  of
 photolytic processes utilized  for detoxification  operations are  also
 reported.

      Keywords;   aqueous   photodegradation,   benzotrifluorides,    photo-
 degradation  modelling, environmental  reclamation,  waste detoxification.
Note  The  present article  was prepared   for the  Fifth  International
Conference of the NATO/CCMS Pilot  Study on "Demonstration of  Remedial
Action Technologies for Contaminated Land  and Groundwater" (Washington,
DC, November 18-21, 1991).  The article  is part and a continuation  of
the experimental project "Sunlight-Induced Inactivation of  Halogenated
Aromatics in Aqueous Media", which was  presented by A. di Domenico  in
the November 1988  Bilthoven meeting  of the  same Pilot  Study and  is
currently in  progress in  our laboratories.   The former  presentation
should be  referred to  as a  background for  most of  the  information
omitted here.
                                  712

-------
INTRODUCTION

     In recent years  the role of  semiconductors in environmental  and
man-induced  photodegradation  of  halogenated  organic  compounds  has
stirred much interest  [Oliver et  al., 1979],   In fact,  it has  been
shown that aqueous suspensions  of finely powdered semiconductors  such
as Ti02, Sn02, and  ZnO—naturally occurring constituents of  sediments
and clays—can catalyze photochemical processes in chemicals  otherwise
not light-labile [Pelizzetti et al., 1985; Barbeni et al., 1985, 1986].
     The range of the economically important applications of photochem-
istry—and photocatalysis—is  very wide  [Bard, 1979;  Bloomfield  and
Owsley, 1982;  Pelizzetti, 1986;  Glatzmaier et  al., 1990a,b].   Among
these applications, we highlight: production of fuels from  inexpensive
raw materials, activation of selective pathways in the field of organic
synthesis (e.g. catalysts determining special steric configuration  and
high chirality),  inactivation of  hazardous waste  or toxic  chemicals
(e.g. dechlorination  and mineralization  of chlorosubstituted  hetero-
aromatics), and  sunlight-induced  transformations  (including  natural
degradation processes) of environmental contaminants.
     The present  article has  been developed  as a  continuation of   a
former study  [see footnote on page  1] and  is composed of two  distinct
sections.  In the first, preliminary   results of a pilot  investigation
on chemical photodegradation carried out  in bur laboratory are  report-
ed.  The second section deals specifically  with examples of   industrial
technology in photochemistry and   two  cases of light-induced  processes
to inactivate toxic  chemicals in diverse  matrices.
 LABORATORY INVESTIGATION

      In the past,  4-chloro-3-nitrobenzotrifluoride (NCTT;  Figure 1)  and
 some other benzotrifluorides were detected in a groundwater system used
 for private and agricultural supplies in northern Italy.   Contamination
 had been provoked by improper disposal of a chemical waste coming  from
 a local factory producing intermediates for pharmaceutical products and
 agrochemicals [Carli et al., 1983].
      Since we are currently investigating sunlight-based  photochemical
 techniques for detoxification of environmentally important  halogenated
 aromatics, we thought it of  interest to use NCTT  in a pilot study  in
 order to define specific experimental methods and criteria, as well  as
 a possible approach to select mathematical models suitable for describ-
 ing photodegradation patterns.
                                    713

-------
  EXPERIMENTAL

  4-Chloro-3-nitrobenzotrifluoride  NCTT  is  a  stable  chemical   exhibiting
  the  following  chemico-physical  properties: molecular weight,  225.6 amu;
  density,  1.542 g/mL; melting   point, -7.5   °C; boiling  point, 222  °C
  [Carli et al.,  1983].  NCTT has a moderate acute toxicity  in  laboratory
  animals and a  mutagenic potential (UDS  test)  [Benigni et al., 1982].
      The  UV spectra  of NCTT (Figure  1) were recorded  in high  purity
  water and in   n-hexane using  a standard quartz  cell.  The   absorbance
  reading for both solvents was negligible over the, full spectral  range.
  Only absorption bands  below 400  nm were visible,  with the  following
  maxima: (a) in water, e = 1.48  x 104 and 3.71 x ,103 L/(mol x  cm) at A =
  215 + 1 and 252 + 1 nm,  respectively;  (b) in .n-hexane, e = 1.86 x  104
  and 2.29 x 10  L/(mol x cm) at  A = 213 ± 1 and 284 ± 1 nm, respectively
  [De Felip et al., 1990,  1991],
      In order  to assess  the  water solubility  of NCTT,   a  saturated
 solution was prepared  by adding an  excess amount of  the chemical  to
 0.51 L high purity water in  a 1 L glass-stoppered Erlenmeyer  flask.
 The mixture was protected from light and stirred continuously for 24 h,,
 and then allowed  to rest for  the next  24 h prior , to sampling.    The
 latter was carried out by immerging the tip of a pipette into the satu-
 rated solution body,  well beneath the surface but away from the  undis-
 solved chemical  on the   flask  bottom;   the pipette  was retrieved  and
 wiped externally  before   releasing   the volume  collected.    For  each
 solubility trial,  1.00 ±  0.01 mL of  the saturated solution was  sampled
 and combined with n-hexane (50.0 ± 0.1  mL)  in a 100 mL   glass-stoppered
 Erlenmeyer flask;  a  mild stirring was   begun,  and   then 5  g   of a  1:4
 NaCl-Na2S04  mixture were  slowly added.    The  organic phase  was left   to
 dry overnight.   Determination of the  NCTT level  in  the  n-hexane   matrix
 was performed by gas  chromatography.  NCTT  was found to  be  soluble at a
 level of 174 ±  12  mg/L (N =  12)  at 20 °C [De  Felip  et al.,  1990,  1991].
      An NCTT mother  solution was obtained  by transferring an  aliquot
 (1/3  to 1/2 of  a liter) of the  saturated solution to a  1 L  (tolerance,
 ±0.4  mL)   glass-stoppered  volumetric  flask,  followed  by  taking  the
 volume up  with high  purity water.   For photodegradation  trials,  an
 amount of  this  solution was added with TiOg (2.0 + 0.1 g/L), or  any one
 of  the other reagents.  All standard  matrices were screened from  light
 with  aluminum foil  and stirred  continuously.   NCTT, from Rimar  Engi-
 neering (Trissino  (Vicenza), Italy),  was >95 %  pure, as confirmed  by
 NMB  (main  impurity: 4-chloro-3-nitrotoluene).

High purity water  High  purity  water was  obtained  by first   refluxing
 alkalinized distilled water in the presence of KMn04 for several  hours^
 in a glass apparatus.   Refluxed water  was then distilled off  and  col-
 lected.   Water purity was  checked spectrophotometrically [De  Felip  et
al., 1990,  1991],
                                  714

-------
Chemicals  Analytical  grade  n-hexane  was obtained  from  Carlo  Erba
(Milan, Italy) and used after distillation in a glass apparatus.  Carlo
Erba also provided analytical grade  anhydrous NaCl, NaOH, Na2S04,  and
KMnO
..
    High purity TiO-  (anatase)  powder,  for use as a
                                                         semiconductive
catalytic support  for NCTT  [Barbeni  et al.,  1985, 1986],  was  from
Aldrich Chimica (Milan, Italy).  Cetylpyridinium chloride was furnished
by Chemical  Market  Research Service  (Eschen,  FRG);  cetylpyridinium
chloroiodide was made in the laboratory [Botre et al. , 1979].

Glassware  Pyrex glassware was  used throughout.  Items were  carefully
cleansed with detergent  and water,  rinsed with  distilled water,  and
kept at 250 °C overnight prior to use.
     The cells  for  irradiation  (Figure 2)  were  dark  flat-bottomed
cylinders (o.d., 3.0  cm; volume, 20  mL) with an  open flanged top  to
accomodate  a  transparent  quartz  disk  (o.d.,  4.0  cm;   thickness,
2.0 ±0.1 mm), later sealed  with high vacuum sealant to prevent  vola-
tilization losses  [De Felip et al., 1990, 1991].

Apparatus  A  Hewlett-Packard  (Palo  Alto, California)  Model 5710  gas
chromatograph, equipped with an electron capture detector (GC/ECD)  and
an HP-5  10 m  long 0.53  mm i.d.  fused silica  column, was  used  for
analytical  assessment.   GC   conditions  were:  (a)  injection  block,
200 °C; (b) oven,  130 °C; (c)  detector, 300 °C;  (d) carrier and  makeup
(for both, argon-10 % methane  mixture)  flow rates, 2.0 and 25  mL/min,
respectively.
     A  "Solarbox"   photoreactor  (Figure   2)  was   purchased   from
CO.FO.ME.GRA.  (Milan, Italy).   Samples, placed  in the  center of  the
metal holder and   irradiated singly, were  kept  at 15  °C by  partially
immerging the  cells in water cooled  by a Haake  (Karlsruhe, FRG)  Model
F3 heat exchanger  [De Felip et al. , 1990, 1991].

NCTT assessment  In general, a portion  (0.50 or  1.00  mL) of the aqueous
matrix containing  NCTT was pipetted out  and mixed with 50 mL  n-hexane
in a 100 mL glass-stoppered  Erlenmeyer flask.   While stirring  mildly,
the organic mixture was  combined with  a 5 g NaCl-NagSC^  (1:4)  mixture
and allowed to dry overnight.  For quantification, a  0.50 mL volume  of
the dry solution was transferred to a  10 mL cone-shaped bottom vial and
diluted to 5.00 mL with n-hexane; then, a  <2 pL aliquot was   injected
into the gas   chromatograph.   Generally,  vials  were  kept  sealed  with
Teflon-lined screw caps.  GC/ECD determination was carried  out with the
external standard  technique  [De Felip  et al.,  1990,  1991].

RESULTS AND DISCUSSION

     The results   of this  photodegradation study  have been  described
elsewhere by De Felip  et al.  (1990,  1991) and  are hereafter  summarized.
                                  715

-------
       NCTT volatilization loss from TiOg-added aqueous media was checked
  by irradiating  samples in  cells which  were completely  covered  with
  aluminum foil.   Irradiation conditions and  lengths (up to 240 min)  in
  the different checks were the same as those later adopted in  photodeg-
  radation trials.    Since the  NCTT concentration  in unexposed  samples
  (60.1 ± 3.2  mg/L;  N = 10) was undistinguishable from the check  samples
  (59.7 ± 3.4   mg/L;  N =  10)  (Table 1),   the cells proved  to be  satis-
  factorily vapor-tight.
       During  photodegradation  trials,   which  took  several   days,   the
  stability of illumination energy was  spot-checked at different  irradi-
  ation times  (from  0 to  240  min)  by placing the  gauge  head  at a fixed
  position next to the sample  where illumination was optimized.   When all
  readings were  compared,  it   was   found that  the energy  had  remained
  constant (28.1+0.7 W/m2; N = 20)  throughout  the study (Table 1).
       Experimental   conditions were  as described  above;   independent
  trials were  repeated three   to ten  times and  provided  the outcomes  of
  Table 2.  For each   time, data expressed   in mg/L  (measured   concentra-
  tions) were  averaged and  then normalized against  the mean concentration
 value of unexposed  (t  =  0)  samples;  standard  deviation of   normalized
 values was estimated via  the  error  propagation  method [Kolthoff et al.,
 1969].  Before normalization,  single data   were checked so   as not  to
 bring any biases into the final data set.
      Normalized data appear   also in Figure  3, where three   regression
 curves are shown.   It  may be observed  that 50 %  of the initial  NCTT
 disappeared within  the first  45—60 min;  however, the  disappearance
 rate became less with increasing irradiation time.  The following  four
 mathematical models  were used  to  fit the  data: (a)  an  exponential
 function [I], and the linear combinations  of (b) I and a steady  state
 function [II],  (c)  I and a linear function [III], and (d) two Is  [IV].
 Aside from the latter, which  did not provide a converging  regression,
 the other  models   reached convergence  yielding  significant  fittings
 (Table 3; Chi-square test) [Hoel,  1971].  Fittings were performed with
 the derivative-free nonlinear regression BMDP-AR Program (BMDP  Statis-
 tical Software,  Inc., 1987,  Los  Angeles,  California).   From  Figure   3,
 it  is evident that  only III  fits  the experimental points closely,  while
 the other models do not,   especially for the longer irradiation  times.
 The regression equation of Model  III is:

           F(t)  =   (0.543  + 0.059)exp[-(0.0350  + 0.0067)t]  +
                  -   (0.00100 ± 0.00034)t   +  (0.464  + 0.060)

     Mathematical modelling   may have;  several  utilizations.    For   in-
 stance, III reflects  two distinct  analytical  trends:  therefore,  it   may
be presumed   that—under the   experimental  conditions  adopted in  this
 investigation—NCTT photodegradation is  associated with at  least   two
processes with different  (virtual) kinetics.  However, as no  data  were
                                  716

-------
taken for irradiation times >240 min, it cannot be excluded that  other
kinetic trends, if any, went undetected: that is, the modelling may  be
considered reliable, only over the  range of the available  experimental
data and additional data should  be gathered to improve the  predictive
quality of the model for longer exposures.  As shown in the caption  of
Figure 3, a specific use of the regression equation may be that of pre-
dicting how  long -it  takes for  given amounts  of NCTT  to  disappear.
Clearly, once the photochemical system has been characterized according
to the interactive components, the modelling approach might turn out to
be a  useful  tool  for describing  and  predicting  the  photochemical
behavior of a compound an both a natural and a man-controlled  environ-
ment.               ... • ' ;
     NCTT photochemical behavior was  studied also in different  media.
As could be predicted  on the basis of   its spectral features, no  NCTT
loss was observed in pure  water.  Similarly, losses were not  observed
when Ti02 was  replaced with  2 %  cetylpyridinium chloride  or with   a
mixture of 2 % cetylpyridinium chloride  and 2 %  cetylpyridinium chloro-
iodide  [Botre  et al.j  1979],  both producing micellar matrices.  As  to
the latter, this finding was  expected.
 INDUSTRIAL PHOTOCHEMISTRY

      It has been reported [see  footnote on page 1]  that  photochemical
 energy generally causes reactions from the potential energy surfaces of
 electronically excited states: because of that,  these reactions may  be
 highly selective and.very useful in organic synthesis.   The  electroni-
 cally excited molecules may react by fragmenting into neutral molecules
 or free radicals, by electron  transfer, by isomerization, or by  addi-
 tion to some  other molecule.   Complicated molecules  may be  obtained
 through one or  a few  steps directly from  relatively inexpensive  raw
 materials and at low temperatures.  In fact, temperatures near absolute
 zero appear  to have  been used,  although the  majority of  industrial
 photoreactions occurs in
 1982].
the range 0—125  °C [Bloomfield and  Owsley,
 LIGHT SOURCES

      To be of interest  for industrial uses, a  light source must  have
 the following characteristics: high  intensity in the desired  spectral
 region (for inactivation processes, generally between 250 and 400  rim),
 long life, emission stability, ease of operation, appropriate  physical
 dimensions",  and  minimum  amount  of  necessary  auxiliary  equipment.
 Man-made  light  sources have  generally been  developed for  production
 purposes; however, as will  be shown later, they  may also be used  for
 inactivation of hazardous materials.
                                    717

-------
  Mercury arc lamps  There  are  several  lamps  available  on  the  market, but
  in general  mercury arc  lamps   meet  all  the  above  requirements   [Roller
  1965; Bloomfield  and Owsley,  1982].   Mercury  is rather inert, does  not
  react with   the glass   and  the electrode   materials, and  produces  an
  emission spectrum which is  rich  in the UV  light component  (Table  4).
  The characteristics of  the lamp   (bulb shape  and material,   electrodes,
•  mercury vapor pressure,  and  starter)  are  chosen to  meet the  overall
  features of the photochemical  reactor and its specific uses.
       The mercury  emission spectrum  contains a  very large  number  of
  lines, whose  relative  intensities  depend upon  vapor pressure  (high
  medium,  or low) of the metal,   arc tube diameter,  and applied  current.
  At medium operating pressures  «2 atm;  1 atm * 0.1 MPa), all  emission
  lines of greater  interest—at 253.7 (mercury  resonance  line),   265.2,
  280.4,   296.7,    302.2,   313.1,   365.4,   404.7,    435.8,  546.1,    and
  578.0 nm  appear distinctly over  the  dim continuous background   radia-
  tion.  However,  the  lines broaden  and the   background brightness   in-
  creases  with  increasing  pressure,  until   the  spectral  distribution
  approaches that  of a  continuum (approximately at   285 atm),  where   all
  traces of the line  character  of the  spectrum disappear.   As pressure
  increases,   the  253.7   nm  line   intensity  diminishes   due  to  self-
  quenching: low pressure  «0.001 atm)   lamps  emit this  radiation  almost
  exclusively.
      In  industry,   many  photochemical  reactions are  carried  out  by
  employing high pressure  mercury arcs operating in the  range 2—110 atm
  These lamps may have a high input wattage (up  to several or tens of kW)
  and require an  efficient cooling  system to dissipate  the heat.   For
  lamps running at lower   pressures  (powers of up  to 30 W), air  cooling
 may be sufficient as  heat evolution during  operation is low.   Medium
 pressure mercury arcs are used to simulate the UV component of sunlight
  (especially the B-section)   by filtering  out any  radiations <280  nm.
 *ow pressure  lamps are mostly employed in sterilization operations.

 Electrodeless mercury arc lamps  These   light sources emit as a  result
 of  the metal  being  subjected to 2450  MHz microwave  irradiation  [Bloom-
 fieldand Owsley,  1982].  As  compared  with conventional mercury   arcs,
 they have the advantages  of longer  lifetimes, better safety,  air-cooled
 operation, and shorter  startup  time.  In addition,  like the  convention-
 al  ones,  their spectral  energy distribution may be  changed  by  using
additives  (e.g. xenon).

Xenon arc  lamps  For many years mercury arcs  have been  the most  widely
used high energy light sources.  However, during the  last  three or four
decades there has been  a notably increasing  tendency  to utilize  high
pressure xenon arcs, particularly in industry and technology.
  ^   Between 10 and 40 atm,  xenon arcs are efficient sources of intense
Visible and UV  radiation with  a brightness  comparable to  that of  a
                                  718

-------
carbon arc.  Although in the 800—1100 nm region some strong lines  are
visible, at  these  pressures the  xenon  spectrum is  substantially  a
continuum extending into  the UV  and IR regions.   The lower  emission
limit of a  lamp is  normally set  at around 170  nm due  to the  fused
quartz envelope.  Lamps  with input power  between 0.15 and  20 kW  are
commonly marketed.  Mixed atmospheres of mercury and xenon produce  arc
spectra in which  the mercury lines,  although preponderant, are  inte-
grated with the IR components of xenon spectrum [Roller, 1965].

Sunlight  As was reported previously [see footnote on page 1], the  sun
may be  presumed to  be  a convenient  light  source in  some  chemical
syntheses and possibly a useful  tool in environmental and  man-piloted
degradation  processes.^   The  wavelength  distribution  reaching   the
earth's surface extends from approximately 300 nm to the far  infrared.
The minor  fraction of  this distribution  at wavelengths  <400 nm  has
great importance in  photochemistry because  of the  high energy  asso-
ciated with it.  A few  facts about sunlight features and  availability
are recalled in the following.
     The earth's atmosphere practically absorbs all of the radiation of
wavelenths <295 nm  which, however,  is available in  space  (Table  5).
Indeed, the solar  UV intensity  at the earth's  surface very  strongly
depends on a number  of factors including: time  of day, time of  year,
latitude, elevation above sea level, atmospheric turbidity,  and  thick-
ness of  the ozone  layer.  The  parameter named  "air mass",  M, is  a
common denominator for the first four factors as it provides a  measure
for the  thickness  of the  layer  of atmosphere  the  solar  radiation
traverses to hit the earth's surface in a given area: the larger M, the
larger the  amount of  energy of  the incident  light absorbed  by  the
atmosphere.  When the  sun is  directly overhead (in  the zenith),  the
length of the path  of the radiation through  the atmosphere is at  its
minimum: at sea  level, M is  set equal to  1 and is  >1 for any  other
position of the sun.  Outside the  earth's atmosphere, M is  0   [Roller,
1965].
     While traversing the  atmosphere, sunlight undergoes  considerable
scattering.  As a result, a significant amount of UV radiation  reaches
the earth's  surface from  the sky  as well  as directly  from  the  sun
(Table  5).  Depending  on the  hour of the  day, the  magnitude of  one
component may be greater or smaller than the other's.  However, over an
entire  day the  two magnitudes are comparable.
     On the whole, it may be envisaged that for  an  eventual  year-round
use in  photochemistry, sunlight would be most effective  in dry  tropical
areas elevated  as much as possible above sea level  and characterized by
a high  incidence  of clear  days.  Of course,  a hole   in the   overhead
ozone layer would also be of help!
     For specific uses  (e.g. laboratory studies, solaria),  sunlight may
be simulated with man-made  light  sources.  It is relatively simple to
                                   719

-------
 match any limited region of the solar spectrum, but the exact  duplica-
 tion in  its  entirety is  a  more difficult  problem.   For   instance,
 mercury arcs are deficient in the  red and infrared regions (Table  4),
 whereas xenon arcs require correction at both ends of the spectrum,  In
 general, sunlight is duplicated with a degree of approximation which is
 good for many  purposes by  using a combination  of arcs,  incandescent
 lamps, and filters.

 REACTOES

      Basically there are two different types of setups for  photochemi-
 cal processes.   In one case,  the  light source is a long cylinder  sur-
 rounded by a cooling jacket and immersed in the reaction medium (gener-
 ally,  in a  liquid form).   In  the other case,  the reaction medium  is
 irradiated by a  lamp placed  outside,  whose  light is  focused on  the
 medium by means of an appropriate reflector to avoid excessive loss  of
 radiated energy.   Further,  the reaction can be either batch or continu-
 ously operated.
      Although developed especially for  production purposes  [Bloomfield
 and Owsley,  1982],  it is appropriate  to provide a short description  of
 the main types  of photochemical  reactors  and of their possible use  in
 the light-induced inactivation or trasformation processes  of  chemicals.
 Figures  46  have been redrawn  from  the originals (cited  authors)   and
 somewhat simplified.

 Batch reactor  A typical   apparatus  (Figure  4)   is   made  of   a   large
 container equipped with one or several  light  sources  isolated  from   the
 atmosphere and ventilated by a nitrogen  flow.   However,  a  larger  amount
 of  heat  is generally  removed by  flowing  water  between  the quartz   inner
 wall and  the  quartz or  filter  glass outer wall  which make up,  together,
 the lamp  jacket.  The  cooling water  may be   replaced  by  a   solution
 optically active  as  to  modify the  overall spectral  quality  of   lamp
 emission.  The  container   can also  dissipate  excess heat  through  a
 heat-exchange outer jacket.  The  batch chemicals are stirred   mechani-
 cally and, if necessary, kept under an inert gas.  On the other hand, a
 gaseous reactant may be insufflated  into the chemical mixture  through
 the same gas inlet.
     As the light output of a lamp is associated with several prominent
 emission lines  or  bands,  different  concurrent  photochemical  reac-
 tions—requiring energy in each  wavelength region—could be  performed
 at the same time by creating appropriate concentric optical niches sur-
rounding the lamp.  However, even when the energy radiated by the  lamp
 is exploited at best,  some 50—65 % of the electric energy supplied  to
the lamp is lost as heat (to be removed by a cooling system).
                                  720

-------
     reactor   Continuous reactors  (Figure 5)  are used  in  reactions
characterized by high  quantum yields.  They  can also be  used in  the
batch mode  and—because  of  their  flexibility—are  very  useful  in
studies at pilot plant level.

Elliptical reflector photochemical reactor  In this type of reactor the
light source  is placed  outside of  the reaction  mixture (Figure  6).
This reactor's design is very simple and, because of that, lends itself
successfully to  laboratory  uses  and  preliminary  investigations  on
photochemical processes  despite  the  fact that  the  radiated  energy
cannot be used as efficiently as  in other reactors.
 INACTIVATION OF TOXIC WASTE

      A  number  of  studies  have  shown  that  hazardous  chlorinated dibenzo-
 dioxins (PCDDs) are  rapidly degraded  by the  UV radiation present   in
 sunlight or from  an  artificial  source  [see footnote   on page 1].    For
 the process to take  place, PCDD   molecules must be  exposed to light   in
 the presence of a substance  acting  as  a  hydrogen-donor.   What  follows
 provides examples of photolytic   processes applied  to the  inactivation
 of two   specific  PCDD   members—the  nontoxic  1,2,3,4- and  the  highly
 toxic  2,3,7,8-tetrachloroidibenzodioxins  (also   known , as  TCDD   or
 "dioxin")—in  waste  and chemical  sludge.  Indeed,  aside from  inciner-
 ation,  the only PCDD destruction method to be employed on a  reasonably
 large scale has been photolysis.

 Solar detoxification  In recent years,  the US Department of Energy  has
 undertaken the development of solar detoxification technologies as part
 of a unified  project managed  by the Solar  Energy Research  Institute
 (SEEI)   in Golden, Colorado,  in conjunction with Sandia National Labora-
 tories.  The SEBI Project Manager,  John V. Anderson, has provided  the
 following synopsis  of  the  overall objectives  of  the  project   (see
 footnote on page  11): "to  identify and advance, new applications  for
 early  utilization  of  solar technologies,  and  to  address   a  growing
 national and  international need   for advanced,  innovative  technologies
 aimed  at environmental cleanup".  The  above technologies  are  concerned
 with (a)   treatment of   hazardous chemicals   in aqueous   media, and,  (b)
 destruction of chemicals in the  gas phase.  Development  of   commercial-
 ized solar detoxification technology is  expected by  the  mid-1990s.
      The cited activities would probably deserve a larger space in,this
 presentation.  However,  owing to the  vastness of  the matter and  espe-
 cially the fact  that   the SERI  endeavor  is still  relatively young,   we
 will postpone the matter to another occasion  and mention exclusively an
 example which also  has  relevance with the issues  of our own  project,
 the "Photox"  detoxification system  applied to the  tracer 1,2,3,4-tetra-
                                    721

-------
  chlorodibenzodioxin [Glatzmaier et al., 1990a,b].  The objective of the
  work, conducted in both laboratory and field tests, was to  demonstrate
  that concentrated solar energy can be used for destruction of hazardous
  chemicals with a very high efficiency  (99.9999 %) due to the  combined
  effect of heat and light.       :
       For the study—which has so far proven successful—a photochemical
  reactor was built and utilized.:   The vaporized chemical,  in the  pres-
  ence of air, was subjected to temperatures between 750 and 1000 °C  and
  flux levels of concentrated sunlight  in the range 500—1000 kW/m2,   an
  energy level  approximately  three  orders of  magnitude  greater  than
  normal sunlight.   The  sample size  of the tracer was chosen to  generate
  a  dilute solution in air which  is typical of processes for  decontami-
  nating TCDD-laden  soils.   Tests   also proved  that the energy in  the
  300—400 nm wavelength region of the solar spectrum—although  approxi-
  mately only 2  % of the total energy of the incoming flux—determined a
  significant enhancement of  the destruction reaction rate.

  TCDD photolysis at Syntex   In 1969,  Syntex Agribusiness, Inc.   (Spring-
  field, Missouri)   bought  a  chemical  plant   in  Verona,  Missouri,  to
  manufacture  animal  feed additives.   Part of the plant and property   had
  been leased  by  the previous   owner to another  company which  produced
  trichlorophenol and hexachlorophene.   In 1971 this  company went out  of
 business and abandoned  the plant.   Only   in  1974 did  Syntex find  out
 that the company had^lso abandoned on the property a large steel  tank
 containing over 16 m    (approximately 21 t) of  a dark chemical  sludge
 laden with some  300—400 mg/kg  (ppm) of  deadly TCDD  (on the  whole,
 approximately 7.3  kg).   After a  cost-to-benefit  evaluation,  Syntex
 decided it would be best if they disposed of the waste and TCDD in  the
 safest manner.   The whole  project  lasted six  years and entailed  many
 different operations which included:  assuring the security and  safety
 of  the storage tank until  a decision could be made on how to handle the
 matter, choice of disposal  method,   and development and application  of
 the photodegradation process [Anonymous,  1980; Worthy,  1983].
      Syntex soon  realized   that,  for  various  reasons,   they could  not
 ship the  contaminated  waste  elsewhere for  treatment  (e.g.   incinera-
 tion),  and started exploring the^possibilities of  on-site   destruction.
 They engaged a  company  known  to be experienced in  environmental   prob-
 lems and management  of hazardous   waste.   A committee   of  experts was
 formed  to obtain independent judgement   and guidance.  The  US  Environ-
 mental  Protection Agency as  well as other US and Missouri state  health
 and  safety agencies  were also involved throughout  the  phases of  the
 project.  After extensive  evaluation,  the  photolytic  method was  cho-
 sen—not  only as   economically feasible to   Syntex at  a cost of  US$

Note  From the article by J.V. Anderson, H. Link, M. Bonn, and B. Gupta
 Development of US Solar Detoxification Technology: An Introduction".
                                 722

-------
500,000 for  equipment  and  installation, but  especially  because  it
seemed to guarantee the highest  level of safety for personnel,  public
health, and  the environment.   In fact,  it was  a low  pressure,  low
temperature, fully closed system presenting minimal human exposure  and
minimal release  into the  environment.  In  addition, several  special
safety procedures were incorporated into the system.
     The approach required that TCDD first be separated from the  waste
by hexane extraction  of 600  L sludge  batches.  For  this step,  some
5.8 m3 of  solvent were employed.   After the mixture  was stirred  for
several hours, the heavy  residue was allowed to  settle and the  TCDD-
containing organic layer was  drawn off.  The  hexane extract was  then
exposed to high intensity UV light generated by an  array of eight 10 kW
mercury arc lamps.  TCDD was photolytically broken  down and the solvent
distilled for recycling  within the  process.  The  sludge and  aqueous
waste  left  from the extraction process  had only very minor amounts  of
TCDD  (overall removal efficiency, >99.9 %) and were disposed of accord-
ingly.  The photodegradation efficiency was estimated at >99.97 %.  The
entire disposal process  lasted approximately  11 weeks.
 ACKNOWLEDGMENTS

      The technical help of Fabiola  Ferri and Susan Holt is  gratefully
 acknowledged.   We are  also indebted  to G.  Briancesco  of the  Drawing
 Unit of the  Editorial Section  of the  Library of  this Institute  for
 preparing the drawings.
 REFERENCES

 Anonymous (1980):  Destroying dioxin:  A  unique approach.   Waste—Age
 11(10), 60—63.

 Barbeni, M., Pramauro, E., Pelizzetti, E., Borgarello, E., and Serpone,
 N. (1985): Photodegradation of pentachlorophenol catalyzed by  semicon-
 ductor particles.  Chemosphere 14, 195—208.

 Barbeni, M., Pramauro, E., Pelizzetti, E., Borgarello, E., Serpone, N.,
 and Jamieson,  M.A.  (1986):  Photochemical degradation  of  chlorinated
 dioxins, biphenyls,  phenols  and benzene  on semiconductor  dispersion.
 Chemosphere 15.  1913—1916.

 Bard,  A.J.  (1979):  Photoelectrochemistry  and  heterogeneous   photo-
 catalysis  at semiconductors.  Journal of  Photochemistry W, 59—75.
                                    723

-------
  Benigni,  R.,   Bignami,   M.,  Conti,   L.,   Crebelli, R.,   Dogliotti,  E.,
  Falconi,  E.,  and Carere,  A.  (1982):  In  vitro mutational studies  with
  trifluralin  and  trifluorotoluene  derivatives.    Annali  dell'lstituto
  Superiore di  Sanita 18,  123—126.

  Bloomfield, J.J.,  and Owsley,  D.C.   (1982):  Photochemical  technology.
  In: Encyclopedia  of Chemical   Technology.   3rd  Edn.,   Vol.   17,  pp.
  540—559.  R.E.  Kirk and D.F.  Othmer, Eds.,  John  Wiley and Sons' (New
  York),

  Botre, C., Memoli,  A., and Alhaique,  F.  (1979):  On the  degradation  of
  2,3,7,8-tetrachlorodibenzoparadioxin .(TCDD) by means of  a new  class  of
  chloroiodides.   Environmental Science and Technology 13,  228—231.

  Carli, G.,  Cosraa,  E.,   di Domenico,  A., Maori,   A., and Young,   C.P.
  (1983): Pollution by halogenated  aromatic compounds at Trissino: A case
  study of groundwater  contamination and   rehabilitation.   Disasters  7,
  266—275.                                                  '	

 De Felip, E.,  di  Domenico, A.,  Volpi, F.,  De Angelis, L.,  Ferri,  F.,
 and Botre,  C.   (1990):  Photodegradation  of  4-chloro-3-nitrobenzotri-
 fluoride in aqueous media.  In;  Organohalogen Compounds - Dioxin 90 and
 EPRI Seminar,  Vol. 4, pp. 211—214.   0.  Hutzinger and H. Fiedler, Eds.,
 Ecoinforma Press (Bayreuth).

 De Felip,  E.,  di  Domenico,  A.,;Volpi,  F.,  De Angelis,  L., Ferri,   F.,
 and Botre, C.  (1991): Photodegradation of a benzotrifluoride induced by
 simulated sunlight in aqueous media.   In: Ecological Physical Chemistry
 - Proceedings  of an International Workshop,  pp. 585—592.  C. Rossi and
 E.  Tiezzi,  Eds.,  Elsevier Science Publishers B.V.  (Amsterdam).

 Glatzmaier, G.C., Nix, R.G.,  and Mehos, M.S.  (1990a): Solar destruction
 of  hazardous chemicals.    Journal of environmental  science. *nH  health
 A25, 571—581.                   ~                         	

 Glatzmaier, G.C.,  Graham,  J.L.,   and Dellinger, B.  (1990b):   Comparison
 of  laboratory  and field experiments for the  destruction of tetrachloro-
 dibenzo-p.-dioxin  usin  concentrated solar energy.    In: Proceedings  of
 the 25th Intersocietv Energy Conversion Engineering Conference.   Vol  6
 pp. 256—261.                                                        '   '

 Hoel, P.G. (1971): Introduction  to Mathematical Statistics,  4th Edn.,
 pp. 226—232.   John  Wiley and Sons (New York).

Roller, L.R. (1965):  Ultraviolet Radiation. 2nd  Edn.  John  Wiley   and
Sons (New York).
                                  724

-------
Kolthoff, I.M.,  Sandell,  E.B.,  Meehan, E.J. ,  and  Bruckenstein,  S.
(1969): Quantitative   Chemical   Analysis.  4th  Edn.,  p.  399.   The
Macmillan Company (London).

Oliver, E.G., Cosgrove, E.G.,  and Carey, J.H.  (1979): Effect of  sus-
pended sediments on the photolysis of organics in water.  Environmental
Science and Technology 13> 1075—1077.

Pelizzetti, E., Barbeni, M., Pramauro, E., Serpone, M., Borgarello, E.,
Jamieson, M.A.,  and Hidaka,  H. (1985):  Sunlight photodegradation  of
haloaromatic pollutants catalyzed by semiconductor particulate  materi-
als.  La Chimica e 1'Industria 67, 623—625.

Pelizzetti, E.  (1986): Homogeneous  and heterogeneous  photocatalysis.
La Chimica e 1'Industria 68, 51—52.

Worthy, W. (1983): Both  incidence, control of  dioxin are highly  com-
plex.  Chemical and Engineering News 61, 51—56.
                                  725

-------
 Figure captions

 Figure 1  UV spectra of  NCTT (15.0 ±0.1 mg/L;  20 °C) in high  purity
 water (a) and n-hexane (b)  were recorded with a Hewlett-Packard  (Palo
 Alto, California, USA) Model 8452A diode-array spectrophotometer and  a
 standard 1  cm lightpath  cell.    Instrumental conditions  for  spectra
 acquisition were: spectral bandwidth, 2 nm; wavelength accuracy, ±1 nm;
 full spectrum  (190—820 nm)  scan  time, 0.1  s.  Absorbance  of  both
 solvents was negligible over the full spectral range.  NCTT appears  to
 exhibit absorption bands below 400 nm only.

 Figure 2  "Solarbox" photoreactor.   The  apparatus was equipped with  a
 1.5 kW  xenon lamp  and  a filter  system to  cut  off light  <300  nm.
 Illumination energy was monitored with a gauge (from CO.FO.ME.GRA.)  re-
 sponding to 295—400 nm light.   When  the lamp was on, the  irradiation
 chamber was  ventilated to  disperse  the heat.   For  photodegradation
 experiments, samples in irradiation cells  were placed one at the  time
 in the  center of  the holding  plate.   This  setup may  be  reasonably
 compared to that shown in Figure 6.

 Figure 3   Photodegradation induced  by simulated  sunlight of  aqueous
 NCTT in the presence of  suspended  TiOg.   Normalized experimental data
 and regression curves of  Models'I,  II,  and III are shown.   With  refer-
 ence to Model  III (mean  estimates),   it might be  predicted that NCTT
 photodegrades by 90,  99,  or 99.9 %  after 360,  450,  or 460  min  irradia-
 tion times,   respectively.    For such  estimates,   it  was  arbitrarily
 assumed that there was no change in  the kinetic  trend for  irradiation
 times >240  min.

 Figure 4 Batch photochemical reactor.    Higher wattage lamps are more
 efficient in producing a  useful   light  emission up  to approximately   an
 input power of  10  kW.   At that   point,  a decrease  in UV emission  effi-
 ciency (einsteins/(kW x hour)) is observed.  Therefore,  if the  reactor
 is  designed to  exceed a 10  kW input power,  an  array of lamps  is  normal-
 ly  arranged.  Lamps   immersed in the reaction medium improve  quantum
 efficiency by avoiding excessive scattering, reflection, or   dispersion
 of  light.   In the  picture,  both  lamps and reaction  medium  are cooled  to
 avoid overheating.

 Figure 5  Flow photochemical reactor.  Continuous reactors are used   in
 high  quantum  yield   reactions   for  vapor,  liquid,   or   mixed  media.
 Contact times are  short and high reaction  rates are  achieved.  The heat
 generated by the light source  is carried  away by appropriate  cooling,
 in the case shown, by ventilating  the lamp well with  air or.  nitrogen.
Owing  to  the continuous  flowing of  the  chemical   medium, cooling   is
 generally required solely to remove heat from  the light source.
                                  726

-------
Figure 6  Elliptical reflector photochemical reactor.  This is a system
in which the light source is outside the reaction medium.  To  maximize
utilization of the emitted light energy, radiation is focused upon  the
reaction vessel by an enveloping elliptical, highly reflective  casing,
lamp and vessel being placed at  the focuses of the ellipse.   Although
heat is removed from  the casing inside  by efficient ventilation,  the
temperature of  the reactant  mixture can  also be  kept under  control
through an additional cooling system.

Figure 7   Syntex  Agribusiness,  Inc.   (Springfield,  Missouri),  used
photolysis to  destroy  highly  toxic TCDD  in  chemical  sludge  waste
contaminated at levels between 300 and 400 mg/kg  (ppm).  For the opera-
tion, over 16  m3 of  sludge were  extracted batchwise  with hexane,  a
common organic solvent.   The organic  layer was  then irradiated  with
high intensity  UV light  to break  down TCDD.    From the  process,  an
overall photo-induced reduction efficiency of >99.9 % was attained; the
detoxified solvent  was distilled  for   recycling.  The  aqueous  layer
(TCDD < 0.002 mg/kg) and the  extracted  sludge  (TCDD < 0.5 mg/kg)  were
later disposed of.
                                    727

-------
Table  1   Stability tests.  The  left   half  of  the  table  shows
benzotrifluoride  concentrations  in samples wrapped with aluminum
foil before  ft=0) and after (t>0) irradiation. The  right half shows
readings of incident light  intensity where this was most intense on
the irradiation plate (middle point)
To/ A i IRRADIA TION TJME (min)
0.0 30 60 90 120 240
Reactant concentration (mg/L)
1 65.1 63.1
2 59.0 59.0
3 62.3 63.5
4 55.2 53.3
5 55.0 56.1
6 60.261.1
7 62.0 62.9
8 62.3 62.0
9 60.1 58.0
10 59.7 57.8
0.0 30 60 90 120
?40
Incident light intensity (W/m 2)
27.9 27.8
28.3 28.5
29.2 28.9
27.5
28.6
28.9 29.2
27.3 27.3
27.7 27.6
29.0 28.3
27.6



27.3
28.0




27.4
Means and standard deviations
/7=f0 60.1 ± 3.2 mg/L
*N=10 59'7 ± 3-4
f>0
AM20 59'9 ± 3'2
28.2 ±0.7 W/m2
28.0 ± 0.7

28.1 ± 0.7




                              728

-------
Table 2   Photqdegradation induced by  simulated  sunlight  of the
benzotrifluoride in aqueous solution in the presence of
   TRIAL
                            IRRADIA TION TIME (min)
            0.0    15     30    45
                           60
                           90   120    180    240
     1
     2
     3
     4
     5
     6
     7
     8
     9
     10
65.1
59.0
62.3
55.2
55.0
60.2
62.0
62.3
60.1
59.7
Mean(mglL) 60.1
 SDb(mg/L) 3.2
   VCC(%)  5.25
     N      10
48.3
50.1
50.0
50.0
48.0
47.5
45.3
45.0
45.0
46.5

47.6
 2.1
4.34
 10
 Reactant concentration (mg/L)
37.5         27.9         21.2
37.8         27.0         21.1
42.0         31.3         23.3
38.4         27.1         21.0
38.9         29.4         23.0
       32.0
       27.4
       29.5
      26.1
      25.2
      25.0
             18.7
             18.0
             16.4
38.4
34.0

38.1
2.4
6.17
 7
                   29.6
                    2.3
                   7.77
                     3
24.0
27.2

27.7
 2.3
8.16
 7
      23.4
      22.0
25.4
0.6
2.30
 3
22.1    17.7
 1.1    1.2
4.86   6.66
 7      3
                          11.3
                          14.0
                          13.2
                          11.9
                          12.0
7.91a
7.80a

12.5
 1.1
8.77
  5
1-000  0.792 0.634  0.493  0.461 0.423  0.368 0.295  0.208
0.074  0.054 0,051  0.046  0.045 0.024  0.026 0.025  0.020
 7.42  6.81  8.10   9.38  9.70   5.73  7.15   8.48  10.6
    SD
   VC(%)
(a) Outliers as per Chauvenet's rule
(b) Standard deviation
(c) Variation coefficient
                                  729

-------
Table 3    Regression  analysis  data  of  TiOa-catalyzed   photo-
degradation kinetics of 4-chloro-3-nitrobenzotrifluoride exposed to
simulated sunlight in aqueous medium
                                                SATISFIED
                                                2.17E-03
                                                3.69E-01 (DOF=6)
                                                3.10E-04
MODEL I  REGRESSION CONVERGENCE CRITERION:
          RESIDUAL SUM OF SQUARES:
          CHI-SQUARE OF REGRESSION:
          MEAN SQUARE ERROR:
    PARAMETER

y1  0.932419
k3  0.009980
                               STANDARD
                               DEVIATION

                               0.052596
                               0.001619
                                               COEFFICIENT
                                               OF VARIATION

                                                0.056408
                                                0.162265
MODEL II REGRESSION CONVERGENCE CRITERION:
          RESIDUAL SUM OF SQUARES:
          CHI-SQUARE OF REGRESSION:
          MEAN SQUARE ERROR:
          V2
          k3
             PARAMETER

              0.721305
              0.276397
              0.023138
                     STANDARD
                     DEVIATION

                     0.036920
                     0.031402
                     0.002830
 SATISFIED
 3.17E-04
 3.67E-02 (DOF=5)
 5.28E-05

COEFFICIENT
OF VARIATION

 0.051185
 0.113613
 0.122330,
MODEL III REGRESSION CONVERGENCE CRITERION:
          RESIDUAL SUM OF SQUARES:
          CHI-SQUARE OF REGRESSION:
          MEAN SQUARE ERROR:
          y2
          k3
          k4
             PARAMETER

              0.542964
              0.463933
              0.035038
              0.001001
                     STANDARD
                     DEVIATION

                     0.059433
                     0.059677
                     0.006729
                     0.000339
 SATISFIED
 1.37E-04
 7.68E-03 (DOF=4)
 2.74E-05

COEFFICIENT
OF VARIATION

 0.109459
 0.128633
 0.192062
 0.338789
                               730

-------
Table  4  Energy  radiated by  spectral region from an  unfijtered
medium pressure mercury  lamp (of the General Electric Uviarc®
type)	^_^__	__
                         LAMP FEATURES
               Arc current                  3.1  A
               Arc power                  0.25 kW
            ;   Arc length                   8.4  cm
               Lamp length                 18.7 cm
               Lamp envelope material     Quartz
Wavelength   Principal
 band (nm)  lines (nm)
  220-230
  230-240
  240-250
  250-260     253.7
  260-270     265.2
  270-280
  280-290     280.4
  290-300     296.7
  300-310     302.2
  310-320     313.1
  320-330
  330-340
  Radiated  Wavelength  Principal
 energy (W)  band (nm)  lines (nm)
    0.45
    0.55
    1.20
    3.00
    1.99
    0.37
    1.82
    1.28
    2.28
    4.43
    0.28
    0.61
340-360
360-370
370-380
380-400
400-410
410-430
430-440
440-540
540-550
550-570
570-580
580-760
365.4
404.7
435.8
546.1
578.0
 Total energy radiated
 input power (over 80%
 singled out)
 Total energy radiated
 input power
in  the 220-400 nm region, 26.2 W
of UV energy carried by the seven
  Radiated
 energy (W)
    0.37
    7.05
    0.20
    0.28
    1.74
    0.26
    4.18
    0.64
    4.65
    0.29
    5.05
    1.09
or 10.5% of
strong lines
 in the 400-760 nm region,  17.9Wor 7.2% of
                              731

-------
Table 5   Results  of  solar energy measurements performed at two
geographical   sites in the US during  midday  on  clear days  in
midsummer
Solar energy (W/m 2) distribution3 at Tucson, Arizona: sun
and without atmospheric absorption (B)b
A (nm) A B k (nm) A B \ (nm)
292
295
300
305
310
315
320
325
0.03
0.08
0.29
0.71
1.1
1.8
2.4
3.5
1.1
2.4
4.1
4.1
4.5
6.1
6.8
8.9
330
340
350
360
370
380
390

3.8
4.2
4.6
5.5
5.9
5.9
7.3

8.7
8.6
8.5
9.6
9.7
9.2
11.1

400
420
450
500
550
600
700

in zenith (A)
A B
10.9
11.7
14.7
15.2
14.2
12.4
8.3
-
16
16.4
20
19.9
18.2
15.7
10.5

(a)  For 10 nm spectral bands centered on the wavelengths indicated
(b)  The effect of the atmosphere  is most visible toward the  shorter
    wavelengths
Distribution0 of total solar energy (A) at Cleveland, Ohio, broken down
into direct sunlight (B) and sky light d(C). Energy in W/m*
A A B C A A B C A/ABC
(nm) (nm) (nm)
300
310
320
330
340
355
370
385
400
410
420
430
440
0.052
0.475
1.25
2.04
2.33
2.59
3.25
3.33
4.33
5.48
6.00
6.17
6.27
0.026
0.24
0.65
1.08
1.26
1.44
1.86
1.92
2.68
3.39
3.77
4.04
4.26
0.026
0.235
0.60
0.96
1.07
1.15
1.39
1.36
1.65
2.09
2.23
2.13
2.01
450
460
470
480
490
500
510
520
530
540
550
560
570
6.69
7.26
7.43
7.43
7.32
7.18
6.92
6.60
6.76
|7.93
6.89
6.71
6.63
4.53
4.92
5.14
5.25
5.25
5.25
5.08
4.92
5.14
5.35
5.35
5.25
5.25
2.16
2.34
2.29
2.18
2.07
1.93
1.84
1.68
1.62
1.58
1.54
1.46
1.38
580
590
600
610
620
630
640
650
660
670
680
690
700
6.46
6.32
6.21
6.05
5.96
5.82
5.66
5.56
5.46
5.51
5.72
5.64
5.49
5.14
5.08
5.03
4.92
4.86
4.75
4.64
4.59
4.53
4.59
4.81
4.75
4.64
1.32
1.24
,1.i8
1.13
1.10
1.07
1.02
0.97
0.93
0.92
0.91
0.89
0.85
(c)  ForSnm spectral  bands centered on the wavelengths  indicated
(d)  Scattered sunlight radiation. Scattering is greater at shorter wave-
    lengths
                              732

-------
  1.25-
                          Wavelength (nm)

Figure 1    UV spectra of NCTT (15.0 ± 0.1 mg/L; 20 °C) in high purity
water (a) and n-hexane (b)  were recorded  with a  Hewlett-Packard
(Palo Alto,   California, USA) Model 8452A diode-array  spectropho-
tometer and a standard 1 cm lightpath cell,  instrumental conditions
for  spectra  acquisition were: spectral bandwidth,  2 nm; wavelength
accuracy,   ±1 nm;  full spectrum (190-820  nm)  scan  time,  0.1 s.
Absorbance of  both solvents  was negligible  over  the  full  spectral
range. NCTT appears to exhibit absorption bands below 400 nm only.
                               733

-------
                    Q_ CD C ,- >s
                    E  ?-B §s

                    «  3 S E fi
                       -i-- •— VW *—

                       ^•q^ o
     cd
o  cd .t

CD  _
X •£ CD

                              CD
                      T3
 _- o
     -g-
      - Q.

     §1
                         to
                           CO
                       c:
                    cd  O
                      7= S3 ,„- Q.
                    +=:
                    CD
     *f

   E|l
   C§ -^  CD
                              CD
                              CO
                         5 S
                    13 C
                    cr a)
                    CD c


                    I'-cd
                    ^ £

                    to'E

                    5 ^
                    cd —

                    cd
                    Q. cr
                    Q. £
                    cd c-

                    0)8
                    -n co
                    H- v
     c
     CD
     JC
                           O
     fl
   cd Q.
      .  CD
      =:  CD
          CD
       ~0 -^
     C? o


     ii
     -_  i_
     o  o
     ^ IJ-

     81
     Q  CD
     p

     O


     CD
                    c^
Sola
                      JC CD
     $



     §

     CD
   o> S  S -E
          CD
O

So

   Q
                      1

                      CO
                       1
   CD
   CO

g CD
O CL
Q.-W
tO "D

£S

•j'TJ


P
q ^

^ S


O a
                              c 
-------
CO
-t~

c
   1.000
   0.750 -
   0.500 -
    0.250 -
Model  I F(t)=y1-e
                                              -K3't
Model II
                                          = y-e
                                              -K't
                              Model III F(t)=y1-e-K3't-K4-t'
                                                            480
     10
                                                 90% DISAPPEARANCE
             SEMI LOG PLOT OF
                 MODEL III
           99  7o DISAPPEARANCE
                                        99.9  % DISAPPEARANCE
                                         .99 % DISAPPEARANCE
                                         300   360   420
                              480
                               Time (min)
Figures    Photodegradation  induced  by  simulated  sunlight of
aqueous NCTT in the presence of suspended  TiOa.  Normalized
experimental data and  regression  curves  of Models  I, II, and III are
shown. With  reference to Model III (mean estimates), it might  be
predicted that NCTT photodegrades by 90, 99, or 99.9 % after 360,450,
or 460 min irradiation times, respectively.  For such  estimates, it  was
arbitrarily assumed that  there was no  change in the kinetic trend for
irradiation times >240 min.
                                735

-------
  Lamp unit
  with A/? cooling
Gas
inlet
 Heat-
 exchange
 input
                                                      Heat-
                                                      -exchange
                                                      output
                          Solution exit


Figure 4     Batch  photochemical reactor. Higher wattage lamps are
more efficient in producing a useful light emission up to approximately
an  input power of 10 kW. At that point, a decrease in UV emission effi-
ciency (einsteinsx kW'1 x hour"1) is observed. Therefore, if the reactor
is  designed to exceed a 10kW  input power, an array  of  lamps is
normally arranged. Lamps immersed in the reaction medium  improve
quantum  efficiency by  avoidling  excessive scattering,   reflection,
or dispersion of light In the picture, both lamps and reaction  medium
are cooled to avoid overheating.
                              736

-------
    Air- or nitrogen-
    cooling tube
    Reactant
    input
                                                   Lamp
                                                   well
                                                   Lamp
                                                   Product
                                                   output
Figures    Flow photochemical reactor.  Continuous reactors  are
used in high quantum yield reactions for vapor, liquid, or mixed media.
Contact times are short and high reaction  rates are achieved. The heat
generated by the light source is carried away by appropriate   cooling-
in  the case shown, by ventilating the  lamp well with air or nitrogen.
Owing  to the continuous flowing of the chemical medium, cooling  is
generally required solely to remove heat from the light source.
                               737

-------
         UVlamp
                                                 Peactants
                                                 Coolant
Product
                                                Coolant
F'gure 6     Elliptical reflector photochemical reactor. This is a system
in which  the light source is outside the reaction medium. To maximize
utilization  of the emitted light energy,  radiation is focused upon the
reaction   vessel by an  enveloping  elliptical,  highly reflective  casing
lamp and  vessel being placed at the focuses of the ellipse.  Although
heat is removed from  the casing inside  by efficient ventilation  the
temperature  of the reactant mixture can also be  kept  under control
through an additional cooling system.
                               738

-------
O  0 H-i  0
                                 0

')













\
»
t
5 •
i
>
>
i
<
^

fc
55 co
IH/GH-/A/7E/V
1 UVLAMP.
-IM
Tl




(23/r?af0.7
ppm TCDD)
till
ria{
zo
/ OD
•• 1 1
^5"
x QQ
// / l\\ \
)TODEGRA
EXTRACTE
sft
a°
AI-
T 2: O :> H £

                    O  E E O T3
                    -1-1    C T3 0
                    CO O 5=;    C -n
     CO 2



     c o
        r— s^p ?S


   ---Fw g


   9? ±i > a rv,
                    T3 CO
                    CO  CO

                    .tf> "0
                         •^•s
             o

             o
                         "co •- .3> Q
 ;sp:
z
1
r WATER
'_l
2

p
E
8:
1
o
v_
                    P  CO

                    & T3
                              -^ co co
                               O — O
                               =3 CO Q.
                    ci§
                    -^ o

                    co" 0}
                    CO tn
                    0 CO    ^



                    1 0 | ^ "5
        r° 0  y -ti
        fc O  fO JO

        "r- ^3  0 m
                            co 'y ~ •
                            > -^ O O)
< co

 x "co
 00
                         CD
                       E  0

                       o
                             -
                            CD CO CD

                    il i—  o
        I III
739

-------

-------
                        NATO/CCMS Fellow:

                   Sjef Staps, The Netherlands

International Evaluation of In Situ Bioremediation
         of Contaminated Soil and Groundwater
           741

-------
 INTERNATIONAL EVALUATION OF IN-SITU BIORESTORATION
 OF CONTAMINATED SOIL AND GROUNDWATER.
 J.J.M. Staps

 National Institute of' Public Health
 and Environmental Protection (RIVM)
 P.O. Box 1,  3720 BA  Bilthoven,  The Netherlands
Final  Fellowship  report for the Third International Meeting
of the NATO/CCMS pilot study on  "Demonstration  of  Remedial
Action  Technologies  for Contaminated Land and Groundwater"
Montreal, Canada, November 6-9, 1989.
                       742

-------
                                                                w
                                                              environmental

                                                            application  of
ABSTRACT

This-paper is the result of the RIVM-project "International evaluation of  in-
situ   biorestoration  of  contaminated soil and groundwater".  As a fellowship
project, it was associated with the international NATO/CCMS pilot  project  on
"Demonstration  of  Remedial  Action  Technologies  for  Contaminated Land and
Groundwater".
The  philosophy  of in-situ biorestoration is to stimulate the indigenous soil
microorganisms  to  degrade  contaminants  by  improving   the
conditions in the soil using a water recirculation system.
The objective of the project is to show the possibilities for
the  technique  in relation with contaminants, soil conditions and other site-
specific circumstances by means of integration and evaluation  of  results  of
in-situ biorestoration projects.
The project  is limited to the Netherlands, West Germany and the  USA.  It  was
implemented  by  visiting 23 relevant projects in these three countries, which
play  a  leading  role  in  the  development  of  remediation  techniques  for
contaminated soil and groundwater.

In-situ biorestoration is a relatively young, developing  technology.  It  has
been  used   at several locations, mainly  in the USA. It can be used especially
for  locations at which both the  unsaturated  zone  and   the  groundwater  are
contaminated with  hydrocarbons. A precondition is a good permeability  of the
soil.
Experience   has   especially  been  gained  with  in-situ  biorestoration  at
hydrocarbon-contaminated petrol stations  and  industrial sites.  The   system
generally    consists  of  a  water  recirculation  system,  aboveground  water
treatment and conditioning of  the  infiltrating water  with  nutrients  and  an
oxygen   source.  However,  there is no one-and-only application  method for in-
situ biorestoration. The remediation, which can  last  from  approximately  six
months   to   several years, can reach residual concentrations below the B-value
of the  Netherlands examination framework  (see table 4). If applicable,  in-situ
biorestoration    is  generally more  cost-effective  than  other  remediation
techniques;  costs  are  approximately between 40-80 US  $/m  .
Recommendations   from   this   evaluation   include a further  stimulation  of the
development of  this  technology,   improvement  of  the  preliminary   research,
expansion  of the  applicability to more recalcitrant  contaminants, research  on
bio-availability and research into  oxygen .supply  and  distribution  in the
 subsoil.

 INTRODUCTION.

 In  behalf  of the Dutch clean-up operation for contaminated soil, development
 of adequate clean-up methods is considered to be of prime importance!.  Besides
 thermal and extraction techniques, which still  account for the greater part  of
 the clean-up operation, biological  techniques   have   been  developed  in  the
 Netherlands.  Landfarming,  a  biological  treatment   technology for excavated
 contaminated soil, is now being used on a practical scale  (Soczd  and  Staps,
 1988).  However,  in  many cases it is impossible or too expensive to excavate
1 in (its original) place
                                  743

-------
 the soil. In-situ techniques are then the most appropriate methods,  and can be
 employed for treating both the soil and the groundwater.
 In the Netherlands, application of in-situ techniques  by  companies  nowadays
 focusses  on  washing, circulating and cleaning of the groundwater (the former
 pump-and-treat method).  Especially in recent  years,   increasing  interest  is
 also  being  shown  in  actual  biorestoration  in the soil.  The environmental
 conditions in the subsoil are optimized by supplying oxygen and nutrients  and
 circulating the water.

 The first Dutch research into in-s,itu biorestoration was  a feasibility  study,
 carried  out  by  Delft  Geotechnics,  to evaluate the scope of an in-situ soil-
 venting  technique  (van  Eyk  and  Vreeken,  1988).   The  RIVM  and  the  TNO
 (Netherlands  Organization  for  Applied  Scientific Research)  are preparing a
 full-scale clean-up by means of a literature study and extensive experiments
 on  laboratory-scale  since 1985 (Verheul et.al.,  1988).  However, a  clear need
 for information from full-scale clean-up projects  and from foreign  experience
 was  still felt.  From literature and international contacts is  was known,  that
 especially in the  USA  experience  with  this  technology had  been  gained.
 Besides,   developments were also under way in West Germany (Nagel et.al.,  1982
 and others).
 While the problem of soil and water contamination  also became evident in other
 countries,   interest  in  remediation  techniques,   and   especially   in-situ
 technologies,   increased.  This emerged at the first international workshop of
 the NATO/CCMS pilot project on "Demonstration of Remedial Action  Technologies
 for  Contaminated  Land   and  Groundwater"  in  spring 1987.   Several western
 countries,  including the Netherlands  and the USA  are  participating  in  this
 pilot project.                                                              .
 This was  sufficient reason for the! RIVM to start this evaluation in  late 1987.
 The  author   was  awarded a fellowship of the NATO/CCMS project.  Because of its
 relevance to  the  development of remediation techniques in the Netherlands,  the
 study is  partly financed by The Netherlands Integrated Soil Research program.

 The project  is  limited to the Netherlands,   West  Germany  and   the   USA.   The
 fellowship  project  was  implemented by visiting  23  projects in this field in
 these three countries, which  play a  leading  role   in   the  development  of
 remediation   techniques   for contaminated soil and groundwater.  An overview of
 the projects  is given in the appendix.
 Information,  results  and data are  directly obtained from  the  experts  involved.
 Total information is  arranged,  and conclusions are  drawn  in this final  paper.
 A   more   comprehensive report,  including detailed  information from the visited
 projects, is  in   print   (Staps,   1989  ).   Information concerning analytical
 procedures is also included in this report.

 EVALUATION OF IN-SITU BIORESTORATION  PROJECTS

 Introduction

Although  not   all organizations  dealing  with  in-situ  biorestoration  are
 included,  the  23  projects  chosen  do provide a good idea of the  feasibility of
 this  technology. The  concerning  group of 23 organizations consisted of fifteen
private   companies, three  institutes, two universities, one co-operation of an
 institute, a university  and a  coast guard, and one air force.
A   schematical  overview  of  the projects,  including several characteristics, is
 given in  the appendix. Projects U8 and U9 cannot really be regarded as  in-situ
                                    744

-------
biorestoration  projects,  because  in both cases biorestoration does not take
place in the original location. The clean-up site in project U8 is  a  lagoon,
and in the case of U9, the clean-up consists of on-site landfarming. These two
projects are not  included  in  this  chapter,  but,  because  of  the  direct
relationship to the other projects, they are included in the general overview.
Other divergent projects are N5 and U5; both are research projects, where  the
contamination  has been caused deliberately. Moreover, project N5 is deviating
because the biostimulation is performed only by venting of the soil, and  thus
is limited to the unsaturated zone.
Projects D5 and D6 differ from the other ones in that they are  still  in  the
conceptual  phase,  and  data  from  demonstration  scale  test are as yet not
available.
A   substantial   proportion   of   the  remaining  group  of  "real"  in-situ
biorestoration  projects  is  .characterized  by  research  aspects,  with  the
majority having been  setrup as a research project (Nl, D2, U3, U5, U6).
Background of the sites at which in-situ biorestoration has been or
applied       .         .
is  being
The locations at which in-situ biorestoration has been or is being applied can
be divided into .two main groups:
- filling stations (service stations, airforce bases,  marshaling  yards,  bus
  stations) with leaking pipelines or storage tanks,
- chemical industry sites, mainly (former) refineries.

All   locations  were contaminated with hydrocarbons, for the most part defined
as  petrol  and/or  diesel.  At  airforce  bases,  also   kerosene   or   JP-4
contaminations   occur.   One-fifth  of  the  projects  concerned  chlorinated
hydrocarbons. The smallest group of  locations was contaminated with PAHs or  a
mixture  of chlorinated hydrocarbons, mineral oil and PAHs. The latter has not
yet been demonstrated.                                                ;

Preliminary site characterization

The  surface area of the sites at which2  in-situ  biorestoration  was  applied,
varies  largely;  from  20 to 75,000 m  . Within  this variety, two clear  groups
can  be  distinguished. The first group   is2 formed  by  filling  stations;  the
surface  area  is  mainly  400  - 1,000  m  . The  second group  consists of large
chemical industry and  (former) jefinery  sites, and here, the  area   is  varying
between 20,000 and  75,000  m  .  The   depth  upto which  the contamination  is
dispersed  is  generally between 3  and 10  meters below surface  level.
It  was  striking   that   the  discovery of  a second contamination during the
cleanup-process occurred  at  several  projects.

 In  relation  with   soil   structure   and  geology, nearly  all locations  can  be
 defined as sandy. At  several places, clay   layers  are  present.  Only   in   an
 exceptional  case,  in-situ biorestoration  is applied at a  site with overburden
 clay and fractured bedrock.
 Concerning  geology,   permeability  is   a  very  important parameter  for  in-sit^
 biorestoration. For the projects^ reviewed,  the Kf-value  varied between  10
 and  10"  m/s5  mainly having 10*  -10"  as  order  of magnitude. Generally, a Kf-
 value of 10"   is regarded as being the   minimum  permeability  for   successful
 application of in-situ biorestoration.
                                      745

-------
 Preliminary biodeeradation research

 In  order   to   decide  whether   in-situ  biorestoration  can  be  applied at a
 contaminated site, microbiological, hydrogeological and chemical aspects  must
 be  regarded.   Hydrogeological   conditions include permeability, dispersion of
 contaminants,  groundwater  level  and  flow.  The  parameters  which  might  be
 considered  before chosing  or designing in-situ biorestoration are:
 -  Microbial parameters    (total   cell   count,   nitrifiers,   denitrifiers
   hydrocarbon  degraders)
 -  Oxygen demand
 -  Nutrient  demand
 -  Contaminant  degradation  rate
 -  Bio-availability.

Total cell  count forms the base for research on populations of microorganisms
Parameters   in relation  with  biological  activity  are an important part of
raicrobial research. For in-situ biorestoration, the number of metabolic active
organisms   and enzyme samples are important as an indicator for biodegradation
in the subsurface. As regards hydrocarbon contamination,  determination of  the
percentage  of hydrocarbon degraders is an important monitoring aspect too.
Besides, there  is a large group of relevant physical and chemical  parameters,
including  permeability,   pH,  oxygen,  redox conditions, temperature,  TOC, DOC,
BOD, Fe-concentration,   Mn-concentration,  concentration  of  (heavy)  metals]
Ntqtal'  a™0111™-concentration,   nitrate-concentration,  nitrite-concentration
ana pHosphate-concentration.
                       is  one  of  the  conditions  for  a successful in-situ
A high   permeability
biorestoration.
Soil pjl may  affect  sorption of  ionizable  compounds  in  addition  to  limiting  the
types  of  microorganisms  in  the  subsurface.  Methanogenes,  which have  been
implicated   in  mineralization  of some aromatic hydrocarbons, are  inhibited at
pH values below 6 (Lee et.al.,  1988).
Biodegradation  of  many  organic pollutants  in the subsurface may be limited by
insufficient concentrations of  oxygen or  unfavourable  redox conditions.
Also temperature influences microbial metabolism of subsurface  pollutants.  The
temperature  of the  upper 10 m of  the subsurface may vary seasonably.  However
in  the   Netherlands,  it   will  not  deviate -much  from 10°C. Also below 10 m'
temperature  will be about  this  value. It  is important  to  keep  this  in  mind
when comparing results from projects in for example Florida (U6) or California
(US) where much higher temperatures (20-25°C) are measured with projects  from
other  regions.
Total  organic carbon  (TOC),  dissolved organic carbon   (DOC),  chemical  oxygen
demand (CJDD.)  and  biological  oxygen demand (BOD) are sum parameters. TOC and
DOC are direct parameters  for the carbon  concentration of  organic  compounds
Decreasing concentrations  of TOC  anjd BOD values indicate mineralization of the
organic contaminants.
Determination   of   Fe	and Mn  concentrations  is  important  because  high
concentrations  of  these  metals  can  cause  precipitation   under   aerobic
conditions,   caused  by  the  infiltration of oxygen during the biorestoration
process.
Other heavy metals can be  important>  especially at contaminated sites, because
at toxic  levels,  they can  inhibit the activity of microorganisms.
Inorganic—nutrients  like  nitrogen  and  phosphorus may be limiting when the
carbon/nitrogen/phosphorus  (C:N:P) ratio  is  unfavourable.   Determination!^
                                   746

-------
ammonium,  nitrate and nitrite gives insight in the stage of the conditions in
the subsoil.

After   the   characterization   of   the   site   regarding  microbiological,
hydrogeological and chemical/physical parameters,  a  first  decision  can  be
taken  whether  in-situ  biorestoration  is  applicable  at  a  specific site.
However,  there  is   no   one-and-only   application   method   for   in-situ
biorestoration.  If the option of in-situ biorestoration is chosen, nearly all
visited organizations perform preliminary biotreatment studies  on  laboratory
scale  to   get insight in the optimal stimulatory actions for a biodegradation
process at  the  site  and  to  choose  the  right  combination  of  microbial,
hydrogeological  and  physical/chemical  actions. Only organizations with very
broad experience in the field of  pump-and-treat  and  in-situ  biorestoration
design  a site-specific in-situ biorestoration system almost directly based on
the site characteristics  (Ul, U7). A large majority of the  projects  included
preliminary  laboratory  research,  both  small-scale  tests  and  percolation
studies in  columns. In  a  few  cases,  field  experiments  in  a  small  area
representative of  the contaminated site have also been performed.

System design

Description of the installation

In-situ biorestoration  involves   the  stimulation  of  the  biodegradation  of
contaminants at   contaminated  sites  without excavation of the  soil.  In this
process,  the soil  of  the  contaminated location is used as   a  bioreactor   (see
figure  1).

The  specifications of the  "bioreactor"   in  the  subsoil   are  based  on  the
characteristics   of the contaminated site,  and the objectives and requirements
of the  clean-up.  They include for example the type   and   distribution  of  the
contaminants  in  the subsoil,  the soil  geology  and  hydrology and the need for
 isolation of the location.
 In  most  cases   a  semi-closed  configuration  is  used  in such  a  way,  that  the
 contaminated location is isolated and controlled;  uncontaminated  groundwater
 can  enter  the  contaminated site,  but  contaminated  groundwater cannot move to
 uncontaminated areas.
 The site can be  isolated using hydrological intervention technologies' or civil
 engineering operations. In general,  a  hydrological   system  is  designed,   in
 which   the   groundwater  is  centrally  withdrawn,  and  after  above-ground
 treatment, reinfiltrated at several points on the periphery of  the  location.
 The  groundwater  is  withdrawn  at  a higher rate than it is infiltrated,  the
 surplus generally being discharged into a sewer.
 To  support degradation in the subsurface, an above-ground treatment system is
 used to  degrade  the  contaminants  in  the  withdrawn  groundwater,  and  to
 condition the water before re-infiltration.
 Biodegradation relies entirely on the contact between the contaminants (in the
 water  phase) and the microorganisms. In the case of highly volatile compounds
 as contaminants, clean-up can  partly  be  achieved  by  vaporization  of  the
 unsaturated  zone  using a soil venting system,  as is shown in project N4.  The
 contaminated exhaust air can be  treated  above  ground  by  adsorbtion  (e.g.
 activated  carbon)  or oxidation in a biological, thermal or catalytic manner.
 Research project N5 describes the design of an in-situ  soil  venting  system,
":• used-'--" both   as   a   physical   (evaporation)   and   a  biological  process
                                      747

-------
  (biodegradations).   This  system  can  be  used  for  contaminations   in   the
  unsaturated zone.

                                                        compost filter
                                                        catalytic oxidation
                  Addition of
                  Addition of nutrients
                  Addition of micro-organisms
                  Heating
   discharge
   well
                I monitoring well
                                                               groundwater-level


                                                             
-------
When required by the legislator, the contaminated air from the air stripper is
oxidized  in  order to bring about degradation of the contamination instead of
moving the contaminants from one compartment (groundwater) to  another  (air).
In  the  Netherlands,  this is performed using a biological compost filter, in
which adapted microorganisms degrade the contaminants. In the  USA,  catalytic
oxidation systems are employed.

Hydrological aspects

In  general,  in-situ  biorestoration  is performed by means of saturating the
subsoil. The main hydrological steps taken consist of  central  withdrawal  of
the  groundwater  and  reinfiltration  at  several  injection  points  on  the
periphery of the location. Groundwater is withdrawn at a higher rate  than  it
is infiltrated. This occurred at about 95% of the locations.

At two projects, in-situ biorestoration was performed without water saturation
(D2  and  D4).  However,   saturating  the soil makes  it easier to  optimize the
environmental conditions in the soil with respect to  other parameters like pH,
oxygen  content,  nutrients,  etc.  It  depends on the site-specific situation
whether saturation  and other optimizations will be chosen,  or  no saturation
and  fewer  other   optimizations.  However,  in  most cases, saturation is the
preferred method.

Oxygen supply

As   far  as   is  known,  in-situ  biorestoration  has  only  been  applied to
hydrocarbon-contaminated   sites.  In  order to initiate hydrocarbon oxidation,
microbial populations utilize oxygen:
                  C6H6
6CO,
3H20
(for benzene).
 As a result of  the  contamination,   the   subsoil   of  contaminated  sites   is
 anaerobic,   or  contains  very low concentrations  of oxygen.  Therefore,  oxygen
 has to be supplied for in-situ biorestoration.  Sources  of oxygen include air,
 pure  oxygen and hydrogen peroxide.  Subsequent  oxidation can also be sustained
 by alternative electron acceptors, for example  nitrate.
 Lack of oxygen or necessary redox conditions will  limit in-situ biorestoration
 of contaminated soil and groundwater. When applying in-situ biorestoration   in
 practice, oxygen is usually the limiting factor.
 The alternatives to oxygen supply used in the projects  visited were:
 - air
 - pure oxygen                                                            •
 - hydrogen peroxide
 - nitrate
 - nitrate / ozone
 - methane / oxygen

 Oxygen sources

 The simplest method of  supplying oxygen is aeration. However, the   amount  of
 oxygen   that  can  be   added  with   air is strongly limited: only 8 mg/1 under
 normal groundwater conditions  (table 1). As a result, very  large  volumes  of
 oxygenated  water  may  have   to  be infiltrated at the contaminated site,  and
                                      749

-------
 because of permeability constraints,  the remediation time  is   then  relatively
 long.

 Table  1.  Available quantities of oxygen from different  sources.
                   air-saturated
                      water	
                                   02-saturated
                                      water	
             H202
         NO,
             200 me/I  200 mg/1
 available oxygen
 fag/1 at 10° O
                     10
40
94
                                                             168
 As shown in table 1,  this problem can be  overcome  in  part  by  using  pure
 oxygen (40 mg 02/1)  or hydrogen peroxide  (100 mg 02/1 from  200 ppm H202):
                             H202  -
                                     H20
 Hydrogen peroxide is  toxic at higher  concentrations and can therefore  only
 be  used  up  to  a  limited  concentration.   In   the  case  of  H202,  the
 bioremediation  is usually  started  with  low  concentrations  (40-50  mg
 HaOa/1),   or even with pure oxygen. The  objective  of this measure  is to let
 the  indigenous population of microorganisms  acclimate  to  the  oxygenated
 environment.  Once the population is acclimated, the peroxide concentrations
 can  be  increased in increments of approximately 50 to 250 ppm  in   intervals
 increasing  from  approximately  one  week to  one  month (U3),  to achieve an
 increased infiltration of oxygen.  Such  a  gradual  increase  of  peroxide
 concentrations  can  be  continued up to a concentration of about 1000 ppm
 H202.
 In  the  initial phase of  biorestoration, the oxygen supplied is utilized by
 the  microorganisms  in  the  vicinity   of  the  infiltration  point.  When
 contaminants  in this  area have beeri degraded,  the  oxygen can be transported
 over larger  distances,  and biodegradation  will  then  occur   in   an  area,
 further   away  from  the   infiltration   point. This process continues until
 oxygen breakthrough at the withdrawal wells.
 An  important  aspect  with  respect  to peroxide  is  its  stability.  As
 remediation  of the site progresses, the  H202 must be  carried  increasingly
 longer   distances.  This  means that H202 must be stable in order to deliver
 the  oxygen to the area where it is [needed. The decomposition of peroxide is
 catalyzed by metals,  such as  iron and manganese. H202 can also be degraded
 by the bacterial  cell,  with the enzyme catalase serving as the catalyst. On
     other hand,  phosphate can stabilize hydrogen peroxide (Britton, 1985).
the
This is actually performed at demonstration projects. The form of phosphate
is  mostly  monophosphate. • To  reduce  phosphate adsorbtion to the soil, a
combination of simple and complex polyphosphate salts can  be  used  (Brown
et.al., 1986). The use of phosphate solutions is twofold: as a nutrient, it
also has a positive influence  on  the  biodegradation  when  the  original
concentration of phosphate is too low.

Nitrate as electron acceptor

Nitrate can  serve  as  an  electron  acceptor.  Comprehensive  fundamental
research  regarding  the  use of nitrate has been performed in West Germany
(Riss et.al., 1987). Here, laboratory research showed that nitrate can only
be  utilized when a first phase with elementary oxygen has passed, and when
                                    750

-------
the nitrate is present under anaerobic conditions.   There  has  not  always
been  given  satisfaction  to these preconditions when applying nitrate for
in-situ biorestoration (see e.g. projects Nl and N4).
As  is shown in table 1, one part of nitrate is equivalent to 0.84 parts of
oxygen. Take, for example, the oxidation of methanol:

                       1.5 02 H- CH3OH -* C02 + 2 H20

   NO; + 1.08 CH3OH + E+ -+ 0.065 C6H7N02 + 0.47 N2 + 0.76 C02 + 2.44 H20

(Brown, 1989).
Until now, application of nitrate has only occurred in a few German  states
and  only  incidentally  in  other  countries.  Application might encounter
licensing problems. In project N4,  nitrate  is  added  to  the  oxygenated
infiltrating  water.  Nitrate  will also be used at project U3 as part of a
research program.  Utilization  of  nitrate  could  not  be  determined  in
research project Nl  (Verheul et.al., 1988).
In project D9, a combination of ozone and  nitrate  has  been  used:  ozone
above  ground,  to   treat the water and oxygenate the organic contaminants;
nitrate in-situ, in  the  subsoil, to  serve  as  an  electron  acceptor  for
subsequent biodegradation by the microorganisms.

Co-metabolism

At one research project (U5), biodegradation of  chlorinated  compounds  by
methane-oxidizing    bacteria    (methanotrophs)    involves   stimulating the
population with methane- and oxygen-containing water. Methanotrophs   obtain
energy  from  the   oxidation of methane.  They  synthesize the  enzyme  methane
monooxygenase, which catalyzes  the first step  in the  oxidation  of methane,
which  they   use   for  energy  and growth.  Monooxygenase  oxidizes  a range  of
hydrocarbons, and  appears to bring about  the   epoxidation  of  chlorinated
 alkenes  (co-metabolism):
                   CHC1-CHC1 + H20
CHC10CHC1 + 2H  -I- 2e
 These epoxides are unstable in water and hydrolyze to a variety of products
 which can be oxidized readily by other heterotrophic bacteria to  inorganic
 end products (McCarty et.al., 1989, Janssen et.al., 1987).

 Comparison of oxygen sources

 Table  2  shows  a  comparison  for  various  oxygen systems for a severely
 contaminated site.

 It  can be concluded, that there is a wide range in both cost effectiveness
 and in treatment effectiveness. For example, venting can only be applied in
 the vadose zone. In terms of cost  effectiveness, the order is:

       venting » peroxide > nitrate > air sparger > water injection

 while in  order  of  treatment effectiveness, the order is

       peroxide  - nitrate > water injection > venting > air sparging.
                                     751

-------
  Table 2' Cost/performance  comparison  for  various  oxygen   systems-  hieh
           decree of contamination. ,                                        6
- t.
System
             ~r~~7 Cos«s   ($> ---- '~  ------ Performance ------ - ---
              Capital    Operation Maintenance  kg/Day S Site  Utilization  Time of  S/kg oxygen
                                        Oxygen Treated Efficiency * Treatment    Used
Air Sparging
Water Injection
Venting System
Peroxide System
Nitrate System
35
77
88
60
120
,000
,000
,500
,000
,000
800/month
1200/nonth
1500/month
10,000/month
6500/month
1200/month
1000/month
1000/month
1500/month
1000/month
3
4
1810
86
96
41
75
60
100
100
70
50
5
15
12.5
858
1B80
132
330
335
days
days
days
days
days
57
62
e
41
49

fm
.
—
..
                                                         (Brown. 1989)


 The  choice  of  an  oxygen supply is most dependent on the contaminant load,
 the mass transfer and   the  ease  of  transport  and  utilization   At  low
                           Systems>   such  as  ^ sparging,  become more cost
 Nutrient supply

 The biodegradation rate will be  limited when inorganic nutrients,   such  as
 nitrogen  and  phosphorus, are present in  limiting concentrations  or mutual
 ratio's.  Regarding  contaminated  sites,   the   presence  of  nitrogen  and
 phosphorus _ should be viewed in  relation with the  carbon concentration from
 the ^ contaminants. In soil, a C:N:P!ratio of 250:10:3  is  considered  to  be
 optimal  for biodegradation. Also other C:N:P-ratios,  e.g.  of 100:10:2 have
 been chosen.
 The need for nutrients is dependent on the  site  characteristics  At certain
 sites,  nutrient addition can be unnecessary.  In  other  cases,  increasing the
 inorganic  concentrations at one time can be  sufficient.  If nutrient supply
 chosen         g     clea*-up, nearly always  batch-mode addition  has  been

 In order to satisfy nutrient requirements,  a  wide  range of   components  can
 be added.  This includes compounds like NH4N03,.Na- and K-orthophospSate anS
 trace elements.
 ?2 !  f6" Pr°Je=ts-  such as N1>  Addition of an easy degradable  carbon source
 (NaAc)  enhanced  the initial degradation of hydrocarbons  during  laboratory
 experiments.   However,  the significance for demonstration seal!  seems to be
 limited.

Addition of microbial populations  '•

Besides stimulating the indigenous  microbial population to degrade  oreanic
compounds   in  the  subsurface,  another option is to add microorganisms with
specific metabolic  capabilities to  the subsurface.  This is demons traced  in
projects  D3,  D4 and D7.  Soil  samples are  taken from the contaminate^ site
at spots where microorganisms   occur,   for   example  at  the  edge  of  the
contamination. The microorganisms which are present at those spots  will be
adapted to the contamination in the soil, and will  be  able Co  degrade  the
contaminants   The  samples  are  taken to the  laboratory,  where selection
occurs by enrichment  culturing,  until  a   suspension  is  obtained  which
contains  the  selected  microorganisms in  high  density. This suspension of
                                     752

-------
microorganisms  is  then  injected  with  the  infiltrating  water  at  the
contaminated  site.  The  objective  of this inoculation is to increase the
number of adapted microorganisms  at  the  site,  in  order  to  accelerate
biodegradation.
Another method to add microorganisms to  the  subsoil  is  applied  when  a
biological groundwater treatment plant is applied above ground (N2 and N4).
The effluent of such an installation will contain large amounts of  adapted
hydrocarbon  degrading  microorganisms which are injected in the soil. In a,
research project, also  inoculation  by  effluent  water  of  a  wastewater
treatment plant was used (project N5).
However, there is much uncertainty about the efficacy of  the  addition  of
microorganisms  to  the  subsoil  and  the  possibilities  of  transporting
bacteria through the soil, in order to get them at the spots where they are
needed.  Generally,  95%  of  the  soil  population tends to adsorb on soil
particles, whereas only 5% can be transported.

Results of the in-situ biorestoration projects

The projects visited differ widely in the clean-up results to be  obtained.
For  example,  some projects do not aim to achieve a given concentration; on
the basis of the   clean-up  progress,  it  is  decided  what  the  residual
concentrations of  contaminants should be.
There  are much differences; in the Netherlands, the objectives set  by  the
legislator  are  generally 50 
-------
corresponding  concentrations   of the Netherlands examination framework for
soil pollutants.
Table  4 shows the  relevant values of this  framework.

Table  3.  Residual  concentrations and remediation time
          biorestoration projects finished.
                   for   a  few  in-situ
Code
N3

N5

D4
D7
D8
D9

Ul
U2

U6
Contaminant
aromatics
mineral oil
BTX
diesel oil
diesel oil
fuel oil
diesel/arom.
aromatics
oil
gasoline
4-chloro-2-
methylphenol
JP 4
Residual
concentration
< 30 pg/1
< 200 pg/1
< 5 mg/kg
150 mg/kg
4600 mg/kg
< 100 mg/kg
30 mg/kg
non- detectable
levels
< 10 mg/kg
> 80% of area
cleaned-up
550 kg he removed
Compartment
water
water
soil
soil
soil
soil
soil
water

soil
water


Remediation time
(months) 	
6

12
18
12
9
31N
10 >

48
24

12
   non-detectable  levels reached  for part of  the contaminated area;  clean-
   up was  continued  for gaining  complete clean-up of the site.

         Relevant  part  of  the Netherlands examination framework for  soil
         pollutants.

 Indicative values: A - reference vtluo
               E - indicative value for further investigation
               C - indieativt velue for eleaning-up
Presence in:
                  •oiKmg/kg dry weight)
                   _A	B      C
groundwater (/ig/1)
•rcnatic compound*
benzene
•thylbenzane
toluene
xylene
phenols
aromatics (total)
I-olycycUe Braat.le
total PAHs
chlorinated orsanic
aliphatic chlor.
coccp. (total)
Chlorophenols
(total)

O.OS(d)
0.05(d)
0.05(d)
O.OS(d)
0.05(d)
-
TjwirffTri I 1 1 I
!
coBtxmds
-

-


0.5
5
3
5
1
7
IB (PAH«)
20

7

1


5
SO
30
SO
10
70

200

70

10


0.2(d)
0.2(d)
0.2(d)
0.2(d)
0.2(d)
.

_

-

-


1
20
IS
20
15
30

10

15

0.5


5
60
SO
60
50
100

40

70

2

mineral oil
                        1000
                               5000
                                              200    600
* " reference value soil quality
d - detection limit
                                      754

-------
As  regards  the  Netherlands  examination  framework  for soil pollutants,
residual concentrations below  B-level,  or  even  undetectable  levels  of
contaminants  have  been reached in the finished projects. Five out of nine
projects reach  the  A-level,  thus  meeting  the  standards  used  in  the
Netherlands.  Venting  (N5)  was  successful as regards volatile components
(petrol), but not for PAHs.

The results of the projects which are underway are generally promising.

Comparison of the;;results of the different in-situ biorestoration  projects
is  very  tricky.  This  is  mainly  caused by the application of different
methods, used in the course of the in-situ biorestoration projects:
- methods of soil and groundwater sampling and analysis,
- determination of physical and chemical site parameters.
Occasionally,  there are also gaps in the total overview of the restoration
course.

The  total overview of the in-situ biorestoration projects, as presented in
the appendix, also shows the results in relation with other  aspects,  such
as soil structure, oxygen source, applied system and nutrients used.

The remediation time varied between 90 days and 4  years,  and  is  largely
dependent  on  the  site  characterization (soil structure) and the kind of
contaminants.

Costs for in-situ biorestoration

A wide range of site- and system characteristics and objectives  influences
total costs for in-situ biorestoration projects. These include:
- geology and soil structure
- type and concentrations of contamination
- distribution of contaminants in the subsoil
- total surface and volume of the contaminated area
- system characteristics: recirculation, water and gas treatment a.o.
Because  these aspects can vary significantly, the costs for completing the
projects can vary considerably.
It must be stressed that these figures should always be seen in relation to
other treatment techniques for a certain contaminated  location,  including
cost for excavation and transport.
The projects can be divided into two main grougs:
- petrol stations  (approximately 400 - 1,000 m ; 1,000
- refinery- a$d industrial sites (approximately 20,000
  - 400,000 m  ).

Petrol stations
5,000 m J
75,000 m
30,000
Costs  for   in-situ  biorestoration  at contaminated petrol stations varied
between 62,000 and 750,000 US §  (40- 250 ys $/m ). Included are  relatively
cheap projects of approximately  60 US $/m  which could be performed without
abovegroundwater treatment (Ul)  or without water recirculation (D4).
                                    755

-------
A  comprehensive  itemization  of  the  different  costs  for using in-situ
biorestoration to treat a specific petrol  station  is  shown  in  table  5
(Fournier, 1988).

It should be noted that, dependent on the situation,  the  contribution  of
hydrogen  peroxide  to the total cost of the operation can be substantially
higher; contributions of 90% have occurred.
Table 5. Estimated  costs  for
         (Fournier, 1988).
in-situ  biorestoration of a petrol station
Capital costs
Groundwatex monitoring wells 5,000
Reinfection well
Nutrient and peroxide addition equipment
Reeirculation equipment
Equipment total
EreHaiTmry site assessacnt costs
Laboratory tests
Field tests
Reports
Total preliminary testing
Total initial expenditures
Annual operating «""* staintecumce costs
Croundwater monitoring
Reinjcction well maintenance
Chemical costs
Total «rmT.«1 COEtS
Present worth factor for 3 years
Present worth of O&M costs
Present worth of ISB option
Total costs over 3 Tears


.2,300
5,000
5.000
17,300

16,300
5,000
2.000
23,300
40,600

8,200
14,200
12.000
34,400
X 2.402
82,630
123,230
$143.800





12




16





72



100
Refinery- and industrial sites

Cost for in-situ biorestoration at refinery- and  industrial  sites  varied
between  330,000  and  16  millioj  US  $.  Again, especially system design
determines total cost: 7.- US $/m  if a relatively simple in-si£u  type  of
landfarming  is  used  (D2)  up  to  approximately 150.- US $/m  for a more
complex system design.

From  the  information from the projects it can be concluded that operating
and maintenance costs account for about 2/3 of the total costs.  Generally,
1/3  of the costs is due to preliminary research and installation costs, in
about equal amounts.

In many cases in-situ biorestoration will be more cost-effective than other
techniques, such as incineration and soil washing of  the  excavated  soil,
possibly  combined  with  groundwater  treatment (approximately 70-170 $/m
excluding excavation and transport costs (Staps, 1989a)).
                                    756

-------
CONCLUSIONS

Application

- The  locations  at  which  in-situ  biorestoration  has  been used can be
  divided into two main groups:
  * filling  stations (service stations, airforce bases,  marshalling yards,
    bus stations) with leaking pipelines or storage tanks (400 - 1,000 m2),
  * chemical industry sites, mainly (former) refineries (20,000-75,000 m ).

- With respect to soil structure and geology, nearly all  locations  can  be
  defined  as  sandy.  Clay layers are present in several areas. Only in an
  exceptional case in-situ biorestoration is used at a site with overburden
  clay and fractured bedrock.

- Regarding hydrology, permeability is a very important parameter  for  in-
  situ  biore§torations  In  the  projects  reviewed,  the  K_-yalue6varied
  between 10   and 10  m/s,  but was mostly of the order of  lo" -10"   m/s.
  In  general,  a  Kf-value  of  10"   m/s is regarded as being the minimum
  permeability   required   for   successful   application    of    in-situ
  biorestoration.

- All locations were contaminated with  hydrocarbons.  Most  contaminations
  are  defined  as  petrol and/or diesel. A few locations were contaminated
  with PAHs or a mixture of chlorinated hydrocarbons, mineral oil and PAHs.
  The  frequent  discovery of secondary sources of contamination points out
  that the characterization is not always sufficiently carried out.

Design

- The approach of in-situ biorestoration at the visited projects  could  be
  characterized  by  either a hydrological or a microbiological background.
  Only rarely, a good integration of both disciplines could be seen.

- The  decision for application of in-situ biorestoration can only be taken
  after    a    comprehensive    site-characterization.     The     specific
  characterization  of  the  contaminated site and preliminary biotreatment
  laboratory studies (if possible followed  by  field  studies)  should  be
  performed   to  determine  optimal  stimulation  actions   and  thus  the
  different forms in which the technology can be applied.

- As  regards  hydrological  measures.  generally  a system is designed, in
  which the groundwater  is  centrally  withdrawn  and,   after  aboveground
  treatment,  is  reinfiltrated at several spots at the outer border of the
  location. In order to support  the  degradation  in  the  subsurface,  an
  aboveground  treatment  system is used to degrade the contaminants in the
  groundwater which  is  pumped-up,  and  to  condition  the  water  before
  reinfiltration.

- As regards the aboveground treatment,  the  first  part  is  generally  a
  sandbox.  Undissolved contaminants are removed in an oil/water separator.
  An air stripper is applied for removal of volatile contaminants. At a few
  projects,  biological  systems, such as a trickling filter, were used for
  degradation of dissolved compounds.
                                    757

-------
 - Reeirculation  of  the  pumped-up groundwater has positive effects on the
  biodegradation in the soil. This  may  be  due  to  the  infiltration  of
  degradation  products,  which ate relatively easy to break down and which
  stimulate the activity of the microorganisms in the subsoil.

 • The  contaminating  vapours  in  the  air  from  the  air stripper can be
  oxidated by means of a biological compost filter or a catalytic oxidizing
  system  in  order  to acquire degradation of the contamination instead of
  moving the contaminants from one  compartment  (groundwater)  to  another
  (air).

 - On demonstration scale, most of Ithe time the limiting factor is  lack  of
  oxygen  or  necessary redox conditions. Hydrogen peroxide is most popular
  as oxygen source. However, for certain applications it can be  relatively
  expensive.  Other  sources  are ;air, pure oxygen and nitrate (as electron
  acceptor).  The  choice  for  a  system  is  based  on   cost-efficiency,
  contaminant load and the ease of transport and utilization.

 - Necessary nutrient addition is fully dependent on the original  available
  nutrients  in  the  soil  and  the uptake by the microorganisms. Usually,
  addition of nitrogen and phosphorus is necessary. In a  few  cases,  also
  trace  elements have been supplied. Other projects could be biorestorated
  without any artificial supply of nutrients.

 - The  effect  of  the adding detergents is still questionable. Fundamental
  research and most  practical  experience  indicate  that  the  effect  on
  degradation  is  negative. Clogging of the soil can occur when detergents
  are supplied, probably due to an  interaction  between  the  oil,  water,
  detergent and solid phase.

 - Addition of microorganisms to the subsoil, with the aim of enhancing  the
  biodegradation,  is  being  used by a few companies. Although such supply
  will always have some beneficial effect, until now,  this  has  not  been
  proved.  Cost-benefit  calculations  are  also lacking. A major objection
  here is, that soil microorganisms tend to adsorb onto  (soil)  particles,
  and  consequently  cannot  be  transported  over  long  distances  in the
  subsoil. This implies that the effect of the inoculation is very limited.

White spots

 - Bottle-necks in relation with in-situ biorestoration can be:
  * insufficient infiltration rates,  mostly caused by clogging,
  * insufficient hydrological isolation,
  * relatively   long   remediation   period,   needed   for  reaching  low
    concentrations of contaminants,

 - When   using   in-situ  biorestoration,  the  precise  fate  of  degraded
  hydrocarbons. such as  gasoline,  is  not  yet  known.  A  proportion  is
  transformed  to  leachable  DOC, another part to DIG, but a large part is
  still unaccounted for.

 - With  the exception of project Nl,  research on in-situ biorestoration has
  not provided knowledge about mass balances. When degradation occurred  in
  project  Nl, the percentages of leached and degraded aromatics were about
                                    758

-------
removed  by  degradation  only,  and  then
  the same; The aliphatics were
  almost completely.

.Results and significance

• As  'regards  feasibility,  in-situ  biorestoration  can  technologically
  compete with  other  technologies  when  it  is  applied  at  a  suitable
  location,  and  the  process  is  well  run.  As  regards the Netherlands
  examination framework for soil pollutants, residual concentrations  below
  B-level, or even undetectable levels of contaminants have been reached in
  most of the  finished  projects.  Contaminants  are  mainly  hydrocarbons
  (gasoline,-diesel, mineral oil).
  The remediation time varies roughly between 3 months and 4 years, largely
  depending  on  the  initial concentrations, the kind of contaminants, the
  soil structure and the requirements which are set.  Concerning  practical
  projects 3without  research aspects, costs can vary between approximately
  40-80 $/m  . This means that in many  cases  in-situ  biorestoration  will
  alsg  be  more cost-effective than other techniques (approximately 70-170
  $/m  excluding excavation and transport costs (Staps, 1989s)).

RECOMMENDATIONS

General policy

- This  evaluation  included  the  visit  of  17  contaminated  sites,  and
  concludes that in-situ biorestoration is a  promising  technology  for  a
  selection  of contaminated sites. However, it is important to notice that
  most spills, and thus damage to the environment and the spending of large
  amounts  of  money  for  remediation,  could  have been prevented by good
  house-keeping.  Therefore,  at  locations  where  spills   might   occur,
  prevention is recommended in the first place.

- The most fundamental recommendation that can be made from this study,  is
  to  stimulate the development of in-situ biorestoration. This study shows
  that the technology has a large potential. At present, it is important to
  collect reliable  (demonstration) data, which can be used in the following
  areas:
  * optimization  of  the technology, mainly regarding oxygen transport and
    utilization,  peroxide  transport  and   stability   and   removal   of
    contaminant residuals from soils (bio-availability).
  * extending the technology's range of applications,  especially  to  more
    recalcitrant contaminants.
  * development of models of (in-situ) biorestoration.

- In-situ   biorestoration  is  expected, to  be  a  promising  technology,
  especially for application at  contaminated  industrial  sites.  This  is
  mainly  because of the minimal physical impact on the environment, caused
  by the process; industrial activities can be continued during the  clean-
  up.

- When demanding certain residual concentration levels,  regulators  should
  not  only  consider  concentrations  in  the groundwater, but also in the
  soil. It should be prevented that an in-situ  biorestoration  project  is
  finished   because  the  contamination  levels  in  the  groundwater  are
   759

-------
sufficiently low, while significant concentrations are still  present  in
the  unsaturated  zone  of the soil. Percolating water from precipitation
will transport  (a part of)  residual  contaminants  and  contaminate  the
clean groundwater again, making; a second clean-up operation necessary.

The approach taken by the experts involved in  several  of  the, ;projects
visited   can   be   characterized   by   either   a  hydrological  or  a
microbiological background. However, in-situ biorestoration is  not  only
pure  biotechnology,  but  is  indeed an integration of biotechnology and
hydrology. Integration of a number of disciplines is indispensable.

Because of the general complexity of soils, the course of the degradation
process  can  never  be  predicted  completely.  Therefore,   preliminary
research,   both  in  the  laboratory  and  in  field tests will always be
necessary. The field  tests  should  include  oxygen  utilization  rates,
possible in-situ peroxide stability and potential clogging problems.
Laboratory  methods  for   predicting   the   course   of   the   in-situ
biodegradation should all be improved.
                           sharing  of  meaningful  site  data  by  those
- There is a need for  more
  experiencing  in  this  technology.  This is especially needed as"regards
  data on peroxide stability and transport,  oxygen  utilization  and  the
  removal of fuel residuals from soils. Therefore, projects like Nl,  U3 and
  U6 are very useful. An open policy of organizations  with  experience  of
  the  technology  can  expose  bottlenecks  concerning  both  practice and
  demonstration,  thereby  directing  the  research  of  universities   and
  institutes and making this research more valuable.

• Knowledge about modelling of transport behavior in the soil seems  to  be
  sufficient. Modelling of biodegradation processes in the soil however,  is
  still a difficult problem and requires further attention. A  precondition
  for  further  development  however  is the availability of representative
  data, which should be  published  by  the  experts  involved  in  in-situ
  biorestoration projects.

System design

- Venting  of  volatile contaminating compounds in the unsaturated zone and
  treatment of  these  components  above  ground  (possibly  combined  with
  recirculation  and  biorestoration  in  the saturated zone) seems to be a
  promising and cost-effective method calling for further attention.

- A combination of chemical treatment above ground and biological treatment
  in  the  subsoil  can  possibly  expand  the   application   of   in-situ
  biorestoration,  especially to compounds which are more difficult to break
  down biologically (such as PAHs)  and  more  readily  biodegraded once   a
  first  oxidation step has taken place. Further research in this field can
  be recommended.

- Stimulation  of  the  biological   activity  by  beating  the infiltrating
  groundwater was   used  at  one .project  only  (D5).   Here,   it  was not
  conclusively shown that this was  a cost-effective method. Measurements  in
  test plots  should be  conducted  to  demonstrate  whether  and  when the
  heating effect is economical.
                                  760

-------
- There  is  much  uncertainty  about  the  efficacy  of  the  supply   and
  distribution of oxygen (-sources) in the subsoil.  Research on alternative
  oxygen sources  (02 ,   H202)  and  electron  acceptors  (NOg)   is  useful.
  Hydrogen  peroxide  is  a relatively expensive oxygen source, the more so
  because only a very limited part of it  can  actively  be  used  for  the
  biodegradation of the contaminants; this is estimated to be approximately
  15% (Brown, 1989).

- In- situ  peroxide  stability must be greatly improved to provide adequate
  oxygen downgradient of injection points.

- As   regards  inoculation,  the  selection  by  enrichment  culturing  is
  especially  performed  by  compounds  of  the   contamination.   A   very
  interesting  possibility would be to expand this technique to a selection
  for the tendency of microorganisms to adsorb  onto  soil  particles.  The
  small percentage of the population that does not tend to sorb, could thus
  be selected,  possibly  resulting  in  improved  biodegradation  in  situ
  because  these organisms can be carried a longer distance in the subsoil.
  This aspect needs further attention.

• Co -metabolism,  such  as  the  biodegradation by methanotrophes , deserves
  more  attention   because   It   nay   broaden   the   applicability   of
  biorestoration.
             could be useful with respect to the following aspects:
  * limitations  caused  by  the  low  availability  of contaminants to the
    microorganisms ,
  * extension  of the applicability of in- situ biorestoration for compounds
    with a low solubility.
  In order to open up possibilities for these aspects, fundamental research
  into the  use  of  detergents  in  this  field  is  necessary.  Not  only
  artificial  supply  of  detergents  in  the in-situ biorestoration system
  should be considered, but also the possible use of  surfactants  produced
  by microorganisms in the soil.

Mass balances

- There  is  a  strong  need for mass balances on both laboratory and pilot
  plant scale .  Mass balances will improve the insight in  the  contribution
  of different processes in the total biodegradation process .

- The limited possibilities to monitor biological activities in the soil is
  partly responsible for the lack of knowledge about the process of in-situ
  biorestoration. The  development  of  methods,  which  can  be  used  for
  monitoring the biological processes in the soil, would greatly contribute
  to a better understanding of  the  processes,  and  thereby,  to  a  more
  selective and economical supply of for example oxygen and nutrients .

- In order to gain a better insight into the contribution of biodegradation
  to the total degradation process in the laboratory, a satisfactory method
  for sterile  experiments  should  be  developed.  The  methods  currently
  available are insufficient.
                                    761

-------
- The precise fate of degradation products  is not yet known.   A  proportion
  is   converted   to leachable  DOC, another  part to DIG, but a large part is
  still  unaccounted  for.   Insight  into    the   quantity,   quality   and
  significance   of  degraded  hydrocarbons,   such  as  gasoline, is needed,
  especially as  regards the  question of "how clean is clean?".

 Specific problems

- More attention should be paid to the problem of clogging in the  subsoil,
  resulting  in   disappointing  infiltration  rates.  This  problem   can be
  related to different factors,  such as geology  (permeability),  excessive
  growth of microorganisms,  or high concentrations of iron or manganese.

- Once relatively low  residual  (threshold)   concentrations  with  in-situ
  biorestoration have been reached, the limiting factor usually becomes the
  availability of contaminants to the microorganisms. This is in the  region
  of,   for  example,  less than 250,mg/kg of dry soil in the case of mineral
  oil.   When  cleaning  up   soil  contaminated  by  mineral  oil   in   the
  Netherlands,   residual  concentrations must always be less than 50  mg/kg.
  This makes the limiting factor in this  case,  principally  availability,
  even  more  important.  Further  fundamental  research   in  this  area is
  recommended.

Overview

An overview of the most important recommendations is given in table 6.

Table  6.  General overview of recommendations.
      [Policy
(System design
(Research
      * stioul.mt.ion of experience
        and sharing of information
      * integration of microbiology,
        hydrology and (soil-)
        chemistry
      * praiiminary research
        including heating and
        Bass balances
      * consideration of both
        •oil and groundw«t«r
 * combination of bioraste-
  ration and venting-
 * problem of clogging
 * oxygen:
  • supply and distribution
  • alternative oxygon
    sources
  - peroxide stability
 * Bonitoring possibilities
 * extension to broadar
  application
 • threshold concentrations
 * co-metabolism
 * addition of mlcro-organismn
 • addition of d«t«rtants
 • sterile experiment®
 • modelling of foiorsstoratio-n
 * combination ef chemical
  and biological tr«atnent
                                        762

-------
ACKNOWLEDGEMENTS

The  author  wishes  to acknowledge the experts visited who are involved in
in-situ  biorestoration  projects.   Without   their   contribution,    this
evaluation  could  never  have been made. The open discussions with many of
these experts gave considerable support to this report.

The  author  wishes  to  thank  Mr.  Donald  Sanning  of US-EPA Cincinnati,
director of the above mentioned NATO/CCMS Pilot study, who  has  played  an
important  role  in contacting key experts in in-situ biorestoration in the
USA.

LITERATURE

Brown,  R.A.  Oxygen  sources  for  biotechnological   application.   Paper
   presented  at  Biotechnology  Work  Group.  Feb.  21-23, 1989, Monterey,
   California.                   ,
Brown,   R.A. ,  Norris,  R.D.  and  Westray,  M.S.  In  situ  treatment  of
   groundwater. Presented at HAZPRO  '86,  The  Professional  Certification
   Symp. and  Exp., Baltimore, Md., April 1986.
McCarty, P.L.,  Semprini, L. and Roberts, P.V. Methodologies for  evaluating
   the  feasibility  of  in-situ  biodegradation  of  halogenated aliphatic
   groundwater  contaminants  by   methanotrophs.   Proceedings,   AWMA/EPA
   Symposium  on  biosystems  for pollution control, Cincinnati, Ohio, Feb.
   21-23, 1989.
Downey,  D.C.   Enhanced Biodegradation of jet fuels. Eglin AFB, USA. A Case
   Study for  the NATO/CCMS Pilot Study on Remedial Action Technologies  for
   Contaminated Land and Groundwater - November 1988.
Eyk, J. van and Vreeken, C. Venting-mediated removal of  hydrocarbons  from
   subsurface soilstrata as a result of  stimulated evaporation and enhanced
   biodegradation. Proceedings of  Forum for  Applied  Biotechnology.  The
   Faculty  of  Agricultural  Sciences.  State University of Gent, Belgium.
   Gent, September 29, 1988.                ,                     -,
Fournier,  L.B.  An effective treatment  for contaminated sites. Hydrocarbon
   Technology International, 1988, p. 207-210. Sterling Publishers, London.
Janssen,  D.B.,  Grobben,  G.  and Witholt,  B.H.  Toxicity of chlorinated
   aliphatic  hydrocarbons and degradation by methanotrophic consortia.  In:
   Neijssel,   ,O.M.,  Meer,  R.R.  van   der  and  Luyben,  K.C.AiM.   (Eds.)
   Proceedings  of  the  fourth European Congress on Biotechnology,  Vol.  3.
   Elsevier Science Publishers, Amsterdam. 1987.
Lee, M.D., Thomas, J.M., Borden,  R.C.,   Bedient,  P.B.;  Wilson,  J.T.  and
   Ward,    C.H.    Biorestoration   of  aquifers  contaminated  with  organic
    compounds. CRC  Critical Reviews in   Environmental   Control,  Volume  18,
            (1988), p.  29-89.
                                        and   Sontheimer,  H.   Sanitation of
                                        of   ozone  treated  water.    GWF-
                                        1982.
Riss,    Gerber    and  Schweisfurth.  Mikrobiologische Untersuchungen  uber
    wesentliche  Faktoren  bei   der  unterirdischen  Beseitigung  organischer
    Altlasten  unter   anaeroben Bedingungen mit Nitratdosierung.  Universitat
    des Saarlandes, Homburg/Saar.  1987.
 Soczd,   E.R.  and  Staps,   J.J.M.  Review   of biological  soil  treatment
    techniques in the Netherlands.  In:  Wolf,  K., van  den  Brink,   W.J.   and
   Issue 1
Nagel,  G.,  Kuehn,  W.,   Werner,  P.
   groundwater    by    infiltration
   wassser/abwasser, 123 (8): 399-407,
                                     763

-------
   Colon,  F.J.  (Eds.), Contaminated Soil '88, p. 663-670. Kluwer Academic
   Publishers, 1988.
Staps,  J.J.M.  European experience in hydrocarbon contaminated groundwater
   and soil remediation. RIVM-report; no. 738708002. 1989a.
Staps,   J.J.M.  International  evaluation  of  in-situ  biorestoratipn  of
   contaminated soil and groundwater,. RIVM-report no. 73708006.  1989   (in
   press).
Verheul,  J.H.A.M.,  van  den  Berg,  R.  and  Eikelboom,  D.H.   In   situ
   biorestoration  of  a subsoil, contaminated with gasoline. In: Wolf, K.,
   van den Brink, W.J. and Colon, F.J. (Eds.), Contaminated  Soil  '88,  p'
   705-716. Kluwer Academic Publishers, 1988.
                                   764

-------


i
i (Probli







(Results

^.
«i
5

I Kutrients

(System
•>
u
o
£
1
o
(Contamination
tructure
I



Location










e
o
e =
V •
w • '
s


o e
O O
0 At
IN PW

» V>
2
rw
M
X
o o
. X 6.
i i
e
recirculatioi


i •§
ry -^
o x
t- o
•D hi
IT S.
o
4J
ffi
Q.





u!



• ^
5 H
I
g .
•H
1 ! g
1.H



;f



46-»9.««
removed in
6 months





n
At
n
5
^1




f->
X
IN
X


li >i
At n •»
2 ° E
553


hi
3
><
1
«
B-value for
aliphatlcs an
mineral oil




o
S i
i u
C
recirculatio




«
domestic fuel
oil
!j
i i 1
• • >H


V £
ili
1 _ s
§-*4 *H
0 P
U • A




Q
X
!
c
o. 5
= f |
is:








o
o
m

M

n O
Sew
K
e
recirculatio




3
condensate;
volatile arom.
a
hi
f Qv
** >1
* ~4
1 1
g

*>
f.
m At




1
X






£

g
s

f f
W hi
s s
a v





u

venting
inoculation




3
petrol and
diesel
i
e
e
*H
V
«
1 3



At
At
5

5
1
a s
I i
£
!
y
•w* C
II
S 3
?, 8
£ -H







e
1 5
|
~

V

»OdE"H
E
Reclrculatic

1
r

u
hydrocarbons
|
^4
hi



Chmical
industry sj
E
•
hi
' S
i
o
i
i










s 1
! =
•-« -H
i!
O **
O 4*
n i
m w

nitrate
fertilizer

unsaturated


«
•>
c
chlorinated
hydrocarbons
M
0
i
i
«
2
?
Solution li
•He

o
o
A
0
X
g

















-

inoculation


e ft
O K
! E
hydrocarbons





5
4*
<•
A>
I



|
*
S






c
^

:
«M

ff
V
^
*>

w

£ e £
11 1
III
*«*
unsaturated
inoculation




*
o
iH
ft
1
1
g


4>
B
I
•H



M
?
O




m

i.
in

f







§
g
recirculatil
heating


*
At
m
hi
c
i
i
«•*,
I



Refinery



,
"
O
















.
s
e
a s
I i
a §


4J
«
hi
-H
C
chlorinated
hydrocarbons /
mineral oil
PAH's
«
SJi
U « f*
&
g
«H
*M
S
hi
1
£


9 *t
hi -H
o- ^
I £
s
765

-------
1
I
1
t
!
1
1
t
1
1
1
1


1
1
• 1 •
£ i 1
o, i *

•* 1 1-1
* I
I? i s
1 I *
V 1 M
2 | y
»
i
t
i
v •> i
i
i
• i
. 1 i «
S! 1 i
*w c i *
IH:
£ £ i £
1
1
I
? % \ -• 1
S 3 ! Tg 5
Using
« 5 ! S J!
S 1 ! t g
P -* 1 ft t.


1
1
1
£ s :
S" •= !
k. o |
T) t* 1
£ U. I
I
I
i
I
J
. 1
CJ !
O t «M t-4
•-• 1 It U
i : i «
j t. s.
"• i
? !
V 1
O 1

**.. 1
tw i
, 5 i
1
1
. i
•> i
-4 1
M 1
K 1 *l
' : £
3 ! S
1 | I
i
j
i
i
; : S
! ! 3 "
' 1 O **
' i J §
: | " 5
i S
1 I
i i
i
i '
j i
! i
• i
| :
J i
i i
j i
M IB
' : :
! g 5-
! " ! -
! ;
i . 1
: s i
! S ! S
\ g . ||
I |
! ! S £
! 1 5 ^
1 J •*"" "**
! !*>•«>
i •
• i
! II
! ! *
! °

1 •>. !
i c i e
1 V IH 1 13 .3
1 *J 3 a* 1 £ 3
! 2 a 5 : z B
i 5 T5 * ! 3 -g
i s s £ ! : s

i o !
1 g j
1 1
i 5 III
II its
1 1
1 1
1 1
1 1
J [
« *
£ ' u i e
JJ | 7? 1 -H
1 S s 1 ! !
- - - \ "
', ;
I 1
i i „
i "• * n i
1 • > 1 1
j . : 5 i
1 •* D1 »H |
J »• K 1
: s i t i
1 0 • (M 1
| |
i i 1 '
1 I 2
10- «
ic i "
13 ! "
j |, ||
i i
i
; I
1 EQ *t \ £
[ ^ 3 J 5
J * ** I C >*
i t s ! £ s
; « - | 1 •&
• i
1 » ! M
10 19
1
1
|:
1
!
1
1
I .
1 .

1 ,
-is
. i : «
: § i '
>i *5 i «
« - j A
1
t
i 1 .
Ill
1 "
-!
0 I
e i
0
n i
V> I

I
i '
i ii
•H 1 «•
V | eQ. S
"° i ,*H o £
^ i , O e. re
g i i i i

i
§ i g
*- -H 1 I ^J
1 1 i |
« u G
3 -H 1


I ,
J ;
1 II
! £ £
•s \ * i'

1 ,,
i o !

S 5- !
! I1 i 2
I tJ 1 t-
! rH
! S
- i 1
•3 J * .
2^3
>, ! • *• •
*» 1 , rt •))
1 ilfg
• i o f 2L
i •
i
i
i

| * *
C 1 D
• O 1 «S ^
1 1 1 I i- g.
s f 1 « ! i
i ja
i 3 s a-
! I » i
t , A> -H
i i* • e
•< I •' 8 ^
j i tc o •
g I S 5 S
sis
1 C 1
i 3 1
i g i
1 O 1
1 t4 |
!> i
• s i
: i i
i i
i i
i t
1. 1
! .
- i : i £•

| ~ If
! p
• i >
i i
i • 'm I m ^
! 3 S I 8 :
1 I : M S
I j " **
^ i i
| i
i i ,
!
i i
! |
I (C I
i *n j
1 5 1
1 O 1
t 4J 1
i n I
i K '
! !
1 J.I J
1 *> C 1 *J
15315
! a I i g
! 1 ? i |
j n * i C
i i
i <
i i
t i
i t
: g « ! :
i ri i 1 1
1 n\ 1 1
i j
i i
i i
1 I
1 j
t i
i 1 is
Is i .
: : ||
i >* i
i S I
! ° ! *
ie i • S
! Si i * S
i I A i s I
* s. * i "^

1 • 44 1
ice
1 *4 m i
1 »* H 1
1*0.1
J & _ 1
! ^L
i ? 3 • » -
> C g 1 h u
! S 1 ! S A
' « • i So
a s : s I
j i
i t» i
i - :
! | g j >•

! ^ ' 1 O h
1 C • I C >
1 I s i 5 1
! !
: s i g
             | 3
             2 5
             S. «

             «
                  e «
                If!

: I
8 £
.S 5
                 s
                 H
BTE. naphta
chloroform
Hatardout
va»te lag
                           .
cl
cs,
ned refi
research
766

-------
                                     SSSATO/CCMS Fellow:

                                       Resat Apak, Turkey

       Heavy Metal and Pesticide Removal from Contaminated
Ground Water by the Use of Metallurgical Solid Waste Solvents
                       767

-------
HEAVY METAL AND PESTICIDE REMOVAL PROM CONTAMINATED

GROUND WATER BY THE USE OP METALLURGICAL SOLID WASTE SORBENTS


Re§at APAK,Istanbul University,Paculty of Engineering,
Department of Chemistry,Avcilar,34840,Istanbul,Turkey

Pinal Report to the NATO/CCMS Pilot Study International
Meeting entitled "Demonstration of Remedial Action Techno-
logies for Contaminated Land and Ground Water"
18-22 Nov.,1991,Washington,D.C.


ABSTRACT
   The prospects for heavy meta^ and pesticide removal
from contaminated ground water by the use of waste metallur-
gical solid adsorbents such as bauxite wastes of alumina
manufacture (red muds),blast furnace slags and coal fly
ashes are assessed.In this regard,the composition of solid
waste sorbents (sws){individual adsorption capacities of
constituents;metal speciation in regard to adsorption/desorp-
tion;design of composite sorbentsjsorption performance of
•sws for metals,phosphate and chlorinated pesticidesfpretreat-
ment of sws;mobilization and ground water transport of
metals;solidification of sorbents after loading with metals;
leachability of metals from the solidified mass;technology
selection and prospects for further studies are discussed.


The Composition of Solid Waste Sorbents

   Red muds are the alkaline leaching solid wastes of bauxite

in the Bayer Process of alumina production.The red muds (rm)

had the following chemical composition: Pe20-z:37.26$,
loss on igntion:7.17$.Red muds,being multicomponent systems,
are composed of sodium aluminosilicates,kaolinite,chamosite,

iron oxides (hematite) and jhydroxides.Basically iron is in

the form of hematite, titaniium is in the form of iron-Ti oxides,

and Al in the form of alumiinosilicates.94$ of rm-s have less
than 10 microns grain size.They are supplied from the Seydi-

§ehir Aluminum Plant.

   The blast furnace slags (bfs) were supplied from lsken.de-

run iron and steel plant,and had the following composition:
                            768

-------
   Jly ash is the yery fine ash produced by combustion of
powdered coal with forced draft,and is often carried off with
the flue gases.Recovered by electrostatic precipitators,fly
ash is a mixture of alumina,silica,unburned carbon and various
metallic oxides including TiO^.
   All these waste solids emerge as unwanted by-products
of widespread metallurgical processes in bulk quantities.
Before discussing the adsorptiye performance of these metallur-
gical waste solids,one needs to know the individual adsorption
capacities of their constituents.
Individual Adsorption gapacities of Constituents
   The essential synthetic adsorbents used for metal removal
-which also take place in the compositions of sws- are
alumina,iron oxides (hematite and goethite),SiOp»hydrous
oxides of Ti and Mn(IV),and carbon.
Activated carbon and carbonaceous materials:
   These are capable of adsorbing a great number of metal
cations and anions.Their highest performance is observed
with soft metal comppunds(Hg,Cd,Pb,etc.) and organics.Acti-
vated C shows an affinity for Cr(VI);adsorption conforms to
a Freundlich isotherm (Watanabe,1982).Powdered actvd.C was
more effective in adsorbing Cd(Il) from electroplating efflu-
ents than granulated actvd, G.The rate of Gd removal* at pH 7
was proportional to Cd concn.,thg C dosage and the available
surface sites.The used C may be generated with strong acids
(Huang,1982).Cd adsorption conformed to the Langmuir iso-
therm (Shirakashi,1984).The activated C species (powder,granu-
lar and felt) as well as C black,graphitized C,graphite,
glassy C and anthracite were useful for Hg adsorption in the
range pH 1-14.The lower valence of Hg is favoured in adsorp-
tion (Koshima,1982).
Activated alumina:
   As(V) is effectively removed from water by adsorption
onto activated alumina.Adsorption isotherms can be developed
by batch tests of 7 days (or more) duration (Rosenblum,1985)•
Alumina is an excellent adsorbent for phosphate (which may
cause eutrophication in lakes and water bodies),arsenate and
a number of metals.
                             769

-------
 Hydrous manganese(I?)oxides
    Adsorption of Zn(II)  occurs at the neutral-type surface
 sites which are predominant at pH 3-7 ((Pamura, 1984) .Maximum
 retention by hydrous Mn02 of 'As(III)  occurs at pH^5,declining
 at higher pH to almost half of the total As at pH 10.The
 adsorption capacity of pyrolusite at  pH 6.5 for As(Y) species
 is 10 mmol/kg.The saturation level by MnOp.xHgO was  10 mrnol/kg
 at an As(V)  concn.of 5-8x10"" M;but the uptake steadily, incre-
 ased  with concn. up to 68 mmol/kg with an As(V) concn.of
 2xlO""5M.Added As(lll)  was oxidized rapidly and most  of the
 product was  retained at  pH 3.As release from the sorbent
 occurs upon exposure to  reducing  and  complexihg agents
 (Thanabalasingam,1986).
    Hg adsorption on MnOg is dependent on electrolyte  concn.;
 all cations  and anions like thiosulfate,thiocyanate,iodide
 suppress adsorption.The  greater the ionic potential of Hg,
 the weaker the  Hg adsorption.A Preundlich type isotherm is
 observed over a wide Hg  concn.  range  (Hasany,1986).
 Hydrous Titanium dioxide:
    Traces  of Fe,Cu,Co,Ni,Mn,Cr and V  ions are adsorbed on
 hydroxylated TiOg  from carbonate  solutions (Koryukova,1984).
 Hydrous Ti02 is an excellent adsorbent for U(VI)  even in
 HCO^~ media;this makes TiOg an efficient  sorbent for  U recovery
 from  sea water.(See  Pig.1)A1though the isoelectric point Of
 Ti02  is at pH 5«5  (Tewari,1975),U adsorption is still signi-
 ficant at  higher pH  showing that  specific chemical (non-elec-
 trostatic) interaction may be  responsible for the U uptake
 (Jaffrezic-Renault,1980).
 Goethite (oc-peOOH):
   Metal adsorption  on goethite strongly  increases with pH
 in  the  interval  pH 4-8.The  pH  5Q  values (corresponding to
 50?S adsorption)  are  4.9,5.6  and 508 for Zn(Il) ,Ni(II)  and
Cd(II),respectively.The differences between the  pH 5Q values
agree with the differences  between  the pK values of the  1st.
hydrolysis constants of these cations.Trace metal adsorption
by  the hydroxylated Pe-oxide surface  is strongly affected
                            770

-------
by hydroxo-complex formation of these metals.Besides
adsorption,some metal sorption occurs within the goethite
lattice as occlusion,and the latter resists to desorption
even in acid leaching (Gerth,1983).
   Kingston et al.(Kingston,1967) observed a linear decrease
in the apparent maximum sorption capacity of goethite ;for
phosphate as pH increased,with a change in slope occurring
at pK  values of orthophosphoric acid (pK2:7.2 and pK^:12.3).
He suggested that the net surface charge and phosphate
speciation determined the max. sorption capacity of goethite
as a function of pH;the high capacity above pH 7 was thought
to "be caused by a higher affinity of goethite for H2PO^""
than for HPO^2"" (Mayer, 1986).
Hydrated iron(III)oxide:
   Transition metals bind strongly to "both iron and manganese
oxides (Murray,1978).The relative adsorption rate constants
for metals on iron oxide follow the order Pb > Cu>Cd,as in
the case of their thermodynamic equilibrium constants of
hydrolysis (Zhuang,1984).Surface complexation constants
for cations on hydrous Fe(III)oxide follows the same order
of their first hydrolysis constants (£1:L) as Pb > Cu > Cd
(Smith,1991).
   In a recent study,Cd sorption on jfegO*.^© (am) in the
presence of alkaline-earth  cations was investigated with
emphasis on the Gd-Ca binary system.Competition was observed
primarily  in Cd-Ca binary mixtures.Solubility limitations
in soil and ground water systems will prevent Sr and Ba
from reaching concentrations high enough to allow them to
compete with Cd on iron oxides.However competitive adsorption
of Cd with Ca can be important in soil or ground water.She
extent of  competition increased with increasing Ca conen.,
and may inhibit Cd adsorption up to 20 
-------
           6        V   '   •
 of Cd (10"  and 10"'M)  increased from 0 to 100$ over the
 pH range from 5.0 to 8.5 (Parley,1985) .The pH 5Q was 6.4-
 for 10"7M Cd and 6.7 for 1Q~6M Cd.As is typical of cation
 adsorption,the pH of 50$ adsorption increased as the concn.
 of the Cd and thus the  ratio of adsorbate (Cd)  to adsorbent
 (PegO^.HgOCam) ) increased (Benjamin,1981),
    Adsorption of uranyl(VI)  on hematite vs.equilibrium
 concn,,of U in solution  at various pH values is  shown in
 Pig.2.The characteristics of the adsorbent hematite sol are;
                                                      ry
 particle size:0.12 microns,specific surface areas34 mf/g,,
 cC-iPegO, content:>98$,and isoelectrie point:pH 7.6 (Ho,1985).
    In general,surface complexation constants of various
 metals on hydrated 3?e and Mn oxides may be calculated by the
 aid of the triple-layer model of the oxide/water interface
 (Smith,1991).These constants help to estimate specific metal
 adsorption on hydrous oxide  surfaces.
 Metal Speciation in regard to Adsorption/Desorption
 One or more metal species may be involved in adsorption
 which occurs  as a result of  specific interaction between the
 adsorbent and the adsorbate.Therefore the distribution of
 metal species with respect to water conditions  such as pH,
 Eh,free  ligand concn.etc. may be important for  adsorption/
 desorption.This will be exemplified with U uptake from water
 by iron oxides.
    In the case of uranyl(Vl),all the Fe-oxide materials
 (well-characterized goethite,amorphous FeOOH,and hematite
 sols)  strongly adsorb dissolved  uranyl species  at pH>5-6
 (with adsorption greatest onto amorphous PeOOH  and least
 onto  well-crystallized  specular  hematite).The presence of
 Ca and Mg at  the 10" 5M  level does  not significantly affect
 U  adsorption.However uranyl  carbonate and hydroxy-carbonate
 complexing severely inhibits  adsorption (desorption is favo-
 ured) .Monodentate  U02OE  and  mono-  to tri-dentate (UOp^OH),^
 surface  complexation are  assumed to  compensate  for the
 observed  specific  adsorption (Hsi,1985).Thus  it may be
 deduced  that  U(VI)  speciation,i.e.,the relative abundance  of
 certain hydrolyzed species in solution,is  of  utmost importance
 in determining the  adsorption of U  on hydrous oxides,e.g.,
hematite  sol.
                             772

-------
The sudden increase in II(VI) adsorption onto hematite
around pH 5 may be attributed to increased relative abun-
dance of (U02)5(OH)5* among the hydrolytic uranyl(YI)
cations.This idea is also supported by IR data,(Ho,1985)
Fig.3 compares the pH dependence of uranyl adsorption
on hematite with the percentage of major hydrolysis pro-
ducts of U(VI) .Since the major species responsible for
U(VI) adsorption  is (TJ02)5(OH)5* which is specifically
adsorbed,desorption is not feasible at pH 6.2.Because iron
(III)orthohydroxide"i>e(OH)5.xH20" has multiple sites for
adsorption,some weakly adsorbed uranyl species at high
adsorption density may be desorbed.However U is  basically
strongly retained at low adsorption density indicating
irreversible adsorption as long as the pH is kept around 6.
                                        p .
At pHH,the adsorbate dissociates to U02   which is relea-
sed  from the oxide surface.In  other words,the equilibrium
   (U02)3(OH)^ ^± 3 U022*  +  5 OH""
shifts  to  the right.
   U(VI) speciation as given in Fig.4  is  based  on  the
assumption of dissolution  of atmospheric  C02 in aqueous
solution at the  indicated  pH.(See Pig.4 for observing
the  percentage  of U(VI) hydrolytic  species and  percentage
adsorption with respect to pH  at  a  total  U  concn,  of  10~ M)
Thus the relative  abundance  of a  given species  is  recorded
vs.  pH.Actually relative  abundance  needs  also  be shown
vs.  p [00,2"]= -  Log[c052~]  so  as  to  discuss  adsorption/:
desorption behaviour in  carbonate and bicarbonate  media
 of a given strength (HG05~ is  abundant in Turkish ground
waters).(See  Appendix l:"The  distribution of species in
uranyl-carbonate-hydroxide complex equilibria",Apak et al.,
 1990)
    An appreciable concn.  of U is still adsorbed by hematite
 in bicarbonate solutions.IR data demonstrate that the
 adsorbed uranyl(Vl) species contain carbonate(Ho,1986).
 Comparison of the pH dependence of adsorption of uranyl
 species on hematite with the relative abundance of uranyl
 species in solution containing 10~5M total HC03" suggests
 that (U02)2C05(OH)3~ may be the active species responsible
 for adsorption (See Fig.5)The uptake of U decreases with
                              773

-------
 increasing solution pHjthe decrease coincides with a
 decrease in the proportion of (U02)2C05(OH)3~ present
 (Ho,1986).A specific chemical interaction between the
 adsorbent and the adsorbate is again indicated.Below
 isoelectric pH of hematite; (pH 7.6),positively charged
 hematite particles adsorb negatively charged U(VI) species.
 Above pH 7.6,the particles are negatively charged,and ad-
 sorption of uranyl-carbonate species makes them more nega-
 tive .Especially above pH 8l5,U02(C03)34" becomes predomi-
 nant, enhancing desorption of U from hematite.However the
 adsorption below 7.6 is not simply of electrostatic origin,
 as TiOg (isoelectric pH 5.5)  shows partial adsorption at
 pH 7.6 (See Fig.l^ Adsorption of U(VI)  on Ti02 hydrate in
 10"" M HCO^"" solution as a function of pH) .The adsorption
 of U is irreversible on hematite as long as the pH is
 maintained  (at 8.1).However fast desorption occurs at pH 9,
 the final amount of U retained being equal to the equi-
 librium adsorption of U at pH 9.0.Thus  adsorption is  rever-
 sible wrt.  pH  in this range.
    The addition of humic acid to a hematite sol alters the
 uhtake of U by the hematite particles.The magnitude of
 this alteration varies  with the  solution pH and with  the
 amount of humic  acid  added.Changes  in U speciation as  well
 as  the blockage  of the  active  surface sites,and the extent
 of  TJ-humic  acid  interactions  are  the dominant factors.
 IR  spectroscopic analysis  suggests  that new bond  forma-
 tion may  be involved  in the uptake  of humic  aeid/U by
hematite  (Ho and Miller,1985).Humic acid enhances U(VI)
uptake  at low pH and humic acid  concn.Once humic  acid
 species cover the hematite surface,the  adsorption pattern
between pH 4.4 and 6.4  is reversed,which is reflected in
an abrupt decrease in U uptake in the presence  of small
amounts of humic acid.In this case,surface blockage and
complex formation between unbound humic  acid  in solution
and the U ions tend to keep the U from being  adsorbed.
   The fact that the relative adsorption rate constants
(Zhuang,1984),surface complexation constants  (Smith,1991)
                           774

-------
and tfee pH values for 50$ adsorption (pHe5Q) of a number
of divalent heavy metals(Gerth,1983) agree with the first
hydrolysis constants of these metal cations reflects the
key role of metal-hydroxo complexes in specific metal
adsorption on the hydrous oxide surfaces (e.g.,iron(III)
oxides).The same trend is observed with red muds
(Pb>Gu>Cd ) when the corresponding slopes of their
adsorption isotherms are compared.Thus the importance of
species distribution in aqueous solution wrt. polynuclear
hydroxo-metal complexes is indicated.The fact that water-
soluble complexing ligands closely control Gu speciation
is reflected in the observed decrease of Cu adsorption
onto Na-montmorillonite with increased pH in the presence
of organics-rich soil extract (Gupta,1983).
Design of Composite Sorbents
   Design of composite sorbents for the purpose of
combining the advantages of individual sorbents is an
important part of sorption studies.Such a task may also
serve to manufacture cost-effective sorbents which can
remove a variety of contaminants  in specific pollution
problems.Some examples are cited  below:
   A combined adsorbent  showing a high sorption capacity
for phosphate can be synthesized  by coprecipitatlon  of
Na-silicate  and A12(SO^)5 solutions resulting in an
aluminosilicate  structure  (Yagodin,1982).A  porous sorbent
prepared fromoc-AlpO* and  containing 15$ C  may be made
by using (0-30$) of moritmorillonite binder  (Komarov,1983).
Thus the properties  of  clay-like  adsorbents are  combined
with those  of  oxide-like adsorbents.A  composite  sorbent
Containing  TiOp  and  activated  G may be used for U recovery
from sea water  (Sakane91982).A triple  combination  of an
inorganic residue,an organic  residue,and  a dehydrating
agent  (salt,acid etc.)  is  mixed  and heat-treated  to  yield
a hydrophobia  adsorbent.Such adsorbents  are made  from!
fly ash,spent  sulfite  liquor,modified  bitumen,etc.
 (Kinder,1984).Adsorbents may be  prepared  from TiOgjAlgO^
 and their mixtures as  porous support  materials,and  chemi-
 cally  treated  by freshly precipitating ferric hydroxide
                             775

-------
 onto their surfaces.Column experiments with the synthe-
 sized adsorbents yielded breakthrough capacities ranging
 between 3.2-8.7 mg As/g-adsorbent working with 0.05 mg/L
 arsenic concentrations(Hlavay,1984).TiOg-AlgO^ adsorbents
 prepared from the hydrolytic precipitation of Ti(IV) and
 Al-hydroxides show a high surface area and Co sorption
 capacity after calcination (Pujita,1985).Carbonaceous
 sorbent materials may be prepared by using a mixture of
 peat with industrial wastes (Ogurtsov,1985).
    A detailed list of unconventional coagulants and sorbents
 used for heavy metal removal from water (with references)
 was presented in the first report(Apak,1988).
    The basic advantage of rm-s and bfs-s is their versati-
 lity. Since they are comprised of a number of adsorbents
 and flocculants-all of which are specific for certain
 treatment procedures-these muds and slags are applicable
 in diverse applications,Po example,in the case of As remo-
 val, Pe(III)  compounds  are especially important,and the
 combined effects  of ferric and Al compounds are  stronger
 than that of  unassisted alum.Lime should be used  in
 combination with  alum  and Ipolybasic  aluminium chloride
 for turbidity removal  (Csanady,1982).Prom the  perspective
 of coagulation  to remove  particulates,orthophosphorus
 exerts a coagulant  demand Ifor  Al and Pe(lll),Ca(OH)2
 precipitates  orthophosphates better,while  aluminum  sulfate
 shows higher  efficiency for polyphosphates  removal.Fly
 ash,when combined with  alumina,is a  more  suitable adsorbent
 for P removal than  alumina alone,The  combined usage  of
 Pe(lII)  and Al sulfates provides  a more  satisfactory
 removal  of U  from drinking water  (Lee,1983).Alum coagu-
 lants of improved efficiency are  produced by adding  small
 amounts  of Ti(IV)sulfate  to Al-sulfate coagulants,
 accelerating  the hydrolysi? of Al-sulfate even at low
 temperature,and improving the growth  of the floes.
 (Pukumori,1974).
   On the other hand,mineral acid-activated red muds (rm)
contain soluble Al,Pe(Hl) and titanyl(lV) sulfates (or
chlorides)which present a valuable combined coagulant,
                            776

-------
while Ca and Na-aluminosilicates of varying composition
in red muds and slags contribute a mixture of adsorbents
for different effluent treatments.Thus the beneficial
properties of oxide-like and clay-like minerals in soils
are combined in a single sorbent which makes use  of both
non-electrostatic specific  chemical interaction and ion-
exchange interaction between the adsorbent and the adsor-
bate.ln the interim report  (Apak,1989) the preparation of
a composite sorbent of bfs  and rm  in  2:1 ratio was repor-
ted. She heat-treated  (at 1100° for 1/2 h) adsorbent was
useful for heavy metal removal.The manufactured  adsorbent
could be used  for  the  treatment of pumped ground  water
which might have been contaminated,e.g.,by seepage  conta-
ining metal  plating and  metal  finishing wastewater at a
pH above  5.5.After the relatively easy separation of
 sludge with or without the use of diatomaceous earth or
 perlite  as filtering aid,the treated ground water may be
 used for irrigation or dilution of contaminated ground
 water (Apak,1989).
 Sorybion Performance of Waste  Solid Sorbents for Metals^
 Phosphates and Chlorinated Pesticides
    Red muds and blast furnace  slags have been utilized
  in the removals of heavy metals,radionuclides,inorganic
  and  organic materials,suspended solids,colour,biological
  and  chemical oxygen demand from water streams as cheap
  adsorbents and flocculants (Sho,1973 and 1980).These
  metallurgical  solid wastes,especially red muds in natural
  form,have a high neutralization  capacity of  acidic waters
  due'to their high  alkali  content.Thus  they  are  good  preci-
  pitating  agents for  heavy metals  via hydrolytic  precipi-
  tation reactions.
     Alumized  red mud  solids (ARMS) obtained by sulfuric
  acid treatment of red muds contain the double salts of
  Na-Al-sulfates as the major constituents,with significant
  amounts  of CaS04,Na2O.Al203.2Si02.2E20,FeS04,Si02,Ti02,
  Fe2(S04)3 and 3?e,,03 (Bayer,1972 and 1974) .The indicated
  composition may serve to act  both as a  flocculant and
  adsorbent for water and waste water.An  ARMS plant for P
                              777

-------
  removal has been shown to:be at least 50$ cheaper in
  operation costs than a conventional plant working with
  alum (Shannon,1975).
     The basic materials subject to this research,i.e.,
  activated or natural rm-s,bfs-s and fly ashes act in a
  four step mechanism: (i)  Gel precipitation (sweep floccu-
  lation),  (ii) Flocculation  by adsorption of hydrolytic •
  products,  (iii)  conventional adsorption,  (iv)  Ion exchange.
     Gel precipitation is the most important contribution
  to  mineral acid-activated irm in water treatment,and  most
 heavy  metals are removed via coprecipitation (Gregory,1978).
 Due  to the  presence  of strongly hydrolyzable  species of
 Al(lII),pe(lH) and  Ti(IV),the  primary effect  of ARMS  in
 particulate removal  results  from the hydroxide precipitate
 initially in the form of a fine  colloidal dispersion.(The
 particles then aggregate to  form hydroxide floes which
 enmesh the originally existing  colloids.The entire process
 is referred to as 'sweep flocculation'.
    Besides sweep flocculation,ferric and Al salts in
 ARMS can cause flocculation by specific adsorption of
 hydrolytic products.The multinuclear hydrolysis products,
 formed as kinetic intermediates,including l'ep(OH)^
 Pe3(OH)4^,Al4(OH)84^ and  A18(OH)204+ are even more
 effective flocculants than their parent ions due to their
 higher charge and strong specific adsorptivities.
    Conventional  adsorption is also a mechanism of  pollu-
 tant removal for metallurgical solid wastes.Natural waste
 adsorbents  have  physical (conforming to  adsorption iso-
 therms) and semi-chemical  adsorption as  their primary
 mode  of action,while  after acid  treatment,gel-precipitation
 may  become  the dominant mechanism.(Apak,1989)
   Especially granulated bf-slags and  fly  ashes  can func-
 tion  as synthetic cation exchangers.The cation  exchangers
 produced from these waste solids after acid treatment
have porous  silicagel-H+ macromolecular structure in
which the chief constituents,Na-  and Ca-aluminosilicates,
act as  zeolites.
                            i

-------
   The lab experiments with solid waste sorbents (sws)
•will be presented in terms of the metal removed.First
some terms need to "be defined.
   A Freundlieh isotherm for adsorption is in the form
   msk'c (1/n)  or Log m^Log k+(l/n) Log Gg
         s
where
mtequil. amt.of solute in the solid phase per unit wt.
  of adsorbent ()ig/g)
G :solution eoncn.of pollutant remaining at equil.(ng/mL)
k and  (1/n):Freundlich parameters which relate to specifie
equilibrium capacity and intensity,respectively.
R, is  a measure of metal sorption efficiency of adsorbent
(i.e.,metal distribution coefficient)
R, = metal(bound)/metal(free)
R^-uff  metal removed per g sorbent/ug metal remaining per mL
 d"^0                                      of  solution
Lead;
   Precipitation  and adsorption  procedures were  compared
for Pb(II) at various pH values.For pptn.experiments,
50 mL  of  300  ppm  Pb  soln-as  Pb(N05)2-  was  equilibrated
at the desired pH by the addition of  appropriate  quanti-
ties  of 1 M HC1 and  0.1 M WaOH solns.For adsorption expe-
riments, 2.0 g of  dried red mud (prewashed with H20,-100
mesh)  was equilibrated with  60 mL of  250 ppm Pb soln. at
the  specified pH.The results are shown in Table 1.
Table  1-  Comparison of pptn. and adsorption for Pb(II)
                                                 (a)
Soln.pH   Pptn.removal,^   Adsorption removal,$
   4.9          7.3                98.6
   6.4         ...                99.5
   6.7        25.6                99.8
   8.0        48.0
   8.2         ...               >99.9
  10.3        99.6               >99.9
         than 0.1$ can be desorbed after drying with a
   ,6:1 water-solid ratio at pH 5.0 after 4 d.
    The chemisorption type of interaction of Pb with the
 -OH groups of the hydroxylated rm surface minimizes the
 leachability of Pb from these surfaces as shown by desorp-
 tion tests.
                             779

-------
2.923
2.894
2.677
0.125
0.093
0.320
 Pig,6  shows Rd vs. 00  curves  of  90 mL Pb(N03)2 influent
 solution contacted with A):l  g rm B):2 g rm C):2 g bfs
 at pH  4 for 6 h (room  temp.).The rm-s were acid-treated
 and bfs-s were washed,as described before.The Preundlieh
 parameters were:
    Oase       Log k      (l/n)
    (A)
    (B)
    (0)
    Sorption of Pb with synthetically prepared hydrous
    g conformed to a'Langmuir isotherm of the type:
    l/mzrl.514xlO~4(l/Cs)    for 10 ^C £310
 PH.50!2 and removal is essentially complete above pH 4.
 Copperi
    Pig.7 shows Rd vs. CQ purves of 100 mL Cu(ll)  soln.
 contacted with (A):l g rmi (B): 1 g bfs at pH 4.5  for 5 h.
 The corresponding Freundlich parameters are:
    Case        Log k      (l/n)
    (A)          3.694     ;0.116
    (B)          3.383      0.154
 Actvd.  rm is generally more effective for Cu(H)  removal
 than  bfs.
    For  Cu(ll)  adsorption with  a synthetic  Pe(OE)3 suspen-
 sion  in 1 M NH5  soln.,Log k and '(!/»)  are  4.135 and  0.130,
 respectively for  the  interval  250 ^ GS ^ 6000 jag/mL.
 Cadmium;
    Pig.8 shows Rd  vs  CQ variation of  Cd(II) contacted
 with 1  g of activated rm at pH 4.5.Log k and(l/n)are
 3.508 and 0.101,respectively.
 Mercury(II)•
    The  fly ashes had  quartz and silica,alumina,loss on
 ignition(mainly amorphous ; carbon) and ferric oxide as  the
 major constituents,Si02 having the highest proportion.She
 variable composition  of fly ash depended upon the origin
 of coal and the combustion  process.The fly as was washed
with dil.HN05,then with HgOjdried and sieved through 100
mesh-sieve.2 h equilibration of fly ash with the specified
                            780

-------
       )2 conen.  was carried out at room temp.The Hg
 adsorption conformed to  a Freundlich eqn.,The optimal pH
 interval was  3.5£pH£4.5.10 ppm or less concentrations of
 Hg showed 100$ adsorption "by using 2 g of fly ash per 100
 mL of  solution.The  adsorption isotherms at different pH
 were (m is in mg/g  and CQ is in mg/L)  :
   Soln.pH    k   (1/n)
    2.1     1.01   0.05
    3.2     1.08   0.34
    4.2     1.30   0.36
 Although the  Freundlich  capacity factor (k)  of fly ash
 for Hg was lower than that of activated G,the intensity
 factor (1/n)  was comparable  at optimal pH.
 Uranium(YI) removal;
    Basaltic ground  water,having a higher pH  and carbonate/
 bicarbonate content than granitic ground water,has the
 ability to  leach out uranium possibly  as the tricarbonate
 complex$ this  may be important for the  mobilization of U
 from uranium  mill and mine dumps and vitaified fission
 products  of nuclear reactors.Aquifers  surrounding the
 U,ore  zone  and related ground waters have been reported
 to  show uranium  contamination incidents at  »in situ leach
 mining1  sites (Vandell,1982).
    Rd  vs*  Go  curves of uranyl(Yl)  with rm and bfs are gi-
 ven in Fig. 9  and Fig.10 ,resp., 100 mL  of uranyl  nitrate
 soln.of a  particular pH  being equilibrated with  1 g of
 adsorbent  in  each case.The corresponding Freundlich para-
 meters  with red  mud (rm) were:
    Soln.pH    Log k     (1/n)   m:ug/g  and 0:ug/mL
     3.0       1.864     0.899
     4.5       3.605      0.217
     7.0       3.265      0.383
 The parameters with blast furnace  slag (bfs)  were:  ,
    Soln pH    Log k     (1/n)
     4.5      3.722      0.051
     7.0      1.098      1.326
 Rm  is a much  better adsorbent  for uranyl  than bfs.The
higher  sorption  of uranyl on rm  at pH  4.5-when compared
                           781

-------
to that of pH 7.0 may "be attributed to specific adsorp-
tion of the hydrous oxide surface,since the former pH
is much lower than the po;int of zero charge of most oxides
constituting the rm.(The ,PZO of the used rm was found to
lie between 7«,5 and 8.5,as observed from acidimetric :
titration curves-See Fig.ii).0n the other hand,the basic
mechanism of removal near hydrolytic pptn.pH values seems
to be gel precipitation.                            r
   At pH 4.5,an uranium loading of 10 mg/g on treated
red muds has been observed in batch equilibration tests
Apak,V.,1987)«                                        -
Phosphorus:
   In batch tests,phospho!rus removal by activated rm
was observed to proceed rapidly at first,then relatively
slower,about 70$ of the total P being removed in the
initial stage.Heat treatment of rm prior to P removal
resulted in dehydration and loss of hydroxyl group
                         i
components which occurred1 to the detriment of phosphate
adsorption.She P removal mechanism comprises adsorptive
flocculation and partly ion-exchange resulting in the
adsorption of A1(OH)2H2PO^ on the rm surface(Apak,V.,1987)
   In conventional treatment .techniques,phosphate removal
is maximum at pH. 4 with 3?e(III) and at pH 6 with Al salts
as coagulants(Recht,1970).For waters of higher alkali-
nity, economics usually dictate the use of additional
metal salt rather than pH adjustment.Theoretically the
min. solubility of AlPO* occurs at pH 6.3fand that of
PePO^ at pH 5o3(Apak,1989).
   Column experiments were performed by using a WajUPOx
solutionjEig.12 shows the experimental set-up for obta-
ining the breakthrough and elution curves of phosphate
from the rm column.The rate of influent percolation thro-
ugh the adsorbent was 15 mL per 8 h,with the purpose of
simulating ground water movement at a hydraulic loading
rate of 1 ft/d.Fig.13 shows the breakthrough curve.
                            782

-------
   In ground water,hydrogen ions are produced in
equilibrium reactions involving Mooxidation of organic
matter and ammonia,and through, oxidative hydrolysis of
       and Mn(II).These H* ions react with HGO," in
natural waters to produce carbonic acid which lowers the
pH.Thus elution possibilities of the retained phosphate
with C0r> saturated water (HUCO^) were investigated so as
to assess the recontamination risks of the aquifer.A rela-
tively low proportion of retained phosphate could be
eluted from the rm column with COp-saturated water (]?ig.l4).
Arsenic:
   Arsenic in an oxidative environment dissolves in
ground water as arsenate(V),and. in weakly reducing envi-
ronments,the predominant species is arsenite(III).At low
pE,As(III)sulfide is stable and As is well beyond the
limits bet for drinking water.Since its main species are
chargeless or negatively charged,As would not be expected
to be retained by adsorption or ion-exchange during ground
water flow.
                 "2
   Arsenate (AsO*  ) removal by coprecipitation was maxi-
mum between pH 6-7 on ferric hydroxide,and between pH
5 and 8 on Al-hydroxide»
   Equilibrated beaker tests showed that 1 g activated
rm almost completely removed 100 mL of 20 ppm As between
pH 5 and 6.Arsenate taken up by the rm could be leached
with alkaline solutions.'
   Arsenic adsorption on rm shows a slow kinetics,and so-
metimes equilibrium could not be attained in column tests.

   Information regarding metal removal with rm and bfs
may be summarized in Table 2.
   Sorption of chlorinated pesticides on red mud is
given in Appendix 2 (Apak,Y.,l§9l).-
                            783

-------
  Table  2- Metal removal with rm and bfs
 Metal  removed

    Pb(II)
    Pb(II)
    'Cu(ll)
    Cu(II)
    Cd(II)
   U02(II)
   U02(II)
   U02(II)
Initial metal eoncn.(ppm)
used in sorption studies
       5-30
       5-30
      100-180  '
       30-70
       30-100
       60^150 (pH3.0)
       60-200 (pH4.5)
  Removal effici-
  ency (4 h contact)
 450-1300tig /g-rm
 450-1250ug /g-bfa
7260-8400ug /g-rm
2700-3700ug /g-bfs
2940-4600ug /g-rm
1570-3300ug /g~rm
5630-10980jug/g-rm
       60-150 (pH7.0)    4800-9100|ig /g-rm
 Pretreatment of Waste Solid Sorbents and Possible Benefits
    Even if no pretreatment of wss is carried out,a
 thorough washing is necessary (with H20) in order not to
 contaminate water any further.In the case of rm the alka-
 line filtrate of rm slurry may later, be used for the fixa-
 tion of heavy metals in the used sorbent-cement mixtures
 in the form of insoluble hydroxides prior to landfill or
 burial.
    In experimental studies9100 g of rm was refluxed with
 2 L of 20$ HC1 for 2 h according to the method of Shiao
 et al.(Shiao,1977).The acid-treated rm was filtered,
 washed with water until the washings were free of acid
 (or chloride),dried at 110°,and sieved through a 80-mesh
 sieve.Heat treatment may or may not follow acid treatment
 depending on the purpose.Although heat treatment can be
 performed at relatively high temperatures,a temperature
 higher than 200-300° is not recommendable for P removal.
. Heat treatment  also makes column experiments easy as the
 physical properties of heat-treated red muds are generally
 improved.
    Acid-activation increases  the surface area of rm
 and related adsorbents up, to  a certain acid concn.
 Although all mineral acids may be used for  that purpose.,
 sulphuric acid  is  cheaper than HOI for treatment.A  combi-
 nation of 10^ HC1-15$ FeCl2 may also be used to obtain
 water treatment coagulants (mixed oxides of Al,Fe,Ti)
 (Zakharova,1985).
                            784

-------
The pulp obtained from HC1 leaching  of  rm is  a hydro-
phobic  colloidal dispersion  of  thixotropic structure.
The addition of an electrolyte,or  an increased concn.of
HOI causes the densification of the  dispersion and  aggre-
gation  of particles.The  treated pulp may  be dewatered  by
filtration and centrifugation (Zakharova,1985).Acid-acti-
vation  of rm generally enhances the  sweep flocculation
and gel precipitation mechanisms of  metal adsorption.
   An alternative acid treatment procedure for the
neutralization of rm slurries may  be the  treatment  with
flue gases (NOX and S02  5) to maintain  a  final pH of
6.5-8.5 (Yersiani91983).If alkaline  rm  is let  to be expo-
sed to atmospheric weathering,C02  is absorbed.CaCO, may
then be precipitated by  the  addition of gypsum,and  a pH
just above 8 may be attained.The resulting neutralized rm
is a good adsorbent for  Cd but  a weak one  for  phosphate
(Barrow,1982).
   Granulated bfs-s,whose active components are calcium
aluminosilicates,were washed  with  water,dried  at 110°,
ground and sieved through a  80  mesh-sieve.When using the
bfs as adsorbent,sulfide,an  environmentally hazardous
precipitating agent for  Cu,As and  most  heavy metals,may
be leached out at acidic pH  from the  adsorbent.Thus the
sulfide problem,which may bring further environmental
risks,may be overcome.On the  other hand,the presence of
S ~" may be advantageous  in the  immobilization  of heavy
metals in the form of their  insoluble sulfides in cement
matrixes before landfill.Sulfide is  also  beneficial in
reducing the toxic hexavalent chromium  of  leather tanning
effluents by the reaction
   Cr207   + 2 PeS + 7H20—*-2]Pe(OH)3+2Cr(OH)3+2S+2  OH~
In any case,sulfide in water must  be  carefully monitored
(Freeman,1988).
   Polymer binding of rm may  be  recommended for converting
it into adsorbent particles  or pellets  of  improved  physi-
cal properties.When a water-soluble polymer (e.g.,Ra-algi-
nate)  is dissolved in a rm slurry  and sprayed as droplets
                           785

-------
into a solution containing a hardenerCe.g.^CaClp soln.),
insoluble pellets are formed which may be separated from
solution,cleaned and fired at 200-850  to form the red mud
adsorbent particles of increased surface area and favourable
usage (Seto,1986).
   In the interim report (Apak,1989),the formulation of
a calcined composite sorbent for water purification was
given.The bfs and rm of identical grain size (-80 mesh)
were mixed in 2:1 ratio and calcined at 1100° for 1/2 h.
The calcined material was denser than water,and gave defi-
nite alkaline reaction in aqueous suspension.! max.amt.
of 0.5$ S compounds as MSn was tolerated in the resulting
material (S coming from bfs).This material successfully
removed the divalent metal cations of Pb,Cd,Cu ant uranyl
from their 10 ppm solutions with an adsorbent dosage of
1 g per 100 mil .The metals taken up by the calcined sor-
bent did not leach out at a pH of 5.4 (EPA test).However
as the pH was lowered,the silicate structure of the cal-
cined material would be affected by acidity and desorp-
tion of metals would start to occur.
MOBILIZATION AED GROUND WATER TRANSPORT 01* METALS
   In general,carbonates and bicarbonates and humic
acids mobilize the sorbed U(VI) from hydrous oxides and
oxide-type mineral soils.As the soln. pH and humic acid
concn. increase,a surface blocking effect and a complex
formation with the free humic acid tend to keep uranium
from being adsorbed onto hematitejdesorption is favoured
(Ho and Miller,1985).Adsorption of U on hematite in HOO^"
medium is irreversible as long as the pH is maintained
at S.ljhowever fast description occurs at pH 9 due to car-
bonate complexation,$he final amt. of U retained on
hematite being equal to the equilibrium adsorption of U
at pH 9.Thus adsorption is reversible with pH for this
range (Ho and Miller,1986).Uranyl-carbonate and hydroxo-
carbonate complexation severely inhibited U(VI) adsorp-
tion on ferric oxyhydroxides,U adsorption being highest
                           786

-------
onto;amorphous PeOOH and least onto.well-crystallized
specular hematite.Presence ,of Ca or Mg -abundant in
ground waters- at the 10"5H level did not significantly
affect U sorption (Hsi,1985).U(VI) penetration in tuff
between pH 3 and 4 was not sufficiently reversible but
was a competitive process consisting of diffusion and ad-
sorption.However the penetration depth was thin above pH 6
due to high adsorption coefficient and rate.(Sato,1986)
   The Eh and pH conditions of soil and ground water
significantly affect As speciation and mobilization
from the adsorbed phase.Greatest mobility of As in arseni-
cal waters codisposed with domestic solid wastes occurs
under mildly reducing conditions at pH 5-9,conditions
typical of the initial stages of leaching of domestic
solid wastes.Under  strongly reducing,e.g.,H2S,environments,
precipitation of As as As2S5 (or copptn. with  PeS) mainta-
ined a leachate concn.of less than  10 mg/L for all wastes
(Blakey,1984).pH  and Bh conditions,presence of other
major elements,specific surface area and electric proper-
ties of colloidal  soil particles*As valence and  speciation
all affect the desorption  and mobility  of As  to  differing
extents  (Hu,1985).
    pH and  complexing  agents  control Cu  mobilization.
Specific  adsorption of Cu(ll)  on  goethite  is  accompanied
by 1-2 moles of  proton release  per  mole of  cation adsorbed.
Desorption experiments revealed two adsorption sites  for
Cu of weak eind  strong binding, corresponding to the  readily
and less-readily desorbed  Cu fractions.The  two kinds  of
binding  are  associated with the cation being coordinated
 to 1-  and 2- surface-OH moieties,respectively.The gra-
 dual interchange of some  readily desorbed Cu(ll) into a
 category that is not readily desorbed after an initial
 time-lag between adsorption and desorption is attributed
 to possible time-dependent isomorphous substitution of
 pe("III)  and Cu(II) in the goethite lattice.(Padmanapham,
 1983).The ratio of the moles of charge of Cu(ll) adsor-
 bed to that of the released protons from synthetic Mn02
                             787

-------
 surface indicated that specific interaction of adsorbent
 and adsorbate takes placje in addition to electrostatic
 forces at a pH less than PZC of Mn02 (Kanungo,1984).Humic
 acid concn. up to 1 ppm in soln. caused a sharp decrease
 in Cu adsorption on kaolin (Gupta,1982).Cu desorption
 occurred from Na-montmorillonite with increased pH in
 the presence of organics-rich soil extract while Od
 adsorption was only slightly affected (&upta,1983).
 Among the metals of Cd,Cu,Ni and ZnjCu moved least readily
 through mineral soil columns.All metals showed least
 mobility in a mineral soil with a relatively high pH,
 cation exchanging capacity,and exchangeable base content.
 The order of mobility of metals was Cu«Zn4Ni£Cd.These
 metals were almost completely extractable  from a limed and
 unlimed acid soil by O.I'M HOI soln.,but were  less extract-
 able from a non-acid mineral soil and an organic soil.
 This behaviour is attributed partly to  irreversible bind-
 ing of the metals in organic matter,a phenomenon inhibi-
 ted by the presence  of Al on organic  complexation sites
 (Tyler,1982),Zn and  Cu supply parameters increased when
 carbonates were  removed  from a sandy  clay  soil(Raikhy,1983)
 Desorption of  divalent metals  from goethite  is  a function
 of  pH  such that  the  50$  adsorption pH(pH 5Q) order of
 metals,i.e.,Zn^Ni
-------
clay minerals (ii) existence of clay mineral-organic
matter complexes in soil make this behaviour more compli-
cated (lamamoto,1985).
Solidification of Sorbents after Loading with Metals
   The test alternative for the disposal of a heavy
metal (and/or pesticide)-loaded waste solid sorbent is
the stabilization/solidification treatment "by mixing with
cement and Ca/silica containing hardener,and curing the
resulting mass to increase the compressive strength and
decrease chemical leachability.Some examples are  cited:
   A plating waste containing Cr(VI) and CN~ was  mixed
with 30$ calcined dolomite and 30$ portland cement to
prepare a sandy product which leached out 0.3 rag  Cr(Yl)
per L and 0.1 rag CN~/L (Onoda Cement Co.,1982).A  slag
solidification agent  contained alumina  and portland cement,
OaO,gypsum and soda  ash (Mitsuboshi Kagaku,1983).Coal  ;
fly ash was mixed with 16$ cement and 35$ water,cast and
cured for 4 h at  80°  for  solidification (Mitsubishi Heavy
Ind.,1982).100 pts of rm  containing 60$ H20 was mixed
with 8 pts of portland cement and 0.6 pt.hardening
promoter to  solidify the  mud.A sludge solidifying agent
(100 pts) may be  synthesized from 30-60 pts  of pulve-
rized  bfs,20-45  P"bs  of gypsum and 15-40 pts  of quicklime
dust.100 pts  of  industrial  sludge may be solidified with
7 pts  of this agent  to yield a  solid mass  of 1 Kg/cm
after  3  d  of curing  (Sumitomo Metal,1985).The compressive
 strength may be  doubled within  1 month*Another sludge
 solidifying  agent was synthesized  from fly ash-quicklime-
 -aluminium sulfate in 70/20/10  proportions,This agent
may be added at 10$ ratio to the sludge to be solidified
 (Autoset,1985).Portland  cement was  used for stabilizing
 hazardous  waste containing Cd,Pb,aldrin and chlordane.
 When the water/cement ratio was 0.5,the amt. of leached
 waste was  minimum and the sample had max. compressive
 strength.
                             789

-------
    In general,a mixture *df industrial sludge (or consu-
 med waste solid sorbent) ,;portland cement,and a silica-rich
 material(e.g.,bfs) can "be solidified into a mass which
 does not leach out the retained metals and can be safely
 disposed of.A compressive stregth at the order of a few
     p         " *
 t/ft  may be achieved within 1 month curing.In order to
 aid stabilization,metals may be immobilized via pptn. as
 their insoluble hydroxide.s or sulfides prior to mixing
 with cement so that the resulting solid mass has better
 leaching resistance.! concentrate of alkaline washings
 of the rm-s and the sulfide from bfs may be used for this
 purpose.
    Some wastes (such as As,which slows the curing of
 cement) can cause problems with the cement-based grout.
 In that case an initial precipitation with Ca(OH)2-or
 with rm slurries- may be necessary (Hass,1984).
 Leachability of Metals from the Solidified Mass

    Since the  cement grouts have a basic  pH ensuring the
 formation of  insoluble metal  hydroxides,most toxic  heavy
 metals  are  immobilized as  hydroxides  or  sulfides  in consu-
 med  sorbent-cement mixtures.Also  normally leachable
 radioactive cations,e.g.,of the IA and HA elements,are
 converted to  insoluble  forms.By using the proper  cement
 and hardener  compositions,,metal leach rates  as  low  as
  «*Q    «—8      P
 10  -10"  g/cm ,d  may  be achieved  (Porsberg,1984).
 These wastes may be  processed into  a  granular form,trans-
ported  in bulk to  a  regional disposal site,mixed with
 special cement-based grouts,and pumped as  a wet,waste-
 cement mixture into underground caverns of a selected
disposal site  (Forsberg,1984).This  is recommendable  esp.
in the case of most  toxic heavy metals,e.g.,Pb,Hg and Gd.
Solid masses containing less toxic metals may be used
in landfill procedures.
                           790

-------
Technology Selection and EPA's Recommendation
   EPA1s Technology Screening Guide*recommends in situ
chemical remediation for heavy metal-contaminated soils
and sludges,the potential applications of which include
treatment of metals and radionuclides (mining mill tailings)
by neutralization,'precipitation, and solidifieation/stabi-
lizationjand treatment of hydrocarbons,metals and radio-
nuclides by oxidation/reduction.The methods worked out
up to this point for simulated purification of heavy
metal-contaminated ground waters mainly focus on the
volume reduction technologies of neutralization/precipi-
tation and adsorption/flocculation with industrial waste
sorbents.Solidification with cement-based grouts is recom-
mended for ultimate disposal.
Evaluation and Prospects
   The proposed technology for ground water decontamina-
tion is constructed from a chemist's view point.This
work must be supported with the proper engineering design
and pilot projects in order to take the form of an appli-
cable technology for ground water treatment.Thus the
cost-efective solid waste sorbents will have taken over
the conventional synthetic sorbents,and a cost minimi-
zation will be achieved.
   Although the presented activity does not introduce
a novel technology of treatment,it may bring a novel
approach -in respect to the materials suggested for use
in the treatment of ground water.
*EPA "Technology Screening Guide for Treatment of CERCLA
 Soils and Sludges",EPA/540/2-88/004,Superfund,Sept.,1988.
                           791

-------
                                                    10
Fig.l- U(VI) uptake by hydrated 0?i02 wrt.  pH
Concn. of hematite sol* 0.2 g/L , HC03~  concn.al.ilO"
                         792

-------
H
O   2   4    6   8   10   42
   Equilibrium Concentration (mg
a :  pH 5.35- 6./9;   o: pH 5.26-5.45;

A:  pH 4.51-4.58;   a ; pH 3.04-3.06

 F±g.2-Adsorption of uranyl(VI) on hematite vs. equilib-
 rium U concn,  in solution at various pH values
 Hematite  sol conca.»0.2 g/Lj contact time»l6 h,ta25°
 Ionic strength  maintained with 5«3d0~^M HaCl
                            793

-------
      9O
    80
o*
^  70
,g
1
 o
•TO
<  5O
 QJ
 cn
c 60
  
                                          20
                                               
-------
   J


   i
100

 90

 80

 70

 60

 50
      30

      20

      10
                                             100
                                                C
                                                o
                                                *)
                                             80 w
                                               -o
                                             70-
                                                o
            4.5  5.0  5.5  6.0 6.5  7.O  7.5  8,
                       PH
50 T(

40 ^

30 «|
   «4-
20  J
                                       O   a.
                                       0
Pig.4- 3?he percentage of U(VT) hydrolytio species and
   percentage adsorption wrt, pH at a total U concn.
   of l.xLO~5mol/L.Carbonate in solution is assumed to
   be formed from dissolved atmospheric C©« .
                           795

-------
  80
  60
c
o
fr
  40
  20
            80
            60 -o
                V
                Q-
                                         VI
                                         
-------
            1000-
             500
                                   25                  50    Fb(II),C0(]ag/iBL)-

         Fig.6- Rd va.Co  curves ©f lead(II) with (A):l g r.m.(B):2 g r.m.

                                                  (C):2 g b.f.s.
1000-
 500
       t
       R,
                   \
                       \ (B)
0
50
100
150
              ,7- Rd vs.  CQ  graph far Cu(II) with  (A):  1  g r.m.(B):l g bfa.

                                      797

-------
1000-
      t
500-
   0                   50          !        100
    - Rfi vs.Ce curve. f»r Cd(II) with 1 g  r.».
                                    798

-------
tooo-
 500
      B.

                        50
                                                                150
           i.C curves fer U00(II)  with 1 g r.m.(A)pH 3.0  (B)pH4.5  (C)pH7.0
             O              (~
                                         799

-------
              50             ,     too
                   U02(II),C0(jug/mL).
                                           150
Rd v3.Ce. far
                              with bfa. (A)pH:4.5 (B)pH:7..o
                                 Contaminant solution feed
                               ••-reservoir (to maintain constant
                                 Hydrostatic pressure o-ver column)
                                Adsorbent-perlite(2:l mass ratio)
                                homogeneous mixture
                               -Perlite layer(to aid percolation)
                               Serous glass plate or glass wool plug


                                Tap
                             — Graduated cylinder for fraction
                                                   collection


Pig.l2.-Egerimentaleset-Up for obtaining breakthrough and

                              800

-------
  -eo -»
  -30-
    o •
e>
o
o
I
O
   80
   120
   150
   180
                   N&Gl Concentration


                       0-OOiSM

                       0.033M

                       0.33M
                 T
T	?	S	T

     Equilibrium pK
       Pig.11- fhe  determination  of point of zero  charge

       (PZC) of red mud as observed from aeidimetric

       titration at different electrolyte concentrations.
                                801

-------
c/o.
                                                         Red mud  grain size

                                                            -144  mesh

                                                            100/144  "

                                                             60/100  "

                                                        -0-  25/60   "
                                  Sorbent: 3 g r.m.+1.5 g perlite
                                  Influent:KH2P04 seln.as 15 ppm P
               1    '    i    i   i    i    i    i    i      »i
           10  20  30  40  50   60   70   80  90  100
        Pie-13-Breakthrough curves !f*r phosphorus adsorption on r.m.
                                   Column material:3 S r.au+1.5  g  perlite
                                   Eluent: COg-saturated water

                                                          -144 mesh
                                                          100/144 "
                                                      -O- 60/100  "
                                                      -O- 25/60   «
    0
            20      40      60      ^      100      120  ^..effluent
        Pig..'L4-Elution ef phespherus from r.m. by C02saturated water
                                   802

-------
Leach rates for some radionuc/ides in several matrix materials
Calcines
Supercalcines
  (sinteredl}
Cements
Grouls

Tailored concretes
Bitumen

Glass: industrial
Glass: phosphate
Glass: borosilicate
Silicates: aqueous
        (clay)
Sili
Metal matrix:   .
unstabiHzea calcine
Metal matrix:
phosphate glass
Lead (punt)
Aluminum (pure)
Metal matrix:.
stabilized calcium
      V///////\     Mixed  fission proefuc-ls fH LW)         l


      I       I     Alkali  anJ alkaline ea.rtli me-tals


       K\\\\\\\VI     Actlnides and rare  tarths


      Mg.l5-Leach rates for  some radionuclides in several

               matrix materials (taken from  Porsberg,C.W.,

               Environmental  Science and  Technology,18(i984)56A).
                                           803

-------
                REFERENCES I
 APAK,R.,Initial  Report to the NATO/CCMS  Pilot Study
 entitled  "Demonstration of Remedial  Action Technologies
 for Contaminated Land and Ground Water",2nd.Intern.Conf.,
 7-11 Nov.,1988,Bilthoven,The  Netherlands.
 APAK,R.fInterim  Report to the NATO/CCMS  Pilot Study,
 3rd.Intern.Conf.,6-9 Nov.91989,Montreal,Canada.
 APAK,R.and Apak  V.,'Chemistry-89r Chemistry and Chemical
 Engineering Symp.,Ege  University,Kusadasi,0ct.,1989
 APAK,V.,and tfnseren,E.,in "Plocculation in Biotechno-
 logy and Separation Systems"(ed.Y.A.A-fctia).Elsevier,
 Amsterdam, 1987 .
 AUTOSET,K.K.,
-------
MJKUMORI.R., Japan Kokai 74 07,178 22 «tan 1974
(C.A.80:137087e).
GERTH,J.»and Bruenmier,G.,Fresenius Z.Aiial.Chem.*
316(1983)616.
GREGORY, J., "The  Scientific Bases  of Floceulation?
Plocculation by  Inorganic  Salts",NATO Science Committee,
Sijthoff and Hoordhoff,The Netherlands,1978,pp.101-130.
GIPTA,G.C.,and Harrison,F.L.,Water,Air,Soil Pollut.,
17(1982)357.
GUPTA,G.C.,Soil  Sci.Soc.Am.
Vol.3),202(C.A.105s 48597p).
HUAUG,C.P.,and Wirth,P.K.,^.Environ.Eng.Div.,
(Am.Soc.Civ.Eng.),108(1982)1280.
JAPPREZIC-RENAULT.H..Poirier-Andrade^.,,and Trang,D.H.,

-------
KANUirGO,S.B.,and Parida,K0M.,«t. Colloid Interface  Sci.,
98(1984)252.
KINDER, R,,Teubel,
-------
SE3?0,H.fYamaguchi,H.fana Uryu,H., Jpn.K0kai fokkyo Koho
JP 6115,735 23 Jan 1986(C.A.104:l5l689y) .
SHIAO,S,«f.,and Akashi,Ke,«f. Water Pollut. Control Fed.,
1977,280.
SHIRAKASHI,T.,Kakii,K.,and Kuriyama,M., Nippon Kagaku
Kaishl, 12(1984) 1997 (C.A.102s33l67n) .
SHO,?., Japan Kokai 73 55,888 06 Aug  1973  (C.A.80:17014g).
SHO,T., Agency of Industr.Sci.and Technol., jpn. Kokai
fokkyo Koho,80 105,687 11 Aug 1980 (C<,A.94:l08746c) .
SMITH, R.W. ,and Jenne,i. A., Environ.Sci«Technol., 25(1991)525
SUMITOMO Metal Ind.,Ltd.ftfpn.Kokai Tokkyo Koho
   60 05,297 11 Jan 1985  (C.A.102:190458a) .
TAMURA, H.f and Nagayama,M.,Prog.Batteries  S01. Cells,
5(1984)143.
TBWARI,P.H.,and Lee, W.,J. Colloid Interface  Sci.,
52(1975)77.
THANABALASINGHAM, P. , and Pickering, W. F. , Water , Air , Soil
Pollut.,29(l986)205.
TYLER,L,D.,and  MCBride,M.B.,Soil Sci.,134(l982)l98.
VANDELL,T.D., Interfacing Technol. Solution Min.,
Proc.SME-SPE,Int. Solution  Min.Symp.,2nd.l981(publ. 1982) 299
VERSIAHI,P.,Light  Met.(Warrendale,Pa)l983,337(CeA.
99:42900t).
WATAHABE,K.,and Kishi ,M. ,Mizu Shori Gijutsu,23(1982)683
 (C.A.97:l80970a).
YAGODIN,G.A.,Gorcliakov,V.B.,and Kir'yanoVjH.A., Deposited
D0c.,1982,VI2iri1}I 3974-82 (C.A.99: 2l9253n) „
YAMAMOTO,K.,Kagakuto Seibntsu,23(l985) 80(0. A.102; 203132s) y
 ZAKHAROVA,V.I.,Kikolaev,I.V.,and Egorov,B0L.,Izv.Vysah.
Ucliebn.Zaved. .Isvetn.Metall. , 5(1985) 33(C.A.104: 55726n) .
 ZHU,Y.,Turang XueT3ao,22(i985)390(C.A.104tl47772c) .
 ZHUAUG,G., Chen, S., and Ai,H.,Haiyang Xuebao, 6 (1984)453
 (C.A.101:2l6015k).
                              807

-------
              APPENDIX-  (1)

  THE DISTRIBUTION OP SPECIES IN URANYL-CARBONATE-
  HYDROXIDE COMPLEX EQUILIBRIA

  E.THtem,M.H.Turgut,R.Apak and V.Apak
  Third International KfK/TNO Conference on Contaminated
  Soil,10-14 Dec.,1990,Karlsruhe9BDR
  "Poster presentation"

     Hexavalent uranium-U02(VI)-as a contaminant may be
  fixed with strong-to-moderate  forces  from uranium mill
  effluents  and leachates of  dumping areas.When ground
  water conditions change as  a result of natural carbona-
  tation processes and pH variance,remobilization may take
  place  in the form of anionlc carbonate- or cationic
  hydroxo-complexes of uranyl.Alkaline soils will also
  yield aqueous leachates containing carbonate and bicarbo-
 nate. Carbonate may also be added deliberately as a comp-
 lexing agent for mobilizing the fixed uranium which has
 got bound to silt- and clay-type soils with high affinity^
    Since carbonatation produces a second ligand,!.©.,
 hydroxide,by virtue of  the hydrolysis reactions!
HCO.
                             and
 a complex mixture of uranyl species is formed,each of
 which will be  adsorbed onio solid mineral  phases  to
 different extents depending on their charges  and  speci-
 fic  adsodbabilities.Cationic poly-nuclear  hydroxo comp-
 lexes are held strongly on  hydrous oxide constituents
 of soils   by specific  adsorption  or chemisorption,and
 anionic carbonate  complexes are either physically
 adsorbed  or ion-exchanged by mineral   surfaces.
   If the  total uranyl(VI)  and carbonate concentra-
 tions  in  the aqueous phase  are known,the distribution
 of dissolved species can be  found by analysis of simul-
 taneous complex equilibria.The calculations reveal the
equilibrium concn.of each species,which is crucial in
estimating the remobilization possibilities of U(VI)
                           808

-------
without considering the adsorption/deaorption kinetics.
   There are two ways of dealing \vith this problems
(i) If M denotes U(>2(VI) and L denotes C032~,then simple
modelling predicts the presence of M0MOH,M (OH)2,
M3(OH)5,L,HL,H2L,ML,B!L2,ML jNa^^.irO^and OH" in aque-
ous mixtures of uranyl nitrate and sodium carbonate. If
the total metal conen. is M^.,and total ligand ( f ree+bound)
 concn. is L ,then the pair of equations holds for clear
            w
mixtures:
2K22[M]
10(5pH-70)
                                                   ...(1)
where the cumulative  stability constants of uranyl
carbonate and uranyl hydroxide complexes are represented
by J3. and K..f respectively, and K ^ and K 2 are the
consecutive acidity constants of carbonic acid. Equa-
tions  (1) and (2) may be solved simultaneously for
[M]and[Ljif the equilibrium pH of mixtures are measu-
red. Subsequently, the concn.of each species may be found,
giving rise to the logarithmio distribution diagram.
(ii) If the electroneutrality principle (stating the
equivalence of total charges of positive and negative
species) is introduced,, equations (1) and (2) may be
solved simultaneously by the aid of a computer program8"
without necessitating pH measurement of mixtures. This
gives rise  to an identical diagram as above (i).
   The possible presence of other species is not inclu-
ded in the model, the variance of pH with ionic strength,
and the slow kinetics of uranyl solutions to attain
equilibrium may account for the small discrepancies bet-
ween the theoretical and practical U(VI) speciation.
      details of the computer program may be obtained
 from the authors upon request.
                           809

-------
     0.0


   -2.0


   _4.0


   -6.0.


   .8.0.


  -10.0.


  -12.0.


0 -14.0.


  -16.0.
era
O
  _20.0.
 -22.0
           X:


           D:
           n—n
              uc
rj.
                :   H
                              +: f
                              .12.0


                              .10.0


                              .8.0


                              .6.0


                              .40


                              .2.0


                               0.0
        Logarithmic Distribution Diagram of Species wrt.the  Free
        Ligand Concentration and pH (experimental and  computa-
                                 i    tional curves)
                                                           concn.
            -1—  1      i      1:1     |

      0.0   2.0   40   6.0   8.0  10.0   12.0
                    V mol/L;  Log C*logarithm of the molar concn.

       of the indicated species;pH:=-log aH+ ;p(L]s-i«gCfree ligand)

       L symbolizes carbonate.
                                     810

-------
           A'PPENDIX(2)
SORPTI01 OF CHLORINATED PESTICIDES pl£ RED MJD
            Y. APAK , A". YVBIKOIr and; S .BAYUIiKEN
Introduction
   Chlorinated pesticides,being persistent pollutants in the
environment,are of major concern with respect to toxicity
to humans.Chlorinated hydrocarbons constitute 70-85% of the
residual pesticides transferred to men through the food chain.
   Chlorinated pesticides (CP's) hare limited aqueous
solubility|they are mostly associated \vith suspended mattes?
such as organic substances,sediments and planktons.The pesticide
enrichment in planktons compared to surrounding waters
(bioaccumulation) assumes the for® of biomagnification toward
 the higher members of the predator chain whereas enrichment
factors as high as ICr may be observed in certain fish.
   Pesticides,although diluted to insignificant concentra-
tions in sea water,are mostly concentrated in ground water.
Possible sorbents for CPfs
   A cold trap filled with activated alumina traps all
hydrocarbons except methane from a dry gas mixture stripped
from a water sample with Ar bubbling^  .Higher molecular
weight hydrocarbons than C,-fractions are retained on a
stationary phase consisting of 20% silicone oil on Chromosorb.
Murray and Riley (1973) hare stripped very low concentrations
of chlorinated hydrocarbons from the water with a stream of
nitrogen,and adsorbed them in a cold trap packed with silicone
oil on Chroraosorb-W prior to analysis with GC equipped with
an electron capture detector   .
   Traces of organic substances in water may be sorbed on
an activated carbon filter and recovered by extracting
                          ( p}
the carbon with chloroformv '.Other interfering pollutants
accompanying chlorinated pesticides may be removed by
chromatography on alumina,and the recovered insecticide may
                              f "*}
be analyzed by IR spectroscopy^'besides GC. •
   Red rauds and other unconventional adsorbents,as received
or pretreated,have been shown to be effective sorbents for
heavy metals and amions such as phosphate and arsenate^  .
                            811

-------
 In this work,they are  tested  for  their efficiency in CP
 removal so  as  to  develop multipurpose  sorbents  for retaining
 a  combination  of  heary me-jbal  ions and  CP's.
 EXPERIMENTAL
 Instrumentation
 Gas chromatograph: Perkin-Elmer 8500
 Detector:Electron-capture detector
 Column: 1.5^0V-17,1.35% 0V 210 on  Chroiaosorb W.HP  100-120 mesh
  packed column (GLT,diameter:l/8inch,length:2 m).
 Column tube temperature: 200°C
 Injector temp.:275°C
 Detector terap.:350°C
 Composition of carrier gas: 35% A.r+5% methane
 .Rat'e of flow of carrier gas: 60 mL/min.
 Materials

    Red muds (raajjor constituents being sodium aluminosilicates,
 kaolinite,chamosite,iron o;rLdes and oxy-hydrates ) were
 receired from Seydifehir Aluminium Plant.Their composition
 was: Fe203:37.26f0,Al203:17.58^Si02:l6.94#,Ti02:5.55#,
 Ha20:8..31#,CaO:4.38#,loss on ignition:7.17$ .
    Although red muds may be  heat-  or acid-treated  for raising
 their  sorption capacity,they were  used as  received for the
 sake of procedure  simplicity and economy.They  were washed
 throughly with water until the leachate was no longer alka-
 line, dried in  an oven at  110°C,ground  and  sieved through a
 100 mesh sieve.Perlite  was mixed with  red   mud to  pi-event
 column blockage by adhering  particles.n-Eexane,aceton®  and
 sodium sulfate were obtained from  E.Mercfc,and  were of extra
 pure grade.The CP1 s,namely aldrin,)T-BHC,endrin,dieldrin,
 p,pf-DDE,p,p'-DDT,were received from EPA.
Methods
    Since common chlorinated  pesticides  were basically
insoluble in water,their concentrated solutions  in acetone
were diluted to relatively higher volumes with water  to yield
 final solutions up to 30 ppm of pesticide.
                            812

-------
       experiments:  3 g of red mud were mixed with. 1 g of
perlite.and the column was filled with this mixture as the •
stationary phase.1 ppm of p,p'-DDT solution was passed through,
 and the concn.of the eluate was continuously controlled,in
5 mL portions.The pesticide was recovered from the eluate by
extracting with an equal volume of hexane (5ml per 5 mL),
dessicated through a J^SO^ column and analyzed by GCUp^'-DDT
was retained by the red mud in the coluram up to a quantity of
8600 ug (Ike value for activated carbon is 10,000 p.g per 3 g).
Since the column experiments were lengthy,the experiments were
continued by agitating the red mud with the pesticide solution.
    Aldrin,y-BHC,endrin,dieldrin and p,p'-DDE were tested
under similar conditions.100 mL of the pesticide solution of
concn.10,15,20,25 and 30 ppm were agitated with 0.500 g of
red mud for 1 hr.2 nL portions of the  original and agitated
solutions were extracted with 8 mL of  hexane,dessicated i»
Na2S04 column and analyzed by GO. The  results are  depicted
in Table 1.
RESULTS  '
   All chlorinated  pesticides,excludingJf-BHC,are  retained by
the  red mud coluran  at several 103 ug/g levels.For  retaining
the  latter, red raud-..activated  carbon  filter combinations may
be required.
   Thus  red muds may be  used  to  remove CP's from  ground water.
In this  respect,they may be used for preeoncentration of  CPfs
prior to  GO analyses as  well  as  for  pollutant  removal from
natural  waters.This finding is in accord with a literature
 observation that the pesticide 2,4-D is held by iron oxides
 (goethite)/5^
    In conclusion,red rauds may serve  as multi-purpose sorbents
 for a variety of pollutants including heavy metals and pesticides*
                               813

-------
 Table 1: Sorption  of  Chlorinated  Pesticides  on Red Mud
Initia
concn.
10

15

20

25

30

1
Cppra) Aldrin
83
(788)
91
(1296)
95
(1805)
91
(2161) •
88
(2516)
If a m
Endrin
80
(784)
86
(1264)
• 88
(1725)
92
(2254)
94
(2763)
e o f p
Dieldrin
99
;(940)
98
(1405) -
98
(1856)
93
'(2218)
: 79
(2248)
esticide
DDE DDT 2f-BNfl
89 78 0
(872) (764) -
98 75 0
(1440) (1102) -
99 75 0
(1940) (1470) -
99 84 0
(2425) (2058) "-
99 90 0
(2910) (2646) -


Removal %

Removal %
jug-removed
Removal %
ug- removed
Removal %
ug- removed
Removal %
ug-r©moved
100 raL of the indicated pesticide solution -.vas agitated with
0.500 g; of red raud for 1 hr.
REFERENCES
1.  .t.P^Riley and  G.Skirrow, Chemical  Oceanography, 2nd ed.,Vol.3,
 Acad. Press, London, 1975, pp. 268-269.
2.  A. AflRosen,P.M.Middleton, Anal. Chem., 27(1955)790.
3.  A. A.Rosen^.M.Middleton, Anal. Chem., 31(1959) 1729.
4.  R.Apak, "Heavy Metal Renoval  from  Contaminated  Ground  'fater
by  the use of Metallurgical Solid Pastes and Unconventional
Materials", Interim Report presented  to the NATO/CCMS Pilot-
Study entitled "Demonstration of Remedial Action Technologies
for Contaminated Land and Ground Water", 3rd. International
Meeting, 6-9 Kov., 1989, Montreal, Canada.
5. «t.R.Watson,A.M.P0sner, J.P.Quirk, J.Soil Science, 24 (l973) 503.
                              814

-------
                                NATO/CCMS Fellow:



                              Aysen Turkman, Turkey



Cyanide Behaviour in Soil and Groundwater and Its Control
                  815

-------
CYANIDE BEHAVIOUR IN SOIL AND fGROUNDWATER AND ITS CONTROL

Assoc.Prof.Dr.Aysen Turkman
Dokuz Eylul University                      .
Department of Environmental Engineering
TURKEY

INTRODUCTION                  j

An important  percentage of total  water  consumption in  Turkey is
supplied from groundwater  (about  34 %).  When only  domestic water
supply is considered (drinking + household use) about 40 % of water
is supplied by dams and the remaining is supplied from groundwater
and springs. Thus, the contribution of groundwater to the domestic
water consumption is about  60 %.  As this  high value  indicates
groundwater pollution control is very important for Turkey.

Although the history of environmental pollution control studies do
not go back very far in Turkey, the environmental problems are not
yet at  an  unsolvable  stage.  At  present, because  of the infra-
structure inadequacy problems  in many places,  there are local water,
sea and soil pollution problems.

The causes  of groundwater pollution  in  Turkey may  be grouped as
follows:

a. Pollution due to the domestic wastes: A considerable percentage
of  inhabited areas  is  unsewered  and septic pools are  used  for
wastewater disposal.  The incidence of waterborne infection indicates
evidence of wastewater infiltration  into the groundwater. Also fecal
coliform and  total  coliform analysis confirm the same result. In
some areas sewerage systems is very old and leaking sewers result in
groundwater contamination.                    ,

b. Fertilizer and pesticide application: Although industrialization
is  taking place  at a  rapid  rate,  Turkey still  keeps  her  main
characteristic of being an agricultural country. Since chlorinated
hydrocarbons (including DDT),  organophosphates, carbamates and many
other types of  pesticides have been  used in Turkey, some studies
reveal that  groundwater contains  some amounts of these chemicals,
especially non biodegredable ones  (Temizer,  1979).

c.  Industrial  pollution:  Because many  industries discharge their
wastes which are not  in compliance with the standards put  forward by
"Water  Pollution Control  Regulation",  it  is inevitable  to   have
groundwater contamination  by  industrial source.

This  study gives  a case  study  of groundwater  contamination by
cyanides  and reveals  the  rate of  natural  decay  under  different
conditions and investigates its control.
                                 816

-------
A GROUNDWATER CONTAMINATION CASE

Kemalpaisa, one of the provinces that belong to Izmir, is established
at.the South East corner of the Kemalpasa plain.  Its surface area is
30  km2  and  height 200  m  from  the sea  level  (Figure 1).  The
settlement is 29  km away from  Izmir and takes  place between the
mountains and the Kemalpasa plain which is valuable and fertile. The
plain is famous for its cherry trees. Transportation is easy with a
main asphalt road.

Kemalpasa has a relatively mild climate with cool temperatures and
fairly  heavy rainfall  in winter and  hot and  dry  summers.  The
deficency of water  in  summer  is compensated by water resources in
the area.  Because the settlement is at the foot of the mountain, the
summers are relatively cool and growing different kinds of crops are
possible (Kalayciocjlu, 1978).

The surface  distribution  of   the  geologic  formation  that underlie
Kemalpa^a Plain area are shown on the hydrogeologic map (Figure 2).
The area  may be studied under  two  separete headings: metarnorphic
mountainous mass and alluvial plain that formed lately.  The plain
remains between Mucuk Menderes River basin and Izmir basins.

Water resources in the area may be divided into two groups. The most
important  water source,  Nif  River  is  fed  by  many  small  creeks
flowing from  the Nif  mountain.  Some of these  small  creeks dries
completely during the summer.

No sewerage system is present in the town.  The cesspools  are rather
primitive and unhealthy.  The immediate need for infrastructure is
pointed out in reports. In 1948, water  is brought from 1500 m away
by a  network.  Because it is  insufficient,  groundwater and spring
water are also used in the area.  Some springs are abandoned due to
the pollution.

The springs  in  the area  issue through the layers of flysh (at the
bottom),  limestone (in between)  and gravel  and debris  (at top).
Rainfall  infiltrates  to  the  bottom along  the cracks of limestone
until  the  impermeable  flysh layer is reached. Then it moves upward
along  the fault to  the surface at two sides of the valley where the
springs are  located (Kalaycioglu, 1987). Hydrogeological survey in
the area  reveals the  following:

l.The  aquifer  formation   in  the area  are;  limestone, alluvial
deposits  that  consist of  sand and gravel, and Neogen series which
consist of  sand, gravel and conglomerates.

2.Safe perennial yield of groundwater  in the area is  25xl06 rn3.

3.The wells  in the area which is indicated as "Area appropriate for
groundwater abstraction" in Figure 2, gives water from about 75-100
m deep, and the yield is about  20 1/s  (DS1,  1979).
                                    817

-------

•'  •• • /:'.  '. V«j -& '
     ;•    • I f'f  i
    Figure 1:  Project  area  location
                                   map
                       818

-------
819

-------
 More than 100 industrial organizations are located in Kemalpa§a.
 Some of  them are shown in  Figure 3.  Cyanide  contamination was
 detected as  follows:

 Kemalpasa Municipality asks-  the   industries   to  analyse   their
 effluents in order to determine industrial wastewater pollution load
 in the area. The samples were brought to the Dokuz Eylul University
 laboratory and many  environmental parameters were tested  including
 cyanide.  0.16 mg  CNT/1  was found  at  the effluent of  a chemical
'industry (situated near no.14 in Figure 3) which mainly  processes
 natural resin. The industry objected to the analytical  result  on the
 ground  that  they  do  not   use any cyanide  compound.  >When the
 wastewater analysis was repeateb, the same result was obtained after
 which groundwater was suspected to contain  the  cyanide.  In  fact,
 groundwater sample analysis indicated the presence of 0.07 mg of CN"
 per  liter. The water  was abandoned for  use as drinking water.  Now  it
 is only used  as process water.

 When groundwater of  the chemical  industry was found  to contain
 cyanide, other nearby industries using groundwater were also curious
 about their  water quality.  The; analysis of groundwater samples  in
 the area revealed that a cyanide contamination of about 0.04 mg/1  of
 CN" existed (Table 1).         ;

 Table Is  Cyanide concentrations of  groundwater samples in the
          study area
 Location of the
 well
Depth of the
  well
Date     cyanide
       concent,  mg/1
Beverage Industry (10)
Zipper Manufacture (A)
Coca-Cola (5)

Chemicals Industry II
Food industry (9)


120 m
135 m
60 m

27.1.1988
13.9.1988
20.1.1988
20.1.1988
14.1.1988
27.1.1988
0.04
0.02
0.04
0.04
0.07
0.03
 CYANIDES  IN WATER

 Cyanide refers  to all  of the CN"  groups in cyanide compounds that
 can be determined as the cyanide ion. The cyanide compounds in which
 cyanide can  be obtained as CN" are  classed as simple and complex
 cyanides.

 In aqueous solutions of  the  simple alkali  cyanides,  the cyanide
 group is  present  as CN"  and  molecular HCN,  the ratio depending on
 pH. In most natural water HCN greatly predominates.  In solutions of
 simple metal  cyanides;  the CN group may occur also in the form of
 complex metal-cyanide anions of  varying  stability.  Many of  the
 simple metal cyanides are sparingly soluble or almost insoluble, but
 they form a variety of highly soluble, complex metal  cyanides in the
 presence of alkali  cyanides.
                                   820

-------
 £


 i?
              t
            > t:
£" n > « w


     ? ™
ItSS-SS ?|2Sa!
w2»iSi°Su"-"
g s= s s s r: a si* -
5S-5if ^*I55
»or** wfsi
                                                              ra
                                                              OJ
                                                              u
                                                              QJ
                                                              n.

                                                              OJ
                                                              Ul
                                                              ai
                                                               in
                                                               o
                                                               ra
                                                               u
                                                               u
                                                               3
                                                               cn
                       821

-------
In this
                              a Variety of formula''  but
           cyanides  normally can be represented by AM (CN)  .
   formula, A represents the alkali present y times, ^ the feavy ma
   (ferrous and  ferric ion,  cadmium, copper,  nickel, silver, zinc or

   22TX* ^X thf KUmbeP °f CN  groups'  The  initial  Association of
   In?l  IK ^ S°!Uble' alkalinnBtelic  complex  cyanides  yields an
   anion  that  is  the  radical  M(CN) /'.  This  may  dissociate to some
   extend depending on several factors,  with the liberation  of cV Ton
   and consequent formation of HCN  (Standard Methods, 1981.),

   The  great  toxicity  to  aquatic  life  of molecular  HCN,'  formed in
  solutions of cyanides  by  hydrolytic  reaction of CI\T  with water is
  well known  The toxicitiy of Qf  is less than that of molecular HCN
  and it usually is unimportant because most of the free cyanide (CN
  group present as CN' or as HCN) exists as HCN.
                 undissociated at PH values of 8 and less.  Hence when
     nh    eXpTSKd  in terms  of the cyanic ion it must  be
  realized that most of the cyanide in water is in the form of HCN  It
  H!S *? f° bS rec°9nized that ;toxicities may vary  markedly with  PH
  and that a  given concentration that is innocuous at pH 8 may become
  detrimental if  the  pH  is lowered  to 6.
                disKsocKiation of  the various metal locyanide complexes
    th         ;     h may "^ ^ attained fot~ a long time, increases
  with decreased concentration  and  decreased pH,  and is inversely
  related  to  their highly variable stability.

  The _ ion-cyanide complexions  are  very stable and not materially
           ^     Acutely toxic  level  of,  HCN is attained only in
           that  are  not very dilute and have been aged for a long
                      ^P1^65 are subJect  to intensive and rapid
           ,   K.   K   tOXiC HCN' °n exP°sut~e of dilute  solution to
             9     he P^todecomposition  is slow in deep, turbid and
       laTTT1?  WatTS*   LOSS °f  HCN  t0  the at»^Phere  and its
  bacterial and chemical  destruction concurrent with its production

                            of HCN
 The WHO International and WHO European Drinking Water Standards both
 set a  maximum allowable limit  of  0.05 mg/1  for cyanides  as
                                  tlme ln 1962' the
                                  limit of °*01 mg/i
 The  biodegradability  of  cyanides  has  been  studied  by   many
 investigators.  The species or ' groups of  organisms that have  teen
•SS ; r  t0 aSSimilate ^nide are listed in a study conducted  by
 Zenon Environmental  (1985).  it has been  shown that microorganisms
 can convert cyanide to less toxic  products (carbondioxide, amm^nS?
 siS'2 °orfmrtrate)-  If Iree ^^ are the only soluble cyani^
 species of  concern,   the  technology  for  the  removal  would  be
 relatively  straigt forward.  However, in  the presence of         *
 cyanides  the chemical attributes (e.g.  decay rate)  of
 cyanide become very important. Both ferro and ferricyanide
                                   822

-------
 exibit slow exchange,  but  their  exchange rate  increases  with
 exposure to UV irradiation (Simovic, 1984). Natural degredation of
 cyanides have been modelled for predicting degredation in gold mill
 effluents (Simovic, 1984). Cyanide degradation was found to follow
 a first order  reaction  with  respect to free cyanide  and .metallo-
 cyanide complexes of Zn, Ni,  Cu,~ and Fe (Simovic, 1985).

 Cyanide is  possible  to  destroy by oxidation. In case groundwater
 contamination by cyanides  takes place, the water must be oxidized or
 passed through an  ion  exchanger.   If  oxidation  is  considered,
 chlorine,  hydrogen peroxide or ozone may be used for this purpose.
 Ion exchange' technique may also be used for removal.

 Chlorination: The following reactions take  place when chlorine is
 added to  a cyanide containing water:
 NaCN +  Cl
CN~  +
2CNO"
      C1
          2
        CNC1
3C1
    20H"
           2
               40H"
- NaCl
 CNO'  +  2Cr
•->   2C02  +  f
SCI'
  (1)
(2)
(3)
 Although the first reaction take place at all pH values,  the second
 reaction occurs at pH values greater than 8. Thus the medium is made
 alkaline  with   the   addition  of   a  base  and  the  reaction   is
 accelerated. Cyanide is destroyed completely as the third  reaction
 indicates,  but  if  the water  is  to be  used for any purpose,   pH
 correction  is required.

 Theoretically, the first reaction requires 2.7 mg chlorine  for each
 mg of cyanide. For the second  reaction,  1  mg cyanide requires  2.7
 mg chlorine and  3.1 mg NaOH. As already known, the toxicity  of
 cyanate  is much lower  than  cyanide.   But,   if  cyanide  is  to   be
 destroyed completely,  the third reaction should take place. Second
 step oxidation requires 2.5 mg of chlorine and 1.9 mg of  NaOH for 1
 mg of CNO".  Totally,  for 1 mg of cyanide,  6.8 mg Cl, is necessary for
 complete dectruction of cyanide.  Generally, the  required  amounts  of
 chemicals found  by calculation is different  than the practice.   In
 this study, experiments were  conducted to  determine the  cyanide
 destruction rates related to the chlorine dose in distilled and tap
 water.

 THE EXPERIMENTAL STUDY

 The study consisted of  four parts:

 1) The origin of  cyanide in groundwater have ben  investigated in the
 study area.
 2) Cyanide containing  groundwater samples have been chlorinated to
 determine the extend of cyanide removal by chemical  oxidation.
 3) A model  .study is conducted to determine  the cyanide  retention
 capacity of soil.
4) Natural  degradation of cyanide have been observed in laboratory
conditions to determine the reaction rate constant and time required
 for cyanide-concentration to drop to a nonharmful level.
                                  823

-------
First series: In the area where cyanide pollution took place there
are many industries that may  discharge cyanide  containing wastes.
These  are   dye  industries  (3),  leather   industries  (3),,  enamel
industries (2), zipper industry (1), chemicals industries (3), metal
industries  (8) and textile industries (8).

In order to  determine the industry or industries causing groundwater
and wastewater samples are analysed for their cyanide content  and
shown in Table 2.

Table 2: Cyanide concentrations in groundwater and wastewater
         samples of some industries in the study area
CN" cone. 1988
mg/1 WW
Textile ind.
Ceramics ind.
Metal pi. ind .18
Zipper ind. .44
Chemicals . 16
Industry
1988
GW


.03
.02
.07
March 189 March 89
ww : GW
0.01.
0.45; 0.02
0.02, 0.01

0.16 0.03
May 89
WW
0.09
0.05
0.19


May 89 •
GW'

trace
0.18


Nif River
0.54
0.03
WW: Wastewater sample, GW: Groundwater Sample

Second series: In this part of  the study the effect of .chlorination
on cyanide content has been determined. When groundwater is supplied
by municipality, in general it  is chlorinated prior to entering the
distribution  system.  Thus, the  cyanide is destroyed  by chemical
oxidation. Although the theoretical amount of chlorine required for
cyanide  oxidation can be  found  with the help  of stoichiometry,
because  of the differences between  the  theoretical  and  practical
values,  a  laboratory study is conducted. Known amounts of cyanides
are added to distilled water and tap water and chlorination results
are shown  in  Figures 4 and 5.

According  to the  experimental  results   of  the  study,  complete
destruction  of 1 mg  of cyanide  requires 14.5-20 mg  chlorine in
distilled  water and  18-30 mg:  in tap water.  The study  has  been
conducted  at  normal pH values.

Third series:  The model  shown  in Figure 6 has been constructed to
observe  the cyanide retention capacity of soil.  Cyanide containing
water has been percolated through the soil and the amount of cyanide
remaining  in water has been measured.  Since  the soil composition
effects  the amount of cyanide retained by the soil, the experiment
was  repeated after the  addition of  iron  salts to  the  soil.  The
experimental  results are given in Table 3 for Kemalpasa soil, Table
4 Bornova  soil and Table 5 for  Kupukdere soil.
                                  824

-------
                       0.1 mg CN  in 1 liter  of  distilled  water
                               CI2 solution  added, ml/I
          0.10
          0.09
          0.08
          0.07
          0.06
          0.05
        £0.04
        e o.03
          0.02
          0.01
z
o
              Q5 mg CN""in 1 liter of distilled water
                                      10
                              C12 solution added ,  ml/1
            0
Figure 4s   Chemical   oxidation
           distilled  water
                              o-'F cyanide   with  chorine   in
                                  825

-------
o
         0.10
         0.09
         0.08
         0.07
         0.06
         0.05
         0.04
         0.03
         0.02
         0.01
              0.1 mg CN~ ih  1 liter of  tap water
                          12            3
                                C12 solution  added, ml/1
      o
      O)
      E
  0.10
  0.09
  0.08
  0.07
  0.06
  0.05
  0.04
  0.03
  0,02
  0.01
     0
                      0.5 mg CN~ in  1 liter of  tap water
                                C12  solution  added,  ml/l
      o
       en
       E
     1
   0.0
   0.80
   0.70
   0.60
   0.50
   0.40
   0.30
   0.20
   0.10
                        1  mg  CN~ in 1 liter  of  tap water
           u               10                      30
                            C\2  solution  added, ml/l


Figure 5s  Chemical  oxidation o-F  cyanide? with chlorine, in tap
            water
                                    826

-------
Table 3: Cyanide Retention by Kemalpa§a Soil
Chemicals added
Kemalpa§a Soil
Soil (3 kg) +3 mg
FeCl3/mg CN"
Soil + 2.5 mg
FeS04
Influent cyanide concentration Average
1 mg/1 2 mg/1 5 mg/1 removal
0.37 mg/1
(63 %)
0.30 mg/1
(70 %)
0.52 mg/1
(48 %)
Soil +• mixture of 0.67 mg/1
FeSQ4+FeCl3 (33 %)
2.8 mg/mg CN"
Table 4: Cyanide retention by
0.78 mg/1
(61 %)
0.68 mg/1
(66 %)
1.1 mg/1
(45%)
1.38 mg/1
(31 %)
Bornova Soil
1.9 mg/1 62 %
(62 %)
1.70 mg/1
(66 %) 67 %
2.7 mg/1
(46 %) 46 %
3.4 mg/1
(32 %) 32 %
Chemicals added
Influent cyanide concentration  Average
   1 mg/1    2 mg/1    5 mg/1   removal
Bornova Soil
3 Kg soil+3 mg
FeClj/mg/Clvr
  0.5 mg/1  1.04 mg/1  2.65 mg/1  48  %
      (50 %) (48 %)     (47 %)

  0.45 mg/1 1.00 mg/1  2.5 mg/1
         )    (50 %)      (50%)      52 %
Soil + 2.5 mg
FeS04/mg CNT
  0.77 mg/1  1;56 mg/1 3.93 mg/1
   (23  %)    (22 %;•   (21 %)      22%
SoiImmixture of
FeS04+FeCl3
2.8 mg/mg CN"
  0.80 mg/1  1.65 mg/1 4.21 mg/1
    (20 %)     (18 %)    (16 %)      18 %
                            25cm
                                         Equal distribution system
Figure 6.  The model used to determine cyanide retention
        capacity of soil.
                                   827

-------
 Table 5.  Cyanide retention by Kucukdere soil.
 Chemicals added
Effluent cyanide             Average
concentration (Infl:5 mg/lj  removal,%
 3  kg soil
 3  kg soil
 +150 mg
 Soil*  150  mg FeClj

 Soil*  150  mg FeSO^
 150 mg FeClj

 Soil*  10 g FeClj

 Soil*  10 g FeS04

 Soil*  5 gr FeSO4
 +5 gr  FeClj
       2.8

       2.5

       2.6

       2.4


       0.9

       1.2

       2.0
AA

50

48

52



83

76

60
Fourth  Series:  Since  cyanides  undergo  biochemical  degradation
cyanideconcentation in polluted groundwater decreases in time. When
enough time  is given cyanide eoncentation drops down to nonharmful
levels. The cyanide reduction experiment was conducted  in two runs.
In the first'run, three 10 liters plastic containers have been used.
One  of them contained  3  ,kg:of soil  +.6 liters  of tap  water
containing 5 mg/1 of CM. the third container contains tap.water with
5 mg/1 of CN. Experimental results are given in Table 6  and the data
obtained are plotted in-Figure  7.

In the second run 8 container^ are filled with the following:
a) 10 mg/1 cyanide solution (2 liters) + soil. (0.5 kg)
b) 10 mg/1 cyanide solution (2 liters)
c) 2 mg/1 cyanide solution (2 liters) + soil (0.5 kg)
d) 2 mg/1 cyanide solution (2 liters)
e) 10 mg/1 cyanide solution (2 liters) + soil (0.5 kg)
f) 10 mg/1 cyanide solution (2 liters)
g) 2 mg/1 cyanide solution (2 liters) + soil (0.5 kg)
h) 2 mg/1 cyanide solution (2 liters)
                                   under
                                   light
                                   conditions
                                   under
                                   dark
                                   conditions
The  decomposition  results  are  given  in  Table  7 and  the  time
concentration curves in dark and light conditions with and without
the addition of soil are shown in Figures 8 and 9. All the cyanide
concentrations were determined according  to colorimetric  method
(Standard Methods,  1981).  The decay  rate  constants  obtained  for
various conditions are summarized in Table 8.
                                  828

-------
                                 •—JO mg/l CN

                                     kg soil
                  4567
                   Time,  days
                     10
0    1    23
5678
Time f days
JO
.Figure 7§  Decomposition o-f cyanide
                                  829

-------
   Table 6. Decomposition of cyanide.
Time, days Cyanide Concentration

1


5
6
7

10
5 mg/1, aerated
2.35
1.90
0.80
0.44

0.21
0.13
0.08
0.0
- 	 „ i1
5 mg/1+soil
0..90
0.70
0.50
0.25

0.12
0.04
0.01
0.005
10 mg/1+soil
1.60
1.20
0.65
0.52

0.20
0.10
0.03
0.025
  Table 7. Variation in cyanide concentration (mg/l) with time.
Hours
Light+S
Dark+S
Light
No soil
Dark
No soil
Light+S
Dark+S
Light
No soil
Dark
No soil
0
10.0
10.0
10.0
10.0
2.0
2.0
2.0
2.0
6
1.18
1.38
2.23
3.06
0.68
0.69
0.78
1.07
30
0.45
0.79
1.18
1.54
0.31
0.40
0.48
0.70
54
0.24
0.41
0.55
0.88
0.15
0.25
0.27
0.37
7
i
0
1'
0
0
0
0,
0.
c
i
0
_.i
8
.04
.14
.26
.45
.03
10
1.16
.24
1
0
0
0.
0.
0.
0.
0
0
02 126
.01 -
.04
.23
37 -
01
05 -
.08
.22
150 174
0.0
0.0
0.0
0.25
0.0
0.0
0.02
0.09

0
0
0.
0.
0.
0.0
0
0
198
.0
.0
.0
14
0
'•;•" 	
.01
.04
Table
Concentration
10 mg/1
10 mg/1
2  mg/1
2  mg/1
Ligth Condition
Light  Dark
0.0500  0.0367
0.0182  0.0176
0.0440  0.0300
0.0242  0.0190
                                        Soil
                                  Present  Absent
                                             X
                                 830

-------
o
c
CJ
c
(0
C
O
•H
•P
•H
•a

§
o
                                                                                                                          in

                                                                                                                          >
                                                                                                                          o
                                                                                                                          s

                                                                                                                          o
                                                                                                                          CO
                                                                                                                          o
                                                                                                                          a>
                                                                                                                          a
                                                                                                                          CO
                                                                                                                           W)
                                                                                                                          •H
                                                                  •z
                                                                   o
                                                            831

-------

                                                                                CO
                                                                                                                         o



                                                                                                                         00
                              p
                              CO'
o
iri
'—  o    q
§  «M    -'
 c
Q
in
 i
 w

 o
•H
•P
•rt
•a
 C
 o
 o
                                                                                                                              . c
                                                                                                                               (u
                          o
                          o
         en
                                                                               £

                                                                               |

                                                                               W
                                                                                                 o
                                                                                                 o
                                                                                 o
                                                                                 u>
u
                                                                                                                              •H
                                                                                                                              •O

                                                                                                                              t,
                                                                                                                              (D
                                                                                                                              •a
                                                                                                                              c
                                                                                                                              3

                                                                                                                              H


                                                                                                                              y)
                                                                                                                              •ri
                                                                                                                              fc
                                                                      o
                                                                      c
                                                                 832

-------
BEHAVIOUR OF CYANIDES IN WATER

Although the  groundwater contamination  by cyanides  is not  very
common, it may take  place especially  in  areas where cesspools are
used for industrial wastewater disposal.  In Turkey, water pollution
by cyanides is a  very important problem. For  example  it has been
reported that about 58 kg of cyanides are  discharged  to water bodies
by industries in the city of  Izmir (Sengul, 1987).

It   is  difficult  to  find   the  origin  of  cyanide  when  the
concentrations are low and the source  is more than one or  diffuse as
is the case in this study. Only experimental results did  not give a
straightforward result but.when it is  combined with  site  histories,
it was possible to reach a conclusion.

That the groundwater contamination was caused by zipper industry was
found  as follows: Textile,  metal plating  and zipper  industries were
discharging their wastes to cesspools and the others to  Nif river.
Ceramics industry  had  its  treatment plant.  Nif  River analysis
indicate that the river  receives same cyanide containing wastes, but
it is  diluted so much that the order of cyanides in Nif River is the
same as in  groundwater in the area. Thus the  idea that  cyanide
containing wastes  discharged  to  Nif   River  have  percolated  to
groundwater   has   been   rejected.   It is  possible  that   cyanide
concentration in Nif River may reach to very high concentrations
 from time to time but it is carried away so rapidly that it is not
very probable that it will cause a groundwater pollution of about 2
 km  diameter. Although  there  may  be some contribution  of  cyanides
 from Nif River,  it cannot be the only source. Wastewater  analysis
 and  discussion  with the  industries who discharge  to cesspools
 revealed that the textile industy wastewaters did not contain more
 than 0.1 mg/1 of  cyanide and  metal planting industry was collecting
 their cyanide wastes in metal containers. Zipper industry  had the
 complete manufacturing units until about  one year ago.  Later  on,
 they  decided  that  the metal  plating of zippers would  be  made  in
 another  place.  Informal information obtained from the personnel
 revealed that they  used to discharge zipper  plating wastes to the
 cesspool   After  the   cyanide  contamination   was  detected  in
 groundwater  within  a  circle of about  2  kilometer  including  the
 zipper  industry,  they found  it  more suitabe to move the cyanide
 baths to  another branch they had  in a small  town. .Although many
 experiments  have been  conducted to determine  the cyanide source in
 the area, it was  possible to  determine the contaminating industry by
 communicating with the people,  not  by scientific judgement that
 arises  by interpreting the experimental results.

 Although chlorination of  cyanide contaminated groundwater- cts^ a
 drinking water  treatment  is not  recomendable  due to the ri^ks
  involved, the effect of chlorine on cyanide concentration have been
  investigated in  this research in connection to the  event occured in
  Izmir.  In some areas,  it is possible that  the cyanide  contamination
 may  not  be  noticed  if   the water is  chlorinated  and cyanide
  concentration is not  very  high.  As  Figures  4  and 5 indicate  the
  cyanide concentration diminish very fast  with the application  of
  chlorine. The chlorine solution used is available in the market  and
                                    833

-------
                chlorine. The solution is prepared in such a way that
      ( dose  was equivalent  to  mg/1.  The concentration  of chloriS
  solution   was  determined   before   each   run  considering
  decomposition reactions.                         ^.R^uering
             thlS reaction is PH. ^pendent and at alkaline conditions
  in         °VS accomPlished>^t: completely and rapidly. However^
  in the laboratory study conducted, water PH was not increased since
  the aim was not the optimize the removal  conditions but to observe
  the change in  cyanide  concentration under  natural  conditions
  Comparison of  Figure 4 and 5 reveal  that  less cyaniJe wS oxidSd
  in  the  case  of   tap  water. Although  the theoretical  amount  of
                        oxidati°:n was  t«Jnd to  be 7  mg   Fe(:C=N)6
bu               i^formed is strongly adsorbed on soil particules,
but addition of  FeH   salt  or combination of  Fe"  and  Fe"'  salts
diminished the retention.       .
                                  834

-------
As Table  5 indicates, at  very high  levels  of iron  addition the
positive effect of ferrocyanide retention was pronounced.

Despite the repeated experimental studies conducted to  determine the
effect of presence of  iron salts, the results were somewhat unclear..
But in any case, it can be said that the presence of iron salts has
a positive effect  clue to the decreased  toxicity of cyanides.

When cyanide contaning water comes  into contact with  the soil, the
cyanide  concentration drops sharply at  the  beginning as shown  in
Figure 7. An 80 % reduction of  cyanide in the first day is caused  by
adsorption by  soil,  later on adsorption and rnicrobial  degredation
should   have  occured  together.   In about  10  days  time  cyanide
concentration  drops to nonharmful  levels.  But  the  concentration  is
expected to fluctuate later on due to the reactions like hydrolysis,
disociation,   etc.  Although  at   a  slower  rate,  the  cyanide
concentration  in  water drops  in the course of  time to nonhamful
 levels  as shown in Figure 7.

 In  fact   the  cyanide   determination  experiments  conducted  in
 contaminated wells in Kemalpa§a  after  six months indicated   that
 cyanide has been removed completely from the water since the source
 has been removed from the area.     .

 Reaction rate constants (microbial decay* volatilization) have been
 calculated  by  assuming  that  cyanide  decomposition  fits  to first
 order kinetics (Figures 8 and 9). Thus, the k values have been found
 to be 0.039 for aerated samples  (5  mg/1), and varying between  0.03
 and 0.05  for soil containing  samples;  and between 0.018 and 0.024
 for samples which do not  contain any soil.  This means  that the
 presence of soil  (i.e. adsoption as a  dominant removal mechanism),
 is very effective in cyanide  removal.

 Table 7 which has been completed after 200 hours of experiment time,
 reveals the following:                                    .

 1  Adsorption  of  cyanide by soil is very effective.  That  is why in
 case  the soil  is present,  the first  values  are  omitted  in decay
 equation.

 2  As can be seen from Table  8, decay rates were found to be  highest
  in case of soil containing samples (0.05 and 0.04 for 10 mg/1  and
 0  04   and  0.03  for 2   mg/1).   As  these values  indicate,   the
 concentration does not have a significant effect. The reason fo this
  is due to the  immediate adsorption, the simultaneous  reduction of
 concentrations to 2.23 and 0.68. Thus the difference  between the  two
 concentrations is reduced  to 1.5 mg/1 from 5 mg/1. Consequently,*  not
  only the concentration, but also the toxicity  is reduced. The effect
  of light  conditions is  considerably smaller than  the effect  of.
  presence of soil. Soil  is a very important factor  on decay rate.
  This shows the presence of cyanide assimilating oganisms in soil  and
  the dominant effect of adsorption on cyanide removal.

  The volatilization rates have not been taken into account. But since
  the pH value  has  not  been  changed  during  the experiments,  the
                                     835

-------
 volatilization is assumed to take place to a certain extent.  Thus
 the actual  decay  rates  must be  smaller than the values  obtained
 here.  In  fact  higher   decay rate. at  high concentration  is  an
 indication of  high volatilization at high concentration.

 Adsorption is  dominant only at  the beginning of  the  experiment and
 reaches to equilibrium.  But volatilization is assumed to be  taking
 place during the course  of  the  experiment.

 In  order  to  eliminate the  effect of adsorption on  decay rate  in
 cases the soil  is  present,  the first values were ommitted and  not
 taken into account in decay  rate calculation.  Besides that,  zero
 values were also omitted also since it is not known exactly at  what
 time the concentration has  reached to zero mg/1.

 The first order kinetics fit the experimental data very well  the R2
 values being around 90 to 98 %,  The cyanide concentrations  decrease
 to  levels  much  less  than  1 mg/1  in about  10  days  time.  These
 findings suggested that  the  natural  degredation of cyanide  is a
 potentially feasible  option for environmental control  unless  the
 source is continous and the concentration is very high.


 REFERENCES

 -   Abidoglu,  A.  Latifoglu,  E. (1981)  :  Su ve Toprak  Kaynaklarimn
 Geli£tj.rilmesindeKurulu§larara:si Koordinasyonun onemi, Su ve Toprak
 Kaynaklanmn  Geliistirilmesi ; Konferartsi,  DSi   Genel  MLjcDrluOj
 Ankara.

 -   DSJ  91979)  .-  izmir Kemalpasja Ovasi  Yeraltisulari  Hidroleolojik
 EtucO, Devlet Su isleri  Genel Md., II.  B51ge Mudurlugj,  Izmir.

 -   Kalaycioglu,  R. (1987)  :  Kemalpaipa (NYMPHAION)   Tarihsel  Kent
 Dokusunun  incelenmesi,   D.E.O.Fen  Bilimleri  EnstitusG, Mimarlik'
 BDlOirii Yuksek Lisans Tezi,  Izmir.

 -  Mckeeand Wolf (1978)  : Water Quality Criteria, California  State
 Water Rescources Control  Board,,  USA.

 -   6zis,  0.  (1981)  : Anadoluda  su  Kaynaklarimn  Durti,   Bugjru,
 Yarini, Su ve Toprak Kaynaklanmn. Gelistirilmesi Konferansi,  DSi
 Genel MudjrlOgCi,  Ankara.

 -   Schippers,  J.C.  (1984)   :  Summary of Standards and Goals  for
 Drinking  Water,  International  Course  for  Hydraulics  and  Env
 Engineering, Delft, Holland.

 -   Simovic,  L.  (1984): Report on  Natural  Degradation of Cyanides
 from the Cullaton Lake Gold Mines Effluent, Environment Canada.

-  Simovic,  L.,  Snodgrass,  W.J.,  Murphy.K.L.,  Schmidt, J.W.   (1984):
Development of  a  Model   to  Describe the  Natural Degradation  of
Cyanide in Gold Mill Effluents, Envionment Canada, Ontario
                                 836

-------
- 'Si-movie,' L.. Snodgrass, W.J.(1985): Water Pollution Res. Journal,
Volume 20, No.2, Canada.

-  Standard Methods (1981) : Standard Methods for the Examination of
water and wastewater, APHA, AWWA, WPCF, USA.

-   SengJl,  F. (1987) :  EncOstriyel  Siyarur Kirlili^i ve Aritirm,
Doga  Mjh. ve Qev. D. 11,3.

-   Temizer,  A.  (1979)  :  Seyhan Baraji Sulama  Bolgesi Yuregin ve
Tarsus Ovalan Sulama,  Drenaj ve Kuyu Sulan  ile Toprakta
Kalici insektisit  Bakiyeleri XJzerine Arastirmalar,  I.  Ulusal  Zirai
MiJcadele  ilaclari  Sempozyumu, Ankara.

     Zenon Environmental  Inc.   (1985):  Biological  Destruction of
Cyanide   in   Canada  Mineral  Processing   Effluents,   File  Number
15SQ.2440-4-9184,  Canada.
                                   837

-------

-------
             NATO/CCMS Fellow:
       Robert Bell, United Kingdom



 Environmental Legislation in Europe
839

-------
                Environmental Legislation in Europe

                          Dr. Robert Bell
                 Environmental Advisory Unit,  Ltd.
                     Liverpool, United Kingdom


      Legislation is the driving force behind much  of the
 environmental activity in the UK and is becoming increasingly
 influenced by laws and directives coming from the  European
 Economic Community.  Indeed it is now almost impossible to
 understand national policy without first understanding the
 European policy into which it fits.

      The European Economic Community (EC)  was established by the
 Treaty of Rome in 1957.  -It now has  some 12  member countries.
 The  Single European Act of 1986 first added  the  word
 "environment" to European policy and gave explicit recognition to
 the  preservation and protection of the environment.   This is a
 powerful tool and all legislation of the EC  now  has to consider
 affects upon  the environment.

      Environmental Action Programmes are used by the EG as a
 policy framework to outline intentions and ideas for legislation
 and  other activities in the year ahead.   We  are  currently in the
 4th  Programme which runs until 1992.

      Before the Single European Act  of 1987,  all EC legislation
 had  to be adopted unanimously;.   Now  it follows the "cooperation
 procedure"  and can be approved by majority vote.   EC legislation
 must be implemented in the member countries,  through national
 legislation,  typically between 18 months and  three years  of  EC
 adoption.   The range of  EC legislation is  now vast,  and includes
 second generation traditional  and new areas of law such as
 environmental auditing,  environmental assessment and eco-
 labelling.  Legislation  items  adopted per  year has increased from
 1 to 5  in the early 1970's,  10  to 15  in  the early  1980's,  to 29
 in 1989.

      The  key  areas  of  EC environmental  legislation are  relatively
 simple  to understand.  Briefly  they  are  as follows:
Wastes

     a)
     b)
     c)
EC law establishes a framework for waste disposal
It requires toxic and dangerous waste to be disposed of
without human health or environmental effects
It establishes a system for transfrontier shipments
Water
     a)   EC law sets quality criteria standards for 62 compounds
     b)   It prohibits and regulates discharges to groundwater of
          listed substances
                               840

-------
          It. sets a framework to eliminate 17 compounds of
          inland, coastal and territorial waters
Air
     a)

     b)

     c)


Chemicals
EC law sets air quality limit values for SO2/  NO2, lead
and dust
It ensures BATNEEC (Best Available Technology Not
Entailing Excessive Costs) for new industrial plants
It sets emission limits for combustion plants for SO2,
NOx and dust
     a)   EC law has limited the production of CFC's and other
          ozone depletors
     b)   It establishes a new system for testing, classifying
          and packaging new chemicals
     c)   It bans or restricts the use of PCB's, PCT's, benzene
          asbestos, lead and organotoxins.

     There are many EC laws coming forward.  Those agreed  by  the
EC but not yet implemented in the UK include:

     a)   Procedures to be followed to allow the release of
          genetically modified organisms
     b)   A freedom of information act which allows the public to
          gain all environmental data      .                 .
     c)   The establishment of a European Environmental Agency

     Proposals currently being considered by the EC include:

     a)   Fixed vehicle emissions
     b)   Cadmium as a dangerous substance
     c)   New systems for introducing pesticides
     d)   Risk assessments for existing chemicals
     e)   A new definition for hazardous waste
     f)   Stopping waste shipments to Africa, Pacific  and  the
          Caribbean
     g)   Minimum standards for  landfills
     h)   Secondary treatment for all sewage discharges.
                                841

-------

-------
             NATO/CCMS Fellow:




 Michael A. Smith,  United Kingdom



                In-situ Vitrification
843

-------
                       IN-SITU VITRIFICATION


       Michael A. Smith, Clayton Environmental Constulants, Ltd.,
        68 Bridgewater Road, Berkhamstead, Hertfordshire HP4 IJB
        United Kingdom                                          '
 1. INTRODUCTION

 In-situ  vitrification  (ISV)   involves  the  in-situ  melting  of
 ~™?minlted solids at temperatures typically in the range 1600 to
 2000 C. The energy required is applied through electrodes inserted
 into the ground.  Organic contaminants are destroyed by pyrolysis
 Inorganic pollutants are immobilized  in  the solidified mass.

 In the United States this technology  is  available (but see below)
 from only one commercial vendor (Geosafe Corporation) and has  been
 under development by Battelle Memorial Institute  since 1980.

 This technology was  to have been the subject of a  separate chapter
 "} Jhe^ain  report,  however,  for  the  reasons explained below,  the
 US Environmental  Protection Agency's SITE demonstration project on
 which it  was  to have been based did not  go ahead. Although it  was
 not possible  to complete  the  chapter  as  planned it  was decided to
 include  an   account  of  the  technology,   based  on  information
 available in  the  published literature, as  an Appendix  to  the main
 report.

 The SITE  demonstration, and the main  clean-up project  that it  was
 to  complement, could not go ahead  because the vendor  of   the
 technology withdrew  it  from the market.

 In July 1991  Geosafe Corp.  infprmed the US Environmental Protection
Agency  (EPA)  that it would no  longer being offering its  in-situ
vitrification technology.  This decision was prompted  by  an accident
that  occurred in March  1991  during  large scale  testing which
resulted  in a fire and destruction of some equipment. The company
took  the  view  that  although  the accident "involved  only non-
hazardous  materials,  ..  such an  event  could have unacceptable
consequences if it were to happen at a hazardous waste site."  The
company did  "not  consider it prudent  to proceed  with  commercial
                               844

-------
large-scale  operations  if there  is  any  reasonable chance  of
recurrence  of this  event.."  Some  work  is continuing  with the
Department of Energy (Anon 1991).

Although the planned demonstration project was not completed, some
background  information on  the  site  where  it  was  to have been
carried out has been included  in this  Appendix,  as a indication of
the type of site to which the technology might be applied,  if the
technical problems can be overcome.


2. SCIENTIFIC AND ENGINEERING BACKGROUND

2.1 The Process

The description  below is  based on  a number of references  (e.g.
Geosafe  1989,  Hansen  &  Fitzpatrick  1989,  Hansen,  EPA   1990)
generally originating from the  Geosafe Corporation.

An array  of four electrodes  is placed to  the  desired  treatment
depth in the volume  to be treated. A conductive mixture of graphite
and glass frit is placed on the surface between the electrodes to
serve  as  an  initial conductive path.  As  electric  potential is
applied between the  electrodes, current flows through the starter
path heating it and  the adjacent soil to the  melting temperature of
the soil (typically 1600-2000°C) . Once molten, typical  soils  become
electrically  conducting  (conductivity tends  to  relate  to the
concentrations of monovalent alkali  cations  (Na, K etc) and in some
circumstances  these may  need  to be added) . Thus  the molten mass
becomes the primary conductor and heat transfer medium allowing the
process  to   continue.   The  molten  mass  grows  downwards  and
horizontally  as  long  as  power is  applied. The process of melt
growth is illustrated in Fig. 1.

With present  equipment a single "setting" treats a zone  8m  (27ft)
square by about 6m  (20ft) deep  (about 900  tonnes at a  bulk density
of 2,200kg/m3)  and takes 7 to 10 days to complete. The  treated mass
takes  several months  to a year  or  more to cool,  although the
collection hood can be removed after a few hours. Once one setting
is completed,  the equipment is  moved  to an  adjacent area (this is
claimed to take about 16hours).  Neighbouring blocks fuse  together.

The temperature  needed to  melt the soil or other materials  to be
treated will depend on their composition. Typically temperatures of
1600  to 2000°C are  reached.  High concentrations of  alkalies or
halides may be expected to  act as fluxes resulting in  lower melting
temperatures.

As the high temperature melt expands  by about 25-50mm/hr (3.6-5.5
tonnes/hour,  4-6  short tons/hour),  a very  steep thermal gradient
 (600 to 1000°C/100mm) precedes the melt. Within the melt a vigorous
                                845

-------
 chemically reducing environment is produced.

 It is  claimed that  vapours migrate  towards  the melt zone. At
 appropriate temperature  regimes within  this  zone  and the  melt
 itself,  solids and  contaminants undergo a change of physical state
 and decomposition/degradation reactions.

 As a  result of the  ISV process individual contaminants  may:

      (a) undergo  chemical and/or thermal  destruction,

      (b) enter the  off-gases from which they can then be removed,

      (c) be chemically or physically  incorporated  into  the
         resultant  solid  product.

 Typically,^organic  materials are destroyed by pyrolysis  (thermal
 decomposition  in the absence of oxygen to  component elements), and
 inorganic   contaminants  are  incorporated into   the   glassy  or
 raicrocrystalline, monolithic residual  product.

 Organic products of incomplete pyrolysis are usually gaseous.  These
 will  move  slowly through  the highly viscous melt  towards the
 surface. Some may dissolve in the melt, the remainder  evolving from
 the surface are captured  in a collection  hood. In  the presence of
 air those that are  combustible will burn  (this clearly  has safety
 implications).  Finally,  evolved volatiles and  the  products of
 surface  combustion  are subjected  to appropriate off-gas treatment
 processes.

 Inorganic   compounds  may  thermally  decompose or  otherwise  be
 retained within  the melt.  For example,  nitrates,  carbonates and
 sulphates yield gaseous products which may dissolve in the melt or
 evolve through it to enter  the off-gas collection  hood.

 As stated above,  it is claimed that vapours will be drawn towards
 the melt. A theoretical explanation for  this  phenomenon has been
 provided  (Geosafe 1989).  However, what happens  in practice must
 relate  closely to  the porosity/particle  sizing  of  the  soil as
 conflicting processes,  capillary  suction  and vapour diffusion are
 involved which respond differently to these parameters.  It is said
 that within the heating zone around the melt an  equilibrium will be
 established  between inward migration  caused by  capillary action
 (for example as fine soils dry out soil suction will increase) arid
 outward migration of vapours  (more rapid through permeable soils).
Vapours will diffuse towardsithe melt and outwards towards cooler
 zones due to concentration and thermal gradients.  However,  those
moving  outwards  will  tend  to condense   being  drawn once  again
towards  the heated  zone. Any that are adsorbed on  the particle
 surfaces will be dealt with by other settings as processing moves1
across the site.
                               846

-------
Within  the dry  zone movement will  be aided  by channelling  and
cracking,  and possibly by the negative pressure sustained within
the off-gas collection hood  (see below).

Movement of organic vapours may also be assisted by the presence of
relatively large quantities  of water vapour -  a process analogous
to steam stripping may occur.


2.2 Applicability

The process may be applied to contaminated natural  soils,  but also
to other naturally occurring soil-like materials  and solid process
wastes, including silts,  sediments, sludges and tailings.

Most soil types can be treated provided there is sufficient silica
and alumina present  to form  a melt at the temperatures achievable
with  the  system. There need to be  sufficient monovalent. alkali
cations  (Na,  K etc)  present to provide  the  degree of electrical
conductivity  necessary  for  efficient  operation.   If necessary,
alkali fluxing agents such as sodium carbonate  may  be added to  the
soil.( In  general alkalis  and  halides  such as   chlorides   and
fluorides will act as  fluxing agents lowering the temperatures at
which a melt will form.                                    ;

While  broadly applicable to  a  range  of contaminants,  present
equipment  is limited to maximum concentrations of organics in  the
5 to  10  wt %  range, depending on the compounds present; and to
metal concentrations not  exceeding 16 wt %. It is also limited by
the presence of buried  metals in excess of 5 wt % and the need to
maintain a negative  pressure within the off-gas collection hood.

The limitations on organics content arise from constraints imposed
by the gas collection and treatment systems in terms of amounts of
heat and gases that  can be tolerated/handled.

If metal concentrations are  too high,  reduction  to the metal may
occur under the  reducing conditions in the melt.  Metals  such as
iron may then collect as a separate phase at the base of the melt.
In  contrast,   low melting metals  such  as  lead   may  partially
volatilize and enter the off-gasses.  It  is to be expected that all
mercury will be lost in this way.

The process is applicable to fully saturated soils, however it is
limited to those  situations in which recharge to the  treatment zone
is reasonable  relative to the rate of growth  of the molten  zone
(limiting  permeability  is  put  at  1  x  icr5cm/sec) .  In  these
situations steps can be taken to limit  the rate  of recharge,  for
example,  by lowering  the water table by pumping. It takes about the
same amount of electrical energy to remove a unit mass of water as
it does to melt  a similar mass of  soil.  There are thus  obvious
                               847

-------
advantages  in  controlling the water  regime  within the treatment
area.

It is claimed  (Geosafe 1989)  that in tests at various scales the
process  is  applicable  to  soils  containing  volatiles,  semi-
volatiles,  and refractory  (non-volatile)  organic  compounds  and
inorganics  (heavy metals), a variety of radioactive materials, and
a  broad  range of  combustible,  metallic,  and  inorgcinic  scrap
materials  (e.g.  paper,  plastic,  wood,  drums,  concrete,  rock,
asphalt) . It  is also claimed that  drummed organic wastes can be
treated.

The  ISV process results in  a  20 to  40% reduction  in volume for
typical soils.  Once  processing  is  completed  clean  backfill is
placed  over the  residual monolith to  restore site  levels.  As
cooling continues additional subsidence may occur.

There  are  some topographic  constraints  to  application of  the
process.  The  hood  is  fitted with  a skirt capable  of accepting
variations in ground levels of about  0.15m (6in). The ground area
supporting the hood should not slope by more than  5%.  Care must be
taken to prevent surface water from entering the treatment zone.

In  general  therefore,  some  prior  preparation of  the  site  to
accommodate the equipment, may be required.

The  process should  be performed  at  sufficient  distance from any
structures  to  avoid damage.  This might  occur  because of thermal
effects or subsidence due to soil densification.

During cooling of  the melt the thermal gradient will tend to spread
out compared to that during the operating stage. For example, the
100°C isotherm may move from about 0.3m to as far as 1.5 to 2.1m (5-
7ft) from the residual monolith.  The rule of  thumb  advice from
Geosafe is that no setting should be placed closer than about 4.6m
(15ft)  to any vulnerable  structure of service run (utility run).
Similarly potential subsidence effects should be taken  into account
and precautions taken if necessary.

There is no evidence  (Geosafe  1989)  that the process has induced
stray  electrical  potential  or  currents  in  nearby  structures,
service lines or pipelines.


2.3 Residual Product

The  residual  product is typically  about 10 times  stronger than
unreinforced concrete  in  both  tension (27-55MPa,  4,000-8,OOOpsi)
and compression (206-276MPa,  3^0,000-40,OOOpsi) .  It is not affected
by either wet/dry or  freeze/thaw cycling.  It  is free of organics
and passes the US Environmental Protection Agency's EP-Tox and TCLP
                               848

-------
leach testing criteria for priority pollutant metals. It also has
acceptable biotoxicity relative to near surface life forms. It is
projected that the residual  product will withstand environmental
exposure for geologic time periods.

Depending upon composition,  silicoaluminate glasses,  such as those
produced  in  the  ISV  process,  may  be  reactive  with  water,
particularly  in  the  presence  of  alkalies. An  example of  this
reactivity is  provided by vitrified blastfurnace  slag which has
been used  for  over one hundred years  as a cementitious material
with Portland cement or alkaline hydroxides as activators. However,
because  of the monolithic  nature  of  the residual  product,  any
reaction would be restricted to the immediate surface of the mass
of material and is therefore very  unlikely to lead to significant
release of toxic materials even over very long periods of time. It
should also be noted,  that the slowly cooling parts of the mass may
undergo some crystallization:  i.e.  it  is unlikely that the final
mass will  be completely  vitrified  in  the true sense  of  being a
uniform glass.

Tests to assess the rate of hydration of the mass have been carried
out  (Geosafe 1989).
2.4 Equipment

The  basic ISV  equipment  system is  shown in  Fig.2.  The maximum
distance between electrodes in the large-scale system is about 5.5m
 (18ft) allowing formation of a maximum treatment zone of about 8.2m
 (27ft) radius.

The  major  part  of  the   ISV equipment  system  is  the off-gas
collection  and  treatment  system.  A 16.8m (55ft) diameter off-gas
collection  hood directs any  evolved gas  to the treatment system
utilizing  quenching,   pH  controlled  venturi  scrubbing,  mist
elimination  (dewatering), humidity control,  particulate filtration
and carbon adsorption, to  ensure clean air emissions. The quenching
and scrubbing solution is  cooled by a self-contained glycol cooling
system eliminating the need for a continuous on-site water supply.

Flow  of air  through  the hood is controlled  to  maintain a negative
pressure  (3.4-6.9Pa, 0.5-l.Oin water gauge). An ample supply of  air
provides  oxygen for  combustion of pyrolysis products  and volatile
organic compounds (VOCs). Typically evolved gases  amount to less
than  about 1% of the total volume of gas processed by the treatment
 system.

Contaminants collected in the scrubber  solution,  filters and or
 carbon adsorption beds may be disposed of by feeding them back into
the  treatment zone.
                                849

-------
 Electric power is typically  taken  from  a public supply at either
 12.5  or  13.8kV.  The  AC power   is  converted  to  2-phase  and
 transformed down to 4000 volts for initial processing. Voltage is
 then progressively reduced during  an  ISV "setting"  to offset the
 increased  conductivity   of   the  growing  melt.   Typical  soil
 applications require 0.8-0.9 kwh/kg (0.35-0.4 kwh/lfo)  of material
 treated. Such electrical services are usually available from public
 supplies  but may  also  be  provided  by  diesel  generators  when
 necessary.

 Means are  required to  determine  the  extent  of  the molten  zone
 (geophysical, optical and thermal methods may be employed) and for
 gas migration.


 2.5 Health & Safety

 The  most  direct hazards would  appear  to  be  the  very  high
 temperatures  reached  in  the; process  and the  large  amounts  of
 electricity that  must be employed.

 It seems  likely  that  hazardous conditions  could  occur in  the
 collection hood  if significant  quantities of  volatile  organic
 compounds are  evolved,  especially  given  the  high temperatures
 involved.
 2,6 Potential  Environmental  Impacts

 Impacts on the environment and public health should be  small given
 the in-situ nature of the process. Operation  may, however, require
 some preliminary earth moving works to prepare the  site within the
 constraints  on site  topography  of the ISV  system.  Thus,  should
 perimeter dust problems  might arise.

 The process requires 24hr operation so in populated  areas there may
 be noise and other operating impacts.
                              i
                              t
 3. THE PLANNED CASE STUDY

 3.1 Introduction

Although the planned  case  study  was  not completed  the background
 information  on the  proposed  study  has  been  included  in  this
Appendix as an indication of the type of site it is believed that
technology of this type might be applicable to.

                                                               '". " ^
3.2 The Case Study Site
                               350

-------
Only one  case study was  to have been  included  in the NATO/CCMS
STUDY: A  US EPA SITE (Super-fund  Innovative Technology Evaluation)
Demonstration  was  to have  been carried  out during 1991  at the
Parsons Chemical site,  Grand  Ledge,  Michigan,  USA. The  site is
contaminated with mercury,  arsenic,  heavy metals, chlordane, DDT
and   TCCDs   (2,3,7,8-tetrachloro-dibenzo-p-dioxins)    from   the
formulation  of  pesticides.  The process applied involved  the
excavation  and  staging  of  the    contaminated  materials  for
processing.


3.3 Background

Following a hydrogeological investigation of  the Parsons Chemical/
ETM Enterprises (Parsons/ETM) site  carried  out  for  the present
owners (ETM) in  1980 limited  action  was taken to remove a septic
tank and "leach  field."

The  near  surface   geology  (to about  5m)  can  be  generally
characterized as a glacial till deposit of stratified clay and loam
with minor sand stringers, a thin loamy top soil ranging from 150 -
600mm in thickness covers the site.  Below this  is found either a
brown loam or  clay  with variable  amounts  of granular material
ranging in thickness  from  2.1 to 3.5m. This material contains thin,
saturated  sand stringers.  Overall  moisture  content varies  from
relatively  dry,  firm clay  to  soft  loam  to the  saturated  sand
stringers. The underlying layer is a grey firm till clay which dips
to the south.

The groundwater appears to be perched and occurs at variable depths
across the  site with only  limited connections . between the water.
bearing units.

Prior to ETM occupying the  property,  Parsons Chemical  Works Inc.
engaged in the  mixing, manufacture  and packaging of agricultural
chemicals.  Materials handled included pesticides,  herbicides,
solvents and mercury based  compounds. Concern arose in 1979 and
1980 when  the  Michigan  Department  of Natural Resources  (MDNR)
collected sediments  from  a  creek and ditch adjacent to the site.
Elevated concentrations  of lead,  mercury, arsenic,  and pesticides,
including chlordane and DDT (dichloro-diphenyl-trichloroethane).

Dioxins were  detected on the site  in 1984  following a  US  EPA
screening operation.  In 1988 a MDNR survey revealed the presence of
on  the  site  of widespread  contamination  with  mercury   (up  to
150mg/kg)   and  high  pesticide  concentrations including 4,4'-DDE,
4,4'-DDT,  dieldrin,  and chlordane.

ETM continued to operate on  the site in 1990 but with the two areas
contaminated with pesticides and TCDD fenced.
                               851

-------
3.4 Treatability Study

Geosafe   Corporation,   in  conjunction   with  Battelle  Pacific
Northwest,  conducted  a  treatability  test  on  soils  from  the
Parsons/ETM site. Both contaminated and uncontaminated soils were
shipped for testing.

Preliminary results available November 1989  (CCMS 1991)  revealed
destruction efficiencies  (DE) for dioxins and furans ranging from
76.3 to  99.9%.  DEs for pesticides ranged from  99.98 to 99.998%.
When  calculated with  the  efficiency of the  off-gas  treatment
system, the demonstrated  removal efficiency (DRE) was 99.9763 to
99.9999%  for  dioxins and  furans and 99.99998  to 99.999998% for
pesticides.  The  calculation of DE  was  limited  by  analytical
procedures and detection limits  achievable for  suspect compounds.

A DRE  for mercury was not  calculated.  Mass balance calculations
accounts  for only 5.4% of the original mercury  concentration: the
assumption was that the greater  part was  retained in the melt (it
seems  odd that this was  not determined  analytically  but  it may
simply be that such results were not available  in November 1989).
4. PERFORMANCE

At  the time  of writing  (December 1991)  only data  produced by
Battelle/Geosafe were available.

In  one  test  on  soil  containing  550  mg/kg  PCBs   the  initial
destruction efficiency (i.e. as a result of the heating associated
with the melting)  was 99.9%. However,  the small amount of PCB that
did escape  thermal  destruction was scrubbed  from  the off-gas to
give an  overall destruction and  removal efficiency  of 99.9999%.
Other data can be found in the listed literature.


The vitrified  waste has  been subjected  to a  variety of leaching
tests  including the EPS-tox and  TCLP.  These  tests show (Geosafe
1989)  a uniformly low  leach rate for heavy metals of cibout 5xlO'7
g/cm2/day or  lower,  and  the  material  appears  to   qualify  for
delisting under US regulations.


5. LIMITATIONS

As  described in  the introduction,  there have been  some  major
operational difficulties withi the process  and unless these can be
overcome  the  technology  is  only  likely  to  receive  limited
application.
                               852

-------
There has  been some
migration of vapours
opinion  is  that  any
practical  importance.
settings for example.
be watched for in any
controversy  over  the question  of possible
away from the hot  zone,  but the balance of
 outward migration  would  not be  of great
  It would  be  dealt with  by neighbouring
However, it is clearly something  that should
practical application of the  technology.
6. COSTS

ISV processing is claimed to be "quite energy efficient." Soil is
a very good insulator and thus little heat is lost to soil around
the  heat zoning  (and  that which  does effectively  preheats the
soil).  The greatest heat  loss occurs  through  the  surface by
radiation and release of hot off-gases.

ISV of typical soils requires 880-llOOkwh/tonne  (800-lOOOkwh/short
ton) total energy input,  most of which is accounted for the heat of
fusion  of  the melt. This  is  in fact less than that required to
incinerate  soils in direct energy applied  (but this  does not
account  for energy  losses during  generation and transmission of
electric power).

Treatability testing (by Geosafe) is said to cost US $35,000-40,000
provided there are no exceptional analytical requirements (e.g. for
the  presence  of  dioxins) and to take about 8 to 10 weeks.

Mobilization  costs  are  about  $50,000-60,000  plus  $30-40 per
kilometre  from  Geosafe's Richland,  Washington base location.

Technical  support is estimated to  cost about $25,000-75,000.

Operational costs will be  influenced by a number of  site-specific
factors including:

      -  the amount of site  preparation  required

      -  properties of the soil/waste to be treated (e.g.  dry
        density,  moisture content)

      -  volume of material  to be processed

      -  depth  of processing

      -  unit price of electricity

      -  season of the year

 Depth of processing has  an important impact because it affects the
 ratio between  operational time and  time devoted  to moving the
 equipment  between  settings  etc.  The  process economics  are
                                 853

-------
 consequently  more favourable  for greater  treatment depths.  For
 shallow  contamination,  say  a imetre or  so,  it  is  likely  to  be more
 economic to move the  material  (stage)  to  form deeper  treatment
 beds .

 Costs can vary by $55-77/tonne ($50-70/short ton)  between dry soil
 and fully saturated soil, in some cases, e.g. wet sediments, pre-
 drying may be justified.                                    ' p

 Electricity is assumed  in the  above costings  to be available from
 public supplies for less than  $0. 07-0. 08/kwh.
  u      Proc.essin9  costs  are put at $275-385/tonne ($250-350 per
 short ton) ,  including all  elements of direct and indirect cost such
 as  labour,   materials,   energy,   equipment   amortization,   and
 extractor  overhead and  profit.  The most  significant variables
 affecting the  cost include  the  price of electricity,  amount of
 water to be removed during processing, and the amount of chemical
 analysis required associated with process control and disposal of
 wastes etc.

 It is  not completely clear whether these costs include any element
    site preparation costs  e.g.  if the  site needs to be regraded to
 meet the limitations of the equipment i.e. surface level variations
 not greater than 0.15m and slopes of no more than 5%.


 7.  PROGNOSIS FOR TECHNOLOGY

 In  view of the decision by Gepsafe  in July 1991 to cease to offer
 the t technology   the  prospects  appear   poor  for   short-term
 application.


 8.  REFERENCES

 ANON 1991: WasteTech News,  12 August 1991.

 CCMS 1991: "In-Situ  vitrification," Appendix  6B  to this  report
 (initial report  from EPA presented 1989) .

 EPA 1990:  "Geosafe Corporation  (In-Situ Vitrification)  " in- "The
 Super-fund  Innovative  Technology  Evaluation  Program:   TechAology
 Profiles,  EPA/540/5-90/006  (us|  EPA,  Cincinnati,  1990) pp 52-53?

 Geosafe  1989:  "Application and Evaluation  Considerations  for In
 Situ vitrification Technology: A Treatment Process for destruction
 and/or Permanent Immobilization of Hazardous Materials," (Geosafe
 Corporation, Kirkland WA,  1989) GSC  1901.

Hansen<& Fitzpatrick 1989: Hansen J and Fitzpatrick V, "In-situ
vitrification:  heat  and   immobilization  are  combined  for soil
                                854

-------
remediation," Hazmat World, 1989,  (December).

Hansen: Hansen J,  "Status  of  In situ vitrification Technology: A
Treatment Process for Destruction and/or Permanent Immobilization,"
(Geosafe Corporation, Kirkland WA, no date).
                               855

-------
Graphite and
Glass Frit
Starter Path
Electrodes
to Desired
Depth
              Subsidence
Backfill Over
Completed
Monolith
       Contaminated
       Soil Region
                Natural
                Soil
                                                 Vitrified Monolith
        (D
                (2)
      (3)
           FIGURE 1.   Stages of ISV Processing
                               856

-------
                       Off-Gas Hood
        Power to Electrodes
                   Electrode
                   Location (typ)
   til
Utility or
Diesel-
Generated
Power
   Clean
Emissions
                       Controlled
                       Air Input
                                                          Backup
                                                          Generator,
                                                          Cooler,  ;
                                                           Filter,; and
                                                           Adsorber^
                                                         Glycql ;•'  ^
                                                         Cooling  ;;
                            ISV  Equipment System
                              857

-------
858

-------
                                        IMATO/CCMS Fellow:



                              James M. Gossett, United States



Biodegradation of Dichloromethane  Under Methanogenic Conditions
                           859

-------
BIOTRANSFORMATION  OF  DICHLOROMETHANE
            IN  METHANOGENIC  SYSTEMS
                              by
              James M. Gossett and David L. Freedman

                School of Civil & Environmental Engineering
                          HoJlisterHall
                         Cornell University
                       Ithaca, New York 14853
                Presented at the Third International Meeting,
               NATOICCMS Pilot Study on Demonstration of
       Remedial Action Technologies for Contaminated Land and Groundwater

                        November 6-9,1989

                        Montreal, Canada
                              860

-------
I.
INTRODUCTION
      The research reported herein is part of a larger, USAF-sponsored effort to investigate the
potential for biodegradation of four chlorinated solvents under methanogenic conditions:
tetrachloroethylene (PCE), trichloroethylene (TCE), chloroform (CF), and dichloromethane
(DCM). In this present paper, we focus on our investigations with DCM.

       A.    Objectives  and  Scope  of  Work

             The broad goal of our research effort is to investigate the fundamental factors
influencing the biodegradation of dichloromethane (DCM) by enrichment cultures grown under
methanogenic conditions.  Gaining a deeper understanding of how DCM is degraded under such
conditions will markedly improve the chances of successfully employing bioremediation
technologies.

             More specifically, our research objectives are:

             1.  To determine if DCM can serve as a growth substrate.

             2.  To elucidate the pathways by which DCM is degraded under methanogenic
                 conditions, including the construction of an oxidation/reduction balance.

             3.  To determine which class of organisms — methanogens or nonmethanogens
                 — is responsible for mediating DCM degradation.

             4.   To develop a kinetic model of DCM degradation in mixed cultures.

             5.   To develop continuous-flow biological reactor systems for treatment of DCM-
                  contaminated waters.

              Surprisingly little information is available in the  literature concerning the
 degradation of DCM under methanogenic conditions.   The progress we have made thus far
 includes: development of a DCM-degrading enrichment culture; correlation of DCM degradation
 to methanogenesis;  demonstration of the pivotal role played by acetogenic bacteria in DCM
 degradation; delineation of degradative pathways; determination that DCM can serve as a growth
 substrate; and successful operation of a fixed-film, continuous-flow reactor which degrades DCM
 to CH4 and COi at 20°C.
 II.     SUMMARY  OF PROGRESS  TO DATE

        A.    Experimental Strategy

              The starting point for this research was the development of a mixed culture capable
 of degrading DCM under methanogenic conditions. This was accomplished using the mixed liquor
 from a laboratory digester. The digester had been started with sewage sludge from the Ithaca, NY
                                           861

-------
  Wastewater Treatment Plant, then operated in a semicontinuous mode with a 10 g COD/L synthetic
  substrate, designed to maintain a diverse population of anaerobes. DCM degradation was achieved
  in a culture derived directly from the lab digester.

               The DCM degrading mixed liquor was then used to inoculate a series of enrichment
  cultures. The purpose of the enrichments was to eliminate as many of the organisms as possible
  which weren't involved in DCM degradation, as  well as to remove significant amounts of
  extraneous, undefined organic matter. Developing enrichment cultures has enabled progress in
  correlating DCM degradation to methanogenesis, and in analysis of the pathway(s) by which DCM
  has been degraded (Gossett and Freedman, 1988).

        B.     Materials  and Methods

              Chemicals and Radioisoiopes.   DCM was obtained in neat form (99 mol %
 pure; Fisher Scientific); Chloromethane (CM) was purchased dissolved in methanol (200 ug/rnL
  1 mL ampule; Supelco, Inc.). [14QDCM (Sigma Radiochemical) was diluted in 150 mL distilled'
 deiomzed water and stored in a 160 mL serum bottle,  capped with a Teflon™ lined rubber septum
 The [14C]DCM stock solution contained 2.93 x 107 dpm/mL (4.68 jimoles DCM/mL)-  GC
 analysis of the [14QDCM stock bottle headspace  indicated the presence of an unidentified
 contaminant, which was shown not to be radiolabeled. There  was also no indication that this
 contaminant interfered (e.g., as an inhibitor) with the  DCM degradation studies.  ScintiVerse-E™
 (Fisher Scientific) liquid-scintillation cocktail (LSC) was employed for [14q  assays .

              Cultures  and  Enrichment  Procedures. With  the  exception of some
 continuous-flow, fixed-film reactor studies, all experiments were conducted at 35°C, under
 quiescent conditions, in 160-mL serum bottles to which 100 mL of liquid was added.  The bottles
 were sealed with slotted grey butyl rubbersepta and aluminum crimp caps (Wheaton Scientific).
 Virtually no loss of DCM was observed from water controls (WC) which used these septa- they
 were less permeable to oxygen than Teflpn™-lined  rubber septa, were easier to puncture, and
 maintained better flexibility following autoclaving.  [They were, not used in PCE/TCE studies
 because significant losses of these compounds (and their reductive dechlorination products) were
 noted in water controls.]

             Autoclaved seed controls (ASC) were  used to evaluate the degree of sorption and
 abiotic  transformations of DCM.  When these phenomena were  consistently shown to  be
 negligible, use of the ASCs was discontinued.

             Semicontinuous  operation Of  the enrichment cultures was often practiced with
 bottles which were actively degrading DCM.  This entailed removal of a volume (usually 4.0 mL)
 of well-mixed liquid and its subsequent replacement by some combination of basal medium and
 DCM-saturated water to yield the desired DCM concentration. When semicontinuous operation;
 was not practiced, the disappearance of DCM was followed by addition of only DCM-saturated
 water.                                 }

             Analytical Methods.   Analysis of  volatile organics was performed by gas
chromatographic (GC) analysis of a 0.5-mL headspace sample, using a flame-ionization detector
                                           862

-------
(FID) in conjunction with a 3.2-mm x 2.44-m stainless-steel column packed with 1% SP-1000 on
60/80 Carbopack-B, as previously described (Gossett, 1987; Freedman and Gossett, 1989).

             Degradative pathways  were examined through semi-continuous addition of
radiolabeled [14C]DCM to sixth-generation enrichment cultures.  Following various periods of
operation, distribution of 14C among suspected DCM degradation products was determined as
previously described (Gossett and Freedman, 1988; Freedman and Gossett, 1989).

             HPLC analysis was employed to  determine  the concentrations of soluble,
nonvolatile fermentation products, similar to the method described by Zinder and Koch-(1984) A
Hewlett Packard 1090 HPLC was used to pump 250-^L samples through a 300-mm HP X-87H
ion exchange column (Bio-Rad Laboratories) and into an LC-25 refractive index detector (Perkin-
Elmer)  The mobile phase (13mM H2SO4) was delivered at 0.7 mL/min. By operating the ion
exchange column at two temperatures (30 and 65°C)  it was possible  to resolve formate,
formaldehyde, acetate, propionate, metlianol, isobutyrate, butyrate, and ethanol.  For example,
although methanol and propionate coeluted at 30°C, they were well resolved at 65°C.

       C.     Results and Discussion

              Pathways of DCM  Metabolism.   Figure  1 summarizes  the  pathways
 involved in biotransformation of DCM by a methanogenic mixed culture. The radiotracer studies
 which led us to this model were presented earlier (Gossett and Freedman, 1988) and will not be
 detailed here. In essence, the model is the result of studies with [WQDCM in which various levels
 of bromoethanesulfonate (BES), a selective inhibitor of methanogenesis, were employed to inhibit
 either acetoclastic methanogens (at low levels of BES) or all methanogens (at high levels of BES).
 Subsequent monitoring of [*4C] species and H2 allowed deductive development of the model.
Tr
'COi \ 2
>a
o
J (J
"1
*CH3CC
Figure 1. Pn
CL
H20
\-^HCl
/" V.
CO- W« CO-. Reducing Methanocens^—
H2O
7
)OH [•"••••i Acetoclastic Methanoeens •"
aposed Model for Biodegradation of DCM by
ilture.
K
N.
— .> CH,
V
mm* y CH4 + CO2
a Methanogenic Mixed
                                           863

-------
              The accumulated evidence suggests that acetogenic bacteria mediate the degradation
 of DCM, not methanogens. This evidence includes the degradation of DCM in the complete
 absence of any methane formation (due to BES inhibition), and the formation of both acetic acid
 and hydrogen as products of DCM degradation.

              Additional indirect evidence which implicates non-methanogens was obtained by
 examining DCM degradation in the presence of a potent antibiotic. Four bottles were set-up
 identically, continuously shaken, and incubated at 35°C. Over the first 29 days of operation, the
 ability of each bottle to repetitively degrade increasing levels of initial DCM doses was established.
 On day 29, two of the bottles received 10 mg of vancomycin (an inhibitor of cell wall synthesis in
 eubacteria), while the other two bottles continued to receive only DCM. As shown in Figure 2, the
 two bottles which did not receive vancomycin were able to repetitively degrade increasing levels of
 DCM (no effort was made to increase the initial DCM dose above approximately 120 jimoles/100
 mL, though it certainly would have been possible).  In the other bottles, DCM degradation was
 sustained for at least one spike after adding the vancomycin, but then DCM degradation began to
 fall off considerably. Assuming that the vancomycin dose used affected eubacteria and not
 methanogens (demonstrated by other investigators), these results add to the evidence indicating the
 central role played by non-methanogens in DCM degradation. Though we cannot be absolutely
 sure, we believe acetogens  are the principal DCM oxidizers, as well as responsible for acetate
production from DCM.
BES additions.
             Figure 3 shows the pathways we have observed to be affected by vancomycin and
                                           864

-------
O
o


1/5
_05

O

E
O
Q
           Bottle #1:
           No Vancomydn Added
o
o

1/5
JD
 O
 E
DC
120


100


 80


 60


 40


 20-j
           Bottle #2:
           Vancomydn Added (10 mg, t=29)
               10
                  20      30

                 Time (days)
 Figure 2.
       Effect of Vancomydn on DCM Degradation.
                        365

-------

^H2^i2 1 A-
J, ' Vancomycin
>f\
0 ' ^-.^ 1—
j| C°2 l=-
/ o
Hcl>/<
1
CH3COOH I.
Figure 3. Pathways in the
Vancomycin.
H20
	 DCM Oxidizers •"•"•B™
YHCI
"C
CO-> Reducing Methanocens •"
H20
>^.
Acetoclastic Methanoaens ••
4
Proposed Model Which are


I/'" 2r
^9^> m.
' \ Y 4
- ' K. ,1
•^— > CH4 + CO,
VK 2
Affected by BES and/or
              Kinetics  of DCM Degradation.   The rate of DCM  degradation at 35*C
 was measured using batch-fed DCM enrichments to which an initial DCM concentration of 2 mM
 was added (170 mg/L). At such a high concentration, methanogenesis was inhibited and hydrogen
 accumulated (Figure 4);  when DCM concentration was sufficiently reduced, methane production
 resumed.  DCM degradation followed an apparently zero-order kinetic model (constant rate = 7.6
 Hmol DCM/hr). The quantity of methane which resulted (about 8 ^mol) was far lower than the
 100 nmol  expected from degradation of 200 ^mol of DCM.  This suggests that most of the
 methane formed resulted from CC-2-reducing methanogens, not from acetoclastic methanogens.
 The latter organisms were likely more severely inhibited by the high initial levels of DCM
 employed.

             DCM as  Growth Substrate.   Significant progress  has  been  made in
 determining whether or not DCM  can  serve  as a growth substrate,  using the  following
 experimental design. Four serum bottles  were set-up.  Each received 90 mL of basal medium
 (with 50 mg/L yeast extract as the only non-DCM carbon source) plus  10 mL of inoculum from
 cultures which were actively degrading DCM, and incubated in an orbital shaker bath maintained at
 35*C. Two of the bottles ("with-DCM") received repetitive additions of DCM. Initially, these
 additions were approximately 10 ^moles/100 mL; over time they were increased to as high as 2.4
mM (204  mg/L).  Two other bottles  ("without-DCM") — serving as controls — were set-up
identically but received no DCM.           !
                                          866

-------
 o
 o
  o

  E
 O
 D
                           DCM = 210-7.65t



                           RA2 = 100.0%
                                            40
                                   50
o £

•oo
  _
          0
10
20       30

Time (hrs)
  Figure 4.   Kinetic Experiment, 35°C.
                            867

-------
              Suspended organic carbon (SOC) was used as the measure of growth.  At approx-
 imately 14-day intervals, 5 mL was removed from each of the bottles, and replaced with 5 mL of
 basal medium containing yeast extract (50 mg/L).  2.3 mL of the sample was used to measure total
 organic carbon (TOC); 2.7 mL was filtered (0.45 (im) and used to measure dissolved organic car-
 bon (DOC). The difference between TOC and DOC was SOC. Net growth on DCM was calcu-
 lated by subtracting SOC in the "without-DCM" bottles from SOC in the "with-DCM" bottles.

              Results after 133 days of operation are shown in Figure 5.  SOC in the  "with-
 DCM" bottles began to rise definitively above SOC in the "without-DCM" bottles on day 49, just
 as cumulative DCM consumption began to rise significantly.  Concern still remained, however,
 that this growth may have been attributable — at least in part — to methanogens (through the use
 of DCM degradation products) and not to the organisms responsible for DCM degradation.
 Nevertheless, even though BES was not added to these bottles, methane output was well below the
 stoichiometric level expected (Figure 6), due to inhibition by high levels of DCM.  This inhibition
 was evident initially, when methane output in the "with-DCM" bottles was slightly below  that of
 the "without-DCM" bottles. Methane'outputin the "with-DCM" bottles did eventually surpass that
 in the "without-DCM" bottles, but it slowed and nearly stopped when DCM additions exceeded
 approximately 1.8 mM (around day 63).  By day 77, average cumulative DCM degradation was
 2565 jimoles; had methanogenesis not been inhibited, cumulative methane output would have been
 1401 (imoles (1283 as a consequence of DCM degradation plus 118  from the yeast extract).
 Instead, only 357 jimoles were produced, indicating severe inhibition of methanogenesis.

             Another, initially unexpected form of inhibition was encountered  in  these
 experiments — that due to low pH resulting from HC1 production in biological dechlorination of
 DCM! Around day 75, DCM consumption stalled. Measurements on day 91 indicated that pH had
 dropped to 5.1 and 5.4 in the two bottles receiving DCM, whereas the two without DCM had pHs
 of 7.4 and 7.6.  Addition of bicarbonate buffer, along  with a 5% reinoculation from an  active
 DCM-degrading enrichment, revived the culture, allowing renewed DCM consumption, followed
 by renewed biomass formation (as defined by SOC).  Note, however, that after this period of
 upset, biomass formation lagged considerably behind DCM consumption.

             Day-91 samples were also analyzed for  HAc and MeOH concentration, using a
 HPLC with a refractive index detector.  HAc concentrations were 6.77 mM and 6.94 mM in the
 two "with-DCM" bottles.  MeOH was not detected in either of the samples. The bottles receiving
 no DCM had no detectable levels of HAc or MeOH.

             The correlation between SOC formed and DCM degraded is quite good (Figure 7).
 SOC formation was calculated by subtracting average SOC in "without-DCM" bottles from the
 SOC in "with-DCM" bottles. These data indicate that approximately 8.5% of the DCM carbon
 degraded was converted to cell carbon, a yield within the range expected for anaerobic micro-
 organisms. The accumulation of SOC in the "with-DCM" bottles was likely due to organisms
 other than methanogens — i.e., from degradation of DCM, rather than from acetate or H2. To
 explain the observed synthesis as resulting solely from methanogenic activity on H2 or acetate
requires that unrealistically high yield coefficients be assumed. Thus, we conclude that DCM does
indeed serve as a growth substrate.
                                          868

-------
     400i
  ^^ ~^
_--J
o £ soo -
J=~
1°
o?
U-- 200
o £
S"5
w£ 100
   =t
       0
                              No DCM Added
             —r~
             40
—r-
60
                               i—
                              80
                                   100   120   140
 DO
 C3^
38
ol
6.0'

5.0'

4.0

3.0

2.0

1.0
         0
                                      Bottle #1
                                      Bottle #2
        20   40    60    80   100
                Time (days)
                                         120  140
  Figure  5.  Growth on DCM.
                          869

-------
       o
       o
       T-
       To
       3
       o
       E
       E
       o>
       c
       CO
        -50 mV).  Effluent from the
pre-column enters the "DCM column" at its bottom. DCM-saturated water is injected into the flow
 (using a syringe pump) prior to entering the DCM column.
                                          870

-------
                    0.3
                               SOC = - 0.000046 + 0.085DCM

                               RA2 = 97.2%
                                                       Bottle #1
                                                       Bottle #2
                       0.0
                        DCM  Consumed  (mmoles/100  mL)

      Figure  7.   Biomass  Formation from  DCM  Degradation —  Day, 26
                   through Day 77.
             Sample ports are spaced at 15.24-cm (6-inch) intervals along the length of the DCM
column. Sample ports are constructed of a hollow stainless-steel cylinder, machined to extend
from the exterior of the sample port to near the column center. The steel is held firmly in place by a
Teflon™ O-ring and bushing assembly.  The exterior of the steel is covered with a Teflon™-lined
rubber septum.  The septum is held in place by an aluminum crimp cap.

             The DCM column was inoculated with 2 liters of an  active, DCM-degrading
enrichment culture. The DCM column was first operated in recycle mode for 22 days (at 35°C) to
facilitate microbial attachment. During this period, 0.1 mM DCM (8.5 mg/L) was added to the
column, allowed to degrade ,(in 2-3 days), then added again. Following this recirculation period,
the pre-column was connected in series and the system was operated in continuous-flow mode.
The DCM column was initially operated at a 2-day hydraulic retention time (HRT) based on void
volume, with a nominal DCM influent concentration of 0.1 mM.  Eight days after commencing
continuous operation, only traces of DCM (<1.2 |iM) were detected at the first sample port within
the column, and none at any higher location. Significant CH4 levels were noted at higher points
within the column, indicating bio transformation.

             Samples are taken from the column routinely as follows:  a 5-mL gas-tight syringe
(Supelco,  Inc.) fitted  with  a custom-made, 4-inch side-port  needle  (Dynatech Precision
Instruments) is purged with 30%CO2/70%N2 and inserted in the effluent (top) sample port. A 4-
mL liquid sample is drawn from the column and injected into a 14-mL vial sealed with a Tefion™-
lined septum and aluminum crimp cap. The volume of sample delivered to the vial is determined
                                         871

-------
gravimetrically.  Samples are taken in this manner from each sample port from top to bottom.
Following equilibration at 35°C, a 0.5-mL headspace sample is analyzed by GC.
                               gas
                               collection
  PRE-COLUMN
                       vent
                peristaltic
                pump

        -S^ refrigerated
            basal salts
            medium +   .
            acetate
                                                             effluent
                                                             collection
                                                DCM COLUMN
                                                   sampling
                                                   ports
syringe pump:
DCM-saturated water
          Figure 8.   Schematic of fixed-film columns.
            After one month of operation, we began lowering the column temperature —
ultimately to 20*C, where it is now operating. Noting successful degradation at 20*C, we stepped
up the influent concentration of DCM to 0.4 mM (34 mg/L).  Figure 9 shows the column profile
for such conditions.  At such a high influent DCM concentration, inhibition of methanogenesis is
expected; die displacement of methane production to higher points within the column suggests that
                                       872

-------
inhibition is-indeed occurring. Due to the relative lack of sensitivity associated with our acetate
method, we had not attempted measurements of acetate within the column. However, as influent
DCM is increased, we expect to be able to detect acetate and H2 at intermediate points of the
column (positions "12" through "24").
       Q)
       C
       to
      O
      Q
•500

 400

 300

 200

 100-
                                                                Methane
                   Influent   6"
                          12"      18"     24"
                          Sampling Port
30"   Effluent
 Figure  9.
    Profile of Fixed-Film, Continuous-Flow Column Receiving DCM at 400
    u.M Concentration. Conditions: T = 20°C;    hydraulic retention time =
    2 days;  0.5 mg/L yeast extract in feed.
 III.    FUTURE  PLANS

       In order to advance the prospect of using methanogenic systems for treatment of DCM
 contaminated water, further research is being carried out in  both  suspended-growth  and
 immobilized-cell reactors.  The suspended-growth studies will allow examination of .more
 fundamental issues:

       «f»     Isolation of the microorganisms responsible for DCM degradation;

       «f»     Characterization of the isolates (including the ability to grow on a variety of non-
 >  •-:•;,         chlorinated substrates, a property of significant practical importance in systems
              where the level of DCM may be too low to alone support growth);
                                          873

-------
       •I*    Delineation of degradation pathways (i.e.,  a more fundamental, approach that
             previously employed; cell-free extracts from isolates will be purified with reversed-
             phase HPLC);

       •t*    Determination of the ability of enrichments to degrade chlorinated compounds other
             thanDCM.

       The fixed-film reactor studies are oriented more towards the practical issues of how best to
accomplish treatment Foremost in this regard is the capability of the DCM degrading organisms to
attach to fixed-film media. Assuming this is possible, the reactors will be used to evaluate a variety
of treatment conditions, including DCM concentration, variation in hydraulic residence time, and
the effect of other chlorinated aliphatics —particularly CF— on the efficiency of DCM removal.

This research was supported by the U.S. Air Force Engineering and Services Center (AFESC),
Tyndall AFB, FL, under contract no. F08635-86-C-0161.
REFERENCES

Freedman, D. L.;  and Gossett, J. M.   1989.  Biological Reductive Dechlorination  of
Tetrachloroethylene and Trichloroethylene to Ethylene under Methanogeic Conditions.  Applied
and Environmental Microbiology 55. 2144^2151.

Gossett, J. M.  1987.  Measurement of Henry's Law Constants for Ci and C2 Chlorinated
Hydrocarbons. Environmental Science &. Technology 21(2): 202-208.

Gossett, J. M.; and Freedman, D.  L.   1988.   Biodegradation of Dichloromethane Under
Methanogenic Conditions. Presented at the Second International Meeting, NATO/CCMS Pilot
Study  on  Demonstration of Remedial Action  Technologies for Contaminated Land and
Groundwater, November 7-11, 1988, at the National Institute of Public Health and Environmental
Protection, Bilthoven, the Netherlands.

Zinder, S. H. and Koch, M.  1984.  Non-Aceticlastic Methanogenesis from Acetate:  Acetate
Oxidation  by a Thermophilic Syntrophic Coculture. Archives of Microbiology 138: 263-272.
                                         874

-------
                            NATO/CCMS Fellow:

                 Merten Hinsenveld, United States

Alternative Physico - Chemical and Thermal Cleaning
                Technologies for Contaminated Soil

    Recent Developments in Extraction and Flotation
  Techniques for Contaminated Soils and Sediments
              875

-------
  ALTERNATIVE  PHYSICO-CHEMICAL  AND  THERMAL  CLEANING
  TECHNOLOGIES FOR CONTAMINATED SOIL

  M.  Hinsenveld1, E.R.  Soczo2, GJ.  van de Leur2, C.W.  Versluijs2 and E.
  Groenedijk^

  1.  Netherlands Organization for Applied Scientific Research
  2.  National Institute of Public Health and Environmental Protection
  3.  Delft University of Technology
  1.  ABSTRACT

  This study has been carried out within the framework of the Netherlands Integrated
  Soil Research Programme. Aim of the study is to make both a systematic evaluation
  as well as a selection of technologies currently developed world-wide which offer
  possibilities for the decontamination of soils. The study has been carried out in three
  phases. This paper describes the first Chaser a survey of alternative techniques and a
  first selection of the most promising ones. In the second phase, a more detailed
  analysis and a further selection of technologies will be made, followed by a research
 programme for the selected techniques in the third phase.


 2.  INTRODUCTION

 In the Netherlands some 15 physico-chemical and thermal cleaning installations are
 readily available. With the present technology a large part of the contaminated soils,
 however, is not or problematically cleanable. This is particularly true for soils that
 are contaminated with halogenated (aromatic) hydrocarbons and/or heavy metals, and
 soils containing a large fraction of fines (< 0.050 mm). Problems arising are,  for
 example:
 - emission of hazardous compounds (e.g. with thermal  treatment);
 - unacceptably high residual concentrations (as is often the case with extraction
   techniques); or
 - production of a large amount of contaminated sludge (e.g. with cleaning of clay in
  extraction installations).
In addition to this, some of tfce techniques used at present, involve qutte high  cleaning
costs. It is, therefore, essential to develop  alternative techniques that either do not
have the above mentioned disadvantages or lead to lower cleaning costs.
                                  876

-------
  3.  CONTAMINATED SOIL IN THE NETHERLANDS; SIZE OF THE PROBLEM
 At present the soil remediation activities are carried out following two lines:
 -  Remediation within the framework of the Interim Law on Soil Remediation (IBS),
    the so-called IBS-sites (IBS = Interim wet Bodem Sanering); and
 -  Remediation carried out by private parties; the so-called non-IBS-sites.
 In 1987, the total number of potential IBS-sites was approximately 7,500. At the time
 it was estimated that approximately 1,600 of them urgently needed remediation and
 should be remediated within the framework of the IBS. A new estimate of the total
 number of contaminated sites (including the non-IBS-sites) was made in 1988 [1], A
 summary of the results of this estimate is given in Table 1.
 In the governmental "Ten Year Scenario  Soil Remediation" [1], it is planned that
 from the  total number of sites (about 110,000)  6,000  very  urgent cases will be
 remediated within the coming ten years, leading to expenses in  the  order of 5
 milliard Dutch guilders (abt. 2.1Q9 ECU).

 TABLE 1: Number of sites needing remediation for five industrial sources.
 Source
Number of sites
 Number of sites
 needing
 remediation
Percentage
needing
remediation
 Gaswork sites
 Dump sites
 Carwreck sites
 Former industrial sites
 Present industrial sites

 Total
        234
      3,290
      2,100
abt. 400,000
abt. 120,000

abt. 530,000
       234
       150
      1,200
 abt. 80,000
 abt. 25,000

abt. 110,000
   100%
    5%
   60%
   20%
   20%

   20%
4.  SHORTCOMINGS   OF   CLEANING   TECHNOLOGIES  IN.  THE
    NETHERLANDS

4.1.  Features and shortcomings per technique
Thermal techniques
In the Netherlands  rotary kiln  ovens are  used. The soil can be directly heated,
indirectly heated or  through combinations thereof. At the moment these techniques'
are only applicable for organic contaminants (including cyanides). Basically, thermal
techniques can also  be used for mercury contaminants. But this is not practised at
present because of emission control problems. The energy need of these  thermal
techniques is rather high and emissions of hazardous contaminants are possible. In the
                                   877

-------
Netherlands it is not  allowed  to  treat  soil  contaminated  with chlorinated
hydrocarbons in  thermal soil cleaning  installations for reasons of  hazardous
emissions.
Extraction and jractionation techniques
Most soluble components are easily removable by flushing the soil with an extractant.
Unfortunately, contaminants are often preferentially sorbed to the fines in the soil.
The extraction techniques available at present make use of this phenomenon by
separating the heavily contaminated fines from the bulk of the soil. But a consequence
hereof is that soils containing many fines cause an excessive sludge production. If, in
some cases, the bulk of the contaminants is  present in the  coarser fraction, this
fraction will be removed from the  soil. At this moment, extraction techniques along
with flotation techniques  are the only techniques that can deal with heavy metals.
There is very limited experience with in-situ extraction techniques.
Biological techniques (not subject of thjs study)
For these  techniques only large-sclale  experience with alifatic and aromatic
hydrocarbons is available. Presently, use of these techniques leads to long treatment
periods (e.g. landfarming). Reduction jof concentrations to acceptable levels (i.e. A-
level) is very difficult. Chlorinated  hydrocarbons are  hardly biodegradable  and
heavy metals can  hardly be removed. Apart from the already mentioned problems,
clogging and channeling of the aquifer may occur when using in-situ techniques,
leading to insufficient cleaning results.

4.2. Features and  shortcomings per contaminant
Heavy metals
Can only be removed by extraction and flotation. There  is no operational technique
that can remove heavy metals from clay or sludges.
Cyanides
Can be removed from all types of soil by thermal treatment. Clay can  pose some
problems in the treatment of the off-gases. Sludges must have a dry-matter content of
at least 50% to avoid handling problems. Cyanides can be removed by extraction as
well as flotation unless the soil contains a large fraction of fines or organics (clay and
peat). Biological techniques may be applicable.
Non-chlorinated alifatic components and simple aromatics
These compounds  are easy to treat  thermally. Extraction and flotation can be applied
for these contaminants, unless they are; present in clay or peat.  These compounds  are
easily biodegradable and are not considered a major problem.
Poly cyclic aromatic compounds
Can be easily treated thermally, resulting in low rest  concentrations. Extraction and
flotation are possible with varying results. Low poiycyclic compounds (less than four
rings) are good biodegradable. Biodegradability of higher poiycyclic compounds is
very low.
                                    1878

-------
 Non-volatile chlorinated hydrocarbons
 In principle thermal treatment is suitable when dealing with these compounds.
 However, it is not allowed to use the presently available thermal soil  cleaning
 techniques for these compounds in the Netherlands. The reason is that this treatment
 may cause hazardous emissions (e.g.  dioxines). Extraction and the presently used
 flotation technique are applicable to these compounds only when they are present in
 sand or sandy soils (not peat or clay).
 Volatile (chlorinated) hydrocarbons and pesticides
 For the above-mentioned reasons, these compounds cannot be treated thermally in the
 Netherlands. Extraction and flotation are applicable  in principle, but are quite
 expensive for these easily removable  contaminants. A better way of treating these
 compounds is by stripping them from  the soil. This technique, however,  is not very
 well developed in the Netherlands. Because of their volatility, the handling of these
 contaminants needs additional  safety precautions. Usually, very low rest concentra-
 tions are required.
 5.  OVERVIEW AND SELECTION OF ALTERNATIVE TECHNIQUES

 An inventory of alternative techniques that offer possibilities for treating soil or
 sludge is listed in Table 2. These techniques can be either emerging techniques for
 soil, or existing techniques for other material (e.g. mining techniques).

 TABLE  2: Overview of alternative techniques.
Techniques
Developer
Features
                             EX-SITU TECHNIQUES
Wet thermal techniques
Supercritical oxidation 1
Supercritical oxidation 2
Wet oxidation 1
Wet oxidation 2
Wet oxidation 3
Dry thermal techniques
Fluid bed oven 1
Fluid bed oven 2
Fluid bed oven 3
Electrical infra-red oven 1
Electrical infra-red oven 2
Plasma reactor
Modar
Oxidyne
Zimpro; Kenneth
Vertech
RISO

Waste Tech
Thyssen
Ogden
Shirco
Thagard
SKF
Thermal immobilization techniques
Ceramic application 1          University Utrecht
mixed and plugflow reactor in series
vertical pipe reactor 3000 m
bubble column
vertical pipe reactor 1600 m
horizontal pipe reactor

stationairy bed
not yet realized stationairy bed
circulating bed
tunnel oven
high temperature fluid wall
free plasma, abt. 2000 °C

sediments for bricks
                                     879

-------
  Table 2. Continued
  Techniques
 Developer
  Features
  Ceramic application 2
  Vitrification 1
  Vitrification 2
  Vitrification 3
  Vitrification 4
 FBI
 Vitrifix
 Westinghouse
 Retech
 Nuclear Research Centre
 Physico-chemical immobilization techniques
 Immobilization based on:
 - cement                        Many developers
 - chalk and/or puzzolanes          Many developers
 - thermoplasts                    Many developers
 - organic polymers
 - waterglass
 Dechlorination techniques
 Hydrothermal decomposition
 Ultraviolet dechlorination
 Radiolytic dechlorination
 Chemical dechlorination
 Sodium detoxification
 Particle separation techniques
 Heavy media separation
 Heavy media cyclonation
 Jig technique
 Wet concentrating tables
 Humphrey spiral separation
 Reichert cone separation
 Pinched sluice process
 Revolving round table
 Tilting frame separation
 Vanner separation
 Bardes-Mozley separation
 Froth flotation
 High gradient magnetic separation
 Extraction techniques
 Extraction with
 - complexing agents
 - crown ethers
-acids
- organic solvents
 Many developers
 Many developers

 Delft University of Techn.
 Atlantic Research Corp.
 Atomic Energy of Canada
 EPA
 Degussa

 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 No developer for soil
 Mosmans
 No developer for soil
PBI
No developer
No developer
SmetJetjSanitex
 cement, fly ash
 asbest treatment
 electric pyrolyser
 centrifugal reactor
 2500 °C, RAD-waste
 bitumen, parrafines,
 polyethylene
 polyesters, epoxides
 hydrogen, 300 C, 18 MPa
 LARC system
 propanol, gammaradiation
 APEG
 pure sodium

 based on density of particles
 based on density of particles
 pulsating waterbed
 shaking tables
 vertical spiral
 perforated plate
 conus-formed stream pipe
 based on friction, inertia
 based on sedimentation velocity
 conveyor belt
 rotating lilting frame
 sorption on air bubbles       ,
 paramagnetic particles
heap leaching
Methylene chloride
                                      880

-------
  Table 2. Continued
  Techniques	
 Developer
 Features
  - aliphatic amines
  - fluidized gases
  Supercritical extraction
 Resources Conservation Co.  ieverse solubility
 CF Systems Corporation    propane, butane
 Critical Fluid Systems       CQz extraction
 Other techniques
 Steam stripping                  Heymans
 Air stripping                    Many developers .
 Chemical and photochem. oxidation Ultrox
                              IN-SITU TECHNIQUES
 Stripping and extraction techniques
 Air stripping
 Vacuum extraction
 Compressed air injection
 Steam stripping
 Extraction with acids
 Immobilization techniques
 DCR-technology
 Silicagel injection
 Vitrification
 Other techniques
 Electroreclatnation
 Adsorbtion by DCR or CAP
 Hydrolysis
 Chemical dechlorination
 High frequency heating
Many developers
Many developers
Many developers
Heymans
Mourik Groot Ammers

Many developers
Many developers
Batelle

Geokinetics
No developer
No developer

HT Research Institute
                          stripping in asphalt mixer

                          H2O2, ozone and UV-light
extraction of cadmium

chalk

electrodes in-situ

electrodes, circulation syst.
adsorption by chalk or foam
raising pH to abt 11
APEG
electrodes in-situ
 6.  CRITERIA FOR A FIRST SELECTION OF ALTERNATIVE TECHNIQUES

 Information on some alternative techniques is limited. Therefore, the possibilities of
 alternative techniques for cleaning soils can only be roughly estimated. It should be
 clearly understood: for new techniques  of which there is only  little or just
 commercial information available, the estimate of experienced scientists in the field is
 indispensable. In order  to be able to make a reliable selection from the large amount
 of techniques, the knowledge of specialists should be used as coherent and objectively
 as possible. An important aid in  this can be the use of a ranking system. Such a
ranking system consists of: a set of criteria, a quantification of these  criteria and a
quantification of the relative importance of the criteria.  Unfortunately, this paper
does not allow a lengthy description of the ranking  system used.  Only the criteria
                                     881

-------
 used and some main features of the ranking system are indicated below. In the study,
 a set of four criteria were chosen for a first selection of techniques:
 A. Applicability of the technique for different matrices and contaminants
 Techniques that are applicable to ai broad range of matrices and contaminants, or
 combinations thereof, score high on applicability. The ranking system expresses the
 relative importance of the combination of organics and heavy metals in more than
 one soil type (sand, loamy sand, peaty sand and clay).
 B.  Priority concerning matrix or contaminant
 In  the ranking system, the relative priority  of removing heavy metals  from most
 matrices and polyaromatic hydrocarbons and non-volatile chlorinated, hydrocarbons
 from clay and sludges is incorporated.
 C.  Status of development
 The study was conducted in order to find alternative techniques that could be
 developed to a practical stage within the near future. For this reason (and some other
 reasons  of  minor importance  not mentioned here) a higher development stage
 influences the score in a positive way.
 D.  Market perspective and costs
 The scoring of market  perspectives is merely  a valuation  in extremes. Most
 techniques score neutral  on this criterium. Only if the techniques are excessively
 expensive, complicated, etc. or excessively cheap,  simple, etc. they score lower or
 higher respectively.
 The present  ranking system is a quick and rough aid in method for selecting the most
 promising techniques. In some cases the indication given by the ranking system was
 overruled by additional  reasons  or information not incorporated  in the ranking
 system.

 7.  SELECTION OF THE MOST PROMISING TECHNIQUES

 A total of 63 alternative techniques are listed. By using the scoring system we were
 able to eliminate 31 techniques from this list. Furthermore, 11 techniques could be
 selected for further study on the basis of the scores. An intermediate group for which
 the scoring system was not decisive, contained 21 techniques. The size of this group
 is an indication for the  difficulty Of selecting alternative techniques  for  which
 information is either very limited or of low credibility. Below a short account of the
 selection of techniques is given.
 Wet thermal techniques (ex-situ)
 In this category, two alternative techniques were found promising; supercritical
 extraction of Modar and wet oxidation of Zimpro. Both techniques are comparable as
to their applicability and problems. The major advantage of supercritical oxidation,
compared to wet oxidation, is the possibility of treating chlorinated  hydrocarbons.
Supercritical oxidation, therefore, is selected for further study.
                                   882

-------
Dry thermal techniques (ex-situ)
In this category the three fluid bed ovens, one of the electric infra-red ovens and the
plasmareactor are promising. The electric infra-red oven is being developed in the
USA. It is recommended to wait for the results of this development. Fluid bed ovens
are suitable for a large range of matrices and contaminants. Furthermore, they can
be used for air stripping. In general, they are somewhat cheaper than rotating ovens,
more flexible as to the material treated and the process conditions can be better
controlled. Therefore, fluid bed  ovens are selected for further research. Plasma
techniques can be used for a variety of matrices using oxidative,  reductive and
pyrolysing process conditions. In  the Netherlands this technique is rather new. It is
therefore recommended to study the technique in more detail.
Thermal immobilization techniques (ex-situ)
These techniques have been incorporated in the study for completeness' sake. Within
the framework of this project we will not take these  techniques into account.
Physico-chemical immobilization (ex-situ)
See remarks under thermal immobilization techniques.
Dechlorination techniques (ex-situ)
Hydrothermal dechlorination does not lead to hazardous emissions and has, therefore,
been  selected for further study. The other techniques  do not lead to sufficient
dechlorination results or lead to hazardous emissions. Most of the listed dechlorina-
tion techniques have a limited field of application. The need for dechlorination taken
into account, however, it was decided that some attention should be given to the other
dechlorination techniques listed as well.
Particle separation techniques (ex situ)
Particle separation techniques in general look very promising. They are cheap and
have proven their applicability in mining industry. Six of them will be studied  in
more detail  (jig,  shaking table, spiral,  tilting-frame,  vanner, Bartles-Mozley).
Techniques that  are expected to be  applicable only to a few soils like heavy media
separation,  heavy media cyclonation,  Reichert cone separation, pinched sluice
process, revolving round table and high gradient separation, will not be considered
for further study.
Extraction techniques (ex-situ)
In this category only extraction with  complexing agents looks promising. This
technique will be further investigated. The other techniques listed either have a lower
status of development or do not solve high priority problems.
Other ex-situ techniques
In this category we find only chemical and fotochemical oxidation as promising.
These, however,  are rather water purification techniques than techniques for soil and
sludge cleaning.  In the framework of this srudy they will not be investigated further.
It is recommended to incorporate these techniques In a research programme for
water purification. Ex-situ stripping  techniques are  fairly expensive for high volatile
                                    883

-------
  and, therefore, in principle easily removable contaminants.  These techniques will not
  be investigated further.
  Stripping and extraction techniques (in-situ)
  In this category compressed air injection can be considered for further study. This
  technique, as well as the other stripping techniques, is frequently used in Germany
  and information on the technique Becomes readily available. There is no need for
  stimulating this development. Within the framework of this study we will, therefore,
  not investigate this technique further. In-situ extraction is expected to be applicable
  only to very sandy soils and highly volatile contaminants; situations that very seldom
  occur.                                   -
 Immobilization techniques (in-situ)
 See remarks for thermal immobilization techniques.
 Other techniques (in-situ)
 In this  category electroreclamation looks very promising. This technique has the
 possibility of cleaning clay contaminated with heavy metals. This technique will be
 further  investigated. The other techniques listed have a rather limited field of
 application and do not solve high priority problems.

 In phase 2 of the project the selected techniques will be studied in more detail. The
 findings will be described in 8 monographs entitled:
 1.  Supercritical oxidation
 2.  Fluid bed incineration
 3.  Plasma reactors
 4.  Dechlorination techniques
 5.  Particle separation techniques
 6. Froth flotation
 7. Extraction with complexing agents
 8. Electroreclamation
 Based on these monographs, a selection of techniques for further development in the
 Netherlands will be made.
8.  REFERENCES

1.  Ten year scenario soil remediation, Stuurgroep Hen Jaren-scenario
    Bodemsanering, Ministerie van VROM, 1989 (in Dutch)
2.  Handbook on Soil Remediation, Staatsuitgeverij, 1988 (in Dutch)
                                   884

-------
  RECENT DEVELOPMENTS IN EXTRACTION AND

             FLOTATION TECHNIQUES FOR

                  CONTAMINATED SOILS

                              AND

                         SEDIMENTS


                        IT. M.Hinsenveld
                         October,  1991
                         University of Cincinnati
               Department of Civil and Environmental Engineering
              741 Baldwin Hall (ML 71), Cincinnati, Ohio 45221-0071
                               USA
Presented at: Fifth International NATO/CCMS Conference on Demonstration of Remedial Action
Technologies for Contaminated Land and Groundwater, 18-22 November, Washington DC, USA.

Phone: + 1 (513) 556-7976 (office) or + 1 (513) 556-3646 (secretary) or + 1 (513) 556-3646
(department), Fax: + 1 (513) 556-2599.
                                885

-------
 I.
       RECENT DEVELOPMENTS IN EXTRACTION AND FLOTATION TECHNIQUES
                      FOR CONTAMINATED SOILS AND SEDIMENTS
                              Merten Hinsenveld, October 1991
INTRODUCTION
 Extraction as presently applied in soil cleaning, refers to a combination of particle separation,
 dissolution and dispersion in an aqueous environment. The terminology "extraction" was adapted
 in the early stages of development, beginning 80's, because it was believed that contaminants could
 easily be extracted from the soil. At present we know that, due to the fact that contaminants are
 heavily adsorbed onto the fines (particles smaller than 63 \im), the cleaning in these installations
 is mainly due to the separation of these heavily contaminated fines from the bulk of the soil. Hence
 the terminology "classification"  techniques would have been more appropriate.
 The process  of extraction-classification can economically be applied to soils with a combined fines
 and organic matter contents of  less than about 30%. There is a simple reason for the limitation of
 30% fines in the current extraction installations: fines are difficult to dewater and lead to huge
 amounts  of sludge. For example: one ton of soil containing 30% fines  leads to about 0.6 ton of
 contaminated sludge (compared to the original soil,  the water content of the fines is roughly
 doubled), and 0.7 ton of clean soil, the waste stream being reduced by only 40%.  Because sediments
 consist mainly of fines, these extraction techniques can not be used for sediments.
 Experience has shown that, in some cases, intermediate fractions, e.g. fractions between  about 30
 and 63 |im, are much cleaner than the fraction below 30 um. Separating these fractions in addition
 to the sand particles  (> 63 fim), would result in a reduction of the amount of heavily contaminated
 sludge. Furthermore, it was discovered that the residual concentrations in  the cleaned sand particles
 could sometimes be attributed to the presence of larger particles of lower density, that showed the
 same behavior in the separation  equipment as the  (clean) sand particles,  e.g. coal  particles
 contaminated with PAHs. It is not surprising, therefore, that current  research  is focussed on
 improving separation performance, e.g. by combining different types of cyclones, and on new
 separation techniques, Chapter  2.

 A technique, which largely circumferences the above mentioned problems,  is the flotation technique.
 Although  strictly spoken a particle separation technique, it is worth a separate  heading. Flotation
 is one of the  most powerful techniques in soil cleaning. Basically any component can be removed
 by flotation.  As opposed to  many  other separation techniques, the surface  properties of the
 contaminants, rather than their particle size governs the separation. The  technique, therefore, can
 be used to separate a mixture of particles of equal size and density, but different surface properties.
 Dissolved material must be flocculated  prior to treatment. As for the process is  contaminant
 specific, it is capable of cleaning a large part of the fine fraction as well.  Hence the production of
 sludge is much smaller than in extraction techniques. Thus far flotation has found little application
 for treatment of sediments. The sludge fraction (particles smaller than about 16 urn) are difficult
 to clean, even by flotation techniques. Application of this technique requires more knowledge than
 extraction. This may well be the reason  that application has  been  limited to a small number of
 installations, mainly in The Netherlands. The flotation techniques are described in Chapter 3.

Emerging extraction techniques, which do iiot rely on particle separation, may use different
extractants, as well as a different types of processes. In aqueous environment, research is focussed
on complexing agents and acids. Biological techniques, using acid producing bacteria, are also focus
                                           886

-------
of current research, Chapter 4.
The use of organic solvents has only recently been considered, but are gaining more and more
attention. These techniques are generally referred to as solvent extraction or supercritical extraction,
dependent on the conditions, Chapter 5. A brief evaluation is given in Chapter 6.
2.
       PARTICLE SEPARATION TECHNIQUES
                                                           Denver/Harz
                                                           Eccentric
                                                                     .Diaphragm
                                                                      or Plunger
                                                                                    — Bed

                                                                                    — Screen
2.1.    Introduction

Particle  separation  techniques can be  applied,  when
contaminants are present as discrete particles or adsorbed
onto a particular particle fraction of soils or sediments.
Mining technology is a major source of information for
these techniques (see f.i. Weiss et al., 1985). Most particle
separation techniques are effective in a certain range of
particle  sizes,  making  a  combination  of  techniques
necessary, to obtain  an effective separation. Because the
techniques are relatively inexpensive, applying more than
one  technique  is  quite feasible.  In this section  some
techniques that are strong candidates for soils treatment
are described. Although the techniques are standard in
mining industry, their application  in soils and  sediments
treatment has only recently been considered.
22.    Jig Technique

In jigging, a bed of particles, supported on a perforated
plate or screen,  is subjected to  an alternate rising and
falling flow of fluid, Figure 1. The objective is to cause the
particles of high  specific gravity to travel to the bottom,
while the particles of lower specific gravity collect at the
top of the bed. The settling conditions can  be  described according to the concept of hindered
settling. By jigging the different fractions again the separation can be improved. Because the
separation is based on size as well as density,  the feed to the system should preferably be sieved or
cycloned into different fractions as to obtain a better result in the process. If fractions of equal size
are jigged separately, the location of the particles in the stratified layers will be a function of density
only. The technique  has already been  incorporated in a number of extraction and flotation
installations in  The Netherlands to remove particles having a diameter larger than 0.15 mm with
a significant different density from the bulk  of the soil.  Practical experience in soil cleaning has
shown that the  technique is very useful for the separation of tar particles. These particles are larger
but also lighter  than sand. This causes the tar particles to show the same behavior as the clean sand
particles in cyclones, thus contaminating the cleaned sand. The jig technique is a standard technique
in mining industry and promises to become one  in soil treatment as well.
                                                                                 Hutch
                                                        Figure 1: Jig separation (Spottiswood, 1982).
                                           887

-------
 2.4.    Humphrey-spiral Separation

 A downward winding gutter used to separate
 particles is called a Humphrey spiral, Figure 2.
 As the water  flows  down the  spiral,  it is
 subjected to centrifugal force which places much
 of the water near the outer rim until the flowing
 film reaches an equilibrium between centrifugal
 force outward and gravitational force downward.
 In such a  curved channel, the bottom layer of
 water,  retarded by  friction,  has  much  less
 centrifugal force and, consequently, will flow
 sideways along  the bottom towards the inner
 edge,  carrying  with  it the heavier  particles.
 Simultaneously with this bottom flow of water
 inward, the  upper mass of water must  flow
 outward to replace it. In fact, the heavy particles
 in the spiral take the shortest way downward,
 while the lighter particles, with the bulk of the
 water at the outer radius, move downward. The
 different fractions can be collected via openings
 in the  spiral at different distances from the
 center. Separation in more than two fractions is
 possible. The technique is applied in the alluvial '
 mining industry  for the winning of heavy metals from river and beach sands. The Humphrey spiral
 has recently  been introduced in the extraction installations in The Netherlands  and seems to
 perform quite well. There is, however, no official information on this performance available (since
 the technique is innovative in soil cleaning, the firms tend to keep silent about the performance).
 Optimum particle sizes range between 0.075 and 3 mm.
                                      Figure 2: Humphrey-spiral separation (Spottiswood, 1982).
3.
FLOTATION TECHNIQUES
3.1.    Chemistry of Flotation

The addition of specific chemicals is one of the major factors in determining the success of this
technique and it is worthwhile to look at the chemicals used in the flotation process. They are
usually  classified  as (1)  Collectors and Extenders, (2) Modifiers  (pH-modifiers,  activators,
depressants, dispersants) and (3) Frothers.  !

Collectors and extenders. When components adsorb on the surface of the soil particles (mostly fine
particles on which the contaminants are adsorbed), the surface energy at the solid-liquid and the
solid-gas interfaces decreases. These interfacial energies are minimal upon saturation of the entire
surface. Many organic components can  in fact adsorb onto  the surface,  making the particles
susceptible to flotation. Components that can do this are called collectors. It is important that the
collectors adsorb onto the contaminated particles that must be removed. Collectors, therefore must
be  selected with care  and be added to the slurry at the right conditions. Establishing these
                                          888

-------
conditions requires extensive laboratory investigation.
An important property of collectors is their polarity. They consist of a "head" and a "tail" with the
polar tail directed away from the particle. In water the result is a decrease of the surface tension
resulting in a hydrophobic particle. Collectors in water are characterized by the charge of the head
that attaches to the particle as anionic (-) or cationic (+). By choosing the charge of the head, the
collectors can attach to specific surfaces; anionic collectors  attach to positively charged surfaces,
whereas cationic collectors attach to surfaces that are negatively charged. The effect of collectors
can be enhanced by extenders which in turn attach to the collector tail to enlarge the it or to give
the tail different  properties. The collectors used in soil treatment can be classified into as: (1)
Anionic collectors for sulfidic contaminants such as xantates, dithiosulphates, thiocarbinilide and
thiocarbamates; (2) Anionic collectors for non-sulfidic contaminants, mostly fatty acids of different
kind; and (3) Cationic collectors for non-sulfidic contaminants, such as alkylamines and quaternary
ammonium salts.
Modifiers.  Different  chemicals
can  be  used  to  modify  the
collector   action   and   the
selectivity  of the collector with
respect  to the  contaminants.
The   most  commonly   used
modifiers  are  pH-regulators,
activators  depressants   and
dispersants.
pH-regulators. These regulators
change the pH of the solution.
They change the  properties of
particles   that   have  hydroxyl
groups or  hydrogen ions at the
surface.  Examples  are:  acids,
chalk and  soda.
Activators. These  chemicals are
added  to  improve  the  bond
between collectors and particle
surface.  Examples are:  sodium
sulfide and copper sulfate.
Depressants.  To  improve the
selectivity  of  the  flotation
process  with  respect  to the
contaminated particles,  it  may
be   necessary   to  depress
flotability  of certain other (non
contaminated)   particles.
Examples are:  dextrine,  sodium
cyanide and starch.
Dispersants.   In   substrates
containing a large amount of
fine   particles   it   may  be
necessary to add chemicals that
Water
    Contaminated soil  Source: M. Hinsenveld (1990), Flotatie
                       tvan verontreinigde grond, Monograph
                       written for the Dutch Ministry of the
                       Environment (second edition)
                      Sand fraction
                      (0.020  to 0.063 mm) to 2 mm
 Chemicals
              Fines
              (0.020 to 0.063 mm)
                                                 Chemicals
  Air
            I  [FOAir            Air
            -^—u—  Froth/Concentrate
       Cleaned soil        Sludge        Cleaned soil
 Figure 3: Mosmans flotation, general flow scheme.

 keep these fines in suspension. Examples are: sodium  silicate
                                              889

-------
  (waterglass) and hexametaphosfate.

  Frothers. Contaminants are captured in a foam that forms at the surface. In order to effectively be
  able to skim this foam and to prevent redisssolution of contaminants a resistant foam is needed
  Examples are: cresylacid and methyl-isobutyl-carbinol. The latter one is a universal frother.


  32.    The Mosmans Process

  In the Netherlands, three full-scale flotation installations are available, with cleaning efficiencies
  generally ranging from 75 to 99%. Lurgi, in Germany is presently developing a flotation plant. One
  of the most advanced flotation installations is the installation of Mosmans Mineraaltechniek (NL)
  the firm that introduced the technique in the field of soil cleaning in 1983. A general scheme of the
  process is  given  in Figure 3. At present, the company is cleaning the famous Sandoz  site in
  Switzerland, that was contaminated by mercury compounds, pesticides  and  herbicides originating
  from a fire [de Boer, 1990]. The company has recently reached extremely high removal efficiencies
  in cleaning soil contaminated with HCH, including the congener lindane: from 1,600 to 0 03 me/ke
  with an efficiency of 99,99% [Mosmans, 1991],                                      '    &/ &'


 33.    Air-sparged Hydrocyclone

 This technique is designed at the University of Utah to achieve fast flotation of very fine particles
 HI a centrifugal field, Figure 4. Compared to  a conventional flotation  cell,  the  air-sparged
 nydrocyclone (ASH) has a high specific flotation capacity (up to 100 tons per day versus 1 to 2 tons
 per day for conventional flotation). The ASH consists of two concentric  vertical tubes with a
 conventional cyclone header at the top and a froth pedestal at the bottom. The slurry is fed
 tangentially through the conventional cyclone header to .develop a swirl flow  of a certain thickness
 in the radial direction, and is  discharged through an annular opening between the insides of the
 porous tube wall and the froth pedestal. Air is sparged radially through the inner porous tube and
 sheared into small bubbles that attach to the hydrophobic particles and  form a froth phase in the
 cyclone axis. The  outer tube serves only as  an  air jacket to provide  for even distribution of air
 through the porous inner tube. The froth phase is stabilized and constrained by the froth pedestal
 at the underflow, moves towards the vortex finder of the cyclone header and is discharged as an
 overflow product. Research on the ASH has also been carried out at the University of Eindhoven
 (NL), but results for cleaning of soils or sediments are not known to me.


 4.     AQUEOUS EXTRACTION TECHNIQUES

 4.1.   Introduction

 Extraction techniques are well developed in the Netherlands and Germany. A total of about ten full-
 scale instaUations is available. Developers of "innovative" extraction techniques often claim magical
 cleaning efficiencies of 99% and better. Generally not indicated is, that  these contaminated soils
hardly contained any fines (particles smaller  than 63 (im) or that only a small clean fraction was
obtained. The techniques described here aim at extraction rather than classification
                                          890

-------
4 2.    Extraction   with
       Complexing Agents
                                                                                  OVERFLOW
                                                                           	(TjJ  HEADER
                                   PARTICLE
                                  •UVBLL
Complexing  agents  have  the
ability  of trapping metals in a
structure and shielding them from
interactions with other  species.
Different types  of complexing
agents can be used,  such  as
EDTA, NTA and citric acid. The
metal-complexes then have to be
separated from the matrix.  FBI
(NL)  applies  a  so-called heap
leaching process for an extraction
of sediments and the fine fraction
in soils with EDTA [Olijve, 1988;
FBI, 1989]. Prior to leaching, the
sludges  are   treated  with  a
flocculant.   The   flocculated
material  is then fed into a basin
and flushed either upward  or
downward   with   a   dissolved
complexing agent in solution. FBI
uses EDTA, but other complexing
agents   could   be   considered.
Essential for the method is that
the percolate is allowed to flush the material homogeneously, without clogging and channeling. In
the FBI  system, the soil is, therefore, slightly consolidated by mechanical vibration. The loaded
complexing agent solution is regenerated by flocculation-sedimentation and/or, depending on the
type of metals separated, electrochemical methods. The metals are being separated as hydroxides,
sulfides or pure metals. The heap leaching method is applicable to metals like Pb, Zn, Cd, Cu and
Co. Arsenic is very difficult to remove. In one experiment researchers reported to have reached the
Dutch A-level for lead and zinc [Kolle; and Van Dijck, 1989]. After the extraction is complete, the
material must be flushed with water to remove residual EDTA.
                                              UHOCHfLOW
                                              ItUMV
                                 Figure 4: Air sparged hydrocyclone (EPA, 1990b).
 43.    Acid Extraction

 Acid extraction is the extraction of contaminants at a pH lower than  3, without significant
 classification. Acid extraction can be used if the metals are bound in a soluble matrix or in an acid
 soluble form to the particles and cannot, or with low  efficiencies, be extracted with sodium
 hydroxide, such as some forms of Pb, Cd, Zn, Cu and Cr(m). Two mechanisms in the acid
 extraction of heavy metals from soil and sediments are important: (1), the rapid exchange of
 reversible exchangeable metals (cations) with H+ and the dissolution of oxides and sulfides and (2)
 a slow exchange from the crystal structure by dissolution of aluminum silicates. In many cases the
 amount of reversible bound metals and the soluble fraction, however, are very small compared to
 the amount that is bound irreversible.
 In the full-scale installations in The Netherlands, acid is applied when tests indicate the positive
                                           891

-------
   influence of its use. In general, however, the pH is not lowered to less than 5 to 4. Since addition
   01 FeCls is often necessary for the subsequent flocculation step in the process, the acidification
   generally is carried out by adding this chemical in the scrubber. An advantage of mild acidification
   is an improvement in settling properties in the solid-liquid separation unit. For more rigorous pH
   values, a protective coating must be applied to prevent corrosion of the installation. Acid extraction

   [Hmse^eld, 1990°]  ***    "^ ** * ** ^ by Heymans V^ but * hardly ever done
   Different  types  of acids can be  used: inorganic  acids such  as HCL,  HNO,  and H,SO, and
   hypochlonc acid, and organic acids such as acetic acid, lactic acid and citric acid. An advantage of
   organic acids is that they not only reduce the pH, but also have complexing properties, and the
   residuals can easily be biodegraded [Joziasse et al., 1990]. Nitric acid has also been considered but
   it was found to create highly undesirable by products, such as tetroxide, and produced end products
   Oat were not amenable to further treatment [[EPA, 1990c]. Florosilic acid, as an extractant for lead
  aSS. ^SOP?SK  ^ 1991b,]' ?f the ln°rganic 3dds' HC1 is usuaUv Considered the mos
  weShlfll? -H '    H r * *   Tely -^ 16SS harmful than °ther in*»»»fc acids> fa™.
  well-soluble chlorides and forms complexes with a number of metals (e.g. Cd, Zn, Cu). Extraction
  efficiencies in the literature range from 5  to 95% (the wide range indicates the importance of
  ^emtion). Extrac ions of lead from sediments was found to be greatly enhanced if,  prior to
  acidification, an oxidation by H2O2 was performed. Large amounts of chemicals were needed to
  obtain the desired pH of 1: about 200 kg HCL and the same amount of NaOH per cubic meter to
  w^rJso rSollft6 ^ ^ ^f V?**™ et *• 19911' Chloride levels after treatmeS
  S '  ™f? ?     ?    ! Pr5lem' Alth°Ugl1 acid extracti°n offers possibilities for the extraction
  of  metals from soils  and sedunents,  some, major  disadvantages, e.g. corrosivity and  handling
  problems, and excessive use of chemicals, may hold its implementation.                "-noting
 4.4.    Microbiological Leaching

 Itoobiologjcal leaching, which involves the use of microorganisms to extract metals from materials
 has recently been introduced in the field of soil treatment. The process is used throughout the worid
 to extract copper from low-grade ores and waste materials via a process called dump leachinelt
     °    d    d t0 Xt™t UraniUm fr°m suffide-rich ^es. The microorganism, kkiaUy Vhough
                         ZnpOTtant ™ the leaching °f metak from  ores ™ the rod-shaS
            ferrooxidans, but many others have now been identified.  They can be classified as
                        ^
   d,/7uOMdrS^™0b^^^ organoparus and the extremely acidophilic
 and thermophylic genus Sutfolobus.  The importance of mixed cultures is now generally accepted
 T.ferrooxidans and T.th ooxidans combined, for example, are more effective in leihiBgi
 Aan either  orgamsm alone. Similarly, the coUination L. ferrooxidans and T. organoparS
                     ?"* ^^ (?**«,  a feat neither species  can a J,mpS
         .  .              mentl?ned endure a low PH  (some enduring up to 7% sulfuric acid)
 and are very tolerant to concentrations of metals that are normally considered toxic. They may use
 SST  -H ^^ i°n°rS ienergy S°UrCe); the CMb011 SOUrce * ""** to bicarbonate^or ?he
 Thiobacdlus thiooxidans, whereas T.ferrooxidans and Sulfolobus may also use organic comDonents
^STS1 aCCePt?r °rf ^ " "°l6CUlar OXygen' bUt Sulf°lobus m^ alsoTe L" or Fe-
Although the organisms degrade mineral structures by directly attacking sulfide and ferrous iron
ST? ^fftong, ferous ion * itself a strong oxidizing agent of sulfide mmerl andLarty
enhances the metal dissolution process [Brierley 1984]. Since natural soils and sediments Jont2
                                          892

-------
                                                  STORE ROOF
                                                                              CUT-AWAY VIE* OF
                                                                              STORE SHOWING ORE
                                                                              FRAGMENTS
limited   amounts  of  sulfur   compounds,
addition of sulfur flower will be necessary.
Remaining  a sufficient concentration  of
oxygen is crucial to the process.
Two  different  leaching  processes   are
generally distinguished: heap leaching (or vat
leaching on a small scale), which consists of
a heap of,  in this case soil  or  sediment,
which is sprayed with the leaching solution;
and  flood  leaching,  where   the  heap is
completely submerged under  the leaching
solution. Stope leaching is a combination of
both flood  and heap leaching and  is  an
intermittent flood and drain, Figure 5.
A  biological   membrane  reactor   was
investigated for use in cleaning of sediments
in The Netherlands by Storm Environmental
Consultancy (SEC). The process consists of two phases. In the first phase, Lactobacillus helveticus,
producing lactic acid from sugar substrate (lactose) is used to leach and complex the metals that are
readily available at a pH of about 3 to 3.5. (L. helveticus is a thermophylic bacterium, used for
example to produce lactic acid from whey ultra filtrate or permeate and does not endure pH values
lower than about 3). In the second stage, T. thiooxidans and T. ferrooxidans are used to extract the
metals bound as sulfides from the sediments. A special membrane reactor was developed to
combine bio-leaching in a slurry with simultaneous separation of fluid from silt particles and
bacteria [Storm van Leeuwen,  1991]. A problem occurring was that L.Helveticus did not produce
sufficient acid. Therefore, lactic acid was added. Extraction efficiency for metals of 85 to 100% are
mentioned.
                                            FILL LINE
                                                                        STOPE FLOOR

                                                                       BULKHEAD


                                                                      DRAIN LINE
                                           Figure 5: Stope leaching (Daniel et al., 1990).
5.
        SOLVENT EXTRACTION
5.1.    Introduction

Solvent extraction has been shown to be effective in treating sediments, sludges and soils containing
primarily organic contaminants such as PCBs, VOCs, halogenated solvents and petroleum wastes
[EPA, 1990d]. Full-scale application is still limited. Some of these systems use solvents that are
either flammable or mildly toxic or both. However, there are long-standing standard procedures
used by the chemical companies, gasoline stations, etc., that can be used to greatly reduce the
potential for accidents [EPA, 1990d].


52.    Extraction with  Organic Solvents

Two main types of solvents can be distinguished.
-  Non water-miscible solvents, such as hexane. An advantage of such a solvent is the relative
   simple separation of solvent and cleaned soil. A disadvantage is that the solvents generally do
   not reach the water  filled pores of the soil particles and, therefore, only remove superficially
                                            893

-------
     present contaminants, unless  some measure is taken (e.g. intensive scrubbing,  ultra-sonic
     vibrations). The latter disadvantage disappears if the solvent can be mixed with a phase
     transmitter that is soluble in both water and solvent.
  •  Water-miscible solvents, such as acetone. An advantage of these solvents is the intense contact
     with contaminants in the pores of the  particles. A  disadvantage  can  be the difficulty in
     separating water, solvent and contaminant. The latter disadvantage disappears if the solvent is
     extremely volatile compared to water.
  Acetone has been successfully used by the US Department  of Defence to remove explosives (TNT,
  DNT, etc.) from sediments. Unfortunately, the process concentrated the acetone dissolved explosives
  in a  closed container, which may be very dangerous. For  this reason, acetone  extraction was
  abandoned in favor of rotary  kiln incineration [EPA,  199 la]. The Sanivan Group (USA) has
  developed a mobile solvent extraction unit, the so-called Extraksol  technology [EPA, 1990b]. The
  extraction is a batch process and is performed in a tumbling vat. No information is available as to
  what solvent is used,  but the solvent is said to be a non-chlorinated and non-persistent organic
  solvent (perhaps acetone?). After the extraction the soil is separated from the extractant and dried
  by heating. The solvent is purified by distillation and recirculated in the process. Some limitations
  mentioned are a maximum clay fraction of 40%, a maximum water content of 30% and a maximum
  size of porous material of 5 cm. The technology has not yet been demonstrated.
  A similar process, the so-called LEEP (Low  Energy solvent Extraction  Process) is subjected by
  Environ-Sciences  (USA)  to the US EPA Emerging  Technologies  Program, Figure 6.  The
  contaminants are leached with a hydrophilic (water miscible) solvent from the soft or sediment and
                                                                      /Dabrla, SlonaiN
                                                                      V^   Ciaual  )
    :
-------
Another process in the Emerging Technologies Program is developed by Harmon Environmental
Services (USA). In this process, a hydrophobic patent solvent blend is used to extract PCBs from
soil. The solvent wash is cleaned by distillation. This process is still on a laboratory scale.
A continuous solvent extraction system using dichloromethane has been developed by Ph0nixMdj0
(DK) The system is marketed under the name CONTEX. The system is specifically  designed for
soil treatment and has a throughput of about 10 tons per hour. The soil is pre-screened down to
about 50 to 80 mm and transported through a tube system by schaftless augers, where it is flushed
counter currently with the extracting agent [Haztech, 1991]. The bulk of the extracting agent is
removed from the soil by gravity separation and regenerated by distillation. The extracting agent
is fully recycled  in the process. Residuals of dichloromethane in  the soil  are  removed  by
volatilization at a temperature of about 100 °C, followed by steam stripping. These residuals are
(after cooling) recovered and also recycled in the process. Contaminants remain in the distillation
unit as an aqueous emulsion, which is sent for off-site treatment. Results reported are (all numbers
in mg/kg): tar, from 270,000 to < 30; jet fuel/gasoline, 15,000 to <  10; heavy fuel/oil sludge, 22,000
to <  10; drilling cuttings, 400,000 to < 60; chlorinated solvents, 3,600 to <  1; BTX compounds,
5,000 to'< 1.5;  naphthalene, 5,300 to < 1.5 and phenanthrene 23,000 to  < 1 mg/kg.


53    Extraction with Aliphatic Amines

Resources Conservation Company (USA) developed a so-called BEST-technology (Basic Extraction
Sludge Treatment), Figure 7. This also is a solvent extraction technique, but the features of the
technique are worth a separate heading. The technology is based on the inverse solubility of a
number of secondary and tertiary amines: complete soluble in water at low temperatures and not
miscible with water at higher temperatures. At present only TEA (triethylamine) is used,  which has
an inversion temperature  of 20 6C. The process is applicable to most  organic and/or oily
contaminants in soil or sludges, including PCBs. In the process, the contaminated mineral matrix
is mixed  with the cold  amines to extract the (chlorinated) hydrocarbons. After the solvent has
broken the oil-water-solid bonds, the solids are allowed to settle from the suspension. Following the
 solid-liquid separation, the fluid phase is heated to separate water and TEA. The organics will be
 transferred to  the organic phase and can, subsequently, be recovered in a stripping column. The
 solvent TEA is recycled in  the process, whereas  the water is discharged to a local waste water
 treatment plant. The oil product fraction is  chemically  unaltered by the process. At large
 concentrations, the oil can be recovered. A full-scale test (100 ton/day) was performed in 1986, on
 a sludge from a refining site. The sludge composition varied considerably, with an oil concentration
 ranging from 0-40%, water from 60-100%, solids from 2-30% and PCBs ranging from 1-13 mg/kg.
 It was reported that separation efficiencies,  defined as the amount of desired product less the
 amount of all  undesired products  times  100, often exceeded 98% [Sudell, 1988]. Research on
 contaminated sediments in The Netherlands indicated that extraction of PAHs with TEA was more
 effective  than  extraction with toluene [van Dillen, 1991]. The presence of detergents can result in
 lower separation efficiency and emulsifiers can affect organics separation from the water fraction.
 Because  TEA  is flammable in  the presence of oxygen, the treatment system must be sealed from
 the atmosphere and operated under a nitrogen blanket. Prior to treatment, it is necessary to raise
 the pH of the  material above 10, to be able to effectively recycle TEA in the process. A research
 and development permit to  operate the BEST pilot plant has recently been issued [Haztech, 1991].
                                           895

-------
  5.4.    Supercritical
          Extraction
         Raw Wail* F«d
        (•ludgi/peat/toll)
                                             TEA
                                       NaOH (racycla)
  Contamlnatad
    Reject
   Malorlal
  Liquified  gasses  can,
  similar  to  extraction to
  organic solvents, be used
  for  the extraction   of
  organic components. An
  advantage of the use of
  liquified gasses  is  the
  high volatility  and  the
  relative  ease with which
  the  solvents   can  be
  recycled. A disadvantage
  is  that  the  extraction
  must be executed under
  pressure. Heated to their
  critical temperature, they
  also   exhibit  the
  properties  of
  supercritical fluids. The
  properties  of
  supercritical fluids, such
  as density, viscosity and
  diffusion coefficients are
  intermediate   between
  those  of the  gas and
 liquid  phases,  and are
 dependent upon the fluid
 composition,
 temperature   and
 pressure.   The
 compressibility   of
 supercritical  fluids   is
 large just above the critical temperature; at this point small changes in pressure result in larce
 changes m the density of the fluid. The liquid-like behavior of a supercritical fluids results in greatly
 enhanced solubihzing capacities compared to subcritical gas, with higher diffusion coefficients; lower
 SWcSf c" e*en.d.eV!r™Perature ra^e Compared to the corresponding liquid [Wright and
 Smith, 1986]  Supercritical fluids possess alhigh solubilizing capacity for organic components and
 an extreme low solubihzing capacity for inorganic components. Many possible supercritical fluids
 are reported to be used, such as ethylene (32.2 «C), carbon dioxide (31.1 »C), ethane (322 »C)

                                   c)> butane (152-° °C)'sulphur dioxide (157-6
Figure 7: BEST extraction (Sandrin and Fleissner, 1990).
An advantage of using carbon dioxide or water to other gasses is, that no environmental hazardous
components are introduced in the system. Supercritical extraction is mostly used for the extraction
?;o^P°ne!f frf fluids' *?ut tests with a0* and sediments have also been reported. Brady et al.
(1987) report on the extraction of PCBs and dioxins from soil, using CO2 as the extracting agent.
They conclude that the extraction of PCBs is relatively simple, whereas the extraction of dioxL is
                                          896

-------
[ 	 j ' -
(T
VI
Column

rr-^ 
-------
 separation techniques is expected.
 Flotation is probably the most powerful technique in soil cleaning. It cleans, rather than separates
 the fines from the soil and is suitable for almost any contaminant. Furthermore, performance of
 flotation keeps improving, whereas extraction techniques have shown little improvement over the
 years. However, the sludge fraction (< 16 jim), is at present considered non cleanable.
 The use of organic solvents and supercritical extraction techniques are being investigated at the
 moment and might further enlarge the applicability of the extraction techniques. This also holds true
 for extraction with acids. However, these techniques, will probably be more expensive than the ones
 used at present.
 Heap leaching processes are likely to be used in future for soils and sediments cleaning. Example
 were given for extraction with complexing agents and microbiological leaching. Heap leaching
 techniques allow more time for the  contaminants, to migrate from the matrix into the leaching
 solution.
 New installations are likely to incorporate the advantages of the different methods into a treatment
 train. Flotation will probably play a larger role in these installations. It is expected that the cost of
 treatment will increase, since the easier techniques have already been tested.
 7.
LITERATURE
 Boer, A. de (1990), Flotatie reinigd bodem bij Sandoz, Ingenieurs krant, 5/8, March 1990 (in Dutch),
 p 17.
 Brierley, C.L. (1982), Microbiological mining, Scientific American, August 1982, pp 44-53.
 Brierley, CJL. (1984), Microbiological mining: Technology status and commercial opportunities in
 The World Biotech Report 1984, Volume 1 Europe.                                       ~~
 Daniel, CJff., PX. Douglas, D.H. Herman, A. Marchbank (1990), Modelling a uranium ore leaching
 process, Canadian Journal of Chemical Engineering, Vol.68, June 1990, pp 427-434.
 Dillen, MJLB. van (1991), Cleaning of sediments contaminated with organic micropollutants,
 Proceedings of CATS  congress on Characterization and Treatment of Contaminated Dredged
 Material, Gent, Belgium (1991).
 EPA (1990a), CF Systems Organics Extraction Process New Bedford Harbor, MA - Applications
 Analysis Report - EPA/540/A5-90/002, August 1990.
 EPA  (1990b), The Superfund Innovative Technology Evaluation Program: Technology Profiles
 EPA/540/5-90/006, November 1990.
 EPA (1990f), Workshop on innovative technologies for treatment of contaminated sediments 13-14
 June 1990 (summary report), EPA/600/2-90/054, November 1990.
 EPA (1990d), Solvent extraction (engineering bulletin), EPA/540/2-90/013, September 1990.
 EPA (1991a), Handbook remediation of contaminated sediments, EPA/625/6-91/028, April 1991.
 EPA (1991b), Treatment of lead contaminated soils (Superfund engineering issue1)  EPA/540/2-
 91/009, April 1991.                                                                '    '
 HazTech News (1991),  Liquid extraction process employs dichloromethane to clean soil  Vol 6
 Nr.12, 13 June 1991.
 Hinsenveld, M. (1990a), Procesmatige extractieve grondreiniging, Monograph written for the Dutch
 Ministry of VROM (2nd. ed.), January 1990 (in Dutch).
 Hinsenveld, M. (1990b), Schuimflotatie, Monograph written for the Dutch Ministry of VROM (2nd.
 ed.), January 1990, or (condensed) in Studie naar alternatieve physisch-chemische en thermische
reinigingstechnieken voor verontreinigde grohd, RIVM Report 736102003, part H Mav 1990 DD
 132-162 (in Dutch).
                                         898

-------
Joziasse, J., H J. van Veen, G J. Annokkeb (1990), Extraction of metals from polluted sediments with
mineral acids in F. Arendt, M. Hinsenveld, WJ. van den Brink (eds), Contaminated Soil '90, Kluwer
Academic Publishers, Dordrecht (1990) pp 1389-1397.
Keldennan, P., GJ. Alaerts, N.H. Nge (1991), Heavy metals in canal sediments of The Hague (The
Netherlands): an inventory and use of acid extraction treatment, Proceedings of CATS congress on
Characterization and Treatment of Contaminated Dredged Material, Gent, Belgium (1991).
Kollee, A., F. van Dy'ck (1989), Eindrapportage proefreiniging op semi-technische schaal van
Helmond-grond,  FBI report R89.052, August 1989 (in Dutch).
Mosmans, S. (1991), Reinigingsgrens HCH-grond doorbroken, Land + Water 1/2, February 1991,
P 13-
Olyve, M. (1988), Reiniging van  met zware metalen verontreinigde grond middels het P.B.I.
reinigingsprocedt* Milieutechniek, Nrs.1/2 (1988) pp 96-101 (in Dutch).
FBI (1989), Werkwijze voor het behandelen van grond, Netherlands Patent Nr.8703081,18 My 1989
(in Dutch).                 .   .
Sandrin, JjL, J. Fleissner,  Case study:  bench-scale solvent extraction treatability testing of
contaminated soils and sludges from  the Arrowhead  Refinery Superfund  site,  Minnesota,
Proceedings of the Second Forum  on Innovative Hazardous Waste Treatment  Technologies
Domestic or International, Philadelphia, USA, 15-17 May 1990, pp 455-464.
Spottiswood, E.G. (1982), Introduction to mineral processing, Wiley Inc. (1982).
Storm van Leeuwen, E. (1991), Biotechnological leaching of heavy metals  from the silt fraction of
dredged material:  experiments using a membrane reactor, Proceedings of CATS congress on
Characterization and Treatment of Contaminated Dredged Material, Gent, Belgium, 1991.
Sudell, G.W. (1988), Evaluation of the  B.E.S.T. solvent extraction sludge treatment technology
twenty-four hour test (project summary), EPA/600/S2-88/051, November 1988.
Weiss, NJL (ed)  (1985), SME Mineral Processing Handbook, Society of Mining Engineers (1985).
Wright, B.W., R.D. Smith (1986), Supercritical fluid extraction  of particulate and adsorbent
materials (project summary), EPA/600/S4-86/017, June 1986.

,                                          ###
                                          899

-------

-------
                                    NATO/CCMS Fellow:

                          Robert Olfenbuttel, United States

 Summary Report: NATO/CCMS Pilot Study on Demonstration
          of Remedial Action Technologies for Contaminated
                                   Land and Groundwater
Not included here; this report was incorporated into Volume 1.
                        901

-------

-------
                     NATO/CCMS Fellow:
         Wayne A. Pettyjohn, United. States



Principles of Ground Water:  Fact and Fiction
        903

-------
    PRINCIPLES OF GROUND WATER
            FACT AND FICTION

            Wayne A. Pettyjohn
             School of Geology
        Oklahoma State University
            Stiilwater, OK 74078

                  Introduction

       All of us have an impression of ground-water
quality or the  manner in which an aquifer system
functions.  Not uncommonly these impressions  are
based on past  experience,  hearsay, published
information, a single sample, perhaps on the basis of
several samples representing a considerable difference
in both space and time, or on an assumption formulated
to simplify a physical phenomenon.

       From these generalized impressions, the next
step to using them as the basis for some plan, effort,
initiative, guideline, or regulation, is a natural one.

       Perhaps one of  the  reasons that  our
generalizations can lead us down a troubled path is that
so many of them are exclusively based on theoretical
concepts or laboratory studies, or are required to simplify
a mathematical model; few have ever been extensively
tested under field conditions. Not only are detailed field
                           studies unusual, but adequate, long term ground-water
                           monitoring, even of contaminated sites, is rare.

                                  The following general principles are classified
                           as "fact" and "fiction". They have been tested over a
                           period in excess of six years at a field site in north-
                           central Oklahoma.

                                               Fiction

                           Ground-water  recharge only occurs when field
                           capacity is exceeded.

                                                 Fact

                                  When a soil or rock unit has been saturated and
                           then allowed to drain by gravity, it is said to be holding
                           its field capacity of water.  This implies that ground-
                           water recharge can not occur if the moisture content is
                           less than field capacity because the earth materials will
                           absorb all of the waterthat infiltrates. Although the idea
                           of field capacity is a useful concept, it ignores flow
                           through fractures and other large  openings called
                           macropores.

                                  A graph of precipitation and water-table
                           elevation is shown in  Figure 1.  This rainfall event
                           occurred in July, 1989 at a field site where the aquifer
                           consists of a rather homogeneous mixture of very fine
                           sand, silt, and clay (silt loam). The soil-moisture content
                           was well below field capacity, and the water table was
                     880
                   +f


                   I   »
                   i
                   I
                   btf
                   o
                   I
                   *l
                   S3
  1
878
                     877
                                               Well AS
                                                  I  I
                                                         I  I
                                                                  lilt
                             10   12    2468
                                      July 14. 1989
                                   1O   12    2    4    6
                                           July 15, 1989
                                             Time, in hours

Figure 1. Relation Between Precipitation and Water-Table Rise.
Prepared for the NATO/CCMS Pilot Study on Demonstration of Remedial Action Technologies for
Contaminated Land and Ground Water, Washington, D.C., November 18-22,1991
                                              904

-------
about 7.5 feet below  land  surface.   Most of the
precipitation, amounting to nearly 3 inches, took place
during a period of an hour. Within 30 minutes of the start
of the rain, the watertable beganto rise, suggesting flow,
through the unsaturated zone at a rate of about 15 feet
per hour. At the A-well cluster, which lies in an open
area, drainage of water from the unsaturated zone
largely ceased after about 7 hours, as indicated by the
downward trend of the water table. At the D-well site,
which lies under a line  of large trees, the water-table
response to the rain was less than the A site, and gravity
drainage continued for several additional hours. This is
probably related to interception by  the trees  and a
greater tillable porosity at the D site.

                     Fiction

The water-table gradient is rather uniform, both in
space and time, if the aquifer is not stressed by
pumping.

                      Fact

        In some suriicial aquifers the  direction and
magnitude of the  hydraulic gradient can change
dramatically from one time to the next even though the
aquifer is not stressed  by pumping.  The April water-
table map, shown in Figure 2, suggests a flow direction
to the southwest with a gradient of .003. By July (tig. 3)
ground water was flowing to the south southeast and
the gradient had doubled. Throughout the period of
record the direction of flow ranges within about 125
degrees each year. There are no water-supply wells in
the entire aquifer.
    GRADIENT: O.OO3 64 SW
                                                      GRADIENT: O.COS  S2SE
 Figure 2.  Water-Table Map for April 2,1986.
                                                   Figure 3.  Water-Table Map for July 26,1986.
        Atthis site, when the watertable is lessthan 7.5
 feet below land surface, ground water discharges into a
 small stream that lies a few hundred feet to the west.
 When the watertable declines below the bottom of the
 stream, the direction of flow changes so that the ground
 water can discharge into a larger and deeper stream
 that lies several hundred feet southward. The major
 cause of the water-table decline, averaging about 8 feet
 from March to October or about .1 feet per day in the
 absence of recharge, is transpiration by large trees.

                      Fiction

 Fine-grained sedimentary strata are impermeable
 or nearly so.
                                                 905

-------
                      Fact
                     Fiction
        Despite it's appearance, the fine-grained alluvial
 material atthefield site is quite permeable, both vertically
 and horizontally. Aquifer and slug tests indicate that the
 hydraulic conductivity ranges from about 30 to 120 gpoV
 fts and averages about 60. This is similar to  many
 fractured sandstones.  Tracer tests, velocity studies,
 the lack of head differences between wells of different
 depths during recharge events, and the response of the
 aquifer to pumping, suggest that the ratio between
 horizontal and vertical permeability is about one.  Wells
 can produce as much as 10 gpm.

        At this site,  the fine-grained alluvial material
 fills a steep-walled valley cut into a massive shale unit.
 The top of two buried soil horizons occurs at about four
 and 27 feet below land surface; they have been dated
 as approximately 1300 and 10,300 years old. Although
 only two A-horizons are evident, soil characteristics,
 including macropores, are  abundantly evident
 throughout the entire 43 feet thickness of the alluvial
 material. These features, however, are apparent only
 through the  scrutiny of cores, and they were entirely
 missed during an examination of cutting derived from
 augering and hydraulic rotary drilling^

                     Fiction

 Laboratory  analyses of hydraulic conductivity of
 cores of unconsolldated fine-grained earth materials
 provide reliable or at least useful results.

                      Fact

        Rather sophisticated laboratory analyses of
 cores from the field site indicate values of hydraulic
 conductivity  that are three to six orders of magnitude
 smaller than values determined from aquifer tests.
 Regardless of the care used in sample preparation of
the cores for the permeameter tests, the macropores
were largely destroyed during packing, which resulted
 in exceedingly small values of hydraulic conductivity.

        This suggests that laboratory analyses of
 unconsolidated material are, at best, open to serious
question.This is of particular concern whenthe analyses
 are driven by regulatory controls. It is suggested that;f ar
 more reliable values can be obtained  by conducting
 aquifertests, particularly by means of discharging wells.
The discharge rate,  commonly only 1 to 5 gallons per
 minute, needs to be merely sufficient to stress the
aquifer. The low rate also hasthe advantage of producing
only a meager quantity of water, which in some cases
mustbe drummed or disposedof by some otherapproved
method.
 Movement of  solutes through  a fine-grained
 unsaturated zone is very slow with residence times
 measured In months or years.

                      Fact

        The rate of movement of chemical constituents
 through the unsaturated zone depends, at least in part,
 on the permeability of the materials.  The migration
 commonly is envisioned in a  manner similar to an
 intermittent wetted front moving through the bulk soil
 matrix (piston flow), in which the residence time may be
 measured in weeks, months, or even years.

        With the exception of a continuous source, it is
 suspected that leachate movement is not continuous,
 but rather one that operates as a series of movements
 in response to changes in soil-moisture  content,
 infiltration, and ground-water recharge. The leeching of
 a contaminant from the unsaturated zone may require
 many years, but the mass flow  rate is caused by
 individual, short periods  of flushing, each of which
 removes a small percentage of the remainder. Some
 preliminary datafrom oil-field brine disposal pits indicate
 that the leaching rate is in the vicinity of 10 percent of the
 remainder per year.

        In fine-grained materials, much of the flushing
 occurs through macropores, in which case the residence
 time may be measured in minutes, hours, or a day or
 two.  The latter can cause significant but short term
 changes in ground-water quality during or immediately
 following a rain, and this phenomenon can have a major
 impact on the interpretation of chemical analyses of
 ground-water samples.

        As Figure 4 shows, the concentration of nitrate
 in ground water increased at least fourfold in less than
 two days following a rain that was immediately preceded
 by the application of fertilizer. Furthermore, the nitrate
 concentration decreased approximately fivefoldto about
 half the pre-rain concentration within another two or so
 days. This event took place in September following a
 relatively long dry period.  A similar event occurred in
 the following March, when the soil-moisture content
was about twice as high as it had been in September
 (fig. 5). This example illustrates the same time related
 concentration distribution pattern but with a decrease in
 concentration with depth.

        In both examples, the water table did not rise
with the original increase in concentration, implying that
the volume of water that infiltrated was relatively small
 but of high concentration. The subsequent water-table
                                             906

-------
 I-  10-
 s
 s
 s
     0-
   •10-
   -20'
      0             10            20
                 Time, in days (September. 1985)


Figure 4.  Relation Between Niitrate
Concentration and Depth to Water Table.
    30
    20-
    10-
                         NitrateinWe!IA-1(8.5(t)
                               Nitrate in Well A-4 (14 ft)
                                               30
                          10

                       Time. In days
                                             20
rise, when compared with the concentration decrease,
suggests infiltration of a more slowly moving but larger
volume of water that was less mineralized.

       The implications relative to ground-water
monitoring are significant.  If the pattern is similar in
other geologic terrains, then the concentration in a
monitoring well sample is related to well depth,
construction, and when the sample was collected with
respect to a rain. To determine a generalized background
concentration, it would appear that samples should be
collected after a week or two of dry weather; but this
suggestion should be based on an understanding of the
reaction rate of the system. The reaction rate can be
determined only by continuous monitoring of the water
table,  which is  best  accomplished by  pressure
transducers/recorders, and by the closely spaced
collection of water samples.

                     Fiction

Ground-water quality Is  nearly constant, both in
space and time.

                      Fact

       The quality of  ground water in  unstressed,
confined aquifers should remain relatively constant, but
in shallow or surficial aquifers, the quality can change
significantly within a matter of days,  hours, or even
minutes, as well as throughout the year.

       Figure 6 shows the  variability of specific
conductance throughout a period of several months in
a well 14 feet deep. In this case extreme values ranged
from about 800 umhos to almost 1175, a difference of
nearly  50 percent.  Superimposed on the graph of
weekly measurements  are quarterly samples, which
ranged from about 920 to 1075 umhos.  This figure
clearly illustrates that quarterly  measurements do not
adequately express the  degree or magnitude of
variability.
                                                     1200
                                                     1100
                                                  u
                                                  O  1000
Figure 5. The Concentration of Nitrate Decreased   «
with Depth Following a March Rain on Day Five.
                                                  u
                                                  £
                                                  ui
                                                      900-
    800
     31400
               31500
                        31600
                                  31700
                                            31 BOO
                    Time, in days
Figure 6.  Electrical Conductivity Ranges Widely
Throughout the Year, But This is Not Evident
from Quarterly Samples.
                                                  907

-------
                      Fiction

  Electrical conductivity can be used to estimate the
  concentration of selected chemical constituents.

                       Fact

         Electrical conductivity provides an acceptable
  measure of the  dissolved solids content of a water
  sample, but it does not allow one to estimate  the
  concentration of  minor  constituents.   Electrical
  conductivity is strongly influencedbythemajorchemical
  constituents, such as calcium, sodium, chloride, sulfate,
  and bicarbonate. Generally, minor constituents, which
  normally occur in concentrations less  than 10 mg/L,
  have little impact on the measurement. Figure 7 shows
 the relation between nitrate and electrical conductivity.
  In this case nitrate, driven by a rainfall event, decreased
 with time as the electrical conductivity increased, then
 tended to follow the same pattern  but at a reduced
 concentration,  and finally nitrate began to increase as
 the electrical conductivity decreased.
     1300
    1200
    1100
    1000
                                               f
                                               s
                                               s"
                                           10
                         10

                      Time. In days
Figure 7. There is No Direct Correlation Between
Nitrate and Electrical Conductivity Following a
Recharge Event.

                     Fiction

Nitrate can be used as an indicator of the presence
of other agricultural chemicals.

                      Fact

       There is a high probability that  a small
percentage of fertilizer and other agricultural chemicals
in some terrains can travel rapidly to the water table by
 means of preferred pathways, as previous examples
 have illustrated.  On the other hand, nitrate can have
 several other sources, none of which are related to
 agricultural activities.

         As one example, in the 1960's several water
 samples were collected from a shallow private well in
 North Dakota that tapped a mass of surficial sand and
 gravel.  Livestock were excluded from several acres
 surrounding the well, the adjacent grasslands were but
 lightly grazed, and there were no other apparent sources
 of contamination in the near vicinity. Nonetheless, from
 one time to the next the nitrate concentration ranged
 from less than 2  to more than 100 mg/L.  The owner
 noticed a relation between rain and high concentration.
 In this case the source of the nitrate was the leaching of
 decaying organic matter that occurs in abundance in
 the many shallow depressions that dot the land su rf ace.

                      Fiction

 The rise in the water table following a period of rain
 can be  used to estimate the amount of ground-
 water recharge.

                       Fact

        One approach to calculating ground-water
 recharge is to multiply the water-table  rise by  an
 estimated value of porosity or specific yield. On the
 other hand, a rain-induced response of the water table
 in a fine-grained,  unconfined aquifer is related to the
 soil-moisture content and the finable  porosity in the
 unsaturated zone. Calculated values of finable porosity
 at the Oklahoma field site range from (ess than 10 to as
 much as 24 percent.

        Dry material can absorb substantially more
 water than can wet material because, in the latter case,
 part of the storage space is already occupied.  The
 infiltration of one inch of water  in a relatively  dry
 unsaturated zone (tillable porosity of 25 percent) results
 in awater-table rise of about fourinches, while the same
 quantity  in a moist situation (tillable  porosity of 10
 percent) will cause a rise of nearly 10 inches (fig. 8).

                     Fiction

 Cuttings of earth materials penetrated during well
drilling provide  an accurate  representation of
subsurface conditions.

                      Fact

       Cuttings from holes drilled by auger, hydraulic
rotary, or air provide a good indication of the lithology of
                                            908

-------
 0
 £
 8
  .
 f
 o
 i
                                 Water-Tabla Rise
     0.0       0.1       0.2       0.3
             Finable  Porosity.  Percent
0.4
Figure 8. The Amount of Water-Table Rise Due to
a Precipitation Event Is Related To Soil Moisture
and Finable Porosity.

the material being penetrated, but they do not furnish
adequate information on the rock structure, which may
control hydraulic conductivity. Cores impart far more
detail than cuttings, but they too can be misleading,
partly because the core represents an infinitely small
sample of an infinitely large area.

       Lithologic logs, based on cores acquired from
a hazardous waste disposal site, are shown in Figure 9.
The alternating layers of very fine-grained sandstone,
siltstone, and mudstone suggest individual water-bearing
zones that are separated by confining units. The cores
display a few fractures, but presumably not enough to
have much influence on hydraulic conductivity.

       Whentheverticalheaddistributionisexamined,
however, it becomes clear that the entire upperSOto 70
feet is so fractured that the entire thickness serves as a
single hydrologic unit, regardless of  rock type (fig. 10).

                  Conclusions

       The conclusions to be reached from the above
described hydrogeologic  misconceptions should be
quite obvious.  The solution, however, boils down to at
least two clear facts.  First, adequate and long-term
monitoring of field sites, representing different geologic
conditions, is essential to justifiable interpretations of
hydrogeologic data, and to the formulationof regulations
and guidance documents.  Secondly, results of the field
studies must be made readily available to and used by
the scientific/regulatory community in orderto avoid the
entanglements brought about by differences between
fiction and fact.
                                                    :•»] Moditeae
         | Sudrtoae

         I SllUtone
      Figure 9. Cuttings and Cores Provide a Good
      Indication of Lithology but not Necessarily
      Hydraulic Conductivity.
      Figure 10. Hydrologic Cross-Sections are Based
      On the Vertical Distribution of Head and Indicate
      Zones of High and Low Hydraulic Conductivity.
                                                909

-------
                  References
                                         i
Acre.T.J., 1989, The Influence of Macropores on Water
Movement in the Unsaturated Zone: Unpublished M.S.
Theses, Oklahoma State University, 129p.

Froneberger, D.F., 1989, Influence of Prevailing
Hydrologic Conditions on Variation in Shallow Ground-
water Quality:  Unpublished M.S. Thesis, Oklahoma
State University, I53p.                     I

Hagen, D.J., 1986. Spatial and Temporal Variability of
Ground   Water Quality in a Shallow Aquifer in North-
Central Oklahoma: M.S. Thesis, Oklahoma State
University, 191 p.

Hoyte, B.L, 1987. Suburban Hydrogeology and Ground-
Water Geochemistry of the Ashport Silt Loam, Payne
County,  Oklahoma: M.S. Thesis, Oklahoma State
University, 277p.

Melby, J.T., 1989, A Comparative Study of Hydraulic
Conductivity Determinationsfor a Fine Grained Alluvium
Aquifer: Unpublished M.S. Thesis, Oklahoma State
University, 148p.

Nelson, M.J., 1989, Cause and Effect of Water-Table
Fluctuations in a  Shallow Unconfined  Aquifer:
Unpublished M.S. Thesis, Oklahoma State University,
194p.

Ross, R.R., 1988. Temporal and Vertical Variability of
the Soil and Ground-water Geochemistry of the Ashport
Silt Loam, Payne County, Oklahoma: M.S. Thesis,
Oklahoma State University, 116p.
                                           910

-------
                         Appendix 2-A



       Thermal Technology Case Studies
Rotary Kiln Incineration, The Netherlands
    911

-------
                  PRACTICAL  EVALUATION OF  A  SOIL  TREATMENT PLANT
 INTRODUCTION

     Since  1982 much experience has  been  gained  in The Netherlands, with soil
 cleanup  techniques.   Thermal  and  extraction processes In particular have
 passed the technological  development  stage and  are applied on a  large scale
 for  various types of soil contamination.

     At present,  the  government, industry and commerce, and the public need,
 first of all,  information on:

     o  how to  select the  most suitable cleanup  technique, depending on the
       type of soil  contamination;

     o  environmental impact assessment and environmental auditing of remedial
       action  technologies; and

     o  costs.

     It is  therefore  essential to  develop a standard methodology  by which soil
 cleanup  projects can be evaluated in  practice.  Within this framework the
 Ministry of Housing,  Physical Planning,  and Environment commissioned DHV,
 Consulting Engineers, in  May  1987 to  carry out  a practical evaluation of a
 thermal  soil treatment plant built  by Ecotechniek.  The objective of the study
 was:

     o  the development of a standard  evaluation methodology for  soil treatment
       plants  in general,  and

     o  the determination  of the environmental aspects of the plant management
       and the cleanup performance  of Ecotechniek's installation in particular.

     The  study was conducted during  the period of August-October  1987.

     This paper will  discuss in turn:

     o  the state of  the art of contaminated soil treatment in The Netherlands,

     o  the thermal treatment plant  where the study took place,

     o  the evaluation study carried out, and
                                  I   i.
     o  conclusions about  the evaluation of soil cleanup techniques.

     Emphasis is laid on the framework and the design of the study.  Only some
of the measured results are presented, to illustrate the monitoring program
which was  undertaken  to determine cleanup performance and emissions into the
air.
                                    912

-------

-------
         Table 1.  Summary of Processing Capacity for contaminated Soil
                      Treatment  in The Netherlands  (ref.  2)
Company
Location
Type of Process
Capacity* in Metric
    Ton/Year**
Broerius
ATM
Ecotechniek I
Ecotechniek II
NBM
BSN
Heidemij
Heijraans
HWZ
Mosmans
Voorthuizen
Moerdijk
Utrecht
Rotterdam
Schiedam
Mobile
Den Bosch
Rosmalen
Amsterdam
Mobile
thermal, <400°C
thermal, <800°C
thermal, <800°C
thermal, <800°C
thermal, <800°C
high pressure fraction
flotation
extraction
extraction
flotation
25,000
60,000
55,000
50,000
40,000
25,000
34,000
14,000
27,000
8,000
* approximately, based upon 8h/day.
    3
**lm  - 1.7 metric tons.
                                   914

-------
    o  drawing up the so-called provisional practical guidelines on sampling
       and analysis in soil contamination research (ref. 5).

    In summary, it can be said that in The Netherlands, the soil cleanup
policy pursued aims increasingly at the total environmental health and
qualitative management of the soil flows released during soil treatment.  Its
background is the soil protection policy which has been taking shape since the
Soil Protection Act came into force early in 1987.  From 1988 on this Act will
also regulate the soil cleanup operation.


A BRIEF DESCRIPTION OF THE ECOTECHNIEK SOIL TREATMENT PLANT

    At the moment Ecotechniek has two thermal soil treatment plants in
operation.  One of these has been in Utrecht (the middle of the Netherlands)
since 1982, and the other one has been operational in the greater Rotterdam
area since early 1987.

    The evaluation study described here concerns the latter installation,
which has a maximum treatment capacity of 50 metric tons per hour.  Figure 1
shows the process scheme of the plant.

    The contaminated soil is treated in the rotary kiln by direct heating at a
temperature of up to 550°C with a residence time of about 7-15 minutes.  The
resulting vapors are afterburned at a temperature of 850-1,000°C and a
residence time of 1-2 seconds.  The afterburner limits the emission of
hydrocarbons, CO, and HCN.  Heat exchangers permit the reuse of the energy
released.  Further measures to limit emission include a wet scrubber (802)
and a fabric filter (dust).  The wash water from the wet scrubber is cleaned
and partly recycled for cooling the cleaned soil.

    The process is suitable for the removal of aromatic and aliphatic
hydrocarbons (up to 10,000 mg/kg), cyanides (up to 400 mg/kg), and polynuclear
aromatic hydrocarbons (PAHs) (up to 800 mg/kg).

    To date, much experience has been gained for the treatment of contaminated
soil from former gasworks sites and soil contaminated by leaking fuel tanks.
A trial cleanup of soil contaminated with chlorinated hydrocarbons is planned
for the end of 1987.

    The installation is suitable for all types of soil.  It is especially the
moisture content, the organically-bound nitrogen and sulfur content of the
soil, and the type of contamination that influence the costs.  These vary
between Dfl 100 and Dfl 190 per metric ton (One Dfl is approximately U.S.
$0.5, October 1987).
EVALUATION STUDY

General
    In 1985, The Netherlands Organization for Applied Scientific Research
(TNO), commissioned by the consultative group on soil treatment of government
and industry, developed a preliminary scheme for a standard evaluation
methodlogy for soil cleanup techniques (ref. 1).  It comprises two phases:
                                    915

-------
                                             COHTJimfUTTP IQH,
                                             CLIAMIO ttHl
                                             COHTMMNATU OAMI)
                                             CLLUttOOAM*
Figure  1.   Process! Scheme for  the Ecotechniek Thermal
                  Soil Treatment Plant
                      916

-------
    o  an Individual evaluation of actual cases, focusing on cleanup
       performance, environmental health and safety aspects, and costs, the
       results of which are to be summarized in public information papers; and

    o  an integral evaluation to be conducted once every 1-2 years by a
       central independent body, comparing the various cleanup techniques with
       each other.

    It was recommended that the proposed scheme be tested in practice, the
main focus of attention being:

    o  sampling and analytical techniques for soil and air emissions;

    o  the way in which the quality of the cleaned soil should be assessed; and

    o  the way in which reliable information can be obtained about the general
       procedure, during the evaluation of the individual cases.

    However, further development of the provisional scheme for a standard
methodology was hampered by a variety of factors.  At the beginning of 1985
relatively little experience had yet been gained from the treatment of
contaminated soil on a practical scale, and it concerned thermal techniques
only.  The proposed testing procedure for cleaned soil met with both practical
and financial objections by the contaminated soil treatment companies.
Furthermore, there was as yet no clarity about the requirements to be set for
cleaned soil.  A number of matters has meanwhile contributed to the fact that
by mid-1987 the time was considered ripe for resuming the development of a
standard evaluation methodology.  They were:

    o  the policy development outlined previously, aimed at an integral
       environmental hygienic and qualitative management of the soil flows;

    o  the greatly increased practical experience gained with various types of
       cleanup techniques, because in the Dutch soil cleanup operation the
       emphasis shifted from investigation and risk assessment to remedial
       action;

    o  the more favorable market prospects for the contaminated soil treatment
       industry; and

    o  the increased interest of government and industry in Environmental .
       Auditing, resulting in pilot projects launched in 1986 in several
       industries.                                                     ;

    A thermal treatment plant of Ecotechniek was chosen for the evaluation
study, since most practical experience had been acquired with this technique
and by this company.  The two main parts of the study will be discussed below
in turn, i.e., the measurement of the cleanup performance and the
Environmental Audit.
                                    917

-------
 Study of Cleanup Performance

 Study design

     The study focused on the treatment of two types of contaminated soil under
 practical circumstances.  The two lots of soil can be characterized as follows:

     o  "easy to clean":  sandy soil, chiefly contaminated with cyanides and
        polynuclear aromatic hydrocarbons (PAHs), derived from a former
        gasworks site; and                                   .

     o  "difficult to clean":  clayish soil with rubble, chiefly contaminated
        with PAHs and mineral oil, derived from a site where various wastes had
        been dumped.

     The study comprised:

     o  the determination of.performance and residual concentrations during the
        cleanup of the soil; and

     o  the measurement of air emissions during the treatment.

     The treatment was carried out under the normal conditions  considered
 optimal by the company.   The average processing capacity was approximately 40  •
 metric  tons per hour.  The  volume of the lots of soil investigated amounted to
 400-600 metric tons each.   The net sampling time was 9 hours for  each  lot.

     Before treatment began  the average  composition of both  lots of soil  was
 determined by sampling and  analysis  of  three mixed samples  per lot.  The
 parameters investigated  were dry  residue,  organic matter content,  grain-size
 distribution,  cyanides (total and free), PAHs (U.S.  Environmental  Protection
 Agency  (EPA)  priority list), and  volatile  aromatics.

     During the  treatment a  1 kg saniple  of  the outgoing cleaned soil was  taken
 every 9 minutes  during the  net sampling period of 9  hours.   A  total of 60
 samples was  thus collected  per lot.   Next,  the testing procedure prescribed in
 The  Netherlands  was  applied in the selection of  the  samples to be  analyzed
 (ref 4).  This will  be discussed  in  more detail  in the next section, "Testing
 procedure oE  cleaned  soil."

     Parallel  to  this,  the ingoing soil  was sampled in  the same way.  Taking
 the  average residence time  of  the soil  in  the treatment plant  into account,  20
 samples of ingoing soil  corresponding to those samples of outgoing soil
 selected for chemical analysis wer^  chosen.   From these,  10  samples were
 chosen at random for  chemical analysis.  For  financial reasons the number of
 samples of ingoing soil  to  be analyzed  was kept as  low as possible.  The
 reason is that in practice  the residual concentration  in  the cleaned soil is
 the  main assessment criterion and to a  lesser extent the cleanup performance.

     The dry residue,  the PAHs, and Cyanide or mineral  oil contents were
 determined in the samples of in-  and  outgoing soil,  in addition,  a mixed
 sample was composed of the 20 selected  samples of cleaned soil, for
 supplementary investigation of several parameters  that play a role in the
possible uses and sales potential of  this material, such as leaching behavior
and ecological integration.
                                   918

-------
    Storage, preservation, and chemical analyses were carried out in
accordance with the provisional practical guidelines on sampling and chemical
analysis in research on soil contamination  (ref. 5).
Testing procedure of cleaned soil

    Testing of the cleaned sbil took place according to the standard procedure
developed in The Netherlands for this purpose (ref. 4).  The procedure for
"once-only" lots of soil, not greater than 2,000 metric tons, was applied.
Another procedure is applicable to larger lots, in which an initial sublet of
between 500 and 2,000 metric tons is intensively analyzed.  If the results of
the first test are favorable, the following sublets are subsequently tested
increasingly less exhaustively.

    The testing procedure used here for "once-only" small lots is multi-stage,
that is, in the first instance the residual concentration is deter ined in 20
of the 60 samples of the outgoing soil.  Next, depending on the results,
another 20 samples or, finally, all 60 samples are included in the test.  The
calculated mean value of the residual concentration and its standard deviation
are examined.  The maximally permissible value for the estimated standard
deviation (Se) is determined on the basis of the requirements set for the
residual concentration:

    Se = (y-x)/2.39
    in which x and y are test requirements regarding the residual
    concentration:
    x = mean value
    y = 99th percentile

    For the first stage of the testing procedure, the lot of soil is passed if:

    o  the calculated value m.20 < 0.8 x

    o  the calculated standard deviation S2o < 0.5 Se

    If this is not the case, then the number of observations is increased by
selecting at random another 20 samples from the remaining 40 and determining
the residual concentration in those.   In the second stage the lot of soil is
passed if:

    m40 <. 0 . 9 x

    S40 < 0.65 Se

    If necessary, all 60 samples may finally be tested,  with as a requirement:

        < x
    S60 < Se

    The sampling procedure used here for in- and outgoing soil during the
investigation can be represented schematically as shown in Figure 2.
                                    919

-------
  SAMPLING
 OUTGOING SOIL !
                               CHEMICAL ANALYSIS :
    60 samples
 1 st stage 	^-20  samples

 2 nd stage	*^20  samples

 3 rd stage 	«»-20  samples

	:	i^l mixed sample
  INGOING SOIL
                                       mixed samples
    60 samples
   20 samples
I camieonding to umoin outgoing 1011
stltcttd for 1 it st«p analysis )
               -10 samples
        Figure  2.   Sampling  and Analysis Procedure
                               920

-------
    pollution, monitoring

    Hydrocarbons, SC>2, NOX, fluorine, chlorine, and HCN  in  the waste  gases
were measured continuously by Ecotechniek, as a standard procedure.   To
supplement  this, during the sampling period DHV measured dust, 02, CO2,
CO, SO2, fluorine, chlorine, mercury, HCN, volatile aromatic hydrocarbons,
nitrous oxides, and hydrocarbons.  PAHs and heavy metals were determined in a
few dust samples.
Results

Cleanup performance

    The principal process conditions during the treatment were:

                                                        Soil From

                                              Gasworks Site     Dump Site
    Rotary kiln temperature (°C)
     .range
     .average

    Afterburner temperature (°C)
510-570
  540

  925
430-570
   510

   850
    The requirements laid down in the specifications of the soil cleanup
projects from which the lots of soil were derived, were:
    Gasworks Site
     .cyanide (total)
     .vol. aromatic hydroc.
     .PAHs (Borneff)

    Dump Site
     .PAHs (EPA priority list,
             individual)
     .benzo(a)pyrene
     .PAHs (EPA, total)
     .mineral oil
                                              Mean Value
                                              (mg/kg d.wt.)
  10
   0.1
  20
   0.1

   0.05
   1
 100
                99th Percentile
                (mg/kg d.wt.)
   20
    2
    0.3

    0.15
    3
  200
    Some of the results are summarized in Tables 2 and 3.
                                   921

-------
             Table 2.  Results of Treatment of Soil From a Gasworks Site
Soil from Gasworks Site  (sandy soil) '
Analyses
Ingoing  Outgoing  Requirements  Removal
 n = 10   n = 20      (mean)       (%)
         Procedure
Ressults     For
Te£3ting  Standard
 Mean    Deviation
 n = 20   s = 20
Moisture %
Organic matter %
PAHs (mg/kg.d.wt.)
fenanthrene
anthracene
f luoranthene
pyrene
benzo(a)pyrene
total Borneff (6)
total EPA-priority
list (16)
Cyanide (total)
(mg/kg.d.wt.)
Volatile aromatic
hydrocarbons
(mg/kg.d.wt.)
benzene
toluene
xylene
Heavy metals
(mg/kg.d.wt.)
chrome
copper
zinc
nickel
cadmium
lead
mercury
arsenic
17.5
5.1
5.41
1.13
8.73
8.45
4.02
33.2
133.04
82

0.03
0.03
0.04
<20
24
110
<20
<20
197
<20
<50
12.4
2.7
0.07 98.8
0.10 91.5
0.08 99.1
<0.01 98.8
0.04 98.9
0.5 20 98.0 accepted
4.88
1.2 5 98.5 accepted rejected

0.04
0.03
o'.oi
<20
28
130:
<20
<20
410
<20
<50
                                    922

-------
                     Table  3.
Soil from Waste Dump Site
(clayish soil with rubble)
Results of Treatment of Soil From
   a Waste Dump Site
                                                                          Procedure
                                                                 Results     For
                                                                 Testing  Standard
                       Ingoing  Outgoing  Requirements  Removal   Mean    Deviation
Analyses                n = 10   n = 20      (mean)       (%)     n = 20   s = 20
Moisture %
Organic matter %
PAHs (mg/kg.d.wt.)
fenanthrene
anthracene
f luoranthene
pyrene
benzo(a)pyrene
total EPA-priority
list (16)
Mineral oil
(mg/kg.d.wt. )
Volatile aromatic
hydrocarbons
(mg/kg.d.wt. )
benzene
toluene
xylene
Heavy metals
(mg/kg.d.wt. )
chrome
copper
zinc
nickel
cadmium
lead
mercury
arsenic
21.1
8.4

13.43
1.65
3.58
3.70
1.09

76.91
2,602




0.03
0.01
0.03


24
36
197
<20
<20
110
<20
<50
11.1
5.0

0.22 0.1
0.03 0.1
0.11 0.1
<0.01 0.1
0.05 0.05

1.86 1
10 100




0.06
0.03
0.02


28
42
200
<20
<20
140
<20
<50



98.4 rejected
98.5 accepted
96.9 rejected
97.3 rejected
95.5 rejected

97.0 rejected
99.5 accepted




















rejected
accepted
rejected
accepted
rejected

accepted
accepted

















                                    923

-------
     The  results  show  that  removal  rates above 95 percent were achieved for
 both  lots of soil.  Residual concentrations of mineral oil and cyanides amply
 meet  the requirements.  For some of the PAHs, the requirements were not met by
 soil  originating from the  dump site,  in the case of the gasworks soil, higher
 residual concentrations of (Borneff) PAHS were allowed.  As expected no
 removal of heavy metals took place.  The testing procedure for cleaned soil
 (first stage:  analysis of 20 samples) resulted in the rejection of the lot of
 soil  from the dump site.   Neither  for the mean value nor for the standard
 deviation were the (quite  strict)  requirements met with regard to various
 PAHs.   Although  the mean value of  the concentration of mineral oil was far
 below the required value,  the requirement set for the standard deviation could
 not be met.  This was due  to one single value just above the required mean
 together with 18 values far below.  This shows that the requirements for the
 standard deviation may need some adaptation in order to avoid this kind of
 illogical result.


 Air pollution

     Some of the main results  are presented in Table 4.

     These results show that the  requirements  set  for emissions  into  the air
 were met during the treatment  of both  lots of soil.


 ENVIRONMENTAL AUDIT

 General

     The  limited  audit  conducted  for this evaluation  study fits into  the
 framework of a number  of pilot projects being  implemented by order of  the
 Ministry  of Housing, Physical Planning, and Environment.  The objective of
 these  pilot projects is to  explore  the possibilities  for setting up  an
 environmental protection system in  various industries.  An environmental audit
 is an  examination of the quality of plant management  in the field of
 environmental protection.   An audit can only succeed when it has the express
 support of the management of the company in question.  It should be convinced
 of the value of the audit for its own company, and where it concerns a pilot
 project, as is the case here, for the entire branch of industry.  All relevant
 information sources should  be placed at the auditor's disposal.  The necessary
 cooperation is also required of the personnel, who must be informed about the
objective and the operating procedure.

    The objective of the audit conducted at the thermal soil treatment plant
of Ecotechniek was:

    o  to gain an insight into the way in which the company meets the
       environmental requirements set by the authorities;  and

    o  to investigate whether the management of the plant,  both from a
       technical and an organizational  viewpoint,  is geared  to implementing
       its own corporate policy,  its own corporate rules regarding the
       environment,  and the regulations laid down  by the government.
                                    924

-------
                 Table 4.  Results of Air Pollution Monitoring

SO2
Dust
Cadmium (in dust)
Mercury (in dust)
PAHs (in dust)
Volatile aromatic hydrocarbons
Chlorine
Fluorine
HCN
Flow (m3*/h)
Measured Value
(mg/m3*)
120-180
1-24
0.02-0.03
0.006-0.0
n.d.
n.d.
<3
<0.3
2.4-3.3
31.000
Requirement
(mg/m3*)
200
75
0.1
0.1
0.002
5
75
5
10
38.000
n.d. = not detectable.
*    = dry gas, temp. 273 K, press. 101.3 kPa.
                                     925

-------
      The audit program focused on the following environmental aspects:

      o  acceptance of contaminated! soil

      o  solid wastes


      o  emissions into air  and soil

      o  waste water


      o  external  safety


      Noise, vibrations, and other nuisances, such as attracting traffic and
 vermin were  not  taken into account, because of the plant's location in an
 industrial area, adjoining a  far bigger solid waste incineration plant.


 Design of the Audit


     The audit was conducted in three phases:

     o  pre-audit activities

     o  audit visits

     o  reporting


     The pre-audit activities  included:


     o  determining the required industrial  information to be  provided by
        Ecotechniek;                                              '      *  .. •


     o  studying the industrial information,  the environmental licenses  In
        force, and the other internal  and external environmental regulations;

     o  drawing up the audit plan;


     o  discussing the plan thoroughly with the management; and


     o holding an information  meeting for the personnel of the plant.
 < «,, the aUdit visits' the audit was conducted on the spot in accordance
with the plan drawn up.  The examination concentrated on both organizational
aspects and the operation and maintenance of the plant.  The efficiency of the
plant management was examined with regard to:


    o  prevention (maintenance, tests, registrations, internal control,
       environmental provisions, safety systems); and


    o  repression (fire extinguishers, scenarios, internal and external alarm
       systems,  handling of complaints).
                                   926

-------
    The information gathered during the audit was evaluated according to the
following criteria:

    o  environmental requirements set by the local authorities and the
       government (Nuisance Control Act, Surface Water Contamination Act,
       Chemical Wastes Act);

    o  corporate policy regarding the environment and safety;

    o  controlability of the activities at the plant; and

    o  involvement of the personnel of the plant with the environmental policy
       of the company.

    No measurements were made as part of the audit.  A summary of the audit,
intended for external use, will form part of the final report of the
evaluation study.  In addition, a more detailed report will be prepared, which
will be confidential and available to the management only.

Results of the audit

    The audit was effectuated with the  full cooperation of the management and
the personnel of the plant.  During the collection of information and the
audits visits, no problems had to be faced.  The audit will result in       •   •
recommendations to the Ecotechniek management with regard to environmental  and
safety policy.  Recommendations to the ministry will refer to the setting up
of an environmental protection system for the soil treatment branch of  the
industry especially.

CONCLUSIONS

    Some of  the main conclusions of the evaluation study are as  follows:

    o  The integral approach, combining the measurement of treatment
       performance and emissions with  an  environmental audit provides a
       valuable  tool  to  assess a soil  treatment  process.  The quality of the
       management  of  the plant in  the  fields of  environment  and  safety
       strongly  influences  the cleanup  performance and the reliability  of  the
       treatment plant process.

    o  It would  be desirable to  determine uniform cleanup  criteria  based upon
       residual  concentrations for those  substances  representative  of  the  main
       contaminants which are present.  Especially in the  case of PAHs, a
       variety of  requirements are laid down by the  authorities  (e.g.,  for
       total EPA-priority list PAHs,  different  individual  PAHs,  as  well as
       total Borneff-group PAHs).

     o  The  standard  testing procedure for cleaned soil which is  used in The
       Netherlands may need some modification with regard  to the requirements
       set  for the standard deviation of  the measured values.

     o  The  thermal treatment plant that was the subject  of this  study
       performed well with respect to cleanup results and  emissions into  the
        air  during the treatment  of both lots of soil. This  shows that thermal
        treatment is a suitable process for the  cleanup of  similar contaminants
        and  concentration ranges  and for different types  of soil  varying from
        sand to clayish soils.
                                     927

-------
REFERENCES

1.  Assink, J.W. and Rulkens, W.H.  "Eerste opzet voor een standaard methodiek
    t.b.v. de evaluatie van bodemreinigingstechnieken" ("Prelimenary Set-up
    for a Standard Method to Evaluate Remedial Action Techniques for
    Contaminated Soil"), TWO, The Hague, The Netherlands, 1985 (in Dutch).

2.  Bavinck, H.F.  "Reiniging van verontreinigde grond en het
    stimuleringsbeleid van de rijksoverheid" ("The Policy of the Dutch
    Government to Stimulate the Treatment of Contaminated Soil"),
    Post-Graduate Course Solid Wastes and Soil Contamination, Technical
    University Delft,  September 1987 (in Dutch).

    "Handboek Bodemsaneringstechnieken"  ('"Handbook for Remedial Action
    Techniques"), Staatsuitgeverij,  1983, The Hague,  The  Netherlands,
    Revised Edition,  1987 (in Dutch).

    INTRON,  Material Testing and Consulting,  "Opleveringscontrole van
    procesmatg gereinigde grond en groundwater"  ("Testing of  Treated Soil  and
    Groundwater"),  Maastricht,  The Netherlands,  1986  (in  Dutch).

    Kooper,  W.F.  and Mangnus, G.A.M.   "Sampling  and Analysis  in Contaminated
    Site Investigations,  Impediments and Provisional  Guidelines in the
    Netherlands,"  in Contaminated Soil,  edited by  J.W.  Assink and W.J.  van
    den Brink, Proceedings of the First  International TNO Conference on
    Contaminated  Soil, Martinus  Nijhoff  Publishers, Dordrecht,  The
    Netherlands,  1986.

    Leer, W.B. de.  "Thermal methods developed in  The Netherlands  for the
    Cleaning of Contaminated Soil," in Contaminated Soil,  edited  by  J.W.
    Assink and W.J. van den Brink, Proceeding's of  the First .International TNO
    Conference on Contaminated Soil, Martinus Nijhoff Publishers, Dordrecht,
    The Netherlands, 1986.

    Soczo,  E.R., Verhagen, E.J.H., and Versluijs, C.W.  "Review of Soil
    Treatment Techniques in The Netherlands," in Environmental Technology,
   Proceedings of the 2nd European Conference on Environmental Technology,
   edited by K.J.A. de Waal, Martinus Nijhoff Publishers, Dordrecht, The
   Netherlands, 1987.
3.
4.
5.
                                         Martein W.F. Yland
                                         DHV Consulting Engineers
                                         P.O. Box 85
                                         3800 AB Amersfoort,  The Netherlands

                                         Esther R.  Soczo
                                         National Institute of Public Health
                                           and Environmental  Hygiene
                                         P.O. Box 1,  3720  BA
                                         Bilthoven,  The Netherlands.
                                  928

-------
                          Appendix 2-B



       Thermal Technology Case Studies
Indirect Heating in a Rotary Kiln, Germany
     929

-------
       Experience with the Decontamination of the
                        I
        Konigsborh Coke-Oven Plant Site using a
                    Pyrolysis Plant
Author: Ass. d. Bergfachs Gerhard Lehmann, Dortmund
Under a research and development project support by the
Federal Ministry for Research and Technology of the
Federal Republic of Germany Ruhrkohle Westfalen AG
operates a pyrolysis plant to contaminate a site of a
former coke-oven plant decontaminated with hydrocarbons.
The plant was erected on the site of Konigsborn 3/4
coke-oven plant in BSnen (district of Unna), which was
closed-down in 1977, by Deutsche Babcock Anlagen AG
following detailed planning according to a Ruhrkohle AG
concept and a simplified approval procedure possible on
the basis of the Federal Act on Emission Control for
such test plants in 1987/88. The initial approval for a
test plant has, in the meantime, been replaced by a
site-specific permanent approval following a procedure
involving the general public. So far, some 85.000 t of
soil have been treated.

The pyrolysis method was selected for this
decontamination project because Ruhrkohle AG disposes
of corresponding expert knowledge from coke-oven
technology and the process moreover offers the
possibility to dispense with the secondary treatment of
the pure heating gases and the fire-resistant lining of
the pyrolysis or carbonisation chamber. As the soil is
not heated to the temperature required for complete
pollutant destruction  it does not ceramilize during the
treatment and is not dead but only biologically
inactive. The R&D-project will be terminated at the end
of this year. The plant will continue its operation at
this site. At the same time it will be investigated at
which site the chamber may be used afterwards.
                          930

-------
Plant Set-Up and Operation

The plant (fig. 1) is designed for a materials'
throughput of 7 t/h with a soil humidity of up to 21 %
and a volatile pollutant content of max. 5 %. It
operates as follows: the material to be treated
undergoes a preparation with crusher, overbelt magnets
and screening plant in which the permissible maximum
grain size ,is limited to 50 mm. Preliminary tests
showed that smaller mesh sizes cannot be used because
of the soil's structure which is in some cases
characterized by  high clay contents. Thus also the
possible purification successes can only,be utilized to
a  limited extent  even if the crushing work is
intensified. The  treated material is supplied  to a  feed
bunker with a downstream twin feed screw delivering to
the  alloyed steel rotary tube with a length of some
26 m, of which  20.6 m are  indirectly heated, and a
diameter of 2.2 m. The  rotary tube rotates with an
infinitely  adjustable speed  of  up to  2.5 rpm.  Staggered
lifting  rails  transport the  material through the rotary
kiln in  1 to  1h h. The  permanently  filled  feed screw
 seals the entire  rotary tube including the downstream
 system in  such a  way that  a  negative  pressure  of
 approx.  1.5 mbar  generated by a fan operated at the end
 of the overall plant can  be  maintained to  avoid the
 escape of  both dust and pyrolysis ,gas. The plant is ,
 heated by means of 18 natural gas-fired burners
 distributed over  3 zones.  They heat the kiln to.a wall
 temperature of 600 to 650°-C so that a lining is not
 required in the kiln.  The observance of thia maximum
 temperature is guaranteed by a limit switch reacting to
 the thermal elongation of the rotary kiln. Given an
 adequate residence time this temperature is sufficient
 to remove the soil humidity first and then the noxious
 PAH's mentioned  in the EPA list which was selected as
 assessment criterium after consultations with the
 approving authority and the involved expert
 institutions (fig. 2). In this procedure the ring-
                           931

-------
 shaped polycyclic aromatics decompose almost completely
 into aromatic chain compounds (fig.  3).  The
 decontaminated soil falls into a discharge chamber with
 a  screening screw arranged below separating the
 material  into the grain sizes > 10 mm and < 10  mm,,  The
 coarse grain size is supplied to a wet ash remover
 where it  is cooled- down in a water  bath to a
                         i
 temperature suitable for the safe further handling and
 transported to an interim storage bunker using  a
 scraper conveyor.  The fine grain size, which cannot be
 cooled directly in the water because of  the complexe
 slurry treatment this would entrain,  is  supplied via a
 cooling screw system to an interim bunker where it  is
 humidified  and cooled-down further by adding water.

 The  pyrolysis gas  is extracted at the top of the
 discharge chamber  and supplied from  below to the
 neighbouring post-combustion chamber with fire-
 resistant lining,  in which 3 natural  gas-fired  burners
 are  used for supporting firing and guarantee a
 temperature for a  reliable combustion of  the pyrolysis
 gases  flowing through here for approx. 3  s  to C02 and
 other  harmless  components  (fig.  4).  Because  of  the
 generally high  water vapour content  an autothermal
 post-combustion process  could not be  implemented so
 far. During the "test approval time"  the  post-
 combustion  chamber was  operated  at a  temperature of at
 least  1.000°  C,  in presence of chlorine above pre-
 defined limits  also  at  a temperature  of some  1.250° C,
 to guarantee  the distruction of  possible  dioxines.
 Subsequently, the  gases are channeled down a  cooling
 tower  (quenching)  after fresh  air addition with approx.
 850° C where  they  are cooled-down to a temperature of
 160° C to 190°  C at  an hourly  water addition of up to
     O
 3.6 m  in short time, i.e.  at  approx. 1.7 s residence
time. Behind  the cooling tower slaked lime is injected
into the flue gas  stream which binds also pollutant
gases, in particular  S02 or  fluorine, on the  surfaces
of the downstream  "system Beth"  filter system with 196

                          932

-------
filter hoses corresponding to a filter area of 216 m2
to separate entrained dust by reaction. The filter
system guarantees the observance of the values defined
by the technical instruction for air pollution control
still effective until the end of 1992 in the outlet
air. Together with the heating gases of the rotary kiln
which are re-used in a heat exchanger to heat the
combustion air to approx. 350° C the burned gases are
vented into the environment via a stack.

In May 1988 the plant was heated for the first time and
initially supplied with weakly contaminated material to
test the mechanical components. After removing the
deficiencies established in this context almost
completely and after finalizing the measurements for a
continous operation defined in the approval certificate
through the "Technische Uberwachungsverein  (TUV)" the
test operation could be started. Initially, a two-shift
operation per working day  (6.00 - 22.00) and from mid
August 1989 a three-shift operation had been approved.
With the granting of the permanent approval time-
related operating restrictions became  ineffective.
 Technical Aspects  and Modifications

 The preparation plant is  so efficient  that  the material
 for a three-shift  operation can be prepared in one
 shift.  Only for the removal of all iron parts a  second
 magnet had  to be retrofitted.  Additional tests and
 modifications related to  different screen covers.  Here
 the retrofitting to rubberized screen  covers is  worth
 mentioning  in particular." They were  used in tests  to
 throughput  smaller grain  sizes than  50 mm but also to
 reduce noise emissions.

 Initially,  non-removed iron parts caused failures  at
 the feed screws in the feed system.  Moreover,
 additional  failures and observations showed that also

                           933

-------
the  screws  need to be wear-resistant  and made  of  a
specially hardened material.  On  the Konigsborn site  to
be treated  matter  which act very corrosively on the
screws  is contained in the crushed material  of former
buildings and  foundations in  addition to the clay parts
in the  soil. Also  for th0 discharge screws which  in
some cases  lacked  the lubricating properties of the
soil humidity very short service lifes of only  a few
weeks were  established initially. By  means of  special
build-up  welding the service  lifes of all screws  could
be extended to  several months.

In serveral inspections of the rotary tube connected
with wall thickness  measurements  signifcant  damages or
changes could not  yet be stated.  In contrast one  of the
two  compensators sealing the  rotary tube
against the environment at the feed and  discharge
openings  had to  be replaced because it had come
defective as a result of frequent flexing work  when
heating-up  and cooling-down during the regular  two-
shift operating  phase.

In the post-combustion chamber modifications were
required  to provide  for the execution of  reliable
temperature measurements. The residual dust  quantities
contained in the pyrolysis gas ceramalized on the
temperature sensors  and caused errornous measurements.
Protective  stones  above the elements  led  to  a
significant improvement here. Another modification had
to be effected in  the  cooling tower.   It was  only after
installing additional  guiding plates  that it became
possible  to design the  water addition  in  such a way
that  an optimal cooling effect could  be reached and
that  the water used  to  this end was distributed also
uniformly over the entire cross-section and thus in the
full  gas  stream and  not  spotwise at individual wall
zones.
                          934

-------
At the connecting pipe between combustion chamber and
cooling tower a flap is installed which is used as
stack when starting-up the plant from the cold
condition. Moreover, it acts as a safety device in form
of a pressure relief valve. It is used for example if
as a result of a failure in the cooling water addition
hot gases with excessive temperatures reach the filter
system and the filter bags may be destroyed by heat. In
individual opening procedures also burned-out red fly
ash deposited under the flap or in the pipeline escaped
leading to irritation and protests in the public. To
avoid repetitions compressed air pulse units known as
"Big Blasters" were arranged at the outlet of the
combustion chamber and below the flap (fig. 5) removing
such dust depositions through regular operation and
providing for a dust-free "opening" of the flap.

The filter systems with the glass fiber tissue filter
bags met, with regard towards the dust emissions, the
requirements of the "TA-Luft" effective at the date of
plant construction. It does, however, not observe the
limits of the 17th Federal Ordinance on air pollution
control and noise abatement (BImSch VO) effective from
1993 onwards. In coordination with the plant
manufacturer the old filter bags were therefore
replaced by bags made of needle felt textile fibers.
The subsequent emission measurements of "TUV" confirmed
that the former dust emissions could be reduced by some
90 % to less than 2 mg/m3 and thus far below the future
limit.
Thanks to the new filters it was also possible to
replace the lime with a surface of approx. 17 m /g used
so far by a more fine grained lime with a surface of'
            O
approx. 36 m^/g. This also improved the binding
capacity for SC>2 and other pollutants.

Particular attention was given to the insulation of the
noise emitters established in noise measurements. As an
                          935

-------
upper noise limit of'40 dBA in the neighbouring
residantial area was set for the night operation both
different electric motors and the fan station had to be
specially encapsulated. As a result of these measures
it was possible to observe the required limits at the
nearest houses - about a 100 m away from the plant -
with unbuilt space in between. All listed improvements
contributed to an increase in plant availability from
some 75 % in the second half of 1988 to 95 % in the
                             t                         i
first half of 1989. This availability could also be
achieved in the first quarter of the three-shift
operation given good scheduled plant maintenance. A
subsequent availability decrease to some 90 % today
(fig. 6) can be explained exclusively by the unexpected
high sulphur concentrations in the feed material for
which the filter system without flue gas scrubber is
not designed. Without this limiting factor a continued
utilization of 95 % would be possible.
Available Operating Experience

Soil Excavation and Preparation

Almost 7 ha of the more than 16 ha site of Konigsborn
3/4 are contaminated and need to be treated or secured
in another form corresponding to the agreement between
the owner and the municipality and the stipulations of
the terminating operations approval procedure required
under the mining law. In test drillings the
contaminations were proven down to a depth averaging
2 to 3 m, in individual cases pollutants had also been
found in 8 m and in one case even in 18 m depth. So
far, 3 part areas were excavated completely, another
2 partly. The excavated material has undergone
comprehensive treatment and, in some cases, was re-
installed after cleaning.
                          936

-------
The material is excavated by using an excavator
applying the undercut method (fig. 7). The excavator is
located at the upper edge of the excavation and loads
the soil from the pit upwards onto a truck. In general,
the personnel does not have to work in the pit and is
thus also not exposed to the emitted volatile
pollutants. Check measurements of the industrial
hygiene agency and of an independent expert at the
mentioned working places showed values in the inhaled
air which range far below the MAC or TRK values. Only
for napthalene the olfactory threshold is occassionally
exceeded leading to initial irretations amongst
observers from a general public which could, however,
be removed  (fig. 8).

Additional measurements were taken in the preparation
area. The air measurements also produced positive data.
In one case - in particularly dry weather - the dust
measurements led to the recommendation to employ the
operators mainly on the windward side, which does not
cause any major difficulties because of the plant
design. Moreover, suitable labour protection devices
are available.
Rotary Kiln Operation
The two-shift  operation  turned  out to be unfavourable
for the plant.  In numerous measuring series taken  by
the Institute  for Thermal Process Technology  of
Dortmund  university  accompanying the R&D project it
could be  proven that an  instable state  for about 2 to  3
hours was created with each  daily plant start-up
 (fig. 9).  During this time the  desired  temperatures in
the soil  could presumably not be reached completely as
the heat  content of  the  already heated  soil was lacking
in the rotary  tube.  Independently from  this it has to
be noted  that  the regular start-up and  shut-down - the
kiln was  operated at reduced capacity during  the nights

                          937

-------
                         I
 at a temperature of approx. 350° C and the combustion
 chamber at approx. 800 to 850° C - resulted in an
 immense waste of energy and thus increased the process
 costs. Another negative 'observation was the freezing up
 of the gas pipe during a frost period after the burner
 at the rotary kiln had been switched off during night-
 time. A retrofitted insulation of the pipe removed this
 failure source. For the normal continous operation the
 Insitute for Thermal Process Technology of Dortmund
 university has established that,  at the rotary tube,
 only 0.8 % of the energy in use is consumed for
 degasification, more than 60 % for water evaporation
 and almost 40 % for the heating of the inert soil
 (fig. 10).

 The three-shift operation significantly improved the
 operating sequence  and  helped  to  reduce failures and to
 improve the operating sequences.  To this  end,  however,
 a  scheduled and adequate maintenance at work-free days
 is required.
Operating Results

The  Institute  for Thermal  Process Technology  of
Dortmund university has  set up a targeted examination
programme and 'collected  the data of this test programme
in a computer. Data were .also evaluated in some cases.
So far, the following results are available:
Figure 11 shows the relation between natural  gas
consumption and soil throughput. The tests were run at
soil humidities of 12 to 13 % with throughputs of 3.8
to 11.3 t/h. At the last mentioned feed quantity, '
however, the measurements had to be terminated
prematurely because the  cooling capacity of the
discharge device was insufficient. However, the data
show clearly that the specific gas consumption
decreases with increasing throughput.  A quantity of 25
                          938

-------
-to 30 m-* STP per tonne of soil should be permanently
required for this plant size.

The soil residence time in the rotary tube is mainly
governed by the speed. Figure 12 shows how the
residence time of the material in the kiln extent's
with reduced speed. Normally the filling rate of the
kiln ranges around 7 to 8 %. The graphs in figure 13
show the relation to the residence time and the hourly
throughput. The extreme values at high filling rates
are only theoretically calculated graphs as the rotary
tube structure is only designed for a filling of
max. 2 2 t.

The specific natural gas consumption was also
established for the post-combustion chamber under
consideration of different parameters (fig. 14). As the
water content of the soil normally amounts to 10 %,
often even 15 %, this straight line shows that an
autothermal operation, i.e. a combustion resulting from
the own energy potential of the pyrolysis gas, is not
possible in general. If, in particular, the combustion
temperature has to be increased to more than 1.200° C
because of the presence of PCB or Cl a considerable
natural gas quantity is required for supporting firing.
However, the figure also shows that a reduction of the
water content in the feed material may lead to
significant energy savings, in particular if also the
consequent savings resulting from the heating of the
soil in the rotary tube are considered. The diagramme
in  figure 15 illustrates how the gas requirement of the
supporting gas firing may vary according to the
pollutant content, in this case by some 20 %.
 In  normal  operation,  i.e.  as  soon  as  the  plant  has  been
 started up and  left the  heating-up phase  completely,
 satisfying cleaning results are  obtained.  The PAHs  are
 decomposed almost  completely,  i.e.  to below the B value
 of  the  Holland  list,  often even  below their A values ,

                          939

-------
 given an adequate residence time in the rotary kiln.  If
 the soil discharge temperature is reduced by"shortening
 the residence time of the soil in the oven or by
 reducing the rotary tube wall temperature residual
 contaminations remain in the soil as shown in figure
 16.  A statement of the cleaning rate in percent seems
 to  be unsuitable because in that case different
 cleaning rates would have to be stated for low initial
 values.  This shall be explained by using two extreme
 analysis series as examples.  In the first sample (fig.
 17)  a very weakly contaminated soil with a load of
 31.3 mg/kg soil was cleaned to a value of 3.9  mg/kg.
 This corresponds to a cleaning rate of 87.5  %.  The
 second sample (fig.  18)  has an initial value of approx.
 17.500 mg/kg and a final value of < 3.2 mg/kg  in a
 coarse grain after cleaning several individual
 components to < 0.1 mg/kg.  In this  case the  cleaning
 rate reaches 99.98 % although the final value  almost
 corresponds  to the value of the first sample.  As  the
 operators  became familiar with the  plant today
 generally  discharge values  are obtained in which
 frequently all residual  values for  the individual
 components range around  0.1 mg/kg and below  (fig. 19).
 Numerous samples were screened at 10  mm so that also a
 first  assessment of  the  cleaning of different  grain
 sizes  is possible.  However, the effect of  grain
 crushing during the  process cannot  be taken  into
 consideration.  More  than 90 %  of the  samples from
 treated  soils  show that  the grain size fraction below
 10 mm  shows  higher contaminations than the coarser
 grains. There  are two explanations  for this:
 Firstly, it  could be  proven in numerous  analyses
 separated  by grain size  fraction that,  in  the  fine
 grain  fraction,  contaminated dust particles are still
 existant which are presumably  extracted  so rapidly from
 the rotary tube directly after evaporation of the soil
 humidity with  the pyrolysis gas  stream that they are
not heated up  to  the  temperature required  for decocting
the pollutants.  In the discharge casing they then

                           940

-------
sediment in the slowed down gas stream and thus reach
the fine grains.  A reduction of the discharge casing
cross-section or an additional heating gas filter at
the end of the rotary tube might help to remove this
residual load.

The second explanation is based on the existing surface
of the soil grains. With its quantity-related larger
surface the fine grain also offers more adhesion
possibilities for pollutants. The coarse grain only
shows a higher contamination in individual special
cases if it is very porous and may be permeated by
pollutants. In a particularly impressive way this is
shown for the already presented sample  (fig. 18) taken
from a former acid sludge store. In the fine grain size
range of this sample the cleaning effect is not quite
so excellent, but also here the individual components
range below the Holland B value. When separating
samples according to grain size ranges  it has to be
considered that the coarse to fine grain ratio may vary
considerably. Individual examples show  20 to 25 %
coarse material, others in contrast up  to 55 %. The
significance  of this phenomenon has not been
investigated  completely. Certainly the  crushing and
preparation of residual foundations have an impact.  In
addition, samples are also examined with regard to
other pollutants in regular intervalls. The BTX
aromatics with their low boiling points are also
decomposed completely in most cases.  It should be
mentioned that the test programme of  Dortmund
university also included a rotary kiln  operation at
lower temperatures. As already the first tests with
lower wall temperatures 'showed insufficient cleaning
rates additional tests were excluded  from the
programme.

Moreover,  samples  are also examined regularly  with
regard  to  their contents of  cyanides  and phenoles
without being able to prove  these  substances in  all

                          941

-------
feed samples.  In presence of  cyanides in the feed they
could never be proven in the  discharge analyses  (fig.
20). Generally, phenoles, if  existing in the feed, can
only be proven in small traces in the discharges but it
has to be mentioned that the  phenole analysis can be
distorted by other components.

At least once  a week also the ashes and filter dusts
produced in the process, together accounting for less
than 0.5 % of  the feed, are included in the examination
programme. As  a result of the secondary treatment in
the higher temperature range  of the post-combustion
chamber the PAH contents of these samples range below
the detectability limits in almost every case
(fig. 21).

Another examination series is to provide information on
the question whether the heavy metal accumulati.ons
detected in individual samples were relocated or
converted through the treatment of the material in the
rotary kiln. First analyses showed that the contents of
heavy metals with low boiling points decrease in the
treated soil as expected. In  contrast accumulations
exist in the ashes and, in particular, the filter
dusts. However, it is too early today to draw
conclusions on the overall behaviour from these
individual results, some of which are shown in figure
22. At present, analyses series are carried out in
which dusts and fine grained material are examined
separately. In this context also the elution behaviour
of the individual components  is determined in "addition
to the absolute pollutant contents as a basis for
further decision making. .      .

In some test series about 6.900 t of soil from other
sites with in  some cases differing contaminations were
treated. Mineral oil contaminated soils of a former
company-owned petrol station, sludges from canal
cleaning and so-called "phenole sands" from a coke-oven
                           942

-------
plant have to be mentioned. The mineral oil
contaminated soil could be cleaned very easily, as also
an analysis (fig. 23) shows. The sludges from the
cleaning of former coke-oven plant sewer canals were,
as expected, characterized by a high fine grain -content
and, in some cases, higher pollutant concentrations.

Individual batches had to be treated twice because the
normal residence time at a rotary tube wall temperatux•
of 600 to 620° C was insufficient to reach values below
the B values in the Holland list. Here the high fine
grain content and the above-average water content were
assumed to be particularly influential.

With the strongly contaminated phenole sands from a
filter system the energy offer in the combustion
chamber was so high that the temperature occasionally
increased to up to 1270° C. In this context fly ash
fusion occurred so that the regular cleaning of the ash
hopper below the combustion chamber became very
complicated. In this case an upstream hot-gas  filter
would have been very helpful. For one batch a  secondary
treatment was required.

Emissions
 In  the  first  approval  certificate  it had been  laid down
 that  a  series of  emission measurements had  to  be
 carried out by "TUV" prior  to  the  final plant
 commissioning for test operation.  In two larger
 measuring  series  the different operating units were
 examined under consideration of the pre-set limits.
 Amongst others "TUV" produced  the  following emission
 measurement results:
                    80  %  of  the  30  mg/m3  limit
                    60  %  of  the  100 mg/m3 limit
                    55  %  of  the  limit
                    70  %  of  the  100 mg/m3 limit,

                          943
Dust
CO
NOX
S09
max.
max.
max.
max.

-------
Additional emission measurements for HCl and HF also
produced values of approx. 30 % below the permissible
values. The individual results are shown in figure 24.
Other measurements related to the mass concentrations
of pulverized anorganic substances according to article
3.1.4 of the "TA-Luft". Here cadmium, thallium, and
mercury from class I, the metals arsenic, cobalt,
nickel, selenium, and tellurium from class II, and
lead, chromium, fluoride, copper, manganese, vanadium,
and tin from class III as well as antimone, palatinum,
palladium, and rhodium in an additional programme were
examined. Also these values did not exceed the pre-set
limits. In the meantime the measurements were repeated
with similarly good results when' treating another site.

Moreover, "TUV" devoted its attention also to the
question of air polluting emissions from the test plant
and connected emissions in the environment by
carcenogenic substances like dioxines and furanes. From
class I of the carcenogenic substances special
measurements were carried out for berylium, benzo(a)-
pyrene (BaP), and dibenz(a, h)antracene (DBahA), from
class II for arsenic, chromium, chromium (VI)-
compounds, cobalt, and nickel - all including the
corresponding compounds - as well as for all pollutants
according to "the list of class III. Also here it was
stated that the values remained far beyond the pre-set
emission limits.   .      '                         ;  •

The measurements for the dioxines and furanes were
carried out in two series over several days by "TUV".
Moreover, a recognized expert institute carried out
measurements for simulated failures following power
drops. The measuring data are listed in figures 25 and
26. They show that the Nato recommended limits were
observed so far with regard to dioxine and furane
emissions.
                            944

-------
In the framework of its measurements ''fUV" Idtthd a
confirmation for an earlier statement Sayiiig* that also
in this plarit benzo(a)pyrene takes on a. ceaftMln
"guiding function". This means that at only" law
concentrations of this substance also bther       •  .'
carcenoganid substances frequently emittetl together
with this substance do not exhibit a significant   •
relevance.       .                                    .

Moreover, "TtJV" produced an emission fdr"§dS§t  for the
neighbouring areas based on its examinatioHlt  In this
context it refered to the spreading clMSS statistics of
the German meteorological service and its ldiif<-year .
measuring series selecting an area with a sidfe length
of 3 km which is larger than the area defindd  under
article 2.6;2.2 of the  "TA-Luft" so that this
examination is even more comprehensive. Moreover, it
assumed that the plant  is operated throughout  the year,
i.e. for  8.760 hours instead of an operation at  five
working days per week  (6.000 hours) according  to the
first  approval certificate. Also in case of  a  permanent
approval  this utilization rate  cannot be achieved as a
result of standstills  on Sundays and bank holidays  or
for  larger revisions.  In its comments  "TUV"  states  that
the  daily take-up  and  the emission concentrations  in
the  neighbouring hazardous  areas range  slightly  more
than 2 %  above the recommended  values which  Menzel
regards as tolerable in his  report  "Assessment aid  for
Dioxines" from the present  view. Given  this  trace
concentration the  approving authority  in  the final
instance  also approved the  operation without
restrictions  in  this direction.

It should be  pointed out  that  in the preliminary
examinations  for hazard assessment  on  the territory to
be treated  only one  area  was established in which
pollutant concentrations  (PCB)  sufficient for the
formation of  dioxines  and furanes might possibly exist.
                          945

-------
Another problem may be caused by the occasional
presence of mercury in the soil to be treated'. Because
of its low boiling point the mercury evaporates in the
rotary tube and forms a constituent of the pyrolysis
gas. In the cooler ash and the filter dust part
quantities are present. To avoid a release of the
remaining mercury into the atmosphere an additional
retaining device has to be provided. At present, we are
undertaking efforts to bind this mercury on activated  .
carbon additions to the lime. To this end also a
specially prepared activated carbon is used in some
cases. Should these tests produce negative results the
use of an activated carbon reactor will have to be
investigated.                                         .
Utilization of the treated Soil
The opponents of thermal decontamination methods claim
again and again that the decontaminated soil is sterile
and unsuitable for a follow-up utilization. To
contradict this assertion different test series - in
some cases in cooperation with university institutes -
concerning the recultivation of. the treated areas are
carried out in Bonen. After the plant constructor had  ,
already proven in small scale tests that in
decontaminated soil shortly after the treatment seed
germinate and plants grow a larger test was started in
cooperation with the municipality of Bonen. The
municipality market garden laid out several beds of  ,
which one with a depth of some 3 m consists exclusively
of treated material. In the other beds compost or 20 cm
and 50 cm top soil were used as top layers. So far, the
thermally treated and normally fertilized and watered ,
soil shows an only slightly weaker groth than the other
beds (fig. 27). In autumn 1989 and spring 1990 the
Institute for Plant Sociology/Ecology of Essen
university laid out several beds on an approx. 5.000 m2
test area in which also an at least 1 m mainly/
                           946

-------
however, more than 2 m thick, thermally treated and
decontaminated soil layer was used. With the experience
gathered with the recultivation of waste heaps in the
hardcoal industry it is tried to revive the growth on
these biologically only inactive soils using special
seed compositions and to accelerate a humus formation
on natural waste through the multitude of the offered
plants. In the meantime
-------
As some quantities were used for the construction of a
noise protection dam samples of the cleaned soil were
moreover submitted to comprehensive eluate examinations
corresponding to the draft directive for the
examination and assessment of wastes - part 2 - of the
Northrhine Westfalian water and waste agency, the so-
called landfill directive (fig. 29). All thresholds for
landfill class 2 are observed. In contrast the eluate
data for landfill class 1 are exceeded generally for
about 5 parameters of which some regress to class 1 in
the course of time under the influence of general
environmental factors. This is particularly pronounced
for the pH value which is positively influenced by the
acid rain and for the chemical oxygen demand (COD
value), which normalizes again under atmospheric
conditions.
Conclusions

The Konigsborn Pyrolysis plant has fully met the
expections in the test period so far. In operation it
was possible to clean also more severely contaminated
soils to such an extent that the treated areas can be
made available for a follow-up utilization. Moreover,
the tests carried out so far also supplied a numbet of
interesting detailed results which will certainly
provide for statements concerning the limits of this
methods in the framework of the R&D project still
extending until the end of 1991. The plant operation
will be continued also after the expiry of the
supported R&D time.
                          948

-------
                                                                                             Fig.  1
                                                                                  1 natural gas
                                                                                  2 rotary kiln
                                                                                  3 incinerator
                                                                                  4 flue gas cooler
                                                                                  5 water
                                                                                  6 flue gas deduster
                                                                                  7 lime
                                                                                  8 stack
RUHRKOHLE
WESTFALEN AQ
                            Degasing  with incineration  of contaminated  soil
Lehmann'
Nov. 1991
                                                949

-------
r Fid. 2

Naphthalln
2-Hethyl-Naphthalln
1-Hethyl-Naphthalln
Fluoren
Phenanthren
Anthracen
Fluoranthen
Pyren
Benz(a)anthracen
Chrysen
Benz(e)pyren
Benz(b)fluoranthen
Benz(k)fluoranthen
Benz(a)pyren
Dlbenz(ah)anthracen
BenzotghDperylen
Indenod, 2, 3/ ctDpyren
ntiunirnui c Boiling points of
n u n mvv/ n L* c
WESTFALEN AG
°C
C10H8 . 218 QQ
C12H12 211
262
C13H10 298 ODD
C11H10 338 Oj
C11«10 310 OX)
C16H10 381 ^J,
ci6Hio 396 <=^::>
r H itto f^YYV
c18H12 138 gyj
C18H12 111 Qj
C20H12 193 ^Q;
C20H12 187
C20H12 181
C20H12 "96 (jrV '
198 I
C22H12 500 JJJ
505
pAu • s Lehmann
Nov. 1991
950

-------
                                Fig. 3

Wa«ar "2° Vo<-%
Sauarstoff 02 Vol-%
Woaentoff H2 Vol-%
Kohlenmonoxid CO Vol-%
Kohlandioxid CO2 Vol-%
Mathan CH4 Vol-%
Ethan C2 H6 Vol-%
Shan C2H4 Vol-%
Ethin C2 H2 vpm
Propan C3 H8 vpm
Propen C3 H6 vpm
iio-Buian C4 H10 vpm
n-Butan C4H10 vpm
trani-Buten C4 H8 vpm )
1-Buten C4H8 vpm )
ijo-Buten C4 H8 vpm
cii-Buten, C4 H8 vpm
Butin C4H6 vpm j
1 ,3-Butadien C4H8 vpm )
n-Pentan C5 H12 vpm
Cyclopentan C5 H10 vpm J
2,2-Dimethylmethan C6 H14 vpm )
2-Methylpentan C6 H14 vpm
3-Methylpentan C6H14 vpm
Hexan C6 H14 vpm
Heptan C7 H16 vpm
Oktan C8 HI 8 vpm
Benzol C6 H6 vpm
Toluol C7 H8 vpm
m-Xylol C8H10 vpm 1
p-Xylol C8.H10 vpm )
o-Xylol C8H10 vpm
rRj/KVyS. ' Pyrolysis Plant Kdnigsfaorn
t~< 1 r^^ /^^ ^V" J
RUHRKOHLE Analyses of Pyrolysis - gas
WESTFALEN AG
1 2 3
65,10 71,60 75,00
2,50 3,60 1,90
0,75 0,37 0,11
0,45 0,23 0,29
1,90 1,00 2,40
0,80 0,36 0,16
0,14 0,08 0,02
0,12 0,06 0,03
8,00 3,00 4,00
436,00 230,00 62,00 .
759,00 400,00 199,00
38,00 21,00 5,00
77,00 42,00 9,00
144,00 77,00 15,00

131,00 70,00 32,00
83,00 44,00 3,00
67,00 33,00 10,00
38,00 20,00 4,00
104,00 55,00 7,00
45,00 28,00 2,00
21,00 16,00 2,00
49,00 32,00 2,00
18,00 37,00 4,00
5,00 5,00 2,00
82,00 77,00 73,OO
64,00 53,00 21,00
7,00 6,00 5,00
5,00 5,00 3,00
3/4 , Lehmann
Nov. 1991
951

-------
cm
=T^^/^
RUHRKOHLE
WESTFALEN AQ
Hn + (m + -J)02«»m C02+4rH20>He;
2 C0 + 02 * 2C0.2
Process in fhe incinerator Lehmi
Nov.'
                                        Fig.  4
952

-------
Fig, 5

-------
                                                                            Fig. 6
 Possible operating
              time
        Stoppages
               by
 - conveyor system
- limitation of SO2
           - filter
   - other  reasons
         including
    • power failure
    • overpressure
discharge chamber
  • limit of  thermal
  elongation of the
              kiln
* stoppages to the
           burner
                              10
                                         100
                                      Time ( h)
1000
           10000
100000
                               Koriigsborn 3/4
                                Thermal Treatment
                           Period: 14^08.1989 - 31.10.1991
                   Lehmann
                   Nov. 1991
                                      954

-------
                                                                            Fig.  7
                              ca. 7m
                                                       contaminated soil
                                                         \   \  \  V
               Refilling           	J I              Excavating

                             ca. 7,0 m  Operating - area
                                    ca. 48,0 m
RUHRKOHLE
WESTFALEN AG
                 System of excavating  and refilling the  soil
Lehmann
Nov. 1991
                                          955

-------
Place of test '
Benzol Toluol . Xylol
/U9/m3 /"S/m3 //9/"'3
Top of excavation area (pit) 17-23 28-29 39-47
Cabin of excavator 15-17 23-38 33-57
Cabin of truck 13-19 17-30 21-24
neighboured area 4 . 5 7
allowed working place concentration 16000 380000 440000
Olfactory threshold (TUV) 2900 8000 350
Olfactory threshold (Hyg. Irut.) 800 2 000
Olfactory threshold by "Roth" 200 200
Olfactory threshold by "Staub"
,-Jc^s\j^ Pollutants content in the air during excavation works
RUHRKOHLE May 1 989, temperature 2:5° C
WESTFALEN AG according to the german norm VDI 3482

Naphtholin
u. g/m3
240 - 440
190-340
35- 58
48
50000
140
30
4
20
Lehmann
Mov. 1991
t
100
80
60
40
20
0
0
2 - shiff operation
June, 13 fh 1989
£ o



v
""^^^
"*•• 1—


i
n
*>
y
]

00 4.00
RUHRKOHLE
WESTFALEN AG

qSw^
f^R



y^




ISwT01-T06
/ /
9^



/
*rl
1
\
8.00 1ZOO 16.00 20DO "&.
Time ( h )
c *,
800 100
700 80
600 60
500 40
400 20
300 0
.00 0.
3 - shift operation
August, 17th 1989
*


\r-
*" 	 "



^yuw--
— v/V^v


»—^k.
^V_X^
s^^**~



ry
-**£>•


a

|Sw

T01 -T06

/
•»—-«_
'

» 4.00 8.00 1200 16.00
Time(h)
Temperature of the wall of the kiln
[ Zone 1 ( intake) -6( outlet )J
/
L^Vl,
«. 	 ,

20.00 24
C
800
700
600
500
400
300
.00
Lehmann
Mov. 1991
956

-------
      Tw=550°C   -
     Tw=600°C   -
                                                                                      Fig.
                            200      400
   Humidity:
   14% - 550° C
   13% - 600° C
                           600       800
                              Energy I kW)
                                                                 1000
1200     1400
RUHRKOHLE
WESTFALEN AG
Used energy for water evaporation, degasification  and
               heating the soil
                                                                                Lehmann
                                                                                Nov. 1991
                                                                                       Fig. 11
spec
45
40
35
30
25
. 20
.gas consumption { m STP/t)

\



3
RUHRKOHLE

^





^
—



r^
-M









-"--s.




«





"^





-s-1
45 6 7 8 9 , 10 11
Throughput
Rotary kiln - specific gas consumption

12
t/h
Lehmann
Nov. 1991
                                             957

-------
                                                                                          Fig.  12
         250
             residence time [min]
         200 -
 RUHRKOHLE
 WESTFALEN AG
Residence time in the rotary kiin




2
tm
V.




0
ann
1991
       Filling rate of the kiln
     221    158     123     100    85     74
                                              65     SB     S3     At
nUHRKOHUE
WESTFA1.EN AG
                Connection of filling rate, throughput  and residence time
                                                                                        Fig. 13
                                                                            residence  time
                                                     Lehrnann
                                                     Nov. 1991
                                             958

-------
                                                                                         Fig. 14
                 Spec, gas -consumption  m STP/t
         60


         50


         40


         30


         20


          10
            1200°C; 0.5% of  PAH's.    r,-£y
                                         8     10    12

                                          Humidify (%)
                                                16     18    20
RUHRKOHLE
               Specific  pas - consumption for heating the  post- combustion  chamber
                r
                                                                  Lehmann
                                                                  Nov. 1991
                                                                                         Fig. 15
        110
        102
         94
              Gas - consumption  m  STP/h
         78
                                                   	post - combustion chamber
                                                                  1000
                                                                   980
                                                                                     960
                                                                                     940
                                                                    920
         70   I        I	1	J	1	1	'	1	1	1   900
             6.00    8.00    10.00   12.00   14.00   16.00    18.00   20.00    22.00   24.00
                                        Time { h )                 date: November, 17 th 1989
 RUHRKOHLE'
, WESTFALEN AG
Influence  to natural gas consumption  in the  post - combustion •
                 chamber  by pyrolyse - gas
                                                                                     Lehmann
                                                                                     Nov. 1991
                                                  959

-------
        PAH
      [mg/kg]
RUHRKOHLE
WESTFALEN AG
                                                                                        Fig 16
                              PAH-concentration in discharged soil
                                        (mean values)
                                     260
                                temperature
                               discharged soil
                                    red
                                                                            fine fraction
                                                                           coarse fraction
                                                300
                                                           340
                                                                       380
Influence of the femperafure of discharged  soil
         fo the cleaning  process
Lehmann
Nov. 1991
                                           960

-------
                              Fig. 17

Naphthalin
2-Melhyl-Naphthalin
1-Methyl-Naphfhalin
Fluoren
Phenanthren
Anlhracen
Fluoranthen
Pyren
Benz[a]anthracen
Chrysen
Benz[e]pyren
Benzo[b]fluoranlhen
Benzo[k]fluoranthen
Benz[a]pyren
Dibenz[ah]anlhracen
Benzo[ghi]perylen
lndeno[l .2.3.cd]pyren
Sum.:
Cleaning efficiency
<=U^r\S^ Pyrolysis Plant
RUHRKOHLE jest August, 2nd
WESTFALEN AQ
Feed Discharge
1,1 0,1
0,5 < 0,1
1,3 < 0,1
0,6 0,2
3,0 0,5
0,8 0,1
4,3 0,6
3,6 0,4
2,4 0,2
1,8 0,2
2,3 0,3
2,6 0,3
1,5 0,2
2,1 0,2
0,5 < 0,1
1,4 0,2
1,5 0,1
31,3 3,9
87,5 %
Konigsborn 3/4 Lehmann
1988 { Results in mg/kg) Nov. 1991
961

-------
                               Fig. 18

Naphthalln
2-Hethyl-NaDhtnalln
1-Methyl-Napnthalln
Fluoren
Phenantnren
Anthracen
Fluorantnen
Pyren
Benz(a)anthracen
Chrysen
Benz(e)pyren
Benzo (b) f luoranthen
Benzo (k) f luoranthen
Benz(a)pyren
Dibenz(ah)anthracen
Benzo(ghl)perylen
Indeno(l,2,3,cd)pyren
Sun
Cleaning Efficiency
Feed Feed
< 10 EH > 10 ED

312,0 3 160,0

153,0 1 600,0
566,0 1 080,0
117,0 151,0
139,0 3720,0
288,0 1 710,0
118,0 1 230,0
78,0 156,0
61,7 660,0
91,9 862,0
76,1 577,0
93,1 821,0
11,8 130.0
56,5 532,0
52,7 518,0
2 551,1 17 570,0

Discharge
< 10 mm

1,1

0,1
2,9
0,2
2,1
1,0
0,9
1,1
1,1
1,9
0,7
0,1
0,5
1,0
0,8
19,3
99,25 Z
={j^£^^ Pyrolysis Plant Konigsborn 3/4
RUHRKOHLE ^^ No.S29, 31th July'89
WESTFALENAG • (Results in ing / kg )
Discharge
> 10 BO

0,7

< 0,1,
0,9
< 0,1
0,1
0,1
< 0,1
0,1
< 0,1
0,1
< 0,1
< 0,1
< 0,1
0,1
< 0,1
< 3,2
99,98 Z
Lehmann
Nov. 1991
962

-------
                                   Fig 19

Naphmolin
Acenaphthylen
Acenaphmen
Fluoren
Phenanthren
Anthracen ,
* Ftuoranthen
Pyren
B«nz[ajanmracen
Chrysen
Benzo[b]Ruoranthen
Benzo[k]fluoranthen
B*nz[a}pyren
lndeno[ 1 .2.3.cd]pyren
", Dibenz(ah]anmracen
Benzo(ghi]peryien

, — lD)/fl\yS£, Pvrnlv<;i<
=y-s^y-\j^^ ryroiysi:
RUHRKOHLE Test Julv
WESTFALEN AG ''
Feed
< 50 mm
36,1
68,0
46,2
131,0
624,0
363,0
602,0
409,0
300,0
458,0
187,0
133,0
204,0
J
) 139,0
89,2
Sum.: 3789,5
Discharge Discharge
< 10 mm > 10 mm
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< ; 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
< 0,1 <
; Plant Kbnigsborn 3/4
17th 1991 (results
in mg/kg)
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
0,1
Lehmann
Nov. 1991
963

-------
1 	 Fin. 20


nig/kg
Cyanide
Phenol

cjRyk^n
RUHRKOHLE
WESTFALEN AQ

Feet' Discharge Ash of quench Dosf of filter
2°<0 < 0,1 < 0,1 < 0,1
4,0 < 0,43 < 0,1 < 0,1

Pyrolysis Planf Konigsborn 3/4 Lehmann
Test on Cyanide and Phenol ( Results in mg/kg ) Nov. 1991
964

-------
                                    Fig. 21

Naphthalin
2-Methyl-Naphthalin
1-Methyl-Naphthalin
Dimethylnaphthaline
Acenaphthylen
Acehaphthen
Fluoren
Phenanthren
Anthracen
Fluoranthen
Pyren
Benz[a]anthracen
Chrysen
Benzo[b]fluoranthen
Benzo[lc]fluoranthen
Benz[e]pyren
Benz[a]pyren
lndeno[ 1 -2.3.cd]pyren
Dibenz[ah]anthracen
Benzo[ghi]perylen
1
945,0
216,0
96,0
160,0
196,0
789,0
480,0
1764,0
649,0
1021,0
589,0
355,0
470,0
189,0
156,0
132,0
165,0
63,8
21,7
51,1
Sum.: 8508,6
1. Feed 2. Discharge
3. Dust 4. Ash of quench
RUHRKOHLE
WESTFALEN AG
Pyrolysis Plant
Test of dust and
2
0,9
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
0,6
1,4
' 0,3
0,4
0,2
0,4
0,4
0,4
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
6,1
3
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
4
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
< 0,1
Date: February, 10th
1989
Kbnigsborn 3/4
ash ( Results in
rng/kg )
Lehmann
Nov. 1991
965

-------










Fiq. 22


Queduilber (Hg)
Cadmium
Arsen
Chrom
Nickel
Kupfer
ZTnk
Btei

gJDVnV^
RUHRKOHLE
WESTFALEN AG
(Cd)
(Ar)
(Cr)
(Ni)
(Co)
|Zn)
(Pb)
mg/kg
mg/kg
mg/kg
mgAg
mg/kg
mg/kg
mg/kg
mg/kg
Feed
7,30
0,71
116,00
97,00
48,00
45,00
151,00
149,00
Discharge
< 10 mm
1,80
0,80
151,00
70,00
37,00
53,00
155,00
236,00
Discharge
> 10 mm
< 0,40
0,44
48,00
88,00
41,00
30,00
121,00
91,00
Dust of filter
174,00
2,70
543,00
30,00
: 22,00
! 30,00
174,00
800,00









Pyrolysis Plant Konigsborn 3/4 Lehmann
Analysis of heavy metals Nov. 1991
966

-------
                                    F/g. 23

Mineroloil-Hydrocarbons
Naphthalin
Acenaphthylen
Acenaphthen
Fluoren
1 Phenantriren
1 Anthracen
Fluoranthen
Pyren
Benz[a]anthracen
Chrysen
Benzo[b]fluoranthen
Benzo[k]fluoranthen
Benz[a]pyren
lndeno[l .2.3.ed]pyren
Dibenz[ah)anthraeen
Benzo[gh!]perylen
Sum.:
Feed Discharge Discharge
< 50 mm < 10 mm > 10 mm
3800,0 < 5,0 < 5,0
7,5 < 0,1 < 0,1
3,7 < 0,1 < 0,1
2,4 < 0,1 < 0,1
6,0 < 0,1 0,1
27,9 0,1 0,3
9,0 < 0,1 0,1
22,6 0,1 0,3
17,3 < 0,1 0,2
17,4 < 0,1 < 0,1
19,5 < 0,1 < 0,1
) 18,2 < 0,1 < 0,1
)
11,9 < 0,1 < 0,1
9,8 < 0,1 < 0,1
4,3 < 0,1 < 0,1
6,4 < 0,1 < 0,1
183,9 0,2 1,0
FD)/^\/S. Pyrolysis Plant Kbnigsborn 3/4 Lehmann
l=^rx^^Y Test wittl soil contaminated by mineraloil - Nov 1991
I WESTFALEN AG hydrocarbons ( Results in mg /kg )
967

-------
                                                                        Fig. 24
      component



     dust old filter
          new filter
         (May-June 1991)


     class I an.


     class I! an.


     class III an.


     CO


     SO2


     C org. (fofal)


     HCI


     HF


     NOx
         limit              measurement results
         (mg/m3 STP)       (nng/m3 STP)
                30        24/22/21/20/19/17
                10    ,    0,7/0,7/1,2/0,4/0,6/0,5/0,5/0,8/1,1
                 0,2       0,00011


                 1,0       0,00259


                 5,0       0,11382


               100        20/35/22/15/13/16


               100        62/26/15/23/39/16


                20        6/ 8/ 5/ 4/ 6/8


                50        16/17/15/14/12/13


                 2,0       0,8/0,6/0,5/0,7/0,6/0,6


         only measurement  140/144/151/161/171/154



                                    m3 STP calculated  on 11 %
RUHRKOHLE
WESTFALEN AG
Measurements  by TUV
January  1989  and  May - Juni  1991
Lehrnann
Nov.  1991

                                      968

-------
Rg. 25



Tetrachlordibenzodioxine
Pentachlordibenzodioxine
Hexachlordibenzodioxine
Heptachlordibenzodioxine
Octachlordibenzodioxin
Summ. Tetra- bis Octachlordibenzodiozine
Tetrachlordibenzoturane
Pentachlordibenzofurane
Hexachlordibenzofurane
Heptachlordibenzofurane
Octachlordibenzofuran
Summ. Tetra- bis Octachlordibenzoiurane
2,3,7,8-Tetrachlordibenzodioxin
1,2,3,7,8-Pentachlordibenzodioxin
1,2,3,4,7,8-Hexachlordibenzodioxin
1,2,3,6,7,8-Hexachiordibenzodioxin
1,2,3,7,8,9-Hexachlordibenzodioxin
1,2,3,4,6,7,8-Heptachlordibenzodioxin
2,3,7,8-Tetrachlordibenzofuran
1,2,3,7,8-Pentachlordibenzofuran
2,3,4,7,8-Pentachlordibenzofuran
1,2,3,4,7,8-Hexachiordibenzofuran
1,2,3,6,7,8-Hexachlordibenzofuran
1,2,3,7,8,9-Hexachlordibenzofuran
2,3,4,6,7,8-Hexachiordibenzofuran
1,2,3,4,6,7,8-Heptachlordibenzofuran
1,2,3,4,7,8,9-Heptachiordibenzof uran


Measurements by TUV
10.10.89
no/m3*)
0,24
0,14
0,12
0,24
0,36
1,10
0,10
0,21
0,12
0,20
0,08
0,71
n.n. **)
n.n.
n.n.
n.n.
n.n.
0,10
n.n.
0,02
0,01
. 0,02
0,02
n.n.
n.n.
0,20
n.n.
0.02
0,013
11.10.89
nq/m3*)
0,26
0,20
0,15
0,24
0,33
1,18
0,16
0,17
0,12
0,17
0,06
0,68
n.n.
n.n.
n.n.
n.n.
n.n.
0,12
0,01
0,02
0,01
0,02
0,02
n.n.
n.n.
0,17
n.n.
0.02
0.02
12.10.89
nq/m3*)
0,15
0,14
0,12
0,16
0,29
0,86
0,13
0,09
0,15
0,05
0,51
n.n.
n.n.
n.n.
n.n.
n.n.
0,10
n.n.
0,01
0,01
0,01
0,01
n.n.
n.n.
0,15
n.n.
0.01
0.01
18.04.90
nq/m3*") ,
0,29.
0,31
0,22
0.54
0,76
2,12
0,82
0,46
1,02
1,24
0,56
4,10
n.n.
0,02
0,01
0,02
0,01
0,29
0,03
0,08
0,05
0,14
0,11
n.n.
0,07
0,84
0,11
0.09
0,09
19.04.90
nq/m3*)
0,17
0,07
0,07
0,06
0,04
0,41
0,21
0,18
0,12
0,25
0,10
0,86
n.n.
n.n. '
n.n.
n.n.
n.n.
0,03
0,02
0,02
0,01
0,02
0,02
n.n.
0,01
0,18
0,02
0.02
0.02
19.04.90
na/m3*)
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
n.n.
	 by GfA
04.10.SO
ng/m3*)
0,02
n.n.
n.n.
0,09
0,21
0,32
0,13
0,09
0,08
<0,29
0,43 .
<0,003
<0,010
<0,010
<0,010
0,047
0,019
0,012
0,016
0,014
0,011
<0,007
<0,042
0,060
<0.005

0.014
ng/m3*)
n.n.
n.n.
n.n.
0,07
0,19
0,26
0.05
0,02
0,04
<0.19
0,17
<0.001
<0,003
<0,003
<0,003
0.042
0,018
0.006
0,008
0.006
<0,004
<0,004
<0,026
0,040
<0.005

0.008
•ibMoMri.utd.jAboisvoliimenlmNortmu.umd 1275 K. 1013 mbir) ~) Die Nicuwelspmnze fur dte ElnzsUsomwe betni9< 0.01 nevm1 rui.= nor evident
RUHRKOHLE
WESTFALEN AG
Contents of dioxines(PCDD) and furans (PCDF.) in the filtered gas
Lehmann
Nov. 1991
                                      Fig. 26



Tetrachiordibenzodioxine
Pentachlordibenzodioxine
Hexachlordibenzodioxine
Heptachlordibenzodioxine
Octachlordibenzodioxin
Summ. Tetra- bis Octachlordibenzodiozine
Tetrachlordibenzoturane
Pentachlordibenzofurane
Hexachlordibenzofurane
Heptachlordibenzofurane
Octachiordibenzofuran
Summ; Tetra- bis Octachlordibenzofurane
2,3,7,8-Tetrachlordibenzodioxin
1,2,3,7,8-Pentachlordibenzodioxin
1,2,3,4,7,8-Hexachlordibenzodioxin
1,2,3,6,7,8-Hexachlordibenzodioxin
1,2,3,7,8,9-Hexachlordibenzodioxin
1,2,3,4,6,7,8-Heptachlordibenzodioxin
.2,3,7,8-Tetrachlordibenzofuran
1,2,3,7,8-Pentachlordibenzofuran
2,3,4,7,8-Pentachlordibenzofuran
1,2,3,4,7,8-Hexachlordibenzofuran •
1,2,3,6,7,8-Hexachlordibenzofuran
1,2,3,7,8,9-Hexachlordibenzofuran
2,3,4,6,7,8-Hexachlordibenzofuran
1,2,3,4,6,7,8-Heptachlordibenzofuran


Measurements by GfA
10.10.90
na/m3*)
0,09
n.n.
n.n.
0,17
0,73
0,99
0,30
0,16
n.n.
0,04
<0,41
0,50
<0,005
<0,006
<0,121
<0,121
<0,121
0,102
0,039
0,009
0,021
<0,021
<0,021
<0,021
<0,083
0,039
<0,009
0.014
0,017
11.10.30
ng/m3*)
n.n.**)
n.n.
n.n.
0,09
0,24
0,33
0,09
0,10
n.n.
0,04
<0,22
0,23
<0,003
<0,004
<0,011
<0,011
<0,011
0,052
0,020
0,009
0,022
<0,021
<0,021
<0,021
<0,105
0,034
<0,010
0,008
0.015
12.10.90
nq/m3*)
0,03
n.n.
n.n.
n.n.
0,25
0,28
0,08
n.n.
0,02
<0,27
0,16
<0,006
<0,019
<0,062
<0,062
<0,062
<0,038
0,020
0,012
0,013
<0,020
<0,020
<0,020
<0,036
0,024
<0,009
0.006
0.010









































RUHRKOHLE
Contents of dioxines ( PCDD) and furans ( PCDF) in the filtered gas
during simulated failures
Lehmann
Nov. 1991
969

-------

970

-------
                        Fig.  28
971

-------
                                          Fig. 29
*•

•
conductor/
KW(n.DEVH181
EOX(CJ)
Anlimon (Sb)
Ari.n (Ar)
Barium (Ba)
Boryllium (Do)
Bio! (Pb)
Bor |B)
Cadmium (Cd)
Chrom (Cr) tola!
Eh«n (Fo)
Koball (Co)
upfor (Cu)
Mangan (Mn)
Nickel (N!)
utckillbor (Hg)
•tin (S»)
IbtrfAg)
hainumffl)
anadlum (VJ
nk (Zn)
uorld (f)
-World |d)
yanW* (CN) total
Irat (N03)
Irit (NO2)
loipot |PJ
Ifot (SO4)
j-J Pl/AV//->>
RUHRKOHLE
WESTFALEN AG
mgO2/
ma/I
mg/l
mg/l
mg/l
mg/l
mg/i
mg/l
mg/l
mg/l
mg/l
ma/I
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/1
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l 1
mg/l
aftor treatment
10,
445
3
< 0,00
< 0,
< 0,0
0,004
<• 0,
< 0,00
< 0,00
< 0,001
< 0,05
< 0,2
< 0,05
< 0,1
< 0,05
< 0,002
< 0,0005
0,004
0,001
< 0,005
< 0,05
: 0,1
: 0,005
0,7
1/4
8,1
0,04
0,5
0,06
0,19
385
Refilled soil
after 20 IT
0 - 0,3 m
1 	 54
1 < 15
< 0,00
1 < 0,05
1 < • 0,1
< 0,0
1 0,012
I 0,08
I < 0,1
< 0,00
< 0,00:
<'
< 0,00
-< 0,0;
< o,:
< 0,05
< 0,1
< 0,05
< 0,001
< 0,0005
0,021
< 0,001
jt 	 0,005
< 0,05
< 0,1
0,006
0,87
< 0,08
< 1
< 0,01
0,1!
: 0,01
: 0,1
255
onlhs
in c
0,3 - 2.0 m
_< 	 _1_
_< 	 0,00
_< 	 0,0
_< 	 0,
< 0,0
	 0,00
	 0,02
_< 	 0,_
< 0,00
< 0,002
_< 	 0,001
_< 	 0,05
_< 	 0,2
< 0.05
Jj 	 0,1
_< 	 0,05
_< 	 0,001
< 0,000i
0,001
_< 	 0,001
< 0.005
< 0,05
_< 	 0,1^
0.005
0,72
_< 	 0,08
	 1,8
	 0,02
	 	 0,27
<^ 	 0,01
	 0,1
' 	 157
depth
_<_ 	 ]S
_< 	 0,01
^ 	 0,001
_< 	 0,00:
_< 	 0,00'
_< 	 0,05
<_ 	 OJ_
_< 	 0,05
< 0,00
< 0,000;
0,002
< 0,001
_< 	 0,OS
_<_ 	 a_
< 0,005
	 u
< 0,08
< 1
	 0,02
0,14
< 0,01
	 0,12
KesuJLts of tests according to the
landfill directive (class I) of Northrhine Westfalia
(Eluat according to the German norm DEV S 4)
._< 	 0,0
< 0,00
_< 	 0,002
<
< 0,001
< 0,05
< 0,2
< 0,1
< 0,001
0,002
< 0,05
< 0,1
< 0,005
045
0,1
1,6
0,01
< 0,1
190


Lehmann
Nov. 1991
972

-------
                     Appendix 2-C



   Thermal Technology Case Studies
      Off-site Soil Treatment/Japan
973

-------
         Thermal Treatment of Contaminated Soil with Mercury
                          Takashi Ikeguchi
                   The Institute of Public Health
                          Tokyo 108 Japan
                                and
                           Sukehiro Gotoh
             National Institute for Environmental Study
                         Tsukuba 305 Japan
                              SUMMARY
     The site of former electro-chemical industry in residential area
 of Tokyo was found to be highly contaminated with mercury and lead
 when the industry stopped its operation  and dismantled facilities in
 order to move the factory to the-suburb and redevelop  the site in
 accordance with the urban planning of Tokyo Metropolitan Government.
 The mercury contamination was mostly limited within surface soil with
 average-  concentration  of  3.68 .% ( max. 15.6 %).  About 56,000 m3
 soils  with mercury level of above 2 mg/kg were considered  to be
 processed  before  redevelopment of  the  site.-    Underground
 containment with and/or without stabilization using with NaoS for
 slightly contaminated soil and thermal treatment. ( roasting ) for
 heavily contaminated soil were taken for  the remedial, technologies of
 the site.    About 840  mj ( ca.1260  tons  ) heavily contaminated soils
 with mercury were  railed to off-site  mercury  roasting plant to
 process and recover mercury.   Total cost of  soil roasting including
 pakaging and transportation was about  $374/ton soil.
1. INTRODUCTION
     A.    Site Description

          The  problem  site,  a former electro-chemical industry site
was situated  in dense residential area on the  bank of the Sumida
River in Arakawa Ward,  northeast part of Tokyo Metropolitan Area
where many small  industries and residential premises  have been
coexited ( see Figure 1 ).    The soils in the  site were  alluvium and
hence soft with a few  organic material  and composed of mainly .sllty
sand.    N  value of soil strata between D m and -21 m ranged from 0 to
4 and cohesive strength was 2-3 tons/m5.  Hydraulic  conductivity was
around 10"^ cm/sec in  silty sand  and  10~5-10~6 cm/sec in  silt
                              974

-------
      ARAKAWA WARD
               Asahi Electro-chemical
               Co.,Ltd.
Figure 1.     Location Map of the Former Asahi Electro-chemical
             Co.,  Ltd. Site
                          975

-------
                                                         0)
                                                        CO
                                                        JZ
                                                        •p
                                                         o>
                                                         o
                                                         01
                                                         o
                                                         S-
                                                        -o
                                                        -o
                                                         c
                                                         fO
                                                        -P

                                                         X
                                                         03

                                                         S_
                                                         en
976

-------
( see Figure 2 ).   Groundwater level  was high (i.e.  0.2-0.8 m  below
surface) and groundwater flow in horizontal direction could not be
recognized.


     B.   Site History

         The industry, Asahi Electro-chemical Co.,  Ltd.  began its
operation at this site in 1917 to produce primarily sodium hydroxide
and breaching powder as a leading industry'in Japan at that time.
Using hydrogen and chlorine, by-products of sodium hydroxide, they
also produced soap, margarine, hydrogen chloride, and  chlorinated
organic  compounds as well as another industrial materials.

         In 1955, the  company developed mercury-electrolysis process
to obtain high  quality sodium hydroxide to meet requirements of
chemical textile industry which had developed rapidly  after World_War
Two in Japan.    By  the time when  closed system was  introduced into
mercury-electrolysis process in  1966,  mercury had spilled out of
plant in the form of gas, wastewater and stains on various  materials.
As a result, the site of 20 ha  was widely contaminated with mercury
and lead.

          In the early 1970's,  Tokyo Metropolitan Government decided
to remove heavy chemical industry located  in residential area to
industrial area or another prefecture.  Asahi Electro-chemical Co.,
Ltd. was included  in this project.  After removing, the site was
planned to be used  for college, pubilic park, municipal wastewater
treatment  plant, and community complex.  'The plant was  closed in
March 1979.

          During dismantling the factory equipments in 1978, highly
contaminated soils  with mercury of max.  15.6 % ( 3.68 % mean ) were
detected under electiroysis vessel.   These soils with a volume of 840
m-* ( ca. 1,260  tons  ) were excavated and packed into  about 6,000
drums and stored on site.    Several alternatives were considered to
treat these highly contaminated soil and  finally these were railed to
off-site mercury roasting plant at Hokkaido to recover mercury in
1979 and 1980.


     C.    Initial Sampling and Analysis Results

          In 1977 the company analysed mercury and lead level of soil
core  samples  and  groundwater in the  site.   Results of  initial
analysis are shown  in  Figure 3 and Table 1.    Among 980 soil samples
at 188 points, maximum mercury concentration was 1,250 mg Hg/kg soil,
while 79 samples showed above  5 mg Hg/kg soil.  Form of mercury was
mostly sulphide  .   A 97 7o  of  groundwater sample  showed the
concentration of below 0.005 mg/1.   Only top soils (i.e.  50 cm below
surface) were severely contaminated and groundwater contamination
outside the site was not recognized.    Maximum lead concentration of
                              977

-------
                                                               
-------
    Table 1.    Initial  Sampling  Results of Soil  and Groundwater



(a)  Soil
Hg Concentration ( mg/kg ]
Depth ( m ) <5 5-25 25-100 100-500 500-1,000
0 109 48 25 4 1
0.5 132 40 11 5
1 172 12 2 2
2 94 2 2 - -
3 84 , 2
4 85 1
5 84 2
7-15 60 - - - - -
No. of
Sample 820 107 40 11 1
( % ) (83.7)^(10.9) (4.1) (1.1) (0.1)
(b) Groundwater
Hg Concentration ( mg/1 )
Depth ( m ) <0.005 >0.005
0 179 9
0.5 179 9
1 181 7
2 97 1
3 85 1
4 85 1
5 85 1
7-15 60
No. of Sample 951 29
( % ) (97.0) (3.0)
)
>1,000 No. of Sample
1 188
188
188
98
86
86
86
- 60
1 980
(0.1) .(100)


No. of Sample
188
188
188
98
86
86
86
60
981
(100)
                             979

-------
  soil  was 2,610 mg Pb/kg soil.     No further information on lead
  contaminaton was available.    Mercury in highly  contaminated soil
  which were  detected  under  the  two electrolysis vessel  during
  dismantling factory'was metal  mercury. Mean concentration of mercury
  in these soils were 3.68 % and mostly distributed between 1 and 7 %
  as shown in Figure 4.


      D.    Technology Selection

           There was no statutory criteria for clean-up decision of
 mercury contaminated soil at that time.  Therefore  Tokyo Metropolitan
 bovernment sampled soils at  71 non-contaminated area through Tokyo

 SLf   nyn?   ^eoCUrytr fc«  Sh°W that most  of fche data distributed
 between 0.02 and 2 mg Hg/kg soil.  Hence they took  2  mg Hg/kg soil as
 a maximum background  level of  non-contaminated area and critera for
 remedial action.


 «f.=K-T   f.Vnd,e?gj':?.UIld   containment   with  or  without
 stabilization/solidification has  been  considered so often as a basic
 treatment method for contaminated soil in Japan.     However,  it was
 ™n?«m*   H S° k.eiep !ihe  11evel Of mercury in solution of highly
 contaminated soil under  the certain  level even if solidified with
 •2  f11? VC12-   Moreover recovering  mercury from these  soils  were
 judged to be economically feasible and preferable from the view point
 of resource conservation.  Fot these reasons,  remedial technologies
 were selected according as the  level of contamination as follows:

     (l)Thermal treatment  (  i.e. roasting )   for  highly
        contaminated soil  with  above 10 mg Hg/kg soil.

     (2) Underground containment  after immobilization for highly
        contaninatead soil with above 10 mg Hg/kg soil.

     (3) Underground containment  without any pre-treatment for
        moderately  contaminated soil with  2-10  mg Hg/kg soil.
2..  THERMAL TREATMENT TECHNOLOGY
„«  f   ,- Im-t.:Lally>  Ashahi Electro-chemical Co., Ltd. planned to
construct on-site rotary kiln with a total  throughput of 0.3  t/day to
roast  contaminated soil  using light oil as  a  fuel.    Due to the
strong opposition  of  nearby residents and immature  technology to
£??££ Saseous mercury under .the level of regulation,  they had to
withdraw this plan and decided to transport contaminated soils to
mercury recovery plant in  mountainous  site of  Hokkaido,  about  1,000
km north of Tokyo.  This plant was owned by mercury refining company
,  Nomura  Kosan Co.  and was used for treatment  of mercury bearing
waste to recover mercury at that time.
                             980

-------
                     3.68  x + o
    2o   x + 3a
     50
 0)

 J3

 E
 o.
 E
 (O
     30
     20
      10
n=  300


x=  3.68%


0=  2.37%
                                                   I
         0-1.0  2.0-3.0  4.0-5.0  6.0-7.0  8.0-9.0    10.0<

           1.0-^2.0  3.0-4.0  5.0-6.0  7.0-8.0 9.0-10.0



                  Mercury Concentration ( % )
Figure  4.     Mercury  Distribution  in  Highly Contaminated Soil
                             981

-------
           This plant is equipped with vertical multistage roasting
 furnace ( called as  Herreshoff Furnace )  with   an annual  gross
 teatment capacity of 3,600 tons.  The plant is primarily composed of
 J°SSte*'. ^denser to recover mercury  and flue gas  cleaning devices.
 Schematic flow diagram of this plant is shown in Figure 5.

 nf  «nn nn5fe^cury7contained  waste  or  soil is  roasted at temperature
 of  600-800 C  using heavy  oil.   Volatiled  mercury  .in flue  gas
 condensed on the inner wall of condenser subsequent to dust removal
 equipment.   Crude  mercury  is   recovered from 'soot at constant
 S (fcTJf      Defined into commercial grade groducts with a purity of
 yy.yy  /8 or  more either  by  thw wet  or the  vaccume distilling.    Trace
 amount of mercury and acid gas components in flue gas are  removed by
 adsorption and neutralization.   Slag from  roasting furnance  are
 disposed of at on-site secure landfill and sludge from flue  gas
 treatment processes,  dust  collected by cyclone, residue from mercury
 recovery are returned  into roaster.        '    ••••••

          Recently this plant has been  partly expanded to recover
 mercury and other metals from used dry cell generated from  individual
 nome.                                            • ., -  .
 3^ DEMONSTRATION RESULTS
         About 6,000 drums packed with highly contaminated soil were
 transported everyday except Sunday during March and July in   1980
 Cost of  soil  roasting  was  65,000  yen/ton  (  $300/ton ),  and
 transportation cost including  package was  16,000 yen/ton (  $747ton )
 Another mercury bearing waste were processed simultaneously at this
 plant,  hence technical data  of soil roasting  were not  available
 separately.
                       O                '
         About 54,500  m*  soils contaminated with mercury of above 2
 mgHg/kg soil  were contained in  on-site underground pit after
 immobilization by Na£S for the soils  with mercury level of above 10
 mg  Hg/kg  soil.       Excavation  and  immobilization  followed by
underground containment were conducted from December 1983  to August
 f-   A  »-   minimize  environmental pollution during clean-up action,
air, dust,  effluent,  noise,  and  vibration  were monitored around the
site and if any environmental  deterioration was recognized, clean-up
had to be stopped or modified.     Groundwater and rainfall inside
site were treated  in the  manner described  in Figure  6.   Groundwater
has been monitored at 4 monitoring wells of 20 m depth  just  outside
tne site and  no groundwater  pollution has been  reported  so  far.
                             982

-------
 L.
 
 i-
 o
                                                                                   OJ
                                                                                   t/1
                                                                                   0)
                                                                                  C£
                       0)
                       U
                       CO
                       c'
                                                                -p
                                                                 o
                                                                 o
                                                                1/1
 >>  s-
 (-  0)
 =J  >
.00
 k.  O
 O)  OJ
S  S_
V.
3
u
t-
0)
                                                                                                                 o
                                                                                                                 KJ
CP
c
•r-
-P
10
Id
o
Di
                                                                                                                  O
                                                                                                                  s_
                                                                                                                  O!
                                                                         13
                                                                          C
                                                                          3
                                                                          O
                                                                          I. 01
                                                                          en CTI
                                                                          J- «3
                                                                          0) V.
                                                                         •O O
                                                                          C -P
                                       
           ra
           -P
           c
           o
                                                           983

-------
iD


1

L '
u 	 Q-) —

, — . 	

\ - E
N tJ o
i ,_; CD
3 10
o>
o
C3


	 '/

/ 	 © 	
/
                                                                                          >>
                                                                                          -P
                                                                                          lO
                                                                                          C
                                                                                          OJ

                                                                                         +J
                                                                                          (0
                                                                                          cu
                                                                                          S-
                                                                                         O)
                                                                                         (1)
                                                                                        •P
                                                                                         M
                                                                                         <0
                                                                                         O)
                                                                                         V)
                                                                                         I
01
s_
3
                                    984

-------
4..  CONCLUSION

          Soil  pollution in Japan was  focused on the agriculture
field so far, however with urban redevelopment at former industry
site or national  institution,  soil pollution at these site has  been
reported  recently.    In  Tokyo,  for  example,  several  soil
contaminations  have been identified at these sites  including the case
of this paper.     Major pollutants are  heavy metals  such as Hg, Pb,
Cd and restoration has been completed at all cases  identified.

          In the light of these facts,  Environment Agency has  been
conducting preliminary study of soil contamination resulted  from
human activities  at residential  area and  clean-up criteria  of these
soils as  well  as clean-up technologies.   Such  criteria for  soil
clean-up has been set up at some local government already.

          Remedial technologies used so far are disposal at secure
landfill  site  or underground containment  with or  without
immobilization  in principle.   These  technologies are however not
detoxification  method  but mere isolation which renders  future
potential hazard  and therefore development  of  both on-site ^and orr-
site detoxification technology of contaminated soil are required now.

          Roasting of contaminated soil  and mercury recovery of this
paper are the only .case of detoxification  that has  been ever  tried in
Japan.      This  example  was  off-site thermal  treatment  and
economically feasible because the mercury  level in soil  was extremely
high,  as  high as mercury  ores.   On-site  roasting, however is
technically and possible as  the  company originally planned,  provided
that nearby residents   accept  such facility.


ACKNOWLEDGMENT

          Authors thank the  Department  of Environment Control,  Tokyo
 Metropolitan  Government  and Arakawa Ward  Office  for  offering
 informations .
 REFERENCES

 (1) Asahi Electro-chemical  Co.,  Inc., Clean-up Plan for Contaminated
     Soil of Former Industrial Site ( in Japanese  ), October IVbJ.

 (2) Hazama-gumi Co.,  Inc.,  Completion of Site Clean-up of Asahi
     Electro-chemical Industry ( in Japanese ), March lytfD.

 (3) Nissaku Co., Inc., Research  on Gas Generation at Restored Site
     of Asahi  Electro-chemical Industry  ( in Japanese ), February 1985.

 (4) Clean Japan Center,  Mercury  and Other Metals  from Used  Dry
     Battery  Cells - Recycling Demonstration Plant -,   February 1V8/.
                              985

-------

-------
                           Appendix 2-D




        Thermal Technology Case Studies
Electric Infrared Incineration, United States
       987

-------
  United States
  Environmental Protection
  Agency
  EPA/540/S5-88/002
  January 1989
   SUPERFUND INNOVATIVE
   TECHNOLOGY EVALUATION
   Technology  Demonstration
   Summary

   Shirco Electric  Infrared
   Incineration  System  at  the
   Peak Oil Superfund  Site
  Under the auspices  of the
 Superfund Innovative Technology
 Evaluation or SITE Program, a critical
 assessment Is made  of the
 performance of  the transportable
 Shirco Infrared Thermal Destruction
 System* during three separate test
 runs at an operating feedrate of 100
 tons per day. The unit was operated
 as part  of an emergency cleanup
 action at the Peak Oil Superfund site
 in Brandon, Florida. The report
 includes a process description of the
 unit, unit operations data and  a
 discussion  of  unit operations
 problems, sampling and analytical
 procedures and data, and an overall
 performance and  cost evaluation of
 the system.
  The results show  that  the unit
 achieved destruction and removal
 efficiencies (DREs)  of polychlo-
 rinated biphenyls  (PCBs) exceeding
 99.99% and destruction efficiencies
 (DEs) of PCBs ranging from 83.15%
to 99.88%. Acid gas  removal
efficiencies were consistently
greater  than 99%. Participate
emissions ranged  from 171 to 358
mg/dscm, exceeding  180 mg/dscm
during two  of the four  tests. The
 Extraction Procedure (EP) Toxicity
 Test on the furnace ash exceeded
 the RCRA EP Toxicity Characteristic
 standard for lead. Small quantities of
 tetrachlorodibenzofuran (TCDF) were
 detected in one of the four stack gas
 samples. Also detected were low
 levels of some semivolatlle organics
 and  a broader range of volatile
 organics, which can Ibe considered
 products  of incomplete combustion
 (PICs).  Ambient  air  monitoring
 stations detected quantities of PCBs,
 which appear to be caused  by the
 transport of ash from the ash pad to
 the ash storage area. Waste feed and
 ash samples were  not mutagenic
 according to the  standard Ames
 Salmonella mutagenicity assay. Unit
 costs  are estimated to range from
 $196  to  $795  per ton with  a
 normalized cost per ton of $425 for
 the Peak Oil cleanup.
  This Summary was developed by
 EPA's  Risk Reduction Engineering
 Laboratory, Cincinnati, OH,  to
 announce  key findings of the SITE
Program demonstration that is  fully
documented in three separate reports
(see ordering information at back).
               988

-------
Introduction
  The SITE Program demonstration test
of the Shirco infrared incineration system
was  conducted  from July 1,  1987  to
August  4,  1987  at the  Peak  Oil
Superfund site in Brandon, Florida during
a removal action by EPA Region IV. The
Region  had contracted  with  Haztech,
Inc., an  emergency removal  cleanup
contractor,  to incinerate approximately
7,000  tons of  waste oil  sludge
contaminated with PCBs and lead after
determining  that  high  temperature
thermal destruction of the nonrecyclable
sludge was capable  of  destroying the
PCBs  in  a  cost-effective  and
environmentally sound  manner. Metals
that concentrated  in  the ash residue
would  be dealt with after the thermal
destruction of the sludge. The removal
action offered an ideal opportunity for the
SITE  program  to  obtain   specific
operating,  design,  analytical,  and cost
information to evaluate the performance
of  the  unit under  actual  operating
conditions. Also,  the  SITE  program
studied  the feasibility of utilizing the
Shirco  transportable  infrared incinerator
as a viable hazardous waste  treatment
 system  at other sites  throughout the
 country. To  this  end, specific  test
 objectives of the Shirco system were:

 • To determine the system's destruction
   and  removal efficiency  (ORE)  for
   PCBs.
 • To  report  the  unit's  ability  to
   decontaminate the solid material being
   processed  and  to  determine  the
   destruction  efficiency (DE)  for PCBs
   based on the PCB  content  of  the
   furnace ash.
  • To evaluate  the ability of the unit and
   its   associated  air  pollution
   control/scrubber  system to  limit
   hydrochloric acid  and  particulate
   emissions.
  • To determine  whether heavy metals
    contaminants in  the  waste feed are
    chemically bonded  or fixated to the
    ash residue  by the process.
  • To determine the effect of the thermal
    destruction  process in  producing
    combustion  byproducts or products of
    incomplete combustion (PICs).
  • To  determine the  impact  of  the unit
    operation on ambient air quality and
    potential mutagenic exposure.
  • To provide  unit cost data for  effective
    development of  a  cost/economic
    analysis for the unit.
• To  document  the  mechanical
  operations  history of the unit and
  analyze and provide potential solutions
  to chronic mechanical problems.

Facility and Process
Description
  Solid waste processed at the Peak Oil
site was incinerated in  a  transportable
infrared  incinerator,  designed and
manufactured  by  Shirco  Infrared
Systems, Inc.  of Dallas,  Texas and
operated by  Haztech,  Inc.  of Decatur,
Georgia. The  overall incineration unit
consists of a waste preparation  system
and  weigh  hopper, infrared  primary
combustion  chamber, supplemental
propane-fired secondary  combustion
chamber  (afterburner),  emergency
bypass stack, venturi/scrubber system,
exhaust system, and data collection and
control  systems,  all  mounted  on
transportable  trailers. The  system
process  flow and the  overall test site
layout are presented  schematically  in
Figure 1.
   Solid waste feed material is processed
 by  waste  preparation  equipment
designed to reduce the waste  to  the
 consistency and particle sizes  suitable
 for processing by  the incinerator. After
 transfer from  the waste  preparation
 equipment,  the solid waste  feed  is
 weighed  and  conveyed  to a  hopper
 mounted over the furnace conveyor belt.
 A feed chute on the hopper distributes
 the  material  across the width of  the
 conveyor belt.  The feed  hopper screw
 rate and the conveyor belt speed rate are
 used  to control the  feedrate  and bed
 depth.
   The incinerator  conveyor,  a tightly
 woven wire belt, moves the solid waste
 feed  material through  the  primary
 combustion chamber where it is brought
 to combustion temperatures by  infrared
 heating  elements.  Rotary rakes  or
 cakebreakers gently  stir the material to
 ensure adequate mixing, exposure to the
 chamber  environment, and complete
  combustion. When the combusted feed
 or ash reaches the discharge end of the
  incinerator, it is cooled with a water spray
  and then  is  discharged  by  a screw
  auger/conveyor to an ash hopper.
   The combustion  air to the incinerator is
  supplied through a series of overfire air
  ports located  at various locations along
  the incinerator chamber; combustion air
  flows  countercurrent to the  conveyed
  waste feed material.
    Exhaust gas  exits .the  primary
  combustion chamber and flows into the
  secondary combustion chamber where
propane-fired  burners  combust any
residual organics present in  the exhaust
gas. The secondary combustion chamber
burners are set to burn at  a  prede-
termined temperature.  Secondary air  is
supplied to ensure adequate excess
oxygen levels for complete  combustion.
Exhaust  gas  from  the  secondary
combustion chamber is quenched by a
water-fed  venturi/scrubber  to remove
particulate  matter and acid gases; the
exhaust gas is then transferred  to the
exhaust stack  by an induced draft fan,
and finally discharged to the  atmosphere.
  The  main  unit  controls and data
collection  indicators comprising the data
collection  and control system are housed
in a specially designed van.
  An  emergency  bypass  stack  is
mounted in the system directly upstream
of the venturi/scrubber for the diversion
of hot process gases  under emergency
shutdown conditions.

 Results and Discussion
  A  detailed summary of  the SITE
 demonstration test  results is presented in
 Table 1.  Based on the  test objectives
 outlined in the Introduction,  the following
 results and conclusions were obtained.

 PCB Destructton and Removal
 Efficiency
   PCBs were analyzed in the solid waste
 feed, furnace ash, scrubber  effluent
 solids, stack gas, scrubber liquid effluent,
 and  scrubber water inlet.  The  ORE
 calculation for PCBs  is based  on  the
 following:
         ORE =
Win-Wout

   W.  ,
     in
                           X 100
 where: Win  = mass rate of PCBs fed to
               incinerator

        wout = mass emission  rate  of
               PCBs in stack gas

 The unit achieved a ORE  for PCBs of
 99.99%.
    It should  be noted  that the unit was
 operated  to produce  an  ash  that
 contained 1  ppm  or less  of  PCB. The
  PCB concentration in  the waste  feed to
 the unit varied from 5.85 to 3.48 ppm
 during the tests. These  low  PCB
  concentrations in the waste feed were the
  result of  mixing the original  oily waste
  having up to 100 ppm of PCBs with the
  PCB-free surrounding  soil,  lime,  and
  sand so that the resulting  material could
                                                        989

-------
                                    Feed Hopper
                                    & Feed Module
                                                            Primary
                                                       Combustion Chamber
                                                        D
                                                                                     Combustion
                                                                                     Air Blower
Material
Stockpile
                          Propane Fuel
                          Forced Air
                           Blower
                                             Secondary
                                        Combustion Chambe
                                                                                         emergency
                                                                                          Bypass
                                                                                           Stack  '
                                                                                                        Fresh
                                                                                                        Water
                                                                                                         to
                                                                                                        Unit
                                           Chemical Chevron
                                           Recycle  Recycle
                                            Pumps   Pumps
                                                         Wate
                                                       Conditioner
                                                                                          Activated
                                                                                           Carbon
                                                                                            Filter
                                                                                     Slowdown 'Water to POTW
  Flgun 1.   Peak Oil Incinerator  Unit.
be handled and processed  as  a solid
waste. It was not possible to calculate the
ORE beyond  two  decimal  places
because  of  the  detection   limits
associated with the analytical  procedures
employed.

Decontamination of Solid Waste
and Destruction Efficiency
  Residual PCBs in the furnace ash were
below the  1 ppm  operating standard,
ranging from 0.007 ppm on August  1 to
                           0.900  ppm on  August 3.  DE  was
                           determined by the formula
                                    DE
W.  - w
  m    out
   W.
                                                    X 100
                           where: W;n  =  mass rate of PCBs fed to
                                         incinerator

                                 Wout =  mass  rate of PCBs  in
                                         stack gas, furnace ash,
                                         and scrubber effluent
A basis for calculating DE was based on
the PCB concentrations in the waste feed
and the furnace ash. The DE or removal
of trie PCBs from the waste feed ranged
from 99.88 wt% (August  1)  to 83 15 wt%
(August 3).

Acid Gas Removal
  Measured  HCI emission  rates ranged
from less than 0.8 to 8.6 g/hr. Since the
chlorine concentration in  the solid waste
feed was below the 0.1% detection limit,
it was impossible to determine actual HCI
removal efficiency.  However,  QO2
                                                      990

-------
Table 1.  Site Demonstration Test Results Summary
                                   811/87
                                              8/2/87
                                                        8/3/87
                                                                  8/4/87
Waste Feed Characteristics
Moisture, wt. %
Ash, wt. %
HHV, Btu/lb
PCB, ppm
Pb, ppm
Chlorine, ppm
Sulfur, ppm
Chlorine (as HCI), kg/hr
Sulfur (as SO2), kg/hr
EP Tox (Pb), mg/L ppm
TCLP (Pb), mg/L, ppm

16.63
69.77
2064
5.850
5900
<7000
25300
<5
200
27.00
8.60

16.06
69.80
T639
3.850
4900
3
>99.9

3287
99.99
93.77
7836
79
7887
>3
^99.7

3626
99.99
83.75
7922
78
7889
>3
>99.9

3600
99.99
84.48
7885
79
7907
>3
>99.9
 emissions were less than 1100 g/hr, with
 an  average 149  kg/hr SOa feedrate
 giving  an average removal  of  SOa  in
 excess of 99%. SOa is more difficult  to
 remove than HCI  in a caustic scrubber,
 and  the  tests  show that  HCI  removal
 should be in excess  of the  99%
 determined for SOa removal.

 Particulate Emissions
   The  parjiculate emissions during the
 first day were 358 mg/dscm.  The unit
 was   cleaned   and   mechanical
 adjustments were made resulting  in  an
 emission rate of 211 mg/dscm during the
 second day. The emissions during the
 third day were 172 mg/dscm (average of
 duplicate measurements).  These values
 exceeded  the RCRA standards during
 two of  the four  sampling  periods.
  Particulate  emissions were  about  60%
  lead, when analyses of all samples were
  averaged.
     Leaching Characteristics
       The solid waste feed, furnace ash, and
     scrubber effluent solids were subjected
     to the EP Toxicity  and proposed TCLP
     tests  to  evaluate  the toxicity
     characteristics of these materials.
       The EP Toxicity and the TCLP data
     present  a contradictory picture regarding
     leaching of metals.  The EP Toxicity data
     did  not indicate  that  the  process
     "encapsulates" 'or ties up heavy metals
     (lead) in the ash to prevent leaching. The
     EP Toxicity data show that  lead content
     in the ash was 30 ppm and exceeded the
     5 ppm  toxicity characteristic standard.
     The measured lead content of leachates
     for feed material and  ash are almost
     equal,  indicating that  the  process
     appears  not  to  affect leachfng
     characteristics for lead.
        In contrast to the EP Toxicity  data, the
     TCLP data show that  the lead content for
     both the feed and ash were  less than the
     proposed toxicity characteristic standard
                                 of 5 ppm. Measured lead concentrations
                                 were an order of magnitude lower in the
                                 TCLP leachate (about 2 ppm compared
                                 to about 30 ppm for EP Toxicity).
                                    The  significant  differences  in  results
                                 from these two  analytical  techniques
                                 have been documented  in a recent Oak
                                 Ridge National Laboratory report  (ORNL,
                                 "Leaching  of Metals  from  Alkaline
                                 Wastes by  Municipal  Waste  Leachate,"
                                 ORNL/TMr11050, March,  1987).  It
                                 appears  that the  differences in the  test
                                 procedures  and alkalinity of  the matrix
                                 provide  a  difference in  the  pH
                                 environment that is sufficient to affect the
                                 solubility  and  leachability  of  heavy
                                  metals, particularly lead.


                                  Products of Incomplete
                                  Combustion

                                    Small  quantities  of  products   of
                                  incomplete  combustion (PICs)  were
                                  identified in the  sampled streams from
                                                          991

-------
  the unit.  No  polychlorinated  dibenzo-
  dioxins (PCDDs)  or polychlorinated
  dibenzofurans (PCDFs) were identified in
  any of the  sampled streams  above
  detection  limits with  the  exception  of
  trace  quantities   (2.1   ng)  of
  tetrachlorodibenzofuran (TCDF) found in
  the slack gas sampled on August 2.
    Low  levels  of some  semivolatile
  organic compounds were identified in all
  streams.  These compounds were
  primarily phthalates, which may be the
  result  of  contamination  from plastic
  components in the process, sampling
  oquipment, or laboratory  apparatus.
  Other semivolatile compounds  included
  aromatic, polyaromatic, and chlorinated
  aromatic hydrocarbons. Low levels of
  pyrene,   chrysene,   anthenes,
  naphthalenes,  and chlorinated  benzene
  were identified in the waste feed stream;
  although possible PICs, their presence
  must  be discounted  to some extent,
  because they were originally introduced
  into the unit with the waste feed.
   Low concentrations of volatile  organics
 were measured in the stack gas and
 included   halogenated  methanes,
 chlorinated organics, and aromatic
 hydrocarbons including BTX compounds.
 No volatile organics were identified in the
 water streams. Low  levels (ppb)  of
 chlorinated hydrocarbons  and BTX
 compounds were  measured in  all solid
 streams. Low levels of  BTX compounds,
 carbon  disulfide, chloroform,  ditri-
 chlorofluoromethane,   and  trichloro-
 fluoromethane, dichloroethane,  and
 trichloroethane, and methylene  chloride
 were identified  in  the waste feed.
 Mothylono  chloride, a  solvent  used
 during testing,  was also  detected in
 laboratory  and field  blanks.  These
 compounds, although  possible PICs,
 must also be discounted to  some extent
 based  on their introduction  to the unit
 from an external source and because of
 possible contamination.

 AmbJent Air Sampling and
 Mutagenlc Testing
  Ambient air monitoring stations placed
 upwind and  downwind of the Shirco unit
 were designed  to collect airborne  PCB
 contaminants. Based  on the downwind
 sampler data, it appears  that the Peak Oil
 site boundaries limited the location of the
 downwind sampler  to an area that was
 significantly  exposed to fugitive
 omissions during the transport  of ash
 from the ash  pad to the ash storage area.
  Samples of the  waste feed and ash
wore collected  on August 2  and
forwarded to the EPA   Health  Effects
  Laboratory,  Research  Triangle  Park,
  North Carolina for mutagenic testing. The
  results  of  these tests indicate  that
  although the samples contain  hazardous
  contaminants, they are not mutagenic
  based on the standard Ames Salmonella
  mutagenicity assay.

  Cost/Economic Analysis
    Several cost scenarios examined were
  based on  a model  for a  Shirco  unit
  operation  equivalent in   processing
  capacity to the unit that operated at Peak
  Oil,  and on  cost  data available  from
  Shirco and other sources. The  economic
  analysis concludes that in using currently
  available Shirco transportable infrared
  incineration systems,  commercial
  incineration  costs  will range   from  an
  estimated $196  per ton for a Shirco unit
  operation at an  80% on-stream capacity
  factor to an  estimated $795 per ton for
  the operation at the Peak Oil  site at  a
  19%  on-stream  capacity factor. A
  normalized total  cost  per ton   of $425
  represents a more realistic interpretation
  of the costs accrued  to the  Peak Oil
  cleanup action  based  on  a 37%  on-
  stream capacity factor.

  Unit Problems
   A  review  of  the  Haztech,  EPA
 Technical Assistance  Team  (TAT), and
 EPA logbooks and progress reports, plus
 discussions  with  unit and   project
 personnel, provided  a summary of
 mechanical  and operating problems
 encountered in this  first application  of a
 full-scale   commercial   Shirco
 incineration system  at a Superfund  site.
 These problems  were categorized  by
 unit operating sections, and a profile of
 the major problem areas within  the unit
 were defined and analyzed to ascertain
 the reasons for and possible solutions to
 these specific operational difficulties. The
 review revealed  that materials  handling
 and emissions control  were the most
 significant problem  areas  affecting
 operation  of  the unit.  Prior  to  the
 operation of such  a  unit,  extensive
 pretest analysis should be conducted on
 the  waste1 feed   matrix.   The
 characteristics of the feed, including the
 nature  of  contaminants plus  the feed's
 effect on incineration system  chemistry,
 must  be defined  to allow appropriate
 assembly of the  unit. The unit must be
 equipped  with  the  proper  feed
 preparation  .system   and   materials
 handling  capabilities  and   adequate
emissions  control , capacity  and
effectiveness.  At  the Peak Oil site,  the
solidified  sludge  feed continually
 agglomerated,  clogged,  bridged,  and
 jammed  feed preparation and  handling
 equipment.  The  high levels  of lead
 contaminant and the excessive carryover
 of calcium and  magnesium salts were a
 continuous  source of  problems  for the
 emissions control system, which  had
 difficulty in meeting  stack  emissions
 criteria.

 Conclusions and
 Recommendations
  Based on  the  above  data and
discussions,  the following conclusions
and  recommendations can  be  made
concerning  the  operation  and
performance of  the transportable  Shirco
infrared thermal destruction system.

  I.The unit achieved DREs  of  PCBs
   greater than  99.99%. Detection limits
   were  used  for this calculation so
   actual DREs were greater.

  2. The unit  achieved DEs of  PCBs
   ranging  from 83.15 to 99.88%! The
   unit was operated to produce an ash
   that contained 1 ppm or less of PCB.

  3.Acid gas removal efficiencies were
   consistently greater than  99%. ,
   Particulate  emissions  during two
   days of  testing were 358 mg/dscm
   and 211 mg/dscm,  which contained
   60%  lead.  The unit's  emissions
   control  system  experienced
   particulate removal  problems due to
   a combination  of  excessive  fines
   carryover from  the waste feed matrix
   and scrubber-washer and an overall
   emissions control system design that
   was not able  to operate efficiently at
   abnormally high particulate loadings.
   As a result, two of the four samples
   taken  exceeded the 180  mg/dscm
   RCRA standard.
    Pretest analysis of the waste feed
   and its combustion and  emissions
   control chemistry and  mechanisms
   must  be performed  to  identify
   potential emissions control problems.
   A  more   flexible  and  adaptable
   emissions control system  should be
   developed that  can  respond  to and
   control a wider range  of particulate
   and stack gas flows.

 4. The  furnace ash failed  to meet the
   toxicity characteristic  standard for
   lead  for  the  EP  Toxicity   Test
   Procedure. Although  the ash passed
  the similar standard for the proposed
  TCLP,  its failure under  EP  Tox
  indicates  that  the  unit  did  not
  immobilize lead in the ash product.
                                                       992

-------
S.Small  quantities  of PICs were
  identified in the sampled  streams
  from the unit. In  addition  to trace
  quantities of TCDF on one  sample,
  low  levels   of  semivolatile
  compounds, including  aromatic,
  polyaromatic,  and  chlorinated
  aromatic  hydrocarbons were
  identified.  Low  concentrations of a
:  broader range of  volatiles  including
,  halogenated methane,  chlorinated
  organics, and BTX compounds were
  also identified.
6. Ambient air  monitoring  stations
  detected quantities of  PCBs, which
  appear to  be caused  by the wind
  transport of ash  resulting from the
  nearby roadway.  Waste feed  and
  ash samples  were  not  mutagenic
  based  on the  standard  Ames
  Salmonella mutagenicity assay.

7. Overall costs  ranged from $196 per
  ton with the unit operating at an 80%
  on-stream  capacity (292 days per
  year)  to $795 per ton  with  the unit
  operating  at a  19%  on-stream
  capacity. (70 days per year).  A
  normalized cost per ton for the Peak
  Oil cleanup was estimated at $425.

8.ln  addition  to  the  particulate
  emissions control system problems,
  waste feed  handling and materials
  handling problems  consistently
  affected the unit's ability  to treat the
  waste  feed  at design  capacity.
  Pretest  analysis of the  waste  feed
  and its handling characteristics must
  be performed to identify  and design
  for any potential materials handling
  or  feeding problems that the waste
  matrix may present at a specific  site.
                                                       993

-------
     The EPA Project Manager. Howard Wall, is with the Risk Reduction Enaineerina
      Laboratory, Cincinnati, OH 45268 (see below).
     The complete report consists of two volumes, entitled "Technology Evaluation
      Report, SITE  Program  Demonstration  Test, Shirco Infrared Incineration
      System, Peak Oil, Brandon, Florida:"
          'Volume I"  (Order No.  PB 89-125 991fAS; Cost: $21.95, subject to
          change) discusses the results of the SITE demonstration
          'Volume II" (Order No. PB 89-116  024/AS; Cost: $42.95. subject to
          change) contains the technical operating data  logs, the sampling and
          analytical report, and the quality assurance project plan/test plan
     These two reports will be available only from:
             National Technical Information Service
             5285 Port Royal Road
             Springfield, VA 22161
             Telephone:  703-487-4650
    A related report, entitled "Applications Analysis Report: Shirco Infrared Thermal
      Destruction  System," which discusses application  and costs,  is  under
      development.
    The EPA Project Manager can be contacted at:
             Risk Reduction Engineering Laboratory
             U.S. Environmental Protection Agency
             Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
      BULK RATE
POSTAGE & FEES PAID
         EPA
  PERMIT No. G-35
Official Business
Penalty for Private Use $300

EPA/540/S5-88/002
                                                        994

-------
                                     Appendix 3-^A

Stabilization/Solidification Technology Case Studies
In Situ Lime Stabilization (EIF Ecology), and Petrifix
                           Process (TREDI), France
                 995

-------
AGENCE  NATIONALS  POUR  LA  RECUPERATION ET  L-ELIMINATION  DES  DECHETS
      Deparcement INDUSTRIE
      RG/BP/MFB/OTAN/1
                                                  NOVEMBRE 1988
                   EVALUATION OF SOLIDIFICATION-STABILIZATION PROCESSES
      I - OBJECTIVES OF THE PROJECT

         ^  The objective of the present study is  to evaluate, some years after
      their application in France, for the treatment of different types of
      contaminated sites, the efficiency of solidification stabilization techniques.


      IX - SAMPLING AND ANALYSIS PROGRAMS

           1 - Sampling

           In order, to get significant  but  rather simple evaluation of the treated
     sites, at was decided to  carry out sampling procedure by digging trenches in
      the treated material by using a backhoe.  This method was prefered to drilled
     boreholes because it is easy to carry  out and moreover it allows easy visual
     observations and can give wide sections showing the treated material,  its
     contacts with the surrounding soil  and possible heterogeneousnesses.

           For every  site,  tnree 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 upper sample, may be considered  as
     representative of the  treated material in contact with biosphere  conditions ;
     freezing,, leaching by infiltrated water,  the middle sample may be considered
     as representative of the  average treated material and lower sample  would give
     an estimation of possible releases of contaminants from the treated material.
     However,  on  the  field, it appeared sometimes that the material underlaying
     the- treated  area  remained much or less contaminated by the orignal pollutant.
     In such  cases  the corresponding sample was not considered as significant.
                                           996

-------
      2 - Analysis program

      For every site it was made three average samples corresponding to the
three sampling levels by mixing the corresponding samples  taken on  the site.
The specific analysis performed on samples included :

      - measuring of physical properties

            . water content
            . permeability
            . compressive strength

      - leaching tests

      In France, at the present time, there are no  specific standardized tests
for the 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 industrial waste landfills. The main features
         of  this test are  :

             - extraction solvent  : demineralized water saturated with C02 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 solutionfor analysis after 16 hours of agitation.

          This test  was choosen although it is going to be replaced as standard
          for waste  acceptation in landfills by a similar one excluding the
          saturation of solvent  extraction by C02 because it has been performed
          for samples  taken from two  of  the sites considered in the present
          study  at  the  time of their  treatment, thus allowing more significant
          comparisons.  For  every sample  two successive extractions have been
          carried out.


       D) In addition,  in order to take  in account  the specific characteristics
          of solidification techniques it  was decided  to perform  the new test
          which is now prepared to be later standardized  for evaluation of
          solidified material.
                                       997

-------
     This  test called oedometric pressure leaching test is based on  the

     followingPnguree Permeater' a Sectlon °f ^ich is represented  on the
                   (£)  0
1 - Sample
2 - Cylinder of confinement
J - Contact resin
4 - Lower baseplate
                                             5 - upper baseplate
                                             6 - Porous stone
                                             7 - Filler joint
                                             8 - Stud bolts

               is performed first to give an evaluation of the permeability of
nroee,,,.  *  ^^1, then the oedometer is operated for leaching  test.  The
pressure is adjusted in order to get a discharge of 0.01 cnfis (36 cm^/hour)
S^fS^VJtracti°rJs. ^e carried out and it is possible to  add separately
the extracted quantities of every contaminant and to represent their  variation
in fuction of the quantity of the liquid discharged through the  sample.
of
                 hyperbolic and its interpretation allows the evaluation
                 f ^e c°nsidered contaminant which can be extracted if
        «*tr«ceion .liquid or the time of extraction was infinite. The

                             aS the ^™ •*«««• ^««t/ of the
                                  998

-------
IJI - EVALUATION OPERATIONS

      1 - SITE A : MARAIS  DE PONTEAU-LAVERA (50 km west of Marseille)

          a - Site history and treatment

              Old salt marsh on the seaside in which were dumped various
industrial residues of the Lavera petro-chemical plant,  About 30 000 m  of
wastes (sludges, sediments) had oeen disposed there for more than 15 years
(1955-1974). ftfter a first estimation of the quantities and nature of the
wastes to treat, many possibilities of treatment were tested :

              - incineration in an industrial waste incinerator,
              - recycling as stocK feed in a cement factory,
              - incinerator in a thermal power plant,
              - recycling in an oil refinery.

              All these treatments  were unsuccessful!, mainly because of the
heterogeneity of the material to treat. Consequently, on site treatment by
stabilization-solidification was decided. The treatment operation was carried
out in September and October 1978 by EIF ECOLOGIE using a solidification
patented process based mainly on the use^f lime as reactive agent. The total
amount of material  treated was  22 000 m .

              Laboratory  tests were carried out, involving lixiviation of
sarnies of  treated  material agitated during 24  hours in water with a pH of 6.
The analysis  gave  figures of 3 to 64 mg/1  for COD and total hydrocaroons < 0,2
mg/1.


          o - Evaluation  of efficiency

              Sampling  operation  :  The  sampling procedure has been carried out
on sept  23, after a rainy night.  The site  is  covered with sandy and  graveled_
material  without vegetation.  It is  shaped  in  a  slignt  dome  the  top of.which.is
 2 or  3 meters above the sea level.  The  sampling has been performed according
 to the procedure described in chapter  II.1.  Two ditches have been digged in
 the central part of the treated  site,  for  both  the  vertical section  was- :

               0 -  15 cm : coverage of sand and gravel
              15 - 100 cm : treated material,  dry and hard,  grey in  the upper
 part and black in the middle  ; slignt  smell of hydrocarbons.  The extracted
 black 'material became rapidly grey after contact with  the air.
                                        999

-------
      K  7    A  t^d ditch has been digged at the limit of the treated zone.  A
 to     ?n?™f?gSfflf? in I^S Wac* UqUld and black soil Beared. According
 to the information given by the representatives of the waste generator and of
 the.local authority who were present during the present sampling operation and
 who took part to the treatment realization ten years ago, this might be
 considered as water and original untreated material which remained outside the
 dikes surrounding the treated area.

             The digging was resumed in the direction of the center of the
 treated area and the treated material appeared in a shallow hard slice located
at a level clearly above which of the untreated material. Because of the'

fromlhishsecTion °f ^ treated layer ii: was declded *° take only one sample

              Results of tests and analysis

            - Physical characteristics
Sample
Al (upper)
A2 (middle)
•A3 (bottom)
water content
(%)
18
17
21
permeability
(m/s)
'
'
7.9 10~6

compressive
(strength)
kg/cm
-
4.7

           - Lixiviation tests

           Landfill lixiviation test (figures given in mg/kg except pHtt
           conductivity, alcalinity)
Sample
Al
A2
A3
Extrac- \ pH I Conduct! -
tion | \vity mS/s
1 1
II
1 112.71 5.6
2 |12.7| 5.4
Total | |
1 1
1 112.71 6.4
2 J12.7I 5.6
Total | |
1 1
1 112.11 5.9
2 \ 9.3\ 1.6
Total \ \
TOO 1 COD [Alkalini-
\mg/Kg\ \ty m mol/
, , *9
900 | 28001 440
450| 20001 400
13201 48001 840
1 1
1 1 70 r 32001 440
41S>| 1400|- 420
158PI 45001 850
1 1
J080I i>100| 280
8451 21001 18
39251 118001 2P8
CU | Pb
1 1
1
1
5.2| 25
3. 7| 21
9.9\ 46
1
11.11 24
4.5| IS*
15.7| 43
1
3.5| <2
1.5| <2
5.11 -
Cd | Co I Va \Ni \HC
III)
1 1 1 1
1 1 II
0.4| <0.5|<0.5|1. 0142.
0.2|<0.5|<0.5|0.8| *
0.5|<0.5K0.5|1.8|eo
1 1 1 1
0.3K0.5K0.5U.3U')
0.4|<0.5K0.5|0.8| 4
0.7|<0.5|<0.5|2.1H?
| | | j
0.4|<0.5K0.5|1.2| f
0.5|<0.5|<0.5|0.5| *
O.S-K0.5K0.5I1.7I ?
                                     1000

-------
Pressure test : maximum releases (long term)
Sample a2\ pH 1 Conduct!- 1 TOC 1
I [vlty ms/s\mg/kg\
111 1
1 1 1- 1
|12.7| 5.6 \1166 1
II II
COD \AlKallnl- \ Fe
\ty mlmol/\
\ \
\ \
3200 \ 850 1 -
1 i
Cu 1 Pb I Cd
1 1
1 i
1 > 1 >
4.8110.510.11
1 1
Cr I Co 1
1 1
1 1
1 1
0.0510.081
1 1
Va\ Ni 1 He
\ \
1 1
- 11.271 n.s
1 1
1 1
                           1001

-------
 2 - SITE B .• BOU1RON MfiRLOTTE - Seine et Marne (70 km south-east of Paris)

       a - Site history and treatment

    _        Ancient sand pit Jocated in BOURRON MARLOTTE at  the  south eastern
 limit of the forest of Fontainebleau which had been used as'dumping site until
 1971.for the wastes generated by an old refinery pland (mainly acid tars and
 filtration residues).  The surface of the lagoon was about 4  200  m'  and the
 dumped material had stratified in three main  layers :

             - upper layer light and viscous
             - middle layer made of aqueous liquid
             - bottom layer of sticky material    .  ,

             Many analysis of these material were performed, including
 lixiviation tests according to the INSA test  utilized  for the present
 investigation.  The following table summarizes the results of analysis carried
 out before treatment.

Ph 	

Resistivity
COD 	 	

Sul fates...
Hydrocarbons \
Upper layer
(lixiviation)
1200 m
3 5

7 900
4 800 ma/kn

260 mg/Kg
15 mg/kg
middle layer
IJOOnr
2nc

232
J O/fn m-i /Isn

1 829 mg/kg
20 mg/kg
1
bottom layer,
(lixiviation)]
27 n
• 1U ;
2J2

-------
      o - Evaluation of efficiency
            Sampling operation : the sampling operation has been performed on
October 4th in the morning by rainy weather. The treated area has a slight
slope-from its southern part besides a surelevated railway to its northern
limit surrounded by the forest, the site is covered with soil without any
vegetation. Three ditches have been digged, two in the center of the treated
area, one out its western limit. The total thickness of the treated section is
about 6 to 7 meters, the treated material is colored in brown, homogeneous and
compact with a very slight smell of hydrocarbons.

            For every ditch, three samples have been taken at the top, in the
middle and under the 'treated section. However, the sandy soil under the
treated material seemed to be still contaminated by untreated tar and in the
third ditch there was a seepage of liquids from an untreated zone remaining at
the western limit of the dumping pit.

              Results of tests and analyses                          ,

            Physical characteristics
Sample
'
Bl (upper)
82 (middle)
B3 (bottom)
water content
(%)
33
1. 24
i' 31
per/neaoility
(m/s)

1.2 10~5

co/rpressive
strength
kg/n?
-
3.8
™"
            - Lixiviation tests

            Landfill lixiviation test  (figures given in mg/kg except pH,
            conductivity, alcalinity)
Sample
Bl
62
83
Extrac-
tion
1
2
Total
1
2
Total
1
2
Total
pH \
8.4
7.8
12.5
11.1
Conduct!-.
vity ms/s
~
1.7
1.6
5.3
1.8
7.8| 2.15
7.7| 1.95
1
TOO 1 COD l/llkalini-
mg/kg\ \ry mimol/
\ 1 kg
\ 1
21001 54001 15
12701 3050 1 60
3370 1 8450! 75
1 1
14401 35001 310
20401 £1001 260
34801 960Q\ 570
1 1
2520 | 77001 66
898\ 3600 \ 68
38181103001 134
Fe | PD | Cd
! 1
1 1
1 1
0.6K0.2I -
0.4K0.21 0.3
1 K0.21 0.3
1 i
Ni
-
-
0.51 4.01 0.4| r-
0.61 - | 0.3| -
1.11 4.0| 0.7| -
1 1
0.5K0.21 0.31 -
0.61 - 1 0.21 -
l.ll - | 0.5| -
HqSfjSfifr
' !
* i
5-. ltoS\-
23, ) 0.6 |< 330
Z* ) - I
1 '
1 - I
A 4 l.illOltJ
4o ^o/l lo 5ft
2.11 - |24#<
1 J
<<. i<0.?l;zr?tf
5- Yo.r| 33i
* \ \ttt>>
                                      1003

-------
I
 D)
 V}
 03
 0)
8
• »

•u
§
JO
          d
            -o "o
                          In'
         Q
                          IN
                          -H
                           •
                          0
VO

-------
3 - SITE C - NESLE - Same (100 km north of Parisfj

      a - Site history, and treatment

            During the first national inventory of hazardous dumps carried out
in 1.978, conta/nination of the chalky aquifer by nitrogen compounds and salts
was pointed ou in the region of NESLE where an industrial dumping site was
found to .be at the origin of the pollution. The waste dumping had occured in
pits digged in silty material laying above the chalk.

            In order to restore the safety of the site and of its environment,
the firm who generated the wastes and owned the dumping site decided to carry
out its tratment by stabilization-solidification.

            The waste consisted mainly of black colored sludges and silty
material. Analysis of leachates indicated COD up to 450 mg/1 iron up to
47 mg/1 amonia up to 80 mg/1 and traces of copper-and zinc. The treatment was
performed by TREDI (Petrifix Process) and a total of about 7 500 tons of
contaminated material has oeen treated beetween mid-october and mid-december
1980.

          •  A control of the mecanical stability of the treated site performed
in October 1981 concluded to a variable resitance to compressive strength,
which was found generaly good or even very good but with weak layers in some
places. These weak parts of the site have been re-treated later.

            In 1985 a new construction has been built at the south-eastern
limit of the treated site.
      b  - Evaluation of efficiency

            Sampling operation  : the sampling operation has been carried out
 on  October  4th,  in  the  afternoon by rainy wather. The site is rather flat,
 covered  with  lush grass.  Three  sampling points have been choosen, each
 corresponding to one of the  ancient lagoons of residues. The two first
 sampling points  showed  a  similar log : down to 1,50 meter, coverage of humus
 soil,  then  about 4  meters of treated material which appeared blackish and
 crumbly  with  a rather strong smell  (of organics and ammonia).

             The  third sampling  point has oeen located near the newly
 constructed ouilding and  its section can be described as follows : 0,80 cm of
 soil,  treated materieal compact and strong down to about 3 meters, then down
 to  about 5  meters,  again  crumbly treated material.

       For  the three points,  the sampling operation has been performed
 according  to  the definition  given in part II.1.
                                       1005

-------
   Results of tests and analyses

 Physical characteristics
Sample
Cl (upper)
C2 (middle)
C3 (bottom)
water content
(%)
45
16
16
permeaoillty
(m/s)
•
_
J.410"8
-
co/Tpressive
(strength)
kg/cm
.
1.1 ,
-
 - Lixiviation tests

 Landfill  lixlviation  test  (figures given in mg/kg except pH.
 conductivity,  alcalinity)                        >  ,
Sample
/
»
Cl
C2
CJ
Extrac-
tion
1
2
Total
1
2
Total
1
2
Total
pH i Conduct! -
Ivity Ms/s
1
1
7.61 l.l
7.6\ 0.57 .
\
\
7.81 0.52
7.5| 0.49
\
\
7.7| 0.54
7.8| 0.41
1
roc
1 mg/kg
411
105
516
61
16
77
35
9
44
COD \filkalini-
1 Iry m mol/
1 kg
\
1000 | 52
280| 31
12801 83
\
120 \ 36
60\ 31
180| 67
\
1JO| 71
70| 3
200 | 74
NH4\ Fe 1 Cu ! Zn |
1 .1 1 1 1
1 1 1 1
1 1 1 1
'5451 0.8|<0.5|<0.;2|,
1441 1.0K0.5I :0.5|
693\ 1.8| | 0.5|
1 II 1
5131 O.SK0.5I 0.6\
108| 0.5K0.5I 0,5|
6211 1.7! 1 l.ll
1 1 1 1
5221 0.5K0.5I 0.5|
1261 0.4K0.5I 0.5|
648| 0.5| | 1.41
- Pressure test : maximum releases (long term)
Sample C2 \ pH
1
1
1
1
1
1 8.5
1
[Con-
1 ducti-
lity
Ms/m
0.4
TOC


.
60
COD \Alka-
Ilinity
\m.mol/
1 kg
1
1
160 | 65
1
NH4
1
t

271
Fe


5.1
Cu


0.06
Zn


O.OJ
                         1006

-------
4 - SITE D ; BELLAY - Ain (80 km east of Lyon)

      a - Site history and treatment

            fit the end of 1980, the lagoons used for the storage of sludges of
wastewater treatment of a tanning plant were full and it became urgent to
proceed to the disposal of these sludges. However, because pretreatment before
disposal was required by the local authorities in charge of environment
protection.

            Analysis performed by the Centre Technique du Cuir fleeter
Technical Center) indicated contents of 8,7 g/kg cromium, 0,475 g/kg sulfide,
0,5 g/kg sulfates and 0,35 % nitrogen.

            Treatment operation was carried out by TREDI using its PETRIFIX
patented process beetween  april and June 1982 about 9 000 tons have been
treated.
      o - Evaluation of efficiency

            Sampling operation s the sampling operation has been carried out
on September 30 th by rainy weather. The site where the treated material was
disposed is a low swampy area with lush vegetation. Three ditches have been
digged, the first one at the supposed limit of the treated area presented a
first layer of about 80 cm of humus soil then about 1,20 meter of treated
material under which some material which seemed untreated appeared until the
original soil was reached at a depth of about .2,50 meters.
            On that point, two samples were taken
and one  from  the middle of the treated material.
one from the upper part
            For  the two other ditches digged in the central part of the
 treated area,  the section was similar but without the appearance of any
 apparently untreated material underneath. Consequently, three samples were
 taken, according to the sampling procedure described in part II.1. Noticeable
 smell  (mainly  ammonia), especially in the first ditch. For the three sections
 of sampling the  treated material presented a noticeable difference beetween a
 first  layer (about -15 cm thick) colored in grey and hard and compact and a
 second much thicker layer rather crumbly and colored in dark blue.
                                        1007

-------
   Results  of tests and analyses

 Physical characteristics
1
Sample I
1
1
01 (upper) I
1
02 (middle )\
. 1
05 (bottom) |
1
water content
(%)
55
57
16
permeability
(m/s)

-
compressive
(strength
_
0.95 + 0.1
- Lixiviation tests

Landfill lixiviacion test (figures given in mg/kg except pH,
conductivity, alcallnity)
Sample
Dl
02
03
Extrac-
tion
1
2
Total
1
2
Total
1
2
Total
pH I Conduct! -
\vicy mS/s
\
\
7.7| 1.1
7.8| 0.72
1
1
9.6\ 0.82
8.2| 0.48
1
1
7.8| 0.52
7.5>| 0.56
\
TOO 1 COD
mg/kg\
1
1
2JO | 800
151 1 220
391 11020
1
1250 13200
250 | 750
1520 13550
1
55 1 250
31 1 150
85 | 400
Alkalini-
ry mimol/
kg
45
35
81
15
9
24
85
37
123
NH4
<1
<1
234
27
251
801
130
531
Fe
Cr ira 1
1 1
II
1 1
0.4 0.8112801
0.5K0.5I1150I
0.51 0.8124301
1 1
0.3| 0.51 5501
0.51 I. 1| 410|
0.5| 2.1| 5701
III
1.4|<0.5| 5201
0.5K0.5I 8501
1.5| |1470|
- Pressure test : maximum releases (long term)
Sample 02
'

pH \ Con-
Iducti-
\vity
1 mS/m
\
\
5.2 I 0.44
1
roc
1200
COD
2580
Alka-
linity
m.mol/
kg
24
NH4
\
131.5
Fe
0.5
Cr
.
<0.05
Ca
315
                         1008

-------
Ill - REMARKS flND CONCLUSIONS

    .  Some remarks may be pointed out aoout these investigations

      - It appears that some of the sites has apparently not  be  completely
treated (A,B,D). This way be understood as the necessity to fix  a  limit  to  the
treatment, but this limit seems rather approximative and may  left  some
significant quantities of untreated material. However this situation  results
from the position of the owner of the site and of the authorities  responsiole
for the control of the rehabilitation project, it is not representative  of  the
treatment process itself. As a consequence, it appears that the  upper and
bottom samples are probably not always realy representative of the treatment
process efficiency at the limits of the treated layer,  therefore our  opinion
is that only the middle sample very be considered without douot  as
representative of this efficiency.

      - With the exception of site B, that has been more recently  treated,  it
was not possible to find figures representatives of the characteristics  of  the
initial contaminant material, and when some datas are available, the
evaluation tests performed (specialy leading tests) are not sufficiently
defines to allow a significant comparison with the result of  the present
investigation. If we considerer the site B and the results of lixlviation test
it should be Kept in mind that the main part of initial material is comparable
to bottom layer characteristics and the test of this material is comparable to
che first extraction performed on the middle sample of treated material.

      - In connection with the previous remarks, it appears that the  treatment
performance required oy the responsible of the sites were far from acurately
defines. For example, for sites C and D, the enterprise who carried out  the
treatment project indicates that the requirement, of the owners  of the sites
were only dealing with the physical stability of the treated  sites.  From the
present sampling operation it appears clearly that the compact ness and  the
mechanical efficency can be adjusted to reach a high level by increasing the
proportion of the reactive agent (and consequently the cost of treatment).

      - The evaluation of the efficiency of a treatment of staoilization
solidification relies greatly upon the choice of tests and analysis.  The
present study, which is based on the use of two different lixiviation tests
shows a relative uniformity of figures of general parameters  of  the ieachates
like pH and conductivity, for the release of the pressure tests  as more
singificant because this test is concerned to give an estimate of  the maximum
release on a long term basis.

      - Comparaison between the two solidification processes, is not  possible
because of the differences of nature and characteristics of the  waste that  has
been treated. However some systematic differences can be mentionned for  the
physical characteristics : water content (generaly higher for Petrifix),
mechanical resistance (higher for EIF), permeability (higher  for EIF), and  for
the chemincal characteristics of Ieachates, the pH and alkalinity  of  which
remains nign for EIF, as a consequence of the use of line as  reactive agent.
                                      1009

-------
      - As a conclusion it may be mentionned :

            . that a significant evaluation of the efficiency of the
stabilization-solidification treatment of these old contaminated site  remains
difficult, mainly because of the lack of knowledge of initial characteristics
of the treated material

            . that the results of the present investigation give an  evaluation
of the present state of the treated sites and a tentative estimation of  their
potential of evolution in the future, fls a whole it appears that the treated
sites are presently and will remain in the future in a satisfactory  state
according to their use and their environmental condition

            . that the present investigation will participate to the' efforts
of adjustement of the methods of evaluation of on site efficiency of
stabilization-solidification.

      This last point might be one of the most positive of our conclusions and
oe usefull for the future projects of treatment.
                                    1010

-------
                                            Appendix 3-B




        Stabilization/Solidification Technology Case Studies
Portland Cement (Hazcon, presently IM-Tech), United States
                      1011

-------
?xEPA
                        United States
                        Environmental Protection
                        Agency	< ,
EPA/540/S5-89/001
March 1989
                        SUPERFUND INNOVATIVE
                        TECHNOLOGY EVALUATION
                        Technology  Demonstration
                        Summary

                        Technology Evaluation  Report,
                        SITE Program Demonstration
                        Test, HAZCON  Solidification,
                        Douglassville,  Pennsylvania
                         The major objective of the HAZCON
                       Solidification  SITE Program
                       Demonstration Test was to develop
                       reliable performance and cost
                       Information. The demonstration
                       occurred at a  SO-acre site of  a
                       former oil reprocessing plant at
                       Douglassville, PA containing a wide
                       range of organic and heavy metal
                       contaminants. The HAZCON process
                       mixes the hazardous waste material
                       with cement, a proprietary additive
                       called Chloranan, and water.  The
                       Chloranan  is claimed to neutralize
                       the inhibiting effect that organics
                       normally have on the hydration of
                       cement
                         The technical  criteria used to
                       develop the effectiveness  of the
                       HAZCON process were contaminant
                       mobility, based  on leaching  and
                       permeability tests; and potential
                       Integrity of solidified soils, based on
                       measurements of  physical  and
                       microstructural properties.
                         Extensive sampling and analyses
                       were  performed  showing (1) the
                       concentration of the organics were
                       the same in the TCLP  leachates of
                       the untreated and treated soils, (2)
                       heavy metals  reduction was
                       achieved,  and (3)  structural
properties of the  solidified  cores
were found to Indicate good long-
term stability.
  This Summary was developed by
EPA's Risk Reduction Engineering
Laboratory,  Cincinnati,  OH, to
announce key findings of the SITE
Program demonstration that Is fully
documented In two separate reports
(see ordering Information at back).

Introduction
  In response to  the  Superfund
Amendments and Reauthorizatipn Act of
1986 (SARA), the Environmental
Protection Agency's Offices of Research
and Development (ORD) and Solid Waste
and Emergency Response (OSWER)
have established a formal program to
acceferate  the   development,
demonstration, and use of new or
innovative technologies as alternatives to
current  containment systems  for
hazardous wastes. This new program is
called Superfund Innovative Technology
Evaluation or SITE.
  The major objective of a Demonstration
Test Program is to develop reliable cost
and performance information. One
technology, which was demonstrated at
the Douglassville, PA Superfund Site, is
the HAZCON proprietary solidification
                                     1012

-------
process. The process involves the mixing
of hazardous waste material and cement
with a patented nontoxic chemical called
Chloranan. The Chloranan  is claimed to
neutralize the inhibiting  effects  that
organic contaminants normally have on
the  hydration  of  cement-based
materials. For this treatment, the wastes
are  immobilized  and   bound  by
encapsulation into  a hardened,  leach-
resistant  concrete-like mass.
  The Douglassville, PA Superfund  Site,
No. 102 on the National Priority List, was
selected  as  the  location  for   the
Demonstration Test. This  is a  50-acre
rural site of  an  oil  recovery facility  that
includes:  two large lagoons once filled
with oily  sludge,  an  oily filter  cake
disposal area, an oil drum storage area,
an  area  where  generated  sludge  was
landfarmed into the soil and the plant
processing  area. More  than 250,000 cu
yd of soil is contaminated.
  The  major objectives of this  SITE
Project were to determine the following:

  1.Ability  of  the  stabilization/
    solidification  technology   to
    immobilize  the site contaminants,
    which included  volatile  organics,
    base neutral/acid  extractables
    (BNAs),  oil    and   grease,
    polychlorinated biphenyls  (PCBs),
    and heavy metals.
  2. Effectiveness  of the technology  for
    treating  soils with  contaminant
    concentrations  varying  over  the
    range  1%-25%  by  wt. oil   and
    grease.
  3. Performance and  reliability of the
    process system.
  4. Probable  long-term  stability  and
    integrity of the solidified soil
  5. Costs   for commercial-scale  appli-
    cations

  Project documentation will  consist of
 two reports. This  Technology Evaluation
 Report describes  the field  activities and
 laboratory  results. An  Applications
 Analysis  will follow  and provide  an
 interpretation of the data and conclusions
 on the results and potential applicability
 of the technology.
   The following technical  criteria  were
 used to evaluate the effectiveness of the
 HAZCON process:

   1. Mobility of the contaminants:
     a. teachability of  the  contaminants
       and oil and grease before  and
       after treatment.
    b. Relative  permeability  of the
      treated and untreated soil.
  2. Integrity of the solidified soil mass:

    a. Physical  properties  -  unconfined
      compressive  strength,  bulk
      density, etc.

    b. Microstructure  of  the hydrated
      matrix.

The above criteria were used to develop
the sampling program.

Procedure
  The  Demonstration  Test  utilized
contaminated soil from six plant areas,
referred  to  as  Lagoon  North  (LAN),
Lagoon South (LAS), Filter Cake Storage
Area (FSA), Drum  Storage Area (DSA),
Plant Facility Area  (PFA), and  Landfarm
Area (LFA). The intent was to process 5
cu yd from each of five areas  and then
perform an extended duration run for the
sixth area. The  purpose of the  extended
run was to confirm the reliability of the
operating equipment. The extended run,
which  was intended   to  process
approximately 25 cu  yd from FSA, was
performed  on LAS feed,  due  to  very
difficult access  to FSA and convenience
of access and high contaminants  level at
LAS. The runs used less  feedstock than
anticipated, and  produced  approximately
5 cu yd from the short runs and 25 cu yd
from the extended run of treated soil.
  The contaminated soil was excavated
and  screened to remove  aggregate and
debris greater than 3 inches  in  diameter.
It then was  fed  to the HAZCON Mobile
Field  Blending   Unit  (MFU)  along with
cement, water, and Chloranan. Cement
was used on an approximately 1:1 ratio
with soil,  and  the   soil-to-Chloranan
ratio  was  10:1.  The  four  feed
components  were  blended in  a mixing
screw  and  fed  into  five  1-cu-yd
wooden molds  for the short tests and
three  1-cu-yd  molds  plus  two  12-
cu-yd pits for the LAS run.
  While the  contaminated soil  was
processed and cured,  the  excavation
holes  were  enlarged,  lined with an
impervious  plastic liner,  and  partially
filled with  clean soil. After  the  1-cu-yd
blocks cured sufficiently to be  moved
(48-96 hours), they were removed from
their molds and placed into the pits. The
blocks then were covered with clean soil.
The blocks  were sampled 28 days later
and will be sampled  at 6 or  12 month
increments  for  5 years,  along with the
surrounding  clean soil,  which is to be
checked for contaminant  leaching from
the blocks.
  Soil  samples  were  taken  in  three
phases: before  treatment,  as a  slurry
exiting  the MFU for analysis after 7 days
of curing, and from  the  buried  blocks
after 28 days of curing. For the first  five
runs, two  untreated  soil  composite
samples, three  sets  of slurry  samples,
and three solidified cores were taken. For
the extended run on LAS feed,  additional
samples were taken.
  Physical  property  measurements
performed were:

• bulk density
• moisture content

• permeability
• unconfined compressive strength of the
  solidified cores
• weathering  tests-freeze/thaw  and
  wet/dry
  Chemical analyses were performed to
identify  the  organic  and  metal
contaminants in the soil. In addition, three
different leacfiing tests were run:

• Toxicity  Characteristic  Leaching
  Procedures (TCLP)-standard leaching
  procedure  used  for  measuring
  leachability of the contaminants.

• ANS  16.1  - simulates  leaching from
  the intact  solidified core with  rapidly
  flowing groundwater

• MCC-1P - simulates leaching from
  the intact  solidified core in relatively
  stagnant groundwater regimes.

  These latter two tests were drawn from
the nuclear industry and modified to suit
hazardous waste analysis.
  In order  to  obtain information  on
potential   long-term    integrity,
microstructural studies were  performed
on the untreated  soil and solidified cores.
These analyses included:

• X-ray  diffractometry-identifying
  crystalline structures in the solid
• Microscopy-scanning  electron
  microscope  and  optical  micro-
  scope-characterizing porosity, hydra-
  tion products,  fractures,  and  the
  presence  of  unreacted soil/waste
  material in the treated soil.

Results and Discussion
  The following observations were made
and summarized in Tables  1 and 2:

• The six plant  areas offered  a wide
  diversity  of feedstock.  The  oil  and
  grease ranged from 1 %  by wt. at  DSA
                                                         1013

-------
 Tail/a 1. Physical Properties  ,
Untreated Soil
Location
DSA
LAN
FSA
LFA
PFA
LAS
2 8 -Day Cores
Bulk Density. Permeability, Bulk Density, Permeability
glml cm/sec g/m cm/sec
1.23
1.40
1.60
1.68
1.73
1.59
0,57
1.8 x 70-3
Impermeable
2.0 X 10-2
7.7 X 10-2
1.5 X 10-S
1.95
1.61
1.51
1.84
2.07
1.70
1.8 x 10-3
4.0 x 1O-9
8.4 X 10-9
4.5 X 10-9
SiO X 10-10
2.2 X 10-9

UCS, psi
1113
523
219
945
1574
889
 Table 2.  Chemical Properties
                               Leachate Concentrations, mg/l
                      Untreated Soil
                                                      28-Day Cores
Location
DSA
LAN
FSA
LFA
PFA
LAS
voc-
0.92
0.02
1.03
5. TO
1.10
0.06
BNA-
NO
1.02
2.81
0.010
0.010
0.010
Pb"
1.5
31.8
17.9
27.7
22.4
52.6
VOC
0.38
0.06
0.72
0.37
0.84 "'
0.11
BNA
ND
1.45
2.79
0.10
0.11
0.73
, Pb
0.007
0.005
0.400
0.050
0.011
0.051
 "VOC  •  Volatile Organic Carbon
 BNA  -  Base Neutral/Acid Extractable
 Pb    -  Lead
  to  25% at  FSA.  Polychlorinated
  Wphenyls (PCBs) were detected up to
  52  ppm by wt., with the  maximum
  value  detected   at   LAS.   Lead
  contamination  concentrations ranged
  from  0.3%  to 2.3%  by wt. at FSA.
  Volatiles and  base  neutral acid
  extractables  (BNAs  -  semivolatiles)
  organics reached levels  in  excess of
  100 ppm by wt. at FSA.

• The volume  of the solidified soil was
  more  than  double  that of the
  undisturbed feedstock. Optimization by
  HAZCON could  reduce  this volume
  Increase, but other physical  and
  chemical properties may change. <
• Permeabilities of  the treated soil, after
  curing for more  than 28 days, were
  very low. in the range of 10-8 to 10-9
  cm/sec.

• The unconfined compressive strengths
  (UCS)  of the solidified  soils  ranged
  from about 220 psi for FSA to 1570 psi
  for PFA, and the values were inversely
  related to the oil and grease level.
• The  wet/dry  and  freeze/thaw
  weathering tests showed small  weight
  losses  (0.5%-1.5%) at the'end  of the
  12 cycle test for the test specimens
and'their  controls.  Unconfined
compressive  strengths,  performed
after  the final  weathering cycle,
showed no loss in strength compared
to the unweathered samples. No other
analyses  were  performed on the
weathered samples.

The TCLP leaching  tests compared
leachate concentrations in  the treated
soil with that from untreated soil. Due
to the addition of cement, Chloranan,
and water, the treated soil contaminant
concentration  levels, on average, was
40%   of  the   untreated   soil
concentrations. The results were as
follows:

-Metals-The  leachate  from  the
  solidified soils  showed metal levels
  at or near the .detection  limits. The
  results for  lead,  the  predominant
  metal, were  lower by a factor of
  about 500, from :20- to 50 mg/l in the
  leachates from the untreated soils to
  less than 0.1 mg/l'in the treated soil
  leachates^ The other five  metals
  were at  or near the  detection  limits
  for both untreated and treated soils.

-Volatile  brganics-The primary
  compounds detected  were  tri-
 chloroethene,  tetrachloroethene,
 toluene, ethyl benzene, and xylenes.
 The leachate concentrations of the
 con.taminants  appear  to   be
 approximately the same in  both the
 untreated  and treated  soils  at levels
 of less  than one  milligram  per  liter
 (mg/l).

-BNAs-The compounds detected  in
 the  leachates were  phthalates,
 phenols,  and  naphthalene.  The
 phthalates were reduced to near their
 detection limits of  10 yg/l  in both the
 treated and untreated soil leachates.
 The total  phenols  in  the leachates
 reach,3-4 mg/l for FSA  where the
 feedstock  had phenol concentrations
 as  high as  400  mg/l  with similar
 C9ncentration levels seen in  both the
 untreated and treated soil leachates.
 The values for naphthalene were less
 than 100  p,g/l  for both the treated
 and untreated soil leachates.

-Oil  and Grease-Leachate concen-
 trations for treated soils were slightly
 greater than for untreated soils   in
 each  case. The values for  the
 untreated  soil were 0.2 to 2.0  mg/l
 and for the treated soil 2 to 4 mg/l.
                                                       1014

-------
  - PCB analyses of all  leachates, both
    for untreated and treated  soil, were
    all below the detection limits of  1.0
    yg/l.
• The special leaching tests,  ANS 16.1
  and  MCC-1P,  which  simulate
  leaching of the solidified soil  cores,
  were  performed  on the treated  soil
  samples from each plant area  except
  DSA and LFA. Experience  with these
  tests  on hazardous waste  is  limited.
  Results  are compared to  the TCLP
  results for treated soils,  but this may
  not be  relevant  as a performance
  measure. The  results are as follows:

    Metals-For  ANS 16.1 the values in
  the leachate  increased with leaching
  time. They were  of the same order of
  magnitude as  for TCLP leachates. The
  MCC-1P leachates also  increased
  with time;  for  the largest time  interval
  (28 days) they were greater by a factor
  of about 10 than the TCLP leachates.
    Volatile Organics-For ANS 16.1,  the
  total concentrations of  all the volatiles
  in the leachates  were  lower than  the
  TCLP leachates by a factor of about 2.
  For  MCC-1P the  leachate concen-
  trations were  approximately the same
  as  for the TCLP leachates.  In both
  tests, no  discernible  trend between
  leachate  concentration  and time
  interval was noted.
     BNAs-The  leachate  concentrations
  for ANS 16.1  were less than for TCLP
  leachates,  while the MCC-1P leachate
  concentrations  were  approximately
  equal. Leachate concentrations for the
  dominant components,   phenols,
  appeared  to  increase with leaching
  time  interval  for  MCC-1P,  but no
  trends for  ANS 16.1 were observed.
     Oil  and Grease-MCC-1P leachate
  concentrations were about the same as
  for TCLP and  ANS 16.1.
     PCBs-For both leaching tests, all
  leachate concentrations were below
  detection limits.

 • Microstructural  analysis are  proven
  methods   for  understanding   the
  mechanism of structural degradation of
  soils,  cement,  and  soil-cement
  mixtures.  However, there  have been
  relatively  few  studies of  complex
  waste/soil mixtures  for   stabiliza-
  tion/solidification processes.  Conse-
  quently, in some cases, interpretation
  of the  observations may  be  difficult.
  The  microstructural studies provided
  the following information:
   -Mixing  did  not appear to be highly
     efficient.
  -There  were  many  pores;  some
    including air bubbles.

  -Soil  components  survived  the
    solidification/stabilization  process
    unchanged and  appeared in the
    cores. The factors that suggest this
    were the  presence in the  cores  of
    unaltered  brownish aggregates, and
    observations  of  x-ray  diffraction
    peaks,  which  cannot be identified
    with minerals constituents of the soil
    and are also  present in the  core,
    suggest that contaminant materials
    are being carried  through  the
    solidification process in an unaltered
    form. Since the two  features being
    referred to are  apparently  major
    components of the  contaminated
    soil, it appears that encapsulation is
    a major part  of the  mechanism  of
    solidification/stabilization.

• The operations for the first five runs (5
  cu yd) required  many  startups due to
  unscheduled  shutdowns  caused
  primarily  by  plugging in the  soil feed
  screw.
    In  addition, the consistency  of  the
  slurry mix was quite variable, running
  the gamut from powdery to a very thin
  slurry.  However,  physical property
  changes due to  this  variation  were not
  observed. For the extended  time  run
  (25 cu yd) at LAS, operation was more
  uniform,  with only a few  short-term
  outages.

• The economic analysis was based on
  the 70%  on-stream  factor  and a 300
  Ib/min operating capacity observed for
  the HAZCON equipment.  A cost  of
  $205/ton  was calculated for the  Mobile
  Field  Blending  Unit  during  the
  Douglassville, PA demonstration. The
  process is very intensive in labor and
  chemical  additives,  with these items
  amounting to approximately 90%  of
  the  total costs. Substantial  cost
  reductions are expected with process
  and chemical optimization.

Summary
  A Demonstration Test on the HAZCON
solidification technology was performed
on  a wide range of  hazardous waste
feedstocks. Test runs producing 5 cu yd
of treated  soil  were performed in five
plant  areas,  and  an  extended run
producing about 25 cu yds of treated soil
in a sixth area. Many samples were taken
and a wide range of laboratory analyses
were performed to obtain a  comparison
of physical properties and  contaminant
mobilities before and after soil treatment.
Highlights of the results are as follows:

• The  volume  of  the  solidified  soils
  compared  to  the  untreated  soils
  increased by  approximately 120%.
  HAZCON could  reduce  the  volume
  increase by optimizing the quantity  of
  cement  and  Chloranan, but  other
  physical and chemical properties may
  change.

• The  unconfined compressive strength
  ranged  from  220-1570  psi  and  is
  inversely related to the  oil and grease
  concentration.

• Permeabilities of 10m8 to  10-9 cm/sec
  were obtained  which surpass the
  generally accepted  permeability  of
  10~7  cm/sec for soil barrier liners.

• The  TCLP  leach test showed that
  heavy metals were  immobilized  over
  the  range  of  oil   and  grease
  encountered.

• TCLP  leach  tests  performed on
  untreated and treated  soils showed
  equivalent concentrations  of volatiles
  organics and BNAs in their respective
  leachates.

• The   leachates  from  MCC-1P
  contained greater  concentrations  of
  metals and organics than ANS 16.1 for
  an equivalent time  interval. There are
  no protocols for  these  tests on
  unsolidified  waste and no  attempt was
  made to  run the  ANS  and  MCC
  procedures on untreated waste.

• PCBs  were  not detected  in any
  leachates, whether the soil was treated
  or untreated.

• The  microstructural  study  of the
  solidified soil showed the following:

  - high porosity

  = brownish aggregates passed through
    the process unaltered
  - mixing was not highly efficient
  - encapsulation is a major part  of the
    mechanism  of  solidification/
    stabilization

• Startup  operating  difficulties  were
  encountered by  HAZCON during the
  Demonstration Test.

• A cost of $205/ton was calculated for
  the Mobile  Field Blending Unit  during
  the.Douglassville, PA demonstration.
                                                         1015

-------
    The  EPA  Project Manager,  Paul de Percin, is with the Risk  Reduction
     Engineering Laboratory, Cincinnati, OH 45268 (see below).
    The complete report consists of two volumes, entitled "Technology Evaluation
     Report,  SITE Program  Demonstration  Test, HAZCON  Solidification
     Douglassville. Pennsylvania:"
         'Volume I" (Order No. PB 89-158 810/AS: Cost: $21.95, sub/ect to
         change) discusses the results of the SITE demonstration
         'Volume II" (Order No. PB 89-158  8281 AS;  Cost: $36.95, subject to
         change)  contains the  technical operating data logs,  the sampling  and
         analytical report, and the quality assurance project plan/test plan
   Both volumes of this report will be available only from:
            National Technical Information Service       i
            5285 Port  Royal Road
            Springfield, VA 22161
            Telephone:  703-487-4650
   A  related  report,  entitled  "Applications  Analysis Report:  HAZCON
     Solidification," which discusses application and costs, is under development.
   The EPA Project Manager can be contacted at:
            Risk Reduction Engineering Laboratory
            U.S. Environmental Protection Agency
            Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
      BULK RATE
POSTAGE & FEES PAID
         EPA
  PERMIT No. G-35
Official Business
Penalty for Private Use S300

EPA/540/S5-89/001
                                                      1016

-------
                                Appendix 4-A




 Soil Vapor Extraction technology Case Studies
In Situ Soil Vacuum Extraction, The Netherlands
           1017

-------
          Soil Vapour Extraction
       of Hydrocarbons  In Situ and
       On Site Biological Treatment
          at a Gasoline Station
S. Coffa,  L.G.C.M. Urlings,  J.M.H.  Vijgen
  TAUW Infra Consult B.V.,  P.O.  Box 479
    7400 AL Deventer, The  Netherlands
              Presented at
    NATO/CCMS  International Meeting
           (Soil Remediation)
   Washington DC, 18-22 November 1991
              1018

-------
         INTRODUCTION
1.1      Background

         In the industrialized countries the extent of soil  contamination
         is enormous.  The  overall costs  of soil remedial  action in  the
         Netherlands is estimated to be  in excess of 30 billion  dollars.
         This  corresponds  with 150  million m3  excavated  soil,  emanated
         from  more  than  100,000  polluted  sites   (Committee   10  Year
         Scenario, 1990).
         The  number  of sites  contaminated  with  hydrocarbons,   in  the
         Netherlands,  is  estimated  to  be  at least  6,600  in 1990.  The
         costs  of soil  remedial  action  (excavation) and  improving  the
         gasoline  stations,   with  regard  to  these  sites,   amounts   to
         approximately  750  million dollars.  Excavating  the  contaminated
         soil is a very effective method of removing the  contamination.

         Treatment  or  disposal  of  the  excavated  soil   is  necessary.
         Nevertheless  excavation  can be  difficult  or  even  impossible
         under certain circumstances e.g.:-

         -  the presence of buildings and civil  engineering works (roads,
            bridges etc.);
         -  certain cables, power lines and pipe lines in the subsurface;
         -  contamination has spread to great depths  (i.e.  > 4 m);
         -  shortage of space and traffic problems (city centres);
         -  irreplaceable  function of  the  site  (e.g.  railway  station,
            production plant).

       '  Generally  the   criteria   favourable   for   applying   in    situ
         techniques  are the following:-

          -   the permeability  of the soil  is reasonable;
          -   less  disturbing layers of clay/peat  appear in the  subsoil;
          -   the contamination is  infiltrated  (i.e. not buried);
          -   only  one contaminant is present  (e.g.  toluene) or comparable
             component (e.g. gasoline);
          -   the quantity of contaminated  soil is substantial;
             the contaminant can be biodegraded;
          -   the contaminant can be  leached and/or volatilized.

          The  presence  of   above  and   underground  infrastructure   is
          unfavourable for conventional  excavating  techniques; this often
          occurs at  sites  situated in city centres  and industrial  areas.
          Therefore  more  effort  has  to  be   put   into   developing   new
          innovative remedial action techniques.

          In this  article  attention is  paid to  the soil vacuum extraction
          (SVE) as a remedial  technique  in general  and in  particular to
          SVE  and the  air based blodegradation of  gasoline  hydrocarbons
          underneath  a  road  in  Raalte.  A  new developed  combined  soil
          vapour/groundwater  treatment   system   (BIOPUR®)   will   also  be
          discussed.
                                   1019

-------
1.2
REMEDIAL TECHNOLOGY
         Soil Vapour Extraction fSVE)

         The mechanism  by which  SVE  operates  is  relatively simple   By
         creating negative pressure gradients in a  series  of zones within
         the unsaturated  soil  a subsurface  air flow  is  induced  (figure
         1).   This   flow  volatilizes  the contaminants  present  in  the
         unsaturated soil. This process,  in  theory,  continues until  all
         volatile   components   are   removed.   The   extraction  wells   are
         individually connected to  the transfer pipes,  then  manifolded to
         a  vacuum  unit  and  the soil vapour  is transported  to  a soil
        vapour  treatment  system.  Figure 1  gives  an  outline  of  three
         different SVE performances.
              ®
       Figure t: Three Different Performances of SVE
       The  withdrawn   soil  vapour   is   often  treated   by  charcoal
       adsorption or catalytic incineration.  The groundwater is usually
       treated by  stripping  and/or charcoal adsorption.  In  order  to
       SAT^T^   ! treatment costs of both  groundwater and soil vapour
       TAW Infra Consult B.V. applied (since 1989)  a biological system
       for combined aerobic treatment (see section 4.2).
                             1020

-------
SVE and Related New Technologies


Meanwhile  experience  with  soil vacuum  extraction  (SVE)  is  so
wide  that  it can be considered as a proven  technology.  A review
on  SVE  techniques  is given  by  Hutzler  (1989).  The  treatment
costs  of  withdrawn soil  vapour  are  substantial,  usually more
than  50%  of the  total remediation  costs (Miller  1990).  The air
based  biodegradation  is  applied  with  the  aim  of  reducing SVE
costs.  This is  a  new innovative  technique  as  indicated  by an
international  review  of  in situ  bioreclamation practice  (Stapp
1990). Only 2  of the 23 studied sites  used air based technology,
most  of the  applied  bioreclamation was water based.  In recent
literature  (Miller, November 1990 and  Hinchee,  October 1990) air
based enhanced biodegradation is described.
Apart   from   the   mentioned   advantage   of   lower   hydrocarbon
concentration  in   the withdrawn  soil  gas,  substantially  less
carrier  media is needed when air is  used as  an  oxygen carrier.
This  is indicated  in Table 1.

Table 1: Mass  Requirements to  Deliver Sufficient Oxygen or nitrate for
       Biodegradation to Mineralization (derived from Hinchee 1990)
Electron Acceptor
                       Carrier Mediun
 Mass Requirements
(kg carrier/kg hydrocarbon)
Oxygen (in air)
Oxygen (pure)
Oxygen (in HjOj)
Oxygen (in H70,)
Nitrate
Nitrate
Oxygen
8
40
100
500
50
300
20.9
mg/l in water
mg/l in water
mg/l »2°2 in Mater
mg/l »2°z in Mater
mg/l in water
mg/l in water
percent in air
400.000
80.000
65.000
13.000
90.000
15.000
13
An  additional advantage of  using air as an  oxygen source in  the
vadose  zone  (gas phase)  is  the fact  that  diffusion  rates  are
much higher in air than in the water filled  saturated zone.

Furthermore one has to realize  that during SVE soil vapour  (with
low oxygen content)  is  continuously replaced  by fresh air  (with
high  oxygen  content).   This  oxygen  source  does   not   require
additional expense!

An  indication of the biological activity in the soil is  given by
several biological parameters such as  TPC (total plate  counts),
OUR (oxygen uptake rate) and NCR (nitrogen consumption  rate).
 ESTIMATED DURATION OF SVE

 To select the  most  suitable  remedial action technique  for  a
 specific  site  it is necessary to estimate  the costs  of 3  or  4
 chosen techniques. In order to give a good estimate  the duration
 of the  SVE  also has  to be calculated.  For  this  purpose  TAUW
 Infra Consult  B.V.  developed a  relatively simple "spread  sheet"
 model. The  most important  input parameter  can be  distinguished
 in groups related to:-
                          1021

-------
  - the contaminants(s)                   ' , 'v  •
     * estimated amount  of contaminants;
     * molecular mass, vapour  pressure;
       solubility of contaminants  (3 maximum);
     * adsorption/desorption coefficients (k  ) •
  - soil                                   oc
     * volume  of contaminated  soil;
     * density,  porosity,  moisture;
     * fraction  organic  carbon in  the soil;              ,
  - biodegradation
     * zero order biodegradation rate;
  - application  SVF.
     * estimated effective air flow in the subsoil.

  In addition  to  the model  and even to  validate the  model it is
  recommended  to  carry  out column tests with  contaminated soil
  from the site.  The application of column tests  can give detailed
  information  on  volatilization,  biodegradation   and  possible
  remedial action limit concentrations.
 SITE CHARACTERISTICS    :                         .

 During  soil  remediation at  a petrol station,  it  appeared that
 contaminants were found underneath a provincial  road.  Excavation
 of this  part of the  contaminated soil was  not  feasible  due  to
 financial and  technical (traffic) reasons.  The  most  favourable
 solution was  a SVE  system in  combination with  biostimulation.
 This system  not  only had to  remove  the volatile compounds from
 the ^gasoline  but   also had  to  stimulate  biodegradation,   of
 particularly  the  non  volatile  components,  by  the   (passive)
 infiltration of air  (oxygen). The unsaturated zone of  the soil
 consisted of fine to gravel sand. The groundwater surface  had  to
 be lowered  from  2 m to 3  m below  ground surface  in order  to
 enlarge the unsaturated zone  and make  the  smear zone  available
 for SVE. Figure 2  is  a diagram of a cross  section of  the site
 On one  side of the  road there are several soil vapour  extraction
 wells   (perforated  2-2.75  m below  ground surface)  and  on the
 other  side  seven infiltration wells (passive).  To prevent  direct
 air infiltration at  the extraction side of  the  road,   a plastic
 liner  was placed between the  road and  the  sheet pile  wall The
 biological   system   applied  for   combined  soil   vapour" and
 groundwater  treatment is, shown on the left hand side of Figure  2
 (above  the excavated  soil) and is  called BIOPUR®.
Figure 2; Cross Section of the site at Raalte
                         1022

-------
          RESULTS  AND DISCUSSION
4.1       SVE and Air Based Biodegradation

          The  under  pressure applied  in the  soil  vapour  extraction wells
          and the corresponding  soil vapour flow are outlined in Figure 3.
                                   UNDER PRESSURE  (mttar)
                                   SOIL VAPOUR FLOW (Nm3/h)
                                     —I	1	1—:—I	1—
                                      60      80      100
                                               WEEK
120
          Figure 3: Under pressure
                               (rcbar) and corresponding soil vapour flow 
-------
   In Figure 5  the  cumulative amount of vapourized and  biodegraded
   hydrocarbons is  given as well as  the  sum of both of  them.  The
   vapourized amount is  the result of the combination of  figures  3
   (.air flow)  and 4  (concentration) .
   The biodegraded  amount is derived from the  withdrawn mass  flow
    fJ!frb0n dl°Xlde'  T*16  concentration  of  carbon dioxide in  the
   withdrawn  soil vapour  ampunted to  approximately 0.5% (v/v)  over
   weeks 0-37   increased till approximately 1% (v/v) in weeks  37-47
   and gradually decreased till 0.25%  (v/v) over weeks 47-118
               TONS OF HYDROCARBONS REMOVED
   1.0 -
   0.5 -
   0.0
                                     T	1	r
                                 80      100     120
                                   WEEK
 .Figure 5; Cumulative amount of vapourized and biodeoraded hvdrocarhnnc


 Figure 6 is an outline of the average hydrocarbon content of the
 soil  during SVE  and biodegradation.  The removal  of  the volatile
 ra"£°Cwas IT "I" aCCT°rdinS t0 -Flotations. However,  tL removal
 rate  was too  slow.   It was therefore  decided to  take  additional
 measures  (commencing   week  78)   to  speed   up  the   in  situ

 Os-^rnu't^-^T^  consisted of (heated> air £•""£
 of nutrieAts                   Sroundwa<-r level  and  infiltration
Figure 6: The average hydrocarbon Content of th» Soii
                        1024

-------
         In Table  2  the biological activity (in week  92)  is expressed as
         TPC  (total  plate  count),  OPC  (oil  plate count),  determined by
         applying  specific  gasoline  degradable  micro-organisms  and  the
         oxygen uptake  rate  (OUR) at different depths.

         Table 2 Concentrations Measured and Biological Activity in Week 92
Depth
(m)


0.5
0.9
1 5
2.0
2.5
2.7
Non Volatile
Gasoline

                        210 B/h
                                 BIOPUR®
Biological
Treatment
System
,


	 ^ exhaust


          Figure 6: Gasoline mass balance at a steady state


          In the purified soil vapour (exhaust gas) no  aromatics or other
          volatile  hydrocarbons  were  detected.  The  detection  limit  for
          these matters  is  0.1 ppm.  In  the  effluent no  aromatics (BTEX)
          could be  detected  (detection limit  0.5 Mg/D •  Discharge of the
          effluent was  suggested by the Water Board.
                                      1025

-------
  In Figure  6 the  input and output  of the on site BIOPUR®  unit are
  outlined.   In   Figure   7   chromatograms   of   GC   analysis   are
  presented.                                                   J
         impurity carrier gas
       Li
                                                                      exhaust gas
       L
                         J[JJ^
                                                                      soil vapour
Figure 7; Chromatogram of soil vapour (untreated) and exhi
        (after biological treatment)
                            1026

-------
CONCLUSIONS AMD DISCUSSION
SVE and Air Based Biodegradation

Within a  period of 1.5 years  approximately 3,200 kg of  gasoline
was  removed.  Initially  based  on  limited  soil  analysis,   the
estimated   amount   at  the   beginning   of  the   soil  vapour
extraction/biorestoration  was about  700  kg! At  the  end of  this
period, the  average gasoline  content  of  the soil  decreased  from
roughly 10 to  2.5  g/kg dw-  T"18  decline was mainly due to  the
removal of the  volatile  fraction  of the  gasoline.  To  speed  up
the  removal  of  the  non  volatile  fraction  several  additional
measure  were  taken.   These  measures  consisted  of (heated)  air
injection  (35-40°C),  fluctuation of  the  groundwater  level  and
infiltration  of nutrients  (N/P).  The results were  significant.  A
further decrease of the average gasoline content of  the  soil was
realized  from  roughly 2.5  to 0.26  g/kg dw within six months.
This  decline  corresponded with the removal of approximately 800
kg gasoline.                                      ,

Thus  some 4,000 kg  of gasoline  was removed  within  two  years.
Approximately 50%  of the removal was due to mineralization  by
the micro-organisms in the contaminated soil.


Combined  Air  and Water Treatment  (On Site)

The  bioreactor  -  BIOPUR  ®  -  has  functioned beautifully. A mass
flow  of  more  than 5  kg  of  gasoline hydrocarbons  per  day was
converted into  carbon dioxide and  water,  with  a  residence time
of approximately  15  minutes for  groundwater and less  than 10
minutes  for  soil vapour.  The degree  of conversion amounted to
more  than 98%!

Related  to  the  entire remedial  action  the  treatment  costs  of
 soil  vapour  and groundwater  were  less than Dfl.  0.40 per m  of
 groundwater.  For calculation purposes  only, the  treatment of the
 soil vapour was free of charge.

 In Table  3 a comparison of the treatment  costs  of four types of
 groundwater purifying plants  is given.

 Table 3; Coroarison of the treatment costs of four types of groundwater purifying plants
Type

1
2
3
4


BIOPUR*
Rotating Biological Contactor
Stripper/active carbon
Stripper/compost filter
Costs CDfl.)
10 m/hour
0.53
0.78
0.85
1.16
per m3 of groundwater
20 m/hour
0.44
0.65
0.54
0.87
40 m/hour
0.35
0.59
0.36
0.72
                          1027

-------
 Estimated Remediation Costs                        	

 On   the   supposition   that   an   in  situ  remedial  action  takes
 approximately two years  and excavation,  including  treating the
 groundwater  takes  one year,  the  estimated costs  for  In situ
 remediation   of   the   site   will  be  approximately  75%  "of  the
 excavation variant.

 Furthermore  one  has  to take; the  two  additional  advantages  of an
 in  situ  remediation  action into account.  These advantages  are
 that no extensive excavation is  required  and that the exact soil
contamination boundaries do  not  have  to be  known.  In the case of
excavation one has to know the exact boundaries.
                         1028

-------
LITERATURE


L.G.C.M. Urlings, F. Spuij,  S. Coffa, H.B.R.J.  van Vree,
"Soil Vapour  Extraction of  Hydrocarbons  - In  Situ and On  Site
Biological Treatment".
Paper presented at:
In Situ  and On Site Bioreclamation, An International  Symposium,
19-21 March, 1991, San Diego, California.

F. Spuij, L.G.C.M. Urlings,  S. Coffa,
"In Situ Soil Vapour Extraction".
Paper presented at:
NATO/CCMS,  1990, France.

F. Spuij, S. Coffa, C.  Pijls, L.G.C.M. Urlings,
"In Situ Soil Vapour Extraction of Contaminated Soil".
Paper presented at:
Third Forum on Innovative Hazardous Waste Treatment Technologies
11-13 June, 1991, Dallas, Texas.

R. Costing, L.G.C.M. Urlings, P. van Riel, C. van Driel,
"Biopur®: Alternative Packaging for Biological Systems".
International   Symposium on  Biotechniques   for  Air  Pollution
Abatement and Odour Control Policies.
27-29 October,  1991, Maastricht, The Netherlands.

"Ten  Year  Scenario,  Soil Remediation",  Stuurgroep Tien  Jaren-
Scenario Bodemsanering,  Ministerie van VROM, 1989.

Miller,  R.N.,   et  al.  "A Field Scale Investigation  of Enhanced
Petrol' Hydrocarbon Biodegradation in the  Vadose  Zone  at Tyndall
AFB,  Florida",  Proceedings   NATO/CCMS  meeting  France,  December
1991.

Hutzel,  N.J.,   et  al.  "State of Technology Review,  Soil Vapour
Extraction System",  EPA/600/2-89/024, 1989.

Staps,  J.J. "International  Evaluation of  in situ Biorestauration
of  Contaminated  Groundwater",  National  Institute  of  Public
Health  and  Environmental   Protection,  Report  738708006,   The
Netherlands,  1990.

Hinchee, R.E.;  Miller,  R.N.  "Bioreclamation  of Hydrocarbons  in
 the   Unsaturated   Zone,    in      Hazardous   Waste  Management
 Contaminated Sites,  and Industrial  Risk Assessment"  et  ed.  by
W. Pillmann and K. Zirm, Vienna 1990, 641-650.
                          1029

-------

-------
                                   Appendix 4-B

    Soil Vapor Extraction Technology Case Studies
Vacuum Extraction of Soil Vapor, Verona Well Field
                    Superfund Site, United States
              1031

-------
          In-Situ Soil Vacuum Extraction System
             Verona  Well  Field Superfund Site
                  Battle  Creek,  Michigan
                       Draft Final
Final Report for NATO/CCMS Pilot Study on Remedial Action
   Technologies  for  Contaminated  Land  and  Groundwater
      Presented at the Third International Meeting
                   November  6-9,  1989
                  Margaret M. Guerriero
                    U.sJ EPA Region V
                        1032

-------
I.  INTRODUCTION
Site Description

The Verona Well Field is located in Battle Creek, Michigan in the
south central portion of the State.  The Verona Well Field (VWF)
site consists of four distinct problem areas within approximately
100 acres.  The municipal well field is located in the northeast
corner of the City and lies on both sides of the Battle Creek
River (Figure 1).  The well field consists of 30 production wells
that supply drinking water for 50,000 residents and several major
food industries.  The Thomas Solvent Raymond Road (TSRR)facility,
the-Thomas Solvent Annex (TSA) facility, and the Grand Trunk
Western Railroad (GTWRR) have been identified as sources of well
field contamination.  Figure 1 shows the location of these
sources relative to the well field.  The site is located^in an
urban setting which is primarily residential with some light
industry.


Site History

The contamination problem at the VWF site was first discovered  in
August 1981, during testing of the City water supply.  Test
results revealed that 10 of the City's 30 supply wells were
contaminated with volatile organic compounds (VOCs).
Concentrations  ranged from 1 to 100 ug/1.  During the same
period, private residential wells were also tested.  Several of
these wells were found to contain VOCs in excess of 1,000 ug/1.
The highest level found in a private well was dichloroethylene
at 3,900  ug/1.  A bottled water program was implemented in this
area and  residents were connected to the City's water supply
system.

In the Fall of  1983, the first phase of a Remedial  Investigation
 (RI) was  initiated to determine the extent of contamination in
the well  field  and potential  sources.  Sample results from the
initial RI work confirmed the existence of a contaminant plume
with VOC  concentrations ranging from 1 ug/1 to  100  ug/1 in the
well field.  The investigation also identified  the  three major
sources of contamination.


Remedial  Measures

In May 1984, U.S. EPA signed  a Record  of Decision  (ROD) calling
for an Initial  Remedial Measure  (IRM)  to implement  a blocking
                              1033

-------
                                                            GRAND
                                                            TRUNK WESTERN
                                                            RAILROAD
                                                            MARSHALLING
                                                            YARD
                                                            THOMAS
                                                            SOLVENT
                                                            RAYMOND
                                                            ROAD
                                                            FACIUTY
                              i  / '  V.
                               / '  ^THOMAS
                              1  I   SOLVENT
                              '   ,   ANNEX
MICHIGAN

• BATTUE CREEK
                                                                FIGURE 1
                                                                VICINITY MAP
                               1034

-------
well system in the well field.  The blocking system consists of a
line of converted supply wells that extract contaminated water
from the southern end of the VWF.  The system prevents
contaminants from migrating further into the well field.  An air
stripper with vapor phase carbon treatment was also installed to
treat" contaminated water prior to discharge to the Battle Creek
River.

In August 1985, U.S. EPA signed a second ROD that addressed the
most contaminated of the three sources, the Thomas Solvent  •
Raymond Road facility.  The ROD included a groundwater extraction
system to remove contaminated groundwater, treatment of_extracted
water utilizing the existing air stripper at the well field,
demolition of the existing warehouse and loading dock, and a soil
vapor extraction system to remove VOCs from contaminated soils.


Site Characteristics and Sampling Results

The Thomas Solvent Raymond Road  (TSRR) facility is a former
solvent repackaging and distribution facility that operated from
1970 to 19S4.  Solvents were  stored in 21 underground storage
tanks, of which 19 were later discovered to be^leaking.  _There
has also been documented reports of surface spillage during
operation.  The site contained an office building, a warehouse
with  a loading dock, and the  21 underground tanks  (see  Figure 2).


Site  geology consists  of a  fine-to coarse-grained  alluvial-
glacial sand with traces of silt, clay, and pebbles underlain by
a  fine-to medium-grained sandstone with minor  lenses of  shales
and limestones.  The unconsolidated sand unit  ranges in
thickness from  10 to 50  feet  and the sandstone varies from  100  to
120 feet  in thickness.  The hydraulic  gradient is  primarily north
to northwest  from the  identified sources toward the VWF.  The
depth to water  is approximately  20 to  25  feet.  The hydraulic
conductivity  of the unconsolidated material ranges from 2.7 x  10"
3  to  4.0  x  10~2 cm/sec.  The  hydraulic conductivity of  the
sandstone ranges  from  7  x  10~3 to  2 x  10~2  cm/sec.        :

Samples  collected  at  the TSRR facility indicate that both  soils_
and groundwater are highly contaminated with  a variety  of  organic
compounds.  Table  1  lists  organic  compounds detected  in soils  at
TSRR.   Groundwater  samples showed  concentrations  as high as
 100,000  ug/1  VOCs.  The  total estimated volume of organics  in
groundwater and soils  was  3,900  Ibs.,  and  1,700  Ibs.,
 respectively.

 It should be  noted that  these total  mass  estimates were based  on
 sample data obtained  using an accepted soil sampling  procedure
 which is now known to produce VOC results lower  than  actual_
 values.   The total mass  in groundwater and soils  is now estimated
                              1035

-------
                                                                      \
                                       OFFICE BUILDING
                                                     FIGURE 2
                                                     LOCATION OP
                                                     UNDERGROUND TANKS
                            TABLE 1
         PRIKCIPAT. SOTT.
                                   S OETEgTBn
 CHLORINATED HVnpQCARBOHS

 - KETHYLSNE CHIflRIDE

 - CHLOROFORM

 ~ 1»2-DIC3TLOROETH\NE

 - 1,1,1-TRlCHLOROETHANE

 - TRICHLOROETHYLEHE

 - TETRACHLOROETHYLENE
 cove,
- TOEDEKE

- 3CYLENE

- ETHYL BENZENE

- NAPHTHALENE



KETOKES


- ACETONE

- METHYL ETHYL  KETONE
   60,000

    2,000

   27,000

  270,000

  550,000

1,800,000
  730,000

  420,000

   78,000

    9,400
  130,000

   17,000
                         1036

-------
to be significantly, greater based on the results of the operating
groundwater extraction system and the soil vacuum extraction
system.

Conventional subsurface soil sampling procedures involving the
use of split spoon samplers require the sampler to be opened and
the sample transferred to a bottle prior to shipment to the
laboratory.  This allows for significant amounts of VOCs to
volatilize before analysis.  This problem, coupled with the lack
of sampling in the capillary zone and beneath the former
warehouse building, resulted in estimates of VOC-contaminated
soils being considerably lower than actual.  This has had a
significant effect on the operation of the soil vacuum extraction
system.  These effects will be discussed in detail in a later
section of this report. ;


Technology Selection

Due to the significant mass of contaminants in the soil and
groundwater at the TSRR facility, alternatives that employed both
groundwater and soil remediation were developed in the
feasibility study.  A two step approach to remedial action was
used at the TSRR facility in which each alternative developed for
the feasibility study included both a groundwater and a soils
portion.  The selected alternative for the site includes a
groundwater extraction (GWE) system and the soil vacuum
extraction (SVE) system.

The groundwater extraction system includes 9 extraction wells
which pump a total of 300 gallons per minute.  The extracted
water is pumped to the existing air stripper at the well field
and discharged to the Battle Creek River after treatment.  Figure
3 shows the layout of the GWE system.  Initially, VOC
concentrations in groundwater were so high that extracted water
was passed through pretreatment carbon units prior to being
pumped to the air stripper.  The system has been operating since
March of 1987.

Several alternatives for soil cleanup were evaluated, including
SVE, excavation of soils with on/off site disposal, site capping,
and soil washing  (flushing water through the unsaturated zone
with subsequent groundwater extraction).

SVE was chosen based on a number of reasons.  Although it was
considered an innovative technology, it was felt that it had a
good likelihood of success given the site conditions and
contaminants.  Excavation was considered unacceptable due to its
potential to release high concentrations of VOCs into the air,
which would create a health hazard to residents in close
proximity to the site, and significantly violate Michigan Air
Quality Standards for VOCs.  Therefore, alternatives that would
                               1037

-------
                                                 TYPICAL
                                                 GROUNOWATER
                                                 EXTRACTION
                                                 WELL
                                                            THOMAS SOLVENT
                                                            BUILDING
                                                            (DEMOLISHED)
MONITORING BUILDING
                                                             LOADING DOCK
                                                             ^(DEMOLISHED)
                                                    SVE PROCESS BUILDING
                                      OFFICE BUILDING
                              1038
                                                      FIGURE 3
                                                      GWE SYSTEM LAYOUT

-------
not disturb soils were favored.  Capping was not considered to be
consistent with future actions because of the high level of
contamination present at the site and because the underground
tanks would eventually have to be removed.

Of the two soil treatment alternatives, SVE was calculated to
take less time to remediate the site than soil flushing.  It was
estimated that SVE, in conjunction with GWE, would reduce the
groundwater contamination to 100 ug/1 within three years.  The
contaminant mass would be reduced to 2% within 1 1/2 years (based
on 1700 Ibs. of VOCs).  Soil washing was estimated to require 8
years to reach 100 ug/1 in groundwater and 8 years to reduce
contaminant mass to 2%.  This was significant in the selection of
SVE because it was not less expensive than soil washing.  The
estimated capital cost of SVE was $413,000, with operation and
maintenance (O&M) of $90,000.  Soil washing capital cost was
estimated at $58,000, with O&M of $6,000.  SVE was, however,
considerably less expensive than the excavation and capping
alternatives.

Since SVE is an innovative technology, the procurement of a SVE
contractor was accomplished using a performance specification
which contained certain minimum requirements but left the major
design details to the discretion of the bidding SVE contractors.
Contract documents .called for the construction, operation, and
maintenance of the SVE system.  The performance standards
require that the SVE would operate until all soil samples showed
VOCs below 10 mg/kg, with no more than 15 percent of the samples
above 1 mg/kg total VOCs.
II. TECHNOLOGY
Process Description

The soil vapor extraction process is designed for use in
removing VOCs from the unsaturated zone in soils.  The mechanism
by which SVE operates is fairly simple and straight forward.  The
system is designed to create negative pressure in the unsaturated
zone using wells that are connected to a vacuum extraction unit.
The vacuum induces a flow of air through the soils, thereby
volatilizing VOCs that are absorbed on soil particles and
extracting the contaminants in the vapor phase.

A vacuum extraction system generally consists of several
extraction wells screened throughout the unsaturated zone or
within discrete units of the unsaturated soils.  The wells are
connected by transfer pipes which are manifolded to the vacuum
extraction unit.  The vacuum applied at the wellhead creates a
negative pressure or vacuum in the subsurface which cause VOCs to
volatilize and migrate to the extraction wells.  A vapor/water
                              1039

-------
 separator  is  also  incorporated to remove water  from the
 contaminated  air stream.      •

 The^materials used for an SVE system are generally readily
 available  and not  specialized.  Equipment requirements include
 PVC/stainless steel wells and piping; conventional vapor phase
 carbon treatment units; a conventional air/water separator; and
 an induction  blower.

 The design of the  SVE system  is critical in order to insure
 adequate cleanup of the soils.  The design requires expertise in
 modeling vapor  flow, understanding site lithology, determining
 contaminant mass and areal extent of contamination.  These
 factors are used to determine system variables  such as well
 spacing, number of wells, and depth of screened interval.

 Site conditions, soil properties, and the contaminant's chemical
 properties are  the important  factors to consider in determining
 whether to use  soil vapor extraction.  Information on soil
 permeability, moisture content, and porosity is needed to make a
 determination on whether the  soils have sufficient air-filled
 porosity.  Insufficient air-filled porosity results from the
 presence of excess water in the pore spaces which reduces the
 effectiveness of vacuum extraction.  Depth to groundwater is an
 important cost  consideration  because if the vadose zone is less
 than 10 feet, it may not be cost-effective to use SVE (excavation
 of less than  10 feet may be less costly).

 In order for  a  contaminant tojbe stripped from  the soil using
 SVE, it must have  a Henry's Constant of 0.001 or greater.  The
 higher the Henry's  Constant,  the easier the compound is removed
 by vacuum extraction.  Figure 4 shows relative  extraction rates
 for compounds found at the TSRR facility.
Design and Construction of the TSRR System

Prior to full-scale construction of the SVE system at TSRR, a
preconstruction investigation was performed.  The investigation
included a geophysical survey, a soil sampling program, and a
soil gas survey.

The geophysical survey was conducted to confirm locations of
underground tanks and to check for additional buried objects in
the effected area.  The soil sampling program was conducted to
investigate the horizontal and vertical extent of contamination,
and to estimate the mass of VOCs in the vadose zone.  As a result
of this activity, a revised estimate for VOCs ranging from 13,000
to 16,500 pounds was calculated.  This did not include estimates
for a^floating product layer that was discovered during the
sampling.  The objective of the soil gas survey was to
investigate the extent of VOC contamination in shallow soils in

                            1040

-------
                                                                        VI
                                                                        0)
                                                                        c
                                                                        o
                                                                        0)
                                                                    tO  -T.
                                                   O .
(p/qi) e.
                       1041

-------
 areas not  previously investigated  at the site.  Results  confirmed
 that the major area  of  contamination was in the vicinity of the
 underground  tanks  and loading dock, however, contamination was
 also detected  along  the northeast  and southern parameters of the
 site.                        !

 Results of the investigation!were  used to determine  locations of
 additional SVE wells, revise|estimates of the mass of VOCs in the
 soils, and to  make determinations  on system parameters.

 Following  the  preconstruction investigation, a pilot phase SVE
 system was installed.   The system  consisted of 11 wells  that were
 operated for a total of 70 hours.  The objectives of the pilot
 phase were to  verify the radius of influence of the  wells and
 determine  the  vapor  flow rate/vacuum pressure relationship at
 each well, investigate  the effect  of the underground tanks on the
 vacuum pressure distribution in the vadose zone, and identify the
 VOC loading  rates  form  individual  wells as a function of vacuum
 pressure and flow  rate.

 Results of the pilot phase were used to determine process
 variables  and  locations of wells.  Extracted airflow rates range
 from 60 to 165 standard cubic feet per minute (scfm) from
 individual wells,  with  wellhead vacuums of 3 to 4 inches of
 mercury.   VOC  extraction rates vary between wells with the
 highest measured concentration at  4,400 Ibs/day during the pilot
 phase.  The  radius of influence for the wells was measured to be
 greater than 50 feet.

 The full-scale system consists of  a network of 23 4-inch diameter
 PVC wells  with slotted  screens from approximately 5  feet below
 grade to 3 feet below the water table.  The wells are packed with
 silica_sand  and sealed  at the screen/casing interface with
 bentonite  and  then grouted to existing grade to prevent  short
 circuiting.  Each  well  has a throttling valve, a sample  port, and
 a vacuum pressure  gauge.  The wells are connected by an  above
 ground PVC piping  manifold system.  Figure 5 shows the location
 of the SVE wells and the piping layout.

 The piping manifold  is  connected to a centrifugal air/water
 separator.    This is connected to the vapor phase carbon
 absorption system  which is followed by the vacuum extraction
 unit (VEU).  The VEU is  responsible for inducing extraction of
 soil vapors  from the subsurface, through the extraction  wells and
 into the treatment unit.  After treatment,  air is discharged
 through a  30 foot  stack located on site.   Figure 6 is a  schematic
 of the SVE system.           ;                                   i

The carbon absorption system [consists of eight stainless steel
 carbon canisters with four iii the primary system and four in the
 secondary  or backup system.   JThe primary carbon system is the
main unit  for absorption of VOCs,  with the secondary carbon
                              1042

-------
                         FENCE UNE
                                           8 INCH HEADER PIPING
MONITORING
   BUILDING
      o
      en
      o

      o
      5
                         CONCRETE
                      DECONTAMINATION
                           SLAB    i
                        TYPICAL SOIL VAPOR
                        EXTRACTION WELL
                                            SVE PROCESS
                                            BUILDING
                       OFFICE
                       BUILDING
                                                                AIR/WATER
                                                                SEPARATOR
                                      1043
FIGURE S
SVE WELLS AND PIPING LAYOUT

-------
 system acting as a backup when breakthrough of the primary system
 occurs.   Each carbon canister holds 1000  pounds of vapor-phase
 granular activated carbon and!are connected to the header piping
 with flexible hoses and couplings that are easily disconnected
 for ease in canister change outs.   Figure 6 shows the various
 sample ports,  pressure gauges;and temperature  probes  located
 before,  between,  and after the carbon units.   A carbon monoxide
 monitor_is installed between the carbon units  to detect
 combustion in the primary carbon unit and triager an  axitomatic
 system shutoff upon detection!.

 The carbon system was installed on the negative pressure  side of
 the VEU to minimize leaks and;eliminate the potential for
 emissions to the atmosphere.  ;During the  pilot phase  of operation
 it  was determined that carbon i adsorption  efficiency wasr,
 equivalent under positive andinegative pressure.

 Breakthrough of the primary carbon system is measured by  an in-
 line HNu^hotoionization detection meter.   Four contaminants are
 used as  indicator compounds,  tetrachloroethylene,
 trichloroethylene,  methylene  chloride,  and benzene.   The
 breakthrough point was determined  using the relationship  between
 total VOCs_measured by the HNvi  and the compound-specific
 concentrations  measured in the  air flow.   When breakthrough      •
 occurs,  the primary carbon canisters are  changed out  and  replaced
 with those from the secondary system and  a new set  of four
 canisters are put into backup in the secondary system.  This
 allows maximum  loading of the primary carbon system prior to
 rotating  the carbon units,  while minimizing the  possibility of
 breakthrough in the backup system.
                              i
 Samples are collected from both  carbon systems  as well  as  at the
 individual wellheads.   Results are used to determine  VOC  loading
 rates and predict rates  of breakthrough.   Sample  analysis  is
 performed on-site using  a gas jchromatograph with  dual  flame
 ionization detectors  and capillary columns.


 III.  RESULTS

The  SVE system  began  full  operation  in March 1988.  Results
discussed in this  report are  for the period of March  1988  through
September 1989.               i     .
Vacuum Extraction System Performance

To date, approximately 40,250 pounds of VOCs have been removed
from the soils.  On-site gas chromatography has been used to
monitor VOC concentrations extracted by SVE.  Off-site analysis
of spent carbon has confirmed that on-site monitoring is accurate
to within approximately 5 percent.
                             1044

-------
1045

-------
The> initial loading rate of total VOCs has declined over the
period of operation from an ibitial level of approximately 45
pounds per hour  (pph) to below 10 pph.  The floating product
      ^hat was detected during- preconstruction and during the
        °Peration Period has not been detected since October

     »WGJeKi' at that Same t^e' a °-5" to 1«°
the water table was recorded.
                                                      increase in
 The average VOC concentration^ measured at the air discharge
 f?aCVVP?r?nimately I'35 mg/1' with an average flow rate in
 the stack of 1060 standard cubic feet per minute (scfm) .   This
 ^/t    SPed 5v°m an initial voc concentration of approximately 23
 mg/l._  over the course of operation of the system,  an average
 efficiency rate of greater than 99.8 percent removal has  been
 measured.
 Technology Evaluation and Performance Monitoring

 «<™« ?VE+-JS an4. innovative technology,  careful consideration was
 given to the method by which the systen's performance would be
 monitored and to confirm that the performance objectives would be
 Sf$; *LQ?   JY/SSUuance Pr°3ect Plan (QAPP)  and Sampling Plan
 were developed for the sampling events.   Three soil  sampling
 episodes were planned.  One prior to startup,  one at the mid-
 operation point,  and the last to confirm that performance
 objectives have been met.

 Since conventional soil sampling methods cause volatilization of
 VOCs prior to analysis,  a  special sampling and analysis
 procedure was developed for collection  of samples.   Samples are
 taken by driving a split spoon sampler  fitted  with four,  3-inch
 brass liners,  through hollow stem augers.   To  prevent excess
 handling,  and thus volatilization of contaminants, one brass
 liner is removed from the  split spoon,  immediately wrapped in
 aluminum foil,  sealed,  and sent to the  laboratory for analysis
 Samples  are analyzed using a core of the undisturbed sample for
 extraction.
x        the Pre-°Perati°nal ^nd mid-operational sampling events
have occurred.  The pre-operational samples verified that the
volume of _ VOCs in the soils had been underestimated during the
remedial investigation at the site.  Based on sample results,
VOC concentrations were estimated to be between 12,800 and 16 500
P°'1^; J? hiS did.not Delude estimates for the floating layer of
product that was identified du[ring the startup work.
                                                         .
Data ^ from the mid-operational Sampling event have not yet been
SSftaMo ?°m ^ laboratory. I If is hoped that this data will be
available for incorporation into the final version of this
report.                      :,
                              1046

-------
Process factors

Carbon handling requirements have been the limiting factor in
performance of the SVE system.  Because the estimate of VOCs
present in the soils was significantly underestimated, the
amount of carbon needed was also underestimated.  The amount of
contaminants extracted to date has resulted in the use of 250,000
pounds of carbon in the treatment system at a cost of $541,000.
It is estimated that a total of more than 400,000 pounds will be
needed to complete the project at a cost of approximately
$886,000.

In addition to the increased costs, the additional carbon
requirements have caused delays in the operation of the system.
Although the system has been operational for more than 18
months, actual'number of days of operation is approximately 100.
This due to the need for frequent carbon change outs and
transporting the spent carbon off-site for regeneration.  It is
estimated that an additional 50 days will be needed to attain the
levels required in the performance objectives.  It is also
expected that carbon change outs will become less frequent1 as the
loading rates decline.

The equipment needed to operate the system has proven to be very
reliable and down time due to equipment failure has not been a
factor in SVE operation.  As discussed, the materials used to
operate the SVE system are conventional and easily replaced if
necessary.  Although Terra Vac, the vendor, has been required to
be on site for 8 to 10 days per month due to the frequent need
to change out carbon and monitor the system, the system was
designed for unattended operation.  It is expected that as
loading rates decline, Terra Vac will be required to spent less
time at the site per month.

Instrumentation and controls have been installed to monitor the
system and trigger shut down if necessary.  These include the
carbon monoxide monitor in the carbon system, a high water-level
shut down in the air/water separator, high temperature shut down
triggers, and an on-line HNu with a shutdown mechanism for
detecting VOC breakthrough of the primary carbon system.  In
addition, the system contains an automatic dialer that contacts
Terra Vac when any of the shutoff mechanisms are 'triggered.


Costs

A summary of the costs to install, and operate the SVE system,
current  and projected future  carbon costs, and  the unit costs  for
operation of the system are listed in Table 2.  It was not
possible to separate out the  cost of the pilot  phase portion from
the  cost of the  full-scale system because the bid was received  as
                              1047

-------
                        TABLE 2
              SOIL VAPOR EXTRACTION COSTS
SVE Lump Sum Bid - For
Construction & Operation
(excluding cost of carbon)
$1,265,535
Cost of SVE/Cubic Yard of
VOC-Contaminated Soil
(excluding cost of carbon)
    $22.50
Unit Cost/pound for Carbon ,
(removal/transport/regeneration)
    $ 2.16
Carbon Costs to Date
(250,000 pounds used)
   $541,000
Projected Total Carbon Costs
(estimated 400,000 pounds)
   $886,000
                        1048

-------
a lump sum for the project.  The original cost estimate was
revised to account for the additional days per month Terra Vac
is required to be on-site, due to the increased contamination at
the site.

As previously discussed, carbon costs have, been quite high due to
the increased level of contamination found at the site.
Initially, it was estimated that 20,000 pounds of carbon would be
needed to remediate the site.  To date, 250,000 pounds have been
used and it is estimated that an additional 150,000 pounds more
are needed to complete the project.  Table 2 lists the actual and
projected future carbon costs for project completion.  No long
term maintenance costs are expected.
Lessons Learned

Vapor Treatment

As has been discussed throughout this report, the underestimation
of the total mass of VOCs in the soils at this site has
complicated the remediation of the site.  The increased levels
have effected the project expense, the time to remediate, and the
operation of the technology.

During evaluation of the treatment options for the project, it
was determined that, based on a total VOC volume of 1,700
pounds, carbon absorption was the least expensive treatment
option.  If contamination estimates were closer to the actual,
carbon absorption would not have been the least costly and would
likely not have been considered.  In addition, the cost of
operating the system is more expensive because Terra- Vac
must be at the site many more days per month than estimated.

The underestimation of VOC mass has also effected remediation
time and the operation of the technology.  Due to the  frequent
number of change outs required during operation, the system is
operational as little as  five to ten days per month.   This has
resulted  in only  100 days of system operation in a period  of  18
months.  '

In order to prevent this  situation,  it  is imperative that
accurate  estimates  of subsurface contamination be obtained prior
to design of  the  system.  Specifically,  accurate mass  estimates
must be  obtained  for amount  of  floating product present •,  and  the
amount  of VOC contamination  in  the, capillary fringe, the  zone
 immediately  above the water  table,  and  in the smear  zone,  the
 zone within  which the water  table  fluctuates.   Based on  data
collected during  operation  of  the  SVE  at TSKR,  it was  estimated
 that  70  percent  of the  mass  of  VOCs occurs  as  floating product
 and in VOC  saturated  soils  in  the  smear zone and  capillary
 fringe.

                              1049

-------
 Radon Gas

 A somewhat unexpected contaminant detected was radon eras  which
 occurs naturally in the site soils.   Measurements of the'carbon
 vessels indicate the presence of radioactivity on the spent
 carbon.   The presence of radon gas is not  too unusual since  it is
 readily volatilized and activated carbon is a good collection
 medium for radon.   Concentrations measured to date at the  TSKR
 facility are not considered to present a public health or  worker
 hazard.   However,  the handling,  transportation and regeneration
 of radioactive spent carbon may need  to be considered for  SVE
 operation in areas where radon occurs at high levels.


 Future Plans

 U.S.  EPA's  contractor,  CH2M|Hill,  is  currently evaluating  the use
 t°rLa  ?atfjYtlc. oxidation (CATOX)  system for the destruction  of
 VOCs  in  the_soil vapor.   This  would replace the carbon absorption
 system.   While other treatment options  have been looked at during
 the life of the project,  the cost  for removing the  carbon  system
 and installing another  treatment  system has not been  shown to be
 cost-effective.  However, cost-effective CATOX systems  have
 recently been  developed that can treat  chlorinated  VOCs  without
 generating  dioxins  or suppressing  catalyst  performance.

 In addition  to the  reduction in"cost  to  treat  contaminants, two
 other ma}or  benefits  from switching to  a CATOX system  are  the
 destruction  of contaminants  on-site,  which  eliminates the
 transporting of wastes  off-site, and  the ability to run  the SVE
 system continuously,  thereby attaining  site cleanup faster
IV. CONCLUSIONS

Since the project is still in operation, certain conclusions
sould not be considered definite.  However, based on evaluation
of the_operating data from the site and on the recent literature
regarding SVE, the following conclusions have been drawn:

   * SVE is a viable technology for the removal of VOPs in
     unsaturated soils.  The|fact that over 40,000 pounds of VOCs
     have been removed from the soils at TSRR indicates that the
     technology works.

   * SVE will operate in a wide range of soil types.   Based on
     work at the site, SVE is very effective in removing
     contaminants from sandy soils.   Recent literature on SVE
     performance indicates that it is effective for soils with
     measured permeabilities;of 10~4 to 10~8 cm/sec.

   * The major considerations in determining the technology's

                             1050

-------
     applicability  are  soil properties, depth to ground water,
     and  the  contaminant's chemical properties.  Soils that have
     a  low air-filled porosity  and high moisture content may not
     provide  adequate conditions.  In  addition, at  sites where
     groundwater is encountered at less than 10 feet, it may be
     more cost-effective  to excavate contaminated soils.
     Chemicals  with a Henry's Constant of  less than 0.001 may not
     be sufficiently volatile  for the SVE process.

   *  The  SVE  system has operated well  in all weather conditions
     at the site.   Cold weather operation  has not proved to be a
     problem.   The  system has operated through an entire winter
     in the midwest with  temperatures  that range from 0 degrees
     celsius  to -26 degrees Celsius.

   *  The  SVE  system can be designed to operate for  the majority
     of the time without  an on-site operator.  Under most
     circumstances, the system  would be  sized to provide
     unattended operation with  vendor  personnel on-site 1 to  4
     days per month depending on the size  of the system and the
     monitoring requirements.

   *  Based on experience  at the TSKR  facility, SVE  appears to be
     the  only technology  that can effectively remove the VOC
     saturation remaining after free product  is removed  from  the
     capillary fringe and smear zone of  VOC-contaminated soils.

   *  The  costs of SVE at  the  site for  1  cubic meter of  soil  is
     approximately $50.00 to  $60.00.   This includes the  cost  of
     treatment of vapors  using  carbon  absorption.   If treatment
     of vapors is not required, costs  could be as  low as  $20.00
     per  cubic meter of soil.


The overall prognosis of  the  SVE process is that  it offers  an
economical, reliable,  and rapid cleanup  technology for
remediating soils contaminated with volatile organics.   The
technology enhances groundwater extraction systems and greatly
reduces the time and cost for groundwater remediation.   The
process works on most soil types and has a limited number of
factors for consideration in determining applicability.   There  is
no limit on size of the site,  or on the level of VOC
contamination  (except in considering the need for treatment of
off-gases).  The system is easily installed and removed,  and does
not require specialized equipment for operation.


V. CONTACTS FOR MORE INFORMATION

Information useful to potential SVE technology users can be
provided by the following sources:
                              1051

-------
 Government^

 Margaret Guerriero
 Remedial Project Manager
 U.S.  EPA
 230 S.  Dearborn  5HS-11
 Chicago,  Illinois  60604
 (312)  886-0399
 Site Engineer:

 Joseph Danko
 Project Manager
 CH2M Hill
 2300 N.W. Walnut
 P.O.  Box 428
 Corvallis,  Oregon
 (503)  752-4271
                   97339
Vendor:

James Malot/Ed Malmanis
Terra Vac Inc.
4897-J West Waters Ave.
Tampa, Florida  33634
(813) 885-5374
VI. REFERENCES
2.
3.
     Danko,  J., soil vapor Extraction at a Superfund Site, CH?M
     Hill, Corvallis, Oregon,; undated.                       2
                             \
     Danko  J., soil Vapor VOC Removal System at the Verona Well
     Field Superfund Site City of Battle Creek, Michigan, CH2M
     Hill, Corvallis, Oregon, March, 1989.

     Tanaka,  J.,  Verona Well :Field Superfund Site Battle
     Creek,  Michigan Soil Vapor Extraction System-,  U.S. EPA
     First International Meeting of the NATO/CCMS Pilot Study
     Demonstration of Remedial Action Technologies  for
     Contaminated Land and Water,  November,  1987.

     U.S.  EPA Office of Research and Development, Terra Vac In
     July   1989™  Extraction system'  Applications Analysis Report,
                             1052

-------
                                 Appendix 4-C

  Soil Vapor Extraction Technology Case Studies


Venting Methods, Air Force Bases, United States
     Bioventing
     Catalytic Oxidation  Emissions Control
     In-situ Soil Venting  of JP-4 Jet Fuels
          (no final paper available)
             1053

-------
           A FIELD SCALE INVESTIGATION OF ENHANCED PETROLEUM HYDROCARBON

              BIODEGRADATION IN THE VADOSE -ZONE AT TYNDALL AFB, FLORIDA


                              Major Ross N. Miller Ph.D, PE, CIH

                                 U. S. Air Force, HSD/YAQE
                                Brooks Air Force Base, Texas                ,


                                Robert E. Hinchee, Ph.D, PE

                                  Battelle Memorial Institute
                                     Columbus, Ohio


                                 Captain Catherine M. Vogel

                               U. S. Air Force, AFESC/RDVW
                               Tyndall Air Force Base, Florida


                                   R. Ryan Dupont, Ph.D

                                   Utah State University
                                       Logan, Utah


                                  Douglas C. Downey, PE

                                 Engineering-Science,  Inc.
                                    Denver, Colorado
                                      ABSTRACT

                  is,.effective forthe Physical removal of volatile hydrocarbons from unsaturated
vtiifronc .f h °i'Ve 3Ia S°"rce °f °Xy9en for biol°9ical degradation of the volatile and non
volat e fractions of hydrocarbons in contaminated soil. Treatment of soil ventinq off-qas is
                n     5/0 °f  0il Ventins remedi*tion costs,  in this r^S
                                        were investisated- with the 90al of
       A seven-month field investigation was conducted at Tyndall Air Force Base (AFB)  Florida
               it510^6 had rSf !ted in contamination of a s'andy soil. Th^ ^con ^amSd a?ea was
      toi30 to^ oSPS!a?y V6 mStf ^ °f TSatUrated SoiL Soil hydrocarbon concentrations
ranged from 30 to 23,000 mg/kg. Contaminated and uncontaminated test plots were vented for 188
                                         1054

-------
days. Venting was interrupted five times during operation to allow for measurement of biological
activity (CO2 production and 62 consumption) under varying moisture and nutrient conditions.
       Moisture addition had no significant effect on soil moisture content or biodegradation rate.
Soil moisture content  ranged from 6.5 to 9.8%, by weight, throughout the field test. Nutrient addition
was also shown to have no statistically significant effect on biodegradation rate.  Initial soil sampling
results indicated that naturally occurring nutrients were adequate for the amount of biodegradation
observed. Acetylene reduction studies, conducted in the laboratory, indicated a biological nitrogen
fixation potential capable of fixing the organic nitrogen, observed in initial soil samples, in five to eight
years under anaerobic conditions.  Biodegradation rate constants were shown to be effected by soil
temperature and followed predicted values based on the van't Hoff-Arrhenius Equation.
        In one treatment cell, approximately 26 kg of hydrocarbons volatilized and 32 kg biodegraded
over the seven-month field test. Although this equates to 55% removal attributed to biodegradation,
a series of flow rate tests showed that biodegradation could be increased to 85% by managing air flow
rate   Off-gas from one treatment cell was injected into clean soil to assess the potential for complete
biological  remediation. Based biodegradation rate data collected at this field site, a soil volume ratio of
approximately 4 to 1, uncontaminated to contaminated soil, would have been  required to completely
biodegrade the off-gas from the contaminated soil.
        This research indicates that proper ratios of uncontaminated to contaminated soil and air flow
management are important factors in influencing total biodegradation of jet fuel, substantially reducing
remediation costs associated with  treatment of soil venting off-gas.


                                      INTRODUCTION
 Background Information

        Approximately 3.6 x 1012 kg (4 billion tons) of hazardous materials are transported annually in
 the United States, and of this amount about 90% consists of gasoline, fuel oil, and jet fuel.
 Massachusetts officials  report, that in 1984, 58% of reported spills in their.northeast region were
 petroleum products, of which 28% were gasoline, diesel, or fuel oil. Assuming that Massachusetts is
 representative of the rest of the United States, transportation and transfer of petroleum products,
 particularly fuels, pose a major risk to the environment (Calabrese et al.,1988a).
        In addition to transportation of fuels, leakage of stored fuel has proven to be a serious
 environmental problem, particularly as a source of ground water contamination. The United States
 Environmental Protection Agency (U.S. E.P.A.) estimates that there are three million underground
 storage tanks in the United States, of which, 78% (2.3 million) are used to store fuel products. Based
 on a random sampling,  EPA estimates that 35%, or approximately 820,000 underground fuel tanks
 are leaking (Calabrese et al.,1988a).
        A recent report indicates that there are three to five million underground storage tanks used
 to store liquid petroleum and chemical substances and that EPA estimates 100,000 to 400,000 of
 these tanks may be or have been leaking.  The majority of these tanks have been gasoline or other
 petroleum distillates (Camp, Dresser,  and  Mckee, 1988). Based on the percentages quoted above,
 the estimate for leaking underground fuel tanks could go as high as 1.4 million. The disparity in
 estimating the number of leaking underground fuel tanks underscores our inability, to date, to
 accurately quantify the magnitude of the problem. Even using minimum estimates, leaking
 underground fuel tanks pose a significant  threat to the environment.


 Hazards Associated With Fuel

       ' Although the general public appears less concerned with fuels than with industrial chemicals,
 regulatory agencies have long been aware of the threat to public health that these fuels pose. The
 U.S. Coast Guard and U.S. E.P.A. have attempted to characterize the toxicological hazards  associated
 with petroleum contamination. They concluded that exposure to petroleum products resulting from
 contaminated soils may occur via the following routes:  inhalation, dermal absorption, ingestion from
 consuming contaminated soil, consumption of plants and animals that have assimilated petroleum
 products, and by consumption of contaminated drinking water (Calabrese et al.,1988b).
                                            1055

-------
  nrinr-h,                 U/,?- Coast Gu*? and U. S. EPA ranking systems resulted in a list of 25
  priority contaminants of public health concern found in petroleum products as shown in Table 1
  Hoag s findings support the U. S. Coast Guard and U. S. E.P.A. He reports finding at least eight listed
  hazardous constituents in gasoline (Hoagetal., 1984).
  Table 1 . Priority contaminants identified in petroleum products
        Heavy Metals
                            Halogenated
Nonhalogenated
    Aromatic
Nonhalogenated
    Aliphatic
Cadmium
Chromium
Tetraethyl lead
Tetramethyl lead
Zinc

1 ,2,-dibromoethane
Dichloroethane
Dichlorobenzene
Tetrachloroethyiene
Trichloroethylene
; PCBs
Benzene
Benzo (alpha) anthracene
. Benzo (beta) pyrene
Phenol
Toluene
Xylene
Heptane
Hexane
Isobutane
• Isopentane
1- Pentene

 Adapted from Calabrese et al. (1988b).                  ~ --- : -

         Jet fuels have received less attention in the literature than has gasoline. The reason for this is
 unknown but may be related to the circumstances under which jet fuel is transported and used
 Millions of liters of jet fuel are transported daily. However, most jet fuel is delivered by underground
 pipeline or rail car directly from the refinery to the user. There have been major jet fuel releases
 although most have not been published in the literature. Approximately half of the chemically '
 contaminated sites on Air Force installations are associated with fuels, most of which  are JP-4
 (Downey and Elliot, 1990).
 i   ^   J~rt^d is not added tp Jet fuels for octane enhancement, but one analysis revealed 0 09 ppm
 lead and 0.5 ppm arsenic (Riser, 1988).  All other metals were below detectable 'levels and no
 halogenated compounds were found. Normal hexane and heptane were measured  at 2 21 and 3 67
    oywf 9ht. respectively, and the benzene, toluene, ethylbenzene, and xylenes (BTEX) fraction'
 constituted 4.5 % by weight. Aromatics totaled 17.6% of the mixture by weight (Riser 1988)
 Seventy-six major components of JP-4 were identified in this analysis, but  as many as 270 different
 components have been reported in other studies (Mason et al., 1985)
 f..»i o^i,Th ore^ay be, s'9unificant environmental health and safety hazards associated with subsurface
 fuel spills.  Pathways for human exposure are through ground water contamination resulting from
 solubiliza ion of normal and substituted alkane, alkene, and aromatic hydrocarbons and exposure to
 toxic levels of vapors trapped in occupied, confined spaces. Explosion from vapors,  which move by
 advection and diffusion to a confined space containing a source of ignition (i.e., basements) is the
 greatest potential safety hazard resulting from subsurface fuel spills (Hoag  and Cliff, 1988).
 BTEX Contamination of Ground Water


        BTEX are the contaminants in fuels which most often result in contamination and
 abandonment of subsurface drinking water supplies.  This fact results from the relatively high solubility
 ?   •  S«5?ma]ICS in water> couP'ed with the low allowable aqueous-phase maximum-contaminant
 levels (MCLs) due to their known or suspected carcinogenicity.
        Dissolved benzene, toluene, and xylene resulting from gasoline contamination have been
 riKSS  T™ J!nnW,at,e/nWne™  at .concentrations of 14, 10, and 10 mg/L, respectively (Hoag and
 Cliff, 1988).  A 38,000 L (10,000 gallon)  release from a gasoline station in Bellview, Florida caused the
 abandonment of the entire Bellview drinking water well field based on the BTEX fraction found in
water samples.  BTEX concentrations in soils collected during construction of monitoring wells ranged
from 894 to 388 mg/kg (Camp, Dresser, and McKee, 1988).
       Conner (1988) indicated that  fuel leaks as large as 1 million L (270,000 gallons) have
occurred but that leaks in the 75,000 to 200,000 L (20,000 to 50,000 gallon) range are more
                                         1056

-------
common.  Considering the damage resulting from the 38,000 L (10,000-gallon) release at Bellview
Florida, the typical 75,000 to 200,000 L (20,000 to 50,000 gallon) spill is environmentally significant.
He also stated that soil can hold up to 70 L of gasoline/m3 (0.5 gallons of gasoline/ft3) and that 3.8 L
(1 qallon) of gasoline can render 3.8 million L (1 million gallons) of water unsuitable for consumption.
This conclusion results from the fact that if 3.8 L (1 gallon) of gasoline containing 1% benzene were
added to 3.8 million, L (1  million gallons) of water, the benzene concentration would be approximately
7 uq/L (ppb) and would be unfit for human consumption based on the current MCL of 5 u,g/L (ppo)
(Pontius, 1990). However, this analysis assumes complete benzene solubilization and ignores
partitioning and kinetics.                                                           '
        One percent benzene in fuel is not uncommon, and much higher levels have been
measured  The American Petroleum Institute (API)/EPA reference fuel, PS-6, contains 1 .7 %
benzene, 4 % toluene, and 9.8 % ethylbenzene and o- m- p- xylene by volume. The total aromatic
fraction was measured at 26.08 % by volume (Calabrese et al.,1988b). The BTEX and total aromatic
concentration in gasoline varies significantly from refinery to refinery and batch to batch.  The fraction
of BTEX in gasoline has been reported to range from 6.4 to 36.4% by weight (Riser, 1 988) .
        Additional research indicates that the Bellview well field and others affected by fuel spills will
be closed for long periods of time unless remediation of the unsaturated-(vadose-) zone is
successful  First, work by Wilson and Conrad (1 984) shows that 1 5 to 40% of the pore space can hold
fuel  This means that 38,000 L (10,000 gallons) of gasoline can be held in a cube 9mx12mx9m
(30ftx40ftx30ft). Malot and Wood (1985) describe a multi-phase transport model by Baehr and
Corapcioglu that predicts benzene from a typical gasoline spill will be leached into water for about 20
years and other components would take several decades longer to  be  removed through water
flushing  Although natural biodegradation may eventually mineralize most fuel contamination, the
process is frequently too slow to prevent ground water contamination. High-nsk sites require rapid
removal of the contaminants to protect drinking water supplies and public health.
 Vadose-Zone Remediation

        The realization that contaminated soil is a long-term source of ground water contamination has
 shifted the focus of remediation from treating contaminated ground water (pump and treat) to treating
 the source of the contamination in the vadose-zone.  The initial remediation method employed by
 consulting firms was excavation of contaminated soil with placement in landfills or for use in asphalt
 plants  Fuel-contaminated soil is not a listed or characteristic hazardous waste, and disposal  in
 sanitary landfills is often recommended to reduced disposal costs. The cost of this alternative ranges
 from $400 to $660 per m3 ($300 -$500 per yd3) of contaminated material (Clarke, 1987). This type of
 recommendation has been made without consideration of the listed hazardous waste components in
 fuels and the future costs associated with being identified as a potentially responsible party (PRP) in
 the clean-up effort of the hazardous waste or sanitary landfill. Increased restrictions by EPA on
 landfilling of hazardous waste and the risk of being identified as a PRP in a hazardous waste or sanitary
 landfill cleanup have' led to the emergence of excavation coupled with incineration technology.
 However this approach is extremely expensive at $1300 to $2600 per m3 ($1000 to $2000 per  y&),
 making this alternative cost prohibitive for large volumes of contaminated soil (Clarke, 1987).
         Excavation is not only expensive but may be impossible if contamination extends beneath
 buildings or across property lines.  If contamination is deep, the size of safe excavations may be
 prohibitive  (Bennedsen et ah, 1987). Numerous failures at hazardous waste landfills together with the
 inability to excavate many sites has sparked increased emphasis for on-site clean-up technologies. In
 many cases, on-site treatment technologies have proven to be less expensive than off-site
 alternatives and, if feasible, are usually preferred by U.S.E.P.A. (U.S.E.P.A., 1989). Technologies for
 in situ  remediation of vadose-zone fuel hydrocarbon contamination include soil washing, radio
 frequency  (RF) heating of soil, soil venting, and enhanced microbial degradation.
         Soil venting is a technology that has been proven effective for the physical removal of volatile
 compounds such as gasoline and TCE from the unsaturated-zone.  However, soil venting produces
 an effluent which may require expensive treatment prior to discharge. This off-gas treatment step
 frequently constitutes a minimum of 50% of total remediation costs.  In addition, volatilization of
 contaminants through soil venting alone is not effective in the removal of nonvolatile or low volatility
 components of jet fuel. This research explored the possibility of reducing or eliminating expensive
                                            1057

-------
                                  '°W
                                                    contamination of vadose-zone soi.s through
                                                                fuel
                           AIR  FORCE  RESEARCH OBJECTIVES

         The Air Force stores and transports 1 1 x 1 09 L (3 x 1 09 gallons) of
                                  boeen identified durin9 Phase " ^
                                  Pr°9ram (|Rp)- Jp-4 ''s less volatile than gasoline and contains a
                        fractlon (Mas°n e' a'- 1985).  This research, funded by the Air Force
                  h    * W!th^SanCTed bioreclama«on through soil venting at Hill Air Force Base
  h,«    (HlI?e,?,eXa!- 1989)'  This research direction resul*ed from the apparent failure of
  hydrogen peroxide H2O2) to adequately deliver oxygen at JP-4-contaminated sites studied at Kelly
  AFB, Texas, and Eglm AFB, Florida (Downey and Elliot, 1990). As an alternative approach Air Force

                                                        as an
                        18 Pr°J   'S a     eValUati°n and demonstration of this in a*/  technology.
         1.
         2.
        3.
 la™ ncc
 large users of
to evaluate the potential for enhanced biodegradation of JP-4 in the
vadose-zone as the result of soil venting and incremental effectiveness
observed with addition of nutrients and moisture,

to evaluate the relationships among air flow rate, biodegradation
and volatilization to determine minimal aeration rates required to maintain
aerobic conditions  for maximizing biodegradation and minimizing volatilization
3 no                         i                                        *

to evaluate the potential for biodegradation of hydrocarbon vapors
(off-gas) in uncontaminated or less contaminated vadose-zone soil as an
alternative to expensive above-ground off-gas treatment.

         result was to develop sufficient information to allow the Air Force and/or other
        f"-' mixtures to progress to full-scale implementation of the technology.
                               MATERIALS  AND  METHODS
 Site Description
               'tu -,field dem°nstration of enhanced biodegradation through soil venting was
                 ' 6 °! an abandoned tank farm located on Tyndall AFB, Florida. The s te is
   undpp  TlSTp^P;4' atn5free Pr°duct has bgen observed floating on the shallow
 ground water table. Tyndall AFB is located on a peninsula that extends along the shoreline of the Gulf
 9 1S£Mem«l£L^t °f the n0*?^™**- The ^est ground on the peninSs 7 6?o
 nr^il nf P.o°Jl!   f u63" S6a level; The uPPermost sediments, at Tyndall AFB, are sands and
 ff?h! l-?l Plei?oc,e"e to.uH°'°ucene a9e (Environmental Science and Engineering, Inc., 1988). Soils
 m«rf!«  .       * dfscribed bv tne Mandarin series consisting of somewhat poorly drained
 moderately permeable soils that formed in thick beds of sandy material (U S DA  1984)
       The climate at the site is sub-tropical with an annual average temperature of 20 5° C  (69° F)
 Average daily maximum and minimum temperatures are 25° C and 16° C (77° F and 610R
 J?5?SmIVely; TemPeratures °f 32° c (90° F) or higher are frequently reached during summer months
 Is  40 cmSspfn^ ^ (1°°°-F) ^ ^^ °n'y rarely  Avera^e annual rainfa» at Tynd^S AFB
2,nfh f«  (   5 mc?es) W^-h aPP,roximate|y 125 days of recordable precipitation during the year. The
Satirtng-cUnHWater on.Tyndal"AFB va"'es from about 0.3 to 3.0 m (1  to 10 ft). The water-table
elevation nses dunng periods of heavy rainfall and declines during periods of low rainfall  Yearly
                                        1058

-------
fluctuations in ground water elevations of approximately 1 .5 m (5 ft) are typical t^"™*^®
and Engineering, Inc., 1988).  Prior to dewatering at the site, the water table was observed to be as
shallow as 46 cm (1 .5 ft).
Field Testing Objectives
        A seven month field study (October,' 1989, to May, 1990) was designed to address the
following areas:
            1.  Does soil venting enhance biodegradation of JP-4 at this site?
            2.  Does moisture addition coupled with soil venting enhance
               biodegradation at this site?
            3.  Does nutrient addition coupled with soil venting and
               moisture addition enhance biodegradation at this site?
            4.  Will the hydrocarbons in the off-gas biodegrade when
               passed through uncontaminated soil?
            5.   Evaluation of ventilation rate manipulation to maximize
                biodegradation and minimize volatilization.
            6   Calculation of specific biodegradation rate constants from a
                series of respiration tests conducted during shutdown of the air
                extraction system.
            7.  Determination of the effects of biodegradation and  volatilization on
                a subset of selected JP- 4 components.                     .  ,   .    '
            8.  Determination of the potential for nitrogen fixation under aerobic
                and anaerobic conditions.
            9.  Evaluation of alternative vent placement and vent configuration
                to maximize biodegradation and minimize volatilization.

  Test Plot DP-sign and Operation
         In order to accomplish project objectives, two treatment plots and two background plots were
  constructed and operated in the following manner:
             1 .  Contaminated Treatment Plot 1  (V1 ) - Venting only for
                 approximately 8 weeks, followed by moisture addition for
                 approximately 14 weeks, followed by moisture and nutrient
                 addition for approximately 7 weeks.
             2.  Contaminated Treatment Plot 2 (V2) - Venting coupled with
                 moisture and nutrient addition for 29 weeks.
             3   Background Plot 3 (V3) - Venting with moisture and nutrient
                 addition at rates similar to V2, with injection of hydrocarbon
                 contaminated off-gas from V1 .
  ;    ... "    4.  Background Plot 4 (V4) - Venting with moisture and nutrient
        ':         addition at rates similar to Vent 2.  .
                                              1059

-------
  Air Flow
  toct«  mAir f'°r W3S ma'n.tained throughout the field test duration except during in situ respiration
  nln Lolo orateS» WH6r? a-djU,Sted t0 maintain aerobic ^^ions in treatment plots, and background
  ffi ivS fnKfHed 3t s'milar air ret.ention times- Off-gas treatment experiments in one background
  plot (V3) involved operation at a series of flow rates and retention times
         Soil gas was withdrawn from the center monitoring well in V1 and V2 and from the onlv
  22S39 h6" ln- V3 and-^' ThiS "rtfe"^" was selected to minimize  leakage oZtsiSe air
  observed when air was withdrawn from the ends of the plots. In all but one plot, V3, atmospheric air
  Scalf°v!d £ Pass!vel enter at both ends.;  Off-gas from V1 was pumped back to the upstream
  ends of V3. Flow rates through  all test plots (were measured with calibrated rotameters


  Water  Flow

  TH* ^ J° a,"°W c°ntro!,of soil moisture, tap water was applied to the surface of the treatment plots
  Sd 2 5 !o 25°mSnS in Th* "1™*™?™ 1 ° <° 1 °° mL/min in the contaminated treatment ptots
  and 2.5 to 25 ml_/mm in the background vents. This corresponds to average annual surface
         0      f 43    30.0"1  (17 10 17° in)- Based on vacuum and oxygen measuremems in the
 Nutrient  Addition  Rates
                       nutrient add]tion was to apply sufficjent inorganic nitrogen (N), phosphorus
   rinn     i  -     LKlt0 en^ure,' as far as P°ssible- that these nutrients would not become limiting
 dunng the b.odegradat.on of fuel hydrocarbons in the test plots. Optimizing nutrient addition rates
 pmmnn-, he VFW JSf'Sl9 °f ^ phase °f the ^^^  Sodium Wmetaphosphate (Na IMP)
 ammonium chlonde (NH4CI), and potassium nitrate (KNO3) were used as sources of P, N, and K,
 r6Sp6CtJV6ly.
                              RESULTS AND DISCUSSION
 Operational Monitoring of
 Treatment Plots V1 and VP

        Treatment plots were operated for 1 88 days between October 4, 1989 and April 24  1 990
 Operat,on was interrupted only for scheduled respiration tests.  Discharge gases were monitored for
 oxygen, carbon dox.de, and total hydrocarbons throughout the operational period  The
 btodegradation component was calculated using the stoichiometric oxidation of hexane. Oxygen

 S?e?SS^S,CalS-teIe?ha8 the differenCS betW6en °Xygen in Back9<™nd Plot V4 and oxygen in
 the treatment plots.  Using the oxygen concentration in the background plot rather than atmosoheric
 ^nn,p2nfCenTa^°n' thhe njiturai biode9radation of organic carbon in u^rtaSnat^sdW
 accounted for. This method ensures that the biodegradation of fuel hydrocarbons was not
overestimated. Biodegradation based on carbon dioxide production was similarly calculated.  As the
more volatile compounds are stripped from the soil, biodegradation becomes increasingly important
overtime as the pnmary hydrocarbon removal mechanism (Figure 1).  As illustrated in Figure 1
                                         1060

-------
               100 -i
                80 -
             o>
             o
60 -
     Moisture
      addition
       to V1
                % Removal by Biodegradation - V1
                % Removal by Biodegradation - V2
Nutrient addition
    to V1
                 20
                                                  150

                               Venting Time  (days)
                                        180  210
Fiqure 1. Comparison of the percent of combined volatilization and biodegradation hydrocarbon
removal rates attributed to biodegradation (oxygen basis) in Treatment Plots V1  and V2 during the
field study.

oercentaqes of combined hydrocarbon volatilization and biodegradation removal rates attributable to
biodegradation were similar in Treatment Plots V1 and V2 throughout the experimental period, and
neither moisture nor nutrient addition appear to have increased biodegradation rates.

Respiration Tests

        Respiration Tests, 1 through 5, were conducted October 24 through 26; November 28
through December 1, 1989; January 3 through 8; March 3 through 11;and Apnl24 through 26,
1990 respectively. In addition, two limited respiration tests, 3A and 4A, were conducted from January
25 through 25 and March 9 through 12, 1990. The respiration tests were designed to determine the
order and rate of hydrocarbon biodegradation kinetics under varying conditions of moisture and
nutrient addition. Treatment Plot V2 received moisture and nutrients throughout the experimental
period and therefore served as a control for kinetic changes due to soil temperature and  other factors
not related to moisture and nutrients.  The respiration tests were conducted by first shutting down the
air delivery system to  both the treatment and background plots, followed by measurement of oxygen
consumption and carbon dioxide production over time.  Biological respiration m Treatment Plots V1
and V2 was most consistently modeled by zero order kinetics during ail respiration tests.  In a sys em
not limited by substrate, such as fuel contaminated soil, biodegradation is likely to be best modeled

by zero'°[d®J'j<2ng|.3phic^Iyri1,lustrates the zero order rate constant data obtained from the respiration
tests. In Treatment Plot V1 , the rate constant showed a significant drop between Test 1  and Test 2
 and between Test 2 and Test 3. The  rate constant significantly increased between Test 3 and Test 4
 in Treatment Plot V1  but did not significantly increase between Tests 4 and 5.  Since moisture was
 addec fto Treatment Plot V1 after T! st 2 and nutrients after Test 4, their addition would seem, without
 further analysis, to be of no benefit and even detrimental, in the case of moisture addition  In
 Treatment Plot V2, there was a statistically significant drop in the  rate constant from Test 2 to Test 3
 and a statistically significant increase  in the rate constant between Test 3 and Test 4. Although a
 depression appears in the rate constant data, there were no other statically significant differences in
          acgn                in respiration rate between Treatment Plots V1 and V2  and
 the Background Plot V4, on all tests, and between Off-Gas Treatment Plot V3 and Background Plot
 V4, on Tests 3, 4A, and 5 are illustrated in Figure 2. From the data presented, it is concluded that
                                           1061

-------
                                                                                    c
                                                                                   I
                                                                                   ^
                                                                                    o
                                                                                    D)
                                                                                    >»
                                                                                    X
                                                                                    o
           Testl
                Test 2'
                      Test 3
                            Test 4'
                                 Test 4A1
                                        Test 5'
                                                          -    0
                                                          Treatment Plot V2
                                                     Treatment Plot V1
                                              'Off-gas Plot V3.
                                         Background Plot V4
 Rgure 2. Average zero order rate constants determined by respiration tests.

 JSe9radati?n °f J'et/VeI in contaminated so|il, and biodegradation of hydrocarbon off-gas, resulted in
 statistically significant increases in respiration over that observed in uncontaminated soil.

 Potential Temperature
 on Respiration


       .lhn*!qUat'C,SySt®ms' the van>t 1°" -Arrhenius equation predicts a doubling of the rate constant
       ?VoaT*Mt"re '"creAase of 10°c- assuming typical activation energy values (Benefield and
 Randall, 1980). Using the Arrhenius constants determined from soil temperature data the rate
 constants for Treatment Plot V1 were corrected to 23 °C, the soil temperature of Test 1 (Figure 3)
 The Arrhenius correction for temperature resulted in insignificant rate constant differences between
 Tests 2, 3, 4, and 5 m Treatment Plot V2. Although a statistically significant difference in rate
 constants remained between Test 3 and Tests 2 and 5 in Treatment Plot V1 , the magnitude of the
 difference is not important from a practical application standpoint.
                                     CONCLUSIONS
       This field scale investigation has demonstrated that soil venting is an effective source of
oxygen for enhanced aerobic biodegradation of petroleum hydrocarbons (jet fuel) in the vadose-
zone. Specific conclusions are:
       1.
Operational data and respiration tests indicated that moisture (6.5 to 9.8% by weight) and
nutrients were not a limiting factor in hydrocarbon biodegradation. Soil and water samples
indicated that nutrients were delivered to the treatment plots and passed throuqh the
vadose-zone to the ground water.
                                          1062

-------
              0.008
                          .j^....95%.Confidence..lntery;aj-V1.,
                              Arrhenius Corrected Minimum k-V1
                               Moisture added   Nutrients added
               0.000
                                 Location/Test No
Figure 3  Temperature corrected (23 °C based on Arrhenius Plot) oxygen consumption rate,
constants (k) determined by respiration tests for Treatment Plot V1. Mean k is at the center of the
95% confidence interval.
        2.  Air flow tests documented that decreasing flow rates increased the percent of
           hydrocarbon removal by biodegradation and decreased the percent of hydrocarbon  •
           removal by volatilization. Under optimal air flow conditions (0.5 air void volumes per day)
           82% of hydrocarbon removal was biodegraded and 18% volatilized. Biodegradation
           removal rates ranged from approximately 2 to 20 mg/(kg day), but stabilized values
           averaged about 5 mg/(kg day). The effect of soil temperature on biodegradation rates
           was shown to approximate effects predicted by the van't Hoff-Arrhenius equation.

        3.  Off-gas treatment studies documented that uncontaminated soil at this test site could be
           successfully used as a biological reactor for the mineralization of hydrocarbon vapors (off-
           gas) generated during remediation of fuel contaminated soil using the enhanced
           biodegradation through soil venting technology investigated in this field study.  The
           average off-gas biodegradation rate was 1.34 (SD ± 0.83) mg/(kg day), or 1.93 (SD ± 1.2)
           g/(m3 day). The percent of off-gas biodegradation was inversely related to airflow rate
           (retention time), and was directly related to hydrocarbon loading rate, at the 95%
           confidence level.  Based on data collected at the field site, a soil volume ratio of
           approximately 4 to 1, uncontaminated to contaminated soil, would be required to
           completely biodegrade the off-gas from a bioventing system operated similar to this field
           project. However, if air flow rates in contaminated soil were designed to maximize
           biodegradation,  the ratio of uncontaminated to contaminated soil required would be
           proportionally less.

        4   Respiration Tests documented that oxygen consumption rates followed zero-order
           kinetics, and that rates were linear down to about 2 to 4 % oxygen. Therefore, air flow
            rates can be minimized to maintain oxygen levels between 2 and 4% without inhibiting
           biodegradation of fuel, with the added benefit that lower air flow rates will  increase the
           percent of removal by biodegradation and decrease the percent of removal by
           volatilization.
                                            1063

-------
         5.  Initial soil samples indicated that naturally available nitrogen and phosphorus were
             adequate for the amount of bipdegradation measured, explaining the observation that
             nutrient addition had an insignificant effect on the rate of biodegradation  Acetylene
             reduction studies revealed an Organic 'nitrogen fixation potential that could fix the
             observed organic nitrogen, under anaerobic conditions, in five to eight years.  '

         6.  Soil moisture levels did not significantly change during the field study.  Soil moisture
             Pin^Jf 9!?xf,o°mNI6--lt° 7-4°/0' and 8'5 to 9'8%' ^ we|9ht' respectively, in Treatment
             Plots V1 and V2. Neither venting nor moisture addition had a statistically significant effect
             on soil moisture at this site.
                       RECOMMENDATIONS  FOR  FUTURE  STUDY


 h,,^,    1° furt£er puursue.the deve'°Pment of an enhanced biodegradation of petroleum
 hydrocarbons through soil venting technology, the following studies are recommended:


         1.  Further investigate the relationship between soil temperature and hydrocarbon
            biodegradation rate.                                                              ;

         2.  Investigate methods to increase hydrocarbon biodegradation rate by increasing soil
            temperature with heated air, heated water, or low level radio frequency radiation.

         3.  Investigate the effect of soil moisture content on biodegradation rate in different soils with
            and without nutrient addition.

        4.  Investigate nutrient recycling to determine maximum C:N:P ratios that do not limit
            biodegradation rates.

        5.  Investigate different types of uncontaminated soil for use as a reactor for biodeqradation
            of generated hydrocarbon off-gas and determine off-gas biodegradation rates

        6.  Investigate gas transport in the vadose-zone to allow adequate design of air deliver/
            systems.                                                                   '.  •
                                      REFERENCES
nroo1 L- D", and C' W" RandalL 198°-  Fundamentals of process kinetics, p. 11-13. In  Biological
process design for wastewater treatment.  Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
  mn,,--y-  1987'  Use of vapor extraction systems for in-situ
removal of volatile organic compounds from soil.  pp. 92-95. In Proceedings of the National
Conference on Hazardous Wastes and Hazardous Materials,  Washington, DC.
                                          1064

-------
Caiabrese, E. J., P. T. Kostecki, and E. J. Fleischer. 1988a. Introduction, p. 1-2. In E.'J., Calabrese,
and P. T. Kostecki (ed.)  Soils contaminated by petroleum - environmental and public health effects.
John Wiley & Sons, Inc., New York, New York.

Calabrese E. J., P. T. Kostecki, and D.  A. Leonard. 1988b. Public health implications of soils
contaminated with petroleum products, p. 191-229. In E. J., Calabrese, and P. T. Kostecki (ed.) Soils
contaminated by petroleum - environmental and public health effects. John Wiley & Sons, Inc., New
York, New York.

Camp Dresser and McKee, Inc. 1988.  Final report for field evaluation of vacuum extraction corrective
technology at the Bellview, FL LUST site. Final Report No. 68-03-3409. U.S. Environmental
Protection Agency, Edison, New Jersey.                                        -

Clarke, A. N., (AWARE Inc.). 1987. Zone 1 soil decontamination through in-situ vapor_stripping
processes.  Final Report No.68-02-4446. U.S. Environmental Protection Agency, Washington, DC.

Conner,  J. R. 1988.  Case study of soil venting.  Poll. Eng. 7:74-78.

 Downey  D. C., and M. G. Elliot. 1990.  Performance of selected in situ soil decontamination
technologies:  a summary of two AFESC field tests.  In Soil and groundwater remediation. Proc. Joint
 DOE/Air Force technology review meeting., Atlanta, GA. 6-8 February 1990. Office of Technology
 Development, Department of Energy, Washington DC., Headquarters Air Force Engineering Services
 Center, Tyndall Air Force Base, FL

 Environmental Science and Engineering Inc. 1988. Installation restoration program -
 confirmation/quantification Stage 2 Volume 1 Tyndall AFB,  FL. Final Report. Headquarters Tactical
 Air Command, Command Surgeon's Office (HQTAC/SGPB), Bioenvironmental Engineering Division,
 Langley AFB, Virginia.

 Hinchee, R. E., D. C. Downey, R. R. Dupont, M. Arthur, R. N. Miller, P. Aggarwal, and T. Beard. 1989.
 Enhanced biodegradation through soil  venting.  Final Report No. SSPT 88-427. Prepared for HQ
 AFESC/RDV  by Battelle Columbus, Columbus, OH.

 Hoag, G. E., and B. Cliff. 1988.  The use of the soil venting technique for the remediation of
 petroleum-contaminated soils, p.301-316. lntE. J., Calabrese, .and  P. T. Kostecki  (ed.) Soils
 contaminated by petroleum - environmental and public health effects. John Wiley & Sons, inc., New
 York, New York.

 Hoag, G. E.( C. J. Braell, and M. C. Marley. 1984. Study of the mechanisms controlling gasoline
 hydrocarbon partitioning and transport  in groundwater systems.  USGS Final Report No. PB85-
 242907. National Technical Information System, Washington, DC.

 Malot J J  and P. R. Wood. 1985. Low cost, site specific, total approach to decontamination. In
 Proceedings of Environmental and Public Health Effects of Soils Contaminated with Petroleum
' Products, Amherst,  Massachusetts. 1985. Also p. 331- 354. In  E. J., Calabrese, and  P. T. Kostecki
 (ed.) Soils contaminated by petroleum - environmental and public health effects. John Wiley & Sons,
 Inc., New York, New York.

 Mason,  B., K. Wiefling, and G. Adams (Monsanto, Co.).  1985. Variability of major organic components
 in aircraft fuels. Engineering and Services Laboratory, Air Force Engineering and Services Center.
 ESL-TR-85-13. Tyndall AFB, FL.

 Pontius, F. W. 1990. Complying with the new drinking water quality requlations. AWWAJ 82(2): 32-
 52.

 Riser, E. 1988. Technology review - In situ/on-site biodegradation of refined oils and fuel. PO No.
 N68305-6317-7115. Naval Civil Engineering  Laboratory.  PO No. N68305-6317-7115. Port
 Hueneme, CA.
                                            1065

-------
 floS;™^ 19I9> • Bioremec|iation of contaminated surface soils. Robert S. Report No.  EPA/600/9-
 89/073. Kerr Environmental Research Laboratory. Ada, OK.
                          °f ^ C°Unty ^^ USDA-SCS"  U'S- Government Printing Office,


           "" and-,a H' ?onrad, 1984, Is physical displacement of residual hydrocarbons a realistic
          in aquifer restoration?  p. 274-298.  In  Proceedings of Petroleum Hydrocarbons and
       h  .l^M1 rGr°,U^Watfr: Prevention. Detection, and Restoration, Houston, Texas.  5-7
 November 1984. National Water Well Association/American Petroleum Institute. Dublin, OH.
                                  Biographical  Sketches
              Ross NT Miller, Ph.D. PF CIH is a Senior Bioenvironmental Engineer with the Air Force
 Human Systems Division Installation Restoration Program Office. He holds a BS in Civil and
 Komto?Sl^Xfun$ fr°mhUtan State University and an MS in Public Health/ Industrial Hygiene
 P?^rlmc OM   Ay-°cUtahnHe ha? mana9ed Occuptional Health  and Environmental Protection
 T, nSToH pS> ^r°et?aSeS^ He nrtuc® environmental contamination investigations and has
 fhPf ^        S 6f?0rtS -at hazardous waste sites.  His research interest is bioremediation of jet fuel in
 536^9001 e'x?1"9 a""aS *" °Xy9en ^^ (USAF HSD/YAQE' Brooks AFB' Texas 78235 (512)
        Bobgrt E Hinchee. Ph,p. PF= is a Senior Engineer with the Battelle Columbus Division of
 Pi3S ^ 6u°   lnStut?' ^holds a PhD in Civil and Environmental Engineering from Utah State
 KSS£     ^P8?'88 '"cludes 'nvestigatations, and remediations at more than 100 petroleum
 424-4698°) C°ntammated S'teS'  (Batte"e' 50^ Kin9 Avenue- Columbus, Ohio 43201-2693 (614)
^  .           Catherine M. Voqftl is a Project Officer with the Air Force Engineering and Services
o M In-  c  •   ds a Bf^m Civil En9ineering from Michigan Technological University and is pursuina
1« «±^"? ronm®ntal Eng-neering from the University of Arizona. Her primary interests included

Flo^rida Iliot^ SSIS?"      6aCt°r deSi9n' ("° AFESC/RDV^ Bldg. 1117, Tyndall AFB,
        P. Rvan Pupont. PhD is an Associate Professor of Civil and Environmental Engineering  and

                     '
  s  i                    n M 3Qh Wa!eo^e^earch Laboratorv a* Utah State University. He holds a
                                   o                                        .
                    u-         " and-PhD degrees in Environmental Health Engineering from the
     ,     f ^ansas' ?icu.rrent actlvlties inv°lve field monitoring and evaluation of soil vacuum
ex raction systems  and the investigation and description of field and laboratory scale vacuum
ffi wUf8,n£anced 'hP ,Sltu blol°9ical treatment of fuels and hazardous waste contaminated soils
(Utah Water Research Labvoratory, Utah State University, Logan Utah 84322 (801) 750-3227)

       Douglas 0. Downey. PE is a Senior Engineer with Engineering-Science, Inc. He holds an MS
in Civil/Environmental Engineering from Cornell University. He has worked in the U S Air Forces's
environmental protection program for the past 13 years. His interests include in-situ remediation
       6"                                           Denver' Co|orado 80204 (303) 825-
                                       1066

-------
                    CATALYTIC OXIDATION EMISSIONS CONTROL
                           FOR REMEDIATION EFFORTS

      Captain Ed Marchand, HQ AFESC/RDW,  Tyndall AFB FL 32403-6001,  USA
INTRODUCTION

     The soils and groundwater under airfield facilities are often
contaminated with jet fuel components, chlorinated solvents, and degreasers.
This contamination has resulted from past disposal practices, leaking storage
tanks, and accidental spills.  As a primary solution to this problem, the Air
Fojrce established the Installation Restoration Program (IRP) to identify
contaminated areas, determine the type and extent of contamination, and
initiate appropriate cleanup actions.  There are now over 3,500 IRP sites at
243 installations with an estimated 60% of the sites requiring cleanup action
(Reference 1).  The Engineering and Services Laboratory (ESL), part of the Air
Force Engineering and Services Center, is responsible for environmental
quality research and development of more effective, cost-efficient remedial
actions.  This research targets the development of chemical, biological, and
physical treatment systems to-meet this challenge.  This paper reports the
findings from several field tests of remediation technology where catalytic
oxidation was used to control or treat the off gasses from the effort.

CONTAMINATED GROUNDWATER REMEDIATION

WURTSMITH AFB STUDIES  ,

      In the late 1970's, trichloroethlyene (TCE), a degreasing agent, was
discovered in the drinking water at Wurtsmith AFB, Michigan.  Chemical
analyses of the groundwater showed levels of TCE exceeding 6,000 micrograms
per liter (ug/L).  The U.S. Environmental Protection Agency maximum   ;
contaminant level for TCE is  5 ug/L.  The source of the TCE was traced  to a
leaking 500-gallon underground storage tank.  Since the leaking tank went
undetected for years, the quantity of TCE leaked could only be estimated.  The
subsequent plume of TCE was determined to encompass approximately  9 million
cubic meters, with a maximum  concentration approaching 10,000 ug/L.

      A  review of the  literature  identified countercurrent packed-bed air
stripping as a possible treatment alternative.  Countercurrent packed-bed air
stripping involves flowing  contaminated water down a packed  column, while
forcing air upward through  the column.  The packing breaks up the  flow  of
water and air,  increasing the air/water contact and enhancing transfer  of the
contaminant  from the  water  into  the  air.  In many states air  emission controls
are  required  to prevent release  of  these volatiles to  the environment.

      The Environics Division of  the  ESL performed laboratory and pilot-scale
tests at Wurtsmith AFB  to verify the operating performance  of packed-bed air
stripping.  As  a  result of  the study Wurtsmith AFB currently has two air
stripping operations  underway removing TCE  from the groundwater from two
separate plumes.   A  third unit,  under construction, will remove benzene from
                                  1067

-------
  another plume of contaminated groundwater.   The initial air stripper does  not
  have any emissions control device while the other two  have (or will have for
  the benzene unit) catalytic oxidation for emissions  control.   Catalytic
  oxidation is a combustion process where the contaminant-laden air stream is
  preheated and passed through a catalyst bed.   Final  products  of the oxidation
  are typically carbon dioxide,  water,  and inorganics.

       Evaluations are underway at  Wurtsmith  AFB on the  catalytic oxidation  unit
  installed to control the  air stream coming  from the  200  gallon per minute  air
  stripper used to remove TCE from  the  groundwater.  The preliminary findings
  are shown in Table 1.  The catalytic  unit is a fluidized bed  reactor.  The
  catalyst particles are spherical  shaped and the  contaminated  air  stream is
  passed  through the reactor at  sufficient velocity  to churn  or  fluidize the
  catalyst bed.   This motion causes jthe particles  to collide  into one another
  which breaks off small pieces  of  the  surface.  Since catalyst  fouling occurs
  on  the  surface,  this type  of reactor  is contiuously self-cleaning.  This
  catalyst attrition is slow and at Wurtsmith AFB  they are still running on the
  same  catalyst  charge from  1988.

       There is  some  concern though because the Wurtsmith AFB catalyst appears
  to be forming  a  small amount of benzene when operating.  Simultaneous sampling
  of  the preheater  effluent  and  the stack emissions show an 40 - 60 percent
  increase  in  the benzene concentrations.  This is based on one sampling effort
  and is a preliminary, and puzzling, finding.  The Engineering and Services
 Laboratory is  looking further  into,the situation to understand the reaction
 mechanisms.  The vendor indicates that the benzene formation is due to a low
  catalyst bed volume  (not enough residence time for the  air to contact the
 catalyst).  This will be verified in the near future.

               TABLE 1.  PRELIMINARY  DATA FROM THE EVALUATION OF A
              CATALYTIC OXIDATION CONTROL UNIT  AT WURTSMITH AFB MI
   AIR STREAM CONCENTRATIONS:
1 part per million Trichloroethylenej 10 parts
per billion of 1,2 Dichloroethane
   CATALYTIC OXIDATION UNIT SPECIFICATIONS:
            CAPACITY:   1200 cubic  feet  per minute
            OPERATING  TEMPERATURE: -700 °F
            NATURAL GAS CONSUMPTION  (ave):   800 cubic  feet per hour
            TIME  ON STREAM:   SINCE JUNE 1988
            DESTRUCTION EFFICIENCY OF TCE (as of Feb 1990)-  >97%
            PURCHASE PRICE:   $113,000

EGLIN AFB STUDIES

     In 1988-1989,  at  a large jet fuel  spill site on Eglin AFB, Florida, we
evaluated (Reference 2) different packing materials for conventional
counter-current  air stripping operations and compared their performance to a
new rotary  air stripper.  In addition several emissions control options were
also evaluated.  The groundwater at! the site contained a large variety of
soluble jet fuel components as well; as inorganic materials that greatly
affected the research  effort.  Table 2 lists some selected parameters from the
Eglin site.
                                  1068

-------
         TABLE  2  SELECT CONTAMINANTS AT THE EGLIN AFB FUEL SPILL SITE
            CONTAMINANT
             BENZENE
             NAPHTHALENE
             TOLUENE
             0-XYLENE
             IRON (as  Fe+2)
             SULFUR (as H2S)
AVE. CONG.
  (ppb)
   78
   70
   60
  220
 8500
 >500
  HENRY'S LAW
    CONSTANT
(atm-m-Vmole)

    .0047
    .00041
    .0059
    .0040
Not Applicable
Not Applicable
     The rotary air stripper  is  a new approach to countercurrent air
stripping.   The unit utilizes a  spinning rotor to move the water out radially
from the center of the unit as shown in Figure 1.  Clean air is forced from
the outside towards the center,  maintaining the countercurrent air/water
flow.  While removal efficiencies were similar, the rotary stripper has the
added flexibility of rotor speed to meet changing feed stream concentrations.
The disadvantage is the added cost to spin the rotor, the increased complexity
and the requirement to have the  rotor perfectly balanced during operation.
                  INFLUENT AIR
                                        EFFLUENT AIR
                                            INFLUENT WATER
                                   ROTATING  PACKING
                 EFFLUENT WATER
                 Figure 1.   Cross  section, rotary air stripper
                                 1069

-------
         Catalytic oxidation,  carbon adsorption and molecular sieves  were
  evaluated for the control of the emissions  from the  air  stripping units  at  the
  Eglin site.   The carbon units had'a very low capacity  for  the  lower molecular
  weight compounds (C4 and below).  'In addition the excess humidity from the
  air stripping effluents further  reduced  the carbon adsorption  capacity   Thus
  a  carbon bed large enough to adsorb the  emissions from the air stripping     '
  operations would have a large capital and operating  cost,  making carbon  a very
  expensive alternative at this site.   Two  molecular sieve materials, Union
  Carbide's type 9102 and 1387-53, were tested  because they  are  not impacted by
  humidity effects and they could be  regenerated  on-site with ozone.  Our data
  showed that  both molecular sieve materials  were unsuccessful' for adsorbing the
  contaminants in the air stripping lemissions.  The unfavorable  performance of
  the molecular sieves may have been  because  their pore sizes were "too small to
  allow  the contaminant molecules access to the active adsorption s-ites.

      Another emission control technique evaluated was catalytic oxidation.  An
 Engelhard pilot-scale catalytic oxidation unit was tested at the Eglin site
 The unit  uses  an electric preheater to raise the inlet gas temperature to
 1000 °F before passing  it through a precious metal fixed bed catalyst
 reaction  chamber.  The  result is on-site destruction of the organic
 contaminants.  Enough of  the hydrogen sulfide (see Table 2) was stripped out
 of the water to  cause a chemical reaction in the catalytic oxidation unit
 which effectively and rapidly deactivated the catalyst.  Total capital
 operations and maintenance cost estimates for a 100  gallon per minute air
 stripping unit, based on 99% removal of benzene from contaminated  groundwater
 are:  $3.19/1000 gallons just for the air stripping  unit, $1.70/1000 gallons '
 for catalytic oxidation of the emissions  (based on other fluidized bed data)
 or  $6.47/1000 gallons for activated carbon emissions  control.

 CONTAMINATED  SOILS REMEDIATION

      There are several methods to remediate  a site contaminated with volatile
 organics such as jet fuel. The ESL tested the efficacy of  using in  situ  soil
 venting to remove JP-4 from a contaminated sandy soil site  at Hill AFB UT
 During  the ten months of operation 1115,000 pounds of hydrocarbons were removed
 from the site.   The emissions from .this effort were sent  through one of two
 catalytic oxidation units.

     The first unit was  a 500 cubic  foot per minute fluidized bed unit that
 operated for  eight months.  The second was a 1000 cubic foot per minute fixed
 bed unit  that used a precious metal  catalyst and was operated for six months'"
 Thus there was  a  period  of four months where the two units operated  together*
 to treat  the  venting  off gases.  The fixed bed was operated between 470 and
 625 °F while  the  fluidized bed unit! was operated between  625 and 700 °F
The results (reference 3) show that1 the fluidized bed unit had an average 89%
destruction efficiency and the fixed bed unit had a 97% destruction
efficiency.  This gives  a cost-per-yolume-treated rate of $23.80/million
air and $29.80/million ft3 air for the fixed and fluidized bed units
respectively.
                                  1070

-------
     While the fixed bed unit appears economically feasible it has it's
limitations.  The unit would not be able to handle a large flow rate of the
initial highly concentrated air stream.  This is because the process is one of
oxidation or burning of the contaminants.  That means releasing heat in the
process.  Fixed beds could get so hot that they actually melt the end of the
bed.  Temperature safety controls prevent this from happening, however it does
limit the amount of contaminant you can treat.  The fluidized bed Unit,
because of the better heat transfer, can handle the higher concentration flow
rates, up to a point.  The draw back is the need to add catalyst.
Approximately 150 pounds of catalyst were added to the reactor over the eight
month operation at Hill AFB UT.                                    "

CATALYST DEVELOPMENT AND TESTING

     Two laboratory studies are now being conducted to investigate catalysts
resistant to deactivation.  The University of Akron is developing a catalyst
that resists deactivation when challenged with a chlorinated air stream.
Akron researchers have found that chromium oxide and vanadium oxide materials
can reach greater than 95 percent conversion of chlorinated organics to water,
carbon dioxide, and dilute hydrogen chloride (Reference 4).  They are  ;
continuing their research to find a superior catalyst that is resistant to
chlorinated organics and sulphonated compounds present in air-stripping
emissions.

     The second study is being done by the Research Triangle Institute (RTI>,
N.G.  They are evaluating off-the-shelf catalyst formulations from five
manufacturers.  The initial step was to create a standard catalyst testing
protocol from which future catalyst formulations can be compared to this
study.  The goal is to find out which catalyst is the best for a given
contaminated air stream.

     After the catalyst has deactivated from constant exposure to a
synthesized air-stripper emissions stream, RTI will determine what caused the
catalyst to deactivate, which operating procedures will minimize deactivation,
and whether the catalyst can be effectively regenerated.  This information
will be used in in an economic comparison of the different catalysts.
Catalyst formulations being tested are the ARI Econocat, a copper chromite
formulation from Harshaw, Carulite from Carus Chemical, three supported noble
metal catalyst formulations from UCI, and a Haldor-Topsoe catalyst.

WHERE WE'RE HEADED - CROSSFLOW AIR STRIPPING WITH CATALYTIC EMISSIONS CONTROL

     Crossflow air stripping is a packed-column aeration process which
involves changing the air flow path of a conventional countercurrent tower.
The main change is the placement of baffles inside the tower which causes the
air to flow in a crisscross pattern up through the packing (Figure 2)., This
forces the air to flow at 90 degrees to the flow of contaminated water rather
than in completely opposing directions, as in a countercurrent tower.  Proper
selection of baffle spacing can produce a marked reduction in gas velocity,
lowering gas-phase pressure drop, and reducing blower energy costs compared to
conventional countercurrent mode of operation.
                                  1071

-------
      Results show the  crossflow tower can greatly reduce  the blower energy
.costs (Reference 5).   However, for the highly volatile  compounds, the blower
 energy cost is not a significant factor in the total  cost;  therefore, a
 countercurrent tower would be just as cost-effective  as a crossflow tower
 Blower energy costs do have a significant impact on the total cost of air
 stripping for the low  and moderately volatile contaminants  such as 1,2
 Dichloroethane and Methyl Ethyl Ketone.  Therefore, the crossflow tower could
 be more cost-effective for removing these compounds from  groundwater.
                      Liquid  In
              Gas Out  j  I  I
                        t  I  I
Gas Out
         Liquid In
               Gas In
                      Liquid Out

                COUNTERCURREN"
 Gas  In
        Liquid Out

      CROSSFLOW
     Figure 2.  Comparisons  of  Grossflow and Countercurrent  Air  Strippers
                               1072

-------
     A field study demonstrating the removal efficiency of crossflow air
stripping for low and semi-volatile organics will be conducted during 1991  and
1992.  During this test field validation of the RTI catalyst selection
procedure and the University of Akron formulations will be carried out as
emissions control from the crossflow air stripping operations.

REFERENCES

1.  Statement of Mr. Gary D. Vest, Deputy Assistant Secretary of the Air Force
(Environment, Safety and Occupational Health) to the Readiness, Sustainability
and Support Subcommittee of the Senate Armed Services Committee, 4 April 1990.

2.  AFESC, Air Stripping and Emissions Control Technologies;  Field Testing of
Gountercurrent Packings. Rotary Air Stripping. Catalytic Oxidation, and
Adsorption Materials, under publication.

3.  AFESC, Field Demonstration of In Situ Soil Venting of JP-4 Jet Fuel Spill
Site at Hill Air Force Base, under publication.

4.  AFESC, Vapor-Phase Catalytic Oxidation of Mixed Volatile Organic .
Compounds. Greene, H. L., ESL TR 89-12, Sep 89.

 5.  AFESC, Laboratory Investigations of Cascade Grossflow Packed Towers for
Air Stripping of Volatile Organics from Groundwater. under publication.
                                  1073

-------

-------
                               Appendix 4-D




Soil Vapor Extraction Technology Case Studies
        Additional Case Studies, United States
          1075

-------
                            APPENDIX 4-D

               ADDITIONAL CASE STUDIES,  UNITED STATES
 Appendix 4-D.i   Groveland Well site

 Located in Groveland, Massachusetts, this site is contaminated
 principally with cutting oils, and chlorinated degreasing
 solvents.  Contamination appeared to be the result of a leaking
 storage tank and improper practices in machine shop operations
 and in handling of cutting oils and degreasing solvents.  The
 total contaminant level was estimated to be 1,350 to 13,500 kg
 (3,000 to 30,000 Ibs) of VOCs, by the Massachusetts Department of
 Environmental Quality Engineering and EPA Region I.

 The site is underlain with a sloping bedrock about 12 to 15
 meters<(40 to 50 ft)  below the higher elevations of the site.
 The soil above the bedrock consists of medium sand and gravel
 near the surface,  a clay layer 0.9 to 2.1 meters (3 to 7 ft)  in
 depth and below this, coarser sand with gravel.

 The results of VOC analysis from 27 soil  borings indicated that
 heavy subsoil contamination existed below the southeastern
 portion of an existing building on the site.   The primary
 contaminants detected in the soil were trichloroethylene (TCE)
 and methylene chloride (MC).   The compounds  1,2  trans-
 dichloroethylene (TDCE),  1,1-fcrichloroethane  (TCA),  and
 tetrachloroethylene (TTCE)  webe detected  in  lesser
 concentrations.  Although thebe compounds were not used at the
 site,  they could have been components  of  the  TCE and MC
 degreasing agent solution or Could have resulted as
 biodegradation products  of these degreasing agents.   The highest
 concentration of VOCs were detected at a  depth of between 1.2 to
 3.6  meters  (4 to 12 ft)  in thfe vicinity of the oil  storage area.
 This area  generally lies  above a clay  lens.   The  clay lens is
 located  approximately 1.5  to  3.7  meters (5 to 12  ft)  below grade
 and  extends  under the same area in which  the  highest
 concentration of VOCs were detected and served as a  barrier
 against  infiltration  of VOC  into  underlying subsoil  strata.

 This site was the location of  a U.S. Environmental Protection
 Agency's Superfund InnovativejTechnology  Evaluation  (SITE)
 demonstration test.   SVE system test objectives were to  determine
 how well the  technology would  remove VOCs  from the vadose  zone-
 to assess effectiveness in various  soil types; to correlate
 declining recovery rates with  time; and> to correlate VOC
 concentration  in soils with those  in extracted vapors.   The tests
were designed to focus on the[periphery of the main  zone of
contamination.  The system consisted of four extraction wells and
                               1076

-------
four monitoring wells.  Each well was installed in two sections,
one section just above the clay lens and one section just above
the water table.  Extraction wells were screened above the clay
and below the clay with each section operating independently of
the other.  The collected vapors were passed through a
vapor/liquid separator and then passed through activated carbon
columns prior to discharge to the atmosphere.  The separated
liquid was stored and periodically removed for appropriate
disposal by a tank truck.

The results of gas sampling indicates a total of 588 kg (1300 Ib)
of VOCs were extracted over the 56-day period of SVE operation.
Subsequently, soil borings were taken and the VOC levels in these
samples were not detectable.

Appendix 4-D.2   Paint Warehouse Fire Site, Dayton, Ohio

In 1987, a fire occurred at a paint warehouse in Dayton, Ohio.
The warehouse had an inventory estimated at 5.7 million liters
(1.5 million gallons) of paint, paint thinners, and associated
products.  Because this facility was located over the drinking
water well field, water was not used to extinguish the fire.
However, uncombusted material still penetrated the soil and
spread through the unsaturated zone.  Several VOCs were detected
in nearby city wells less than six weeks after the fire.

The site is made up of glacial outwash material and the water
table is approximately 14 to 15 meters (45 to 50 ft) below grade.
The soil types were principally sand and gravel with occasional
thin strata of cobbles or clay.  The ground water flows toward
the Great Miami River, however, the river contributes to the
aquifer recharge by the pumping the municipal wells.

Soil contamination was comprised mainly of acetone, methyl
isobutyl ketone  (MIBK), methyl ethyl ketone (MEK), benzene,
toluene, xylenes, and other volatile aliphatic and alkylbenzene
compounds.  More than 45 VOCs were identified in gas samples from
the soil.  Due to infiltration, the presence of the water soluble
VOCs such as acetone, MEK, and MIBK, represented a threat to the
groundwater supply.

The soil vapor extraction system, employed at this site,
consisted of injection and extraction wells, heated headers from
extraction wells, 8 blowers, various plumbing, valves, gauges and
sampling ports, an air/water separator, and monitoring wells.
The system was installed as four identical modules with 2 blowers
each — one for extraction and one for injection.   Concrete,
which was already in place and a clay cover were used to provide
an impermeable cap.  Initially, vapor treatment consisted of
burning the extracted vapor.  Vapor treatment was discontinued
when the emissions rates decreased to State of Ohio approved
levels.  An air/water separator had to be installed when
excessive water was pulled from the soil along with the vapors.
                               1077

-------
 After 56 weeks of operation, over 3600 kg (8000 Ib) of VOCs had
 been recovered.  By April 1988, composite off-gas VOC
 measurements had fallen to less than 1 ppm.   All perched water
 extracted from the vacuum wells yielded VOC concentrations below
 action levels accepted by Ohio EPA.  At the same time, VOC
 concentrations in the unconfined aquifer had, for almost two
 months,^been at nondetectable levels for all hydrocarbon species
 identified prior to treatment.  As closure levels,  developed in a
 risk assessment during the fall months of 1987, were met, the SVE
 operations in that area of the site was halted.  (Ground water
 VOC action levels established were: acetone, 810 ug/L; MIBK, 260
 ug/L; and, MEK 450 ug/L.)   Thus in April 1988,  two  of the four
 systems  were shutdown.  By June 1988,  all off-gases extracted
 from the shallow, unsaturated soil were below 1 ppm.  Extraction
 operations were then halted and the site was subsequently closed.

 Appendix 4-D.3   Twin Cities Army Ammunitions Plant Sites

 This case study covers two sites (Site D and Site G) located at
 the Twin Cities Army Ammunitions Plant (TCAAP), in  New Brighton,
 Minnesota.  Site D was a solvent leaching pit/burn  area and Site
 G x^as an active landfill from the 1940's to  the 1970's,,

 Site D covers an area of 0.2 hectare (0.5 acres)  and the surface
 soils are composed of Arsenal sand,  stained  sediments,  and
 residues from burning activities.   The Arsenal  sand extends below
 the site to a depth of approximately 36 meters  (120 ft).   Below
 this is  a layer of Hillside sand.   The groundwater  lies
 approximately 49.5 meters  (165 ft)  below ground surface.   Two
 separate pilot systems were installed  to test several  design and
 performance variables.   System 1 was designed to evaluate TCE
 removal  from soils with relatively low VOC contamination (less
 than 2.3  mg/kg).   This system operated at an extraction rate of
 1.1 to 1.6 mVmin  (40 to 55 cfm) and had a vent pipe spacing of 6
 meters (20 ft).   The second system was designed to  remove TCE at
 concentration levels up to 5,000 mg/kg.   System 2 operated at an
 extraction rate of 5.7  to  6.2  m3/min (200 to 220 cfm) and had a
 vent spacing of 15 meters  (50  ft).   The SVE  process successfully
 demonstrated removal of 750  kg (1650 Ib)  of  TCE.  Soil  sampling
 and analysis indicated  that  TCE  removal from the  stained,  less
 porous soils were not as effectively removed as from the
 unstained soils (soils  in  close  proximity to burning activities).

 Site  G consisted  of  landfill material  such as cinder, slag,  tar,
 brick, glass, metal  and wood.  Contamination consisted primarily
 of VOCs,   although some  metals  (lead, chromium,  and  cadmium)  were
 also  detected.  A full-scale SVE system was  designed and
 installed  at  Site G  consisting of  89 air  extraction  vents,
ranging in depth  from 9.6  to 10.2 meters  (32 to 34  ft); an air
manifold;  four  centrifugal blowers;  and,  a building  to house the
blowers and motor  controls.  Operation  of  the system was
periodically  interrupted for carbon  change out.  Sixteen batches
 (119,700  kg or  264,000  Ib) of activated carbon were  used for off-
                                1078

-------
gas treatment.  The removal rates showed a sharp decline during
the first few months of operation and they gradually declined
during later operations.  In April 1987, the activated carbon
off-gas treatment was discontinued (due to the lower removal
rates) and the extracted gases were vented to the atmosphere.  As
of May 17, 1990, approximately 45,000 kg (100,000 Ib) of VOCs
have been removed from Site G since remediation began in January
1986.
                                1079

-------

-------
                                     Appendix 5-A




Physical/Chemical Extraction Technology Case Studies
      High Pressure Soil Washing (Klockner), Germany
                1081

-------
           FINAL REPORT ON THE AFU (ANWENDUNGSGESELLSCHAFT FOR
             UMWELTTECHNIKEN), BERLIN,  MOBILE PLANT FOR SOIL
              DECONTAMINATION BY HIGH-PRESSURE SOIL WASHING
             Contents


             1. Introduction

             2. The High-Pressure Soil Washing Process

             3. Chemical Analyses by the "Bundesanstalt fiir
                Materialforschung und -prtifung (Referat
                Uraweltschutztechnologien)"

             4. Conclusion and; Prospects
 (Distributed at the Third International Meeting 6-9 November 1989,

Monteal, Canada)
                                1082

-------
io introduction

Since 1986 AFU (Anwendungsgesellfjehaft £Ur Usaweltechniken ffibH) is
running a mobile  plant for the cleaning of  contaminated soil by
use of high-pressure jets  of water  in Berlin.  First located on a
former scrap  yard on Treidelweg, the high-pressure soil washing
plant is now in its third location in Berlin*

During the  pilot  project on Treidelweg  1986  through 1988, 7,000
tons of  contaminated soil from three different industrial sites
have been  successfully cleaned.  Th©  noil  had  been polluted with
chlorinated   hydrocarbons,   polycyclic  hydrocarbons,   phenoles,
benzene, mineral oil,  lindan, lead, and cyanides.

Permanent  tests by order  o£ AFU  show that  the limits which had
been set for  the  reduction of pollutants have  been reached in
most cases.  These results  have  been  supported by expertises on
random  samples ordered  by  the  independent  "Bundesanstalt fur
Materialforschung  und  -prtifung"  (Federal  Agency  for  Material
Research and  Material  Testing).
 2. The High-Pressur®  Soil Waahlng;PxqpeaB.

 The  high-pressure soil washing process has been daveloped  in  the
 Netherlands   and   adapted   to  German  requirementst  by  KlSekner
 Umwelttechnik,  Duisburg,   FRG.  The  process  is  aiming  at  the
 complete   separation  of  pollutants  from  contaminated  soil   by
 by means  of  washing.  Th© remaining  fraction®  are clean gioil and a
 small  amount of  concentrated contaminants,  the latter  having to
 be disposed  in  a  toxic waste dump or incinerated.

 Prior  to the  development  of  the  high-pressure  soil  washing
 process,  pollutants adherent to soil eonld only b@ seperatsd with
 the  help of  wash active  additives, i.e. tensid@g, which created
 new  threats  to  the environment. In  the high-pressure soil washing
 process,  the soil  first gets broken up by high-pressure jets of
 water,  and  then  is  cleaned from  the  adherent pollutants.  This
 separation takes  place within a high-pressure jetjpipe.

 Within the  jet pipe, water is shooting from  a ring  of nozzles
 with pressures  of  up to   350 bar.  The  jets  conically  focus,
 creating suction by which  the  homogenised  soil  is sucked through
 the  jets'   focal  point.  During the  acceleration,  the soil is
 chruehed  to grains  and  the pollutants are  separated  from  it.
 Easily   volatile   pollutants  are   exhausted   and   absorbed   by
 activated carbon filters.  Th© other pollutants  are  dissolved in
 the  process water,  dispersed and  emulsified.  The water   gets
 processed and recirculated  into the washing process.
                               1083

-------
            hum°Be particles arp separated from the  process water.
             "8£lUablf, P°llutant* together with the  fin© particle
                    diamet,er < 0 03 am) are removed. The process

                           JSjSSE* oSES,  t5SSii:E21

 non-soluabl© stage and removed.
 3. Chemical Analyses by the "Btodesanstalt  fttr MaterialforBchuna
     und -prilfung  (Referat TfeftmltBchutztechnologien)            '
            /?.f^ thf   "Bundesanstalt  fttr  Materialforsc'hung   und
            n^deral Agency  for  Material  Research and  Material
                         0*  C°ntarainated  "d <*•»« -oil  have
 3.1  Preparation of the samples

 Samples  of  about 30  kg  were taken  either  from  the
 conveyor belts prior to and af4.r  the washing Tproc^B ol fr   th
 dump,  and  then  automatically  divided into portions  of  abou? 1.2


 In order to prepare the required number of portions , each

 2?^Ji?BO?fni"d  !irst' The  *°mogeniZationPwas  Carried  ouin
 minimal time  so that  the  easily  volatile substances would  not
 evaporate. Then the samples were*  dried and the content of wa?er
 was  analysed. Finally  the  samples were extracted by  extraction
          *"* thS  Bolution ^  concentrated  i
3.2 Analytical results

Contaminated and cleaned soil are classified into  three groups i

Class   I  -  usable
Claae  II  -  polluted, to be disposed of
Class  IV  -  heavily polluted,: to be disposed of:
                             1084

-------
Sample I

Three samples of contaminated (con.) and three samples ©f cleaned
(el.)  soil  were  taken   from   th©  running  sonvayor  belt  ©n
KanalstraB© on May 11, 1987 and analyiad.
                        Lcon,
        6 con,
     'con,
      Lel.
Jel.
                                                             3cl.
Lead
Cadmium
Phenole
AHC
MOHC
org. Cl
18.
0.
3.
474
< 1
3
2
1


18
0
2
250
2
.1
.1
.7
0.1


for


1.5
all
341
.1
<
1


samples
18
<• 1 <
4.0
<0.1
<0.1
16
1



16
< 1
PAHC;

Naphtaline
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+k)fluo-
ranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylen
 Indeno(l,2,3-cd)-
pyrsne
 Benzo(ghi)perylen

 Sum of  all P&HCss
1.6
8.6
6,0
2.5
5.9
31.3
25.7
9.7
8.0
17.0
8.9
11.5
2.0
5.4
6.2
1.7
7.0
6.2
6.3
7.9
24.6
21.0
11.2
8.2
17.9
9.4
11.8
2.2
5.4
5.2
2.1
14.5
12.2
14.8
12.8
37.5
27.8
16.0
9,9
20.1
12.0
13.9
3.1
6.7
7.4
0.4
0.3
0.2
0.7
0.4
1.9
1.4
0.8
0.6
1.1
0.6
0.7
0.1
0.2
0.2
0,3
0.1
0.2
0.5
0,3
1.8
1.5
0.7
0.6
1.1
0.5
0.6
0.1
0.2
0.3
0.4
0.2
0.2
1.3
0.5
2.3
1.9
0.7
0.8
1.2
0.6
0.7
0.1
0.2
0.4
150,3   140.8   210.8
               9.6
               8.5   11.5
 PCS (in ng/kg » ppb)

 No.  28
 No.  52
 No. 101
 No. 153
 Ho. 138
 No. 180

 Sum of all PCBsi
=
1.0
48.0
62.0
*•
0.1
0.2
0.1
2,0
8.0
5.0
1.0
m,
1.0
5.0
4.0
0.1
0.1
0.1
0.1
0.1
1.0
11.0
4.0
112.0
0.5
16.0   10.0
   0.2   16.0
                               1085

-------
 For  these  samples,  the  cleaning results  can only  be
 S??«<5?  f° thS Pf^T3110 iroMtio hydrocarbons ?PA¥C    caue
 the  initial concentrations  of other pollutants have aJreadv baen
 very low before the washing  process. Referring  to tht IIScs the
 Sn  SiS? rfSUlt  ^tettez than 99%. Also referring to the mineral
 oil  content a result better than  90% is achieved.

 While the  contaminated  samples  are 'Clase  II' soil (polluted, to
 r I JX?£?MJL   ^'uthe, 8amPlee  iof cleaned soil belong to 'Class I'
 (usable).  Thus the cleaning has been successful.



 Sample 2

 Three samples of contaminated (con.) and three samples of cleaned
 (Cl.)  soil  were  taken  from   the   running  conveyor  belt  on
 Xanalstrafle on May 27,  1987 and analysed,
                        '•con.
                 'con.
•ol.
'cl
Lead
Cadmium
Phenol®
AHC
MOHC
org. Cl


1,8
<0,1
414
<0,1
15,3
X0,l
^2,1
:420'
*^0 i 1


1,8
<0,1
346
<0,1
&•»
, *
0,1 0,1
<0,1 <0,1
16 IS
«0,1 <0,1


0,1
<0,1
11
 PAHCi

 Naphtalin©
 Acanaphtane
 Fluorene
 Phenanthrene
 Anthracene
 Fluoranthene
 Pyren©
 Ben 20(a)anthracene
 Chrysene
 B©nzo(b+k)fluo-
 ranthene
 Benzo(e)pyrene
 Benzo(a)pyrene
 Perylen
 Indeno(1,2,3-cd)-
 pyrene
Benzo(ghi)perylen

Sun of all PAHCst
1,3
0,9
0,8
6,0
1,2
11,9
27,1
S , 6
6,7
28,6
13,5
18,0
3,1
10,0
9,4
0,7
1,7
0,8
6,0
0,9
10,4
14,8
5,8
6,1
30,7
14,5
19,5
4,4
9,3
9,6
3,9
14,5
i , 3
20,5
7,8
38,4
31,8
14,6
11,6
25,0
10,9
15,8
2,4
7,0
7,1
0,3
0,3
0,3
1,0
0,3
1,8
1,4
0,8
0,6
1,9
0,9

oj2
0,6
0,6
0,3
0,1
0,1
1,0
0, 2
1,3.
1,1
0,7
0,6
1,9
(1,9
1,2
Ci,2
o-, ($;••'
0,5
0,4
0,8
0,7
1,6
1,0
3,7
2,7
1,6
1,1
2,3
1,1
1,4
0,2
0,7
0,6
144.1   135.2   220.S   12.2   10,7   19.9
                              1086

-------
PCS (in ng/fcg "

No.  28
No.  52
No. 101
No. 153
No. 138
No. 180

Sum of all PCBs:
                        10
                        15
                         5
                         7
                         9
                         5

                        51
 3
 4
 3
 4
 4
 3

21
 4
 4
 4
 5
 4
 2

23
1
1
1
1
1
1
1
1
1
1
1
1
 1
 1
 1
 2
 3
 2

10
While the  contaminated soil  belongs  to 'Class  II',  the cleaned
soil is ranking  'Class  I'.  Thus the cleaning has been successful
with a ratio of more than 90% reduction of pollutants.
Sample 3

Three samples of contaminated  (con.) and three samples of cleaned
(ol.)  soil  were   taken  from  the  running  conveyor  belt  on
Kanalstrafie on June 05,  1987 and analyzed.
 Phenols
 AHC
 MOHC
 org.  Cl
                        Lcon,
                                •con.
        'con.
                                                       2cl.   3cl.
Lead
Cadmium
11.0
< 0.1
5.0
< 0.1
                          1.5     2.2     2.3    0.1    0.1    0.1
                                  < 0.1  for all samples
                        436     449     472     34     23     42
                                  < 0.1  for all samples
 PAHCt

 Naphtaline               2.5
 Acenaphtene              2.0
 Fluorene                 1 • 7
 Phenanthrene             6.2
 Anthracene               1.8
 Fluoranthene            12.0
 Pyrene                  21.9
 Benzo(a)anthracene       6.9
 Chrysene                 7.0
 Benzo(b+k)fluo-
 ranthene                30.3
 Benzo(e)pyrene          14.3
 Benzo(a)pyrene          20.0
 Perylen                  4.4
 Indeno(l,2,3-cd)-
 pyrene                   9.2
 Benao(ghi)perylen        8.1
 Sum of all PAHCBt      148.3
                                 12.0
                                 23.7
                                 23.5
                                 47.9
                                 20.6
                                 62.2
                                 46.5
                                 22.0
                                 15.8

                                 33.9
                                 16.4
                                 20.8
                                   3.7

                                   9.9
                                   9.8
          1.5
          2.6
          2.1
          6.2
          1.8
         28.7
         44.5
          9.0
          9.8
         34
         16
         20
   ,7
   .7
   ,7
          3.9

          9.1
          8.0
 1.9
 1.1
 1.1
 2.9
 1.4
 4.0
 3.2
 1.6
 l.*3

 2.4
 1.1
 1.4
 0.2

 0.6
 0.6
                                 368.7    199.3    24.8
 1.1
 1.5
 1.6
 4,3
 1.7
 5.6
 4.0
 2.2
 1.7

 2,8
 1.3
 1.8
 0,3

 0.7
 0.6
 31.2
  0.8
  1.4
  1.3
  3.2
  1.6
  5.2
  3.7
  2.0
  1.5

  2.6
  1.3
  1.7
  0.2

  0.6
  0.6
 27.7
                                1087

-------
 PCS (in fig/kg - ppb)
 No.  28
 No.  52
 No. 101
 No. 153
 No. 138
 No. 180

 Sum of all PCBe;
7
5
3
4
6
3
6
2
3
5
5
3
2
4
1
2
3
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
•b
1
1
2
1
28
24
                                          14
Sample 4
Three  samples  of  flotation  concentrate  (flo.)   in which   the
pollutants  should be concentrated  and three  simple^ of cleaned
i   I ^S°«1  were  taken  from  the  running  conveyor  belt  on
Kanaletrafle on June 10,  1987 and analyzed.     conv®yor  belt  on
                       Lcl.
Lead
Cadmium
Phenole
Benzene
Toluene
Xylen©
MOHC
org. Cl
0.1
0.1
17
1.0
12.4
0.1
0.3
0.1 foi
0.1 foi
0,1
24
1.0
                                      0.1
                    6.0
                                      0.1     0.2
                                       46
                                      1.0
                   2350
                    1.0
                   'flo.
                   •MUMM^MM

                   94.5
                    0.5
                    4.7
                            0.3
                    1870
                    1.0
                                                             sflo.
                                                              7.4
                            0.2
2110
 1.0
                             1088

-------
                        •flo.
                          'tlo.
               3flo.
PAHC i

Naphtaline
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Ben20(a)anthracene
Chrysene
Benzo(b+k)fluo-
ranthen©
Benzo(e)pyrene
Benzo(a)pyrene
Perylen
Indeno(l,2,3-cd)-
pyrene
Benzo(ghi)perylen
Sum of all PAHCii

PCB (in ng/kg «  ppb)
No,
No,
23
52
 No.  101
 No.  153
 No.  138
 No.  180

 Sum of all PCBst
                    21
                    II
                    ??
                   U*
                    75
                   20?
                    91
                    13
                    !•'
                    112
                    29

                    40
                  1538
1
1
1
1
2
1
        104
         to
        179

        217
        201
         Sf
         69
        140
         60
         to
         IS
         34
         32
       1500
 1
 2
 2
 4
 5
 3

17
         31
         112
         160
         111

         33?
         263
         112
         .13
         112
          17
          3S
          37
        1784
 2
 3
 2
 2
 3
 1

13
       < §S1
         r,o
         6,1
         1,5
        ' 1,5

         o!i
 39
 57
 28
 28
 28
 14

194
         0,1
         §,1
         1,4-

         lit
         1,4
         •0,i

         ill
         1,0
         1,1
 30
 14
 32
 26
 24
 14

140
        0,1
        0,1
        §64
        1,1
        1,5
        1,8
        5,1
        1,1
        2,3
        11.3    13,6    19.2
 32
  9
 21
 34
 48
 17

159
 The  separation of  pollutants  and  their  concentration  in  the
 flotation concentrate have  obviously been achieved. Referring  to
 the 15 PAHCs, there are on  average 15 mg contained in  each  kg^of
 cleaned  soil  and 1200  mg  contained  in  on©  kg  of  flotation
 concentrate. These hardly volatile  PAHCs  have been concentrated
 by flotation.

 But  not • all of  the  pollutants do  accumulate  in the  flotation
 concentratej otherwise their  overall concentration would  be much
 higher.  Therefore  part  of  the  contamination  must  have  been
 removed with the exhaust air and the waste  water.
                                1089

-------
  Smuple 5
                 °B1* °°"tamin;^d  (con.)  .oil  and  three eamples  of
                  ff1  frora the  runnin9  conveyor belt,  and  three
                                               takan
                          'con.
"con.
'con.
Lcl.
cl.   3cl.
Lead
Cadmium
Phenole < 0 . 1
AHC
MOHC 60
org. Cl <.0,i
PCS (in |.ig/kg * ppb)
No. 28 1,3
No. 52 5,5
No, 101 7,0
No, 153 9,0
No, 138 12,0
No, 180 5,5
Sum of all PCBs: 40,3

Phenole
Benzene
Toluene
Xylene

MOHC
org. Cl
PCB (in ug/kg - ppb)
No. 28
No. 52
No. 101
No. 153
No. 138
NO. 180
26.0
••** i v
0,1
0,6 0,2 40,1
bei alien Proben 0,1
63 225 9,5

2,3 1,3 0,5
1,7 3,5 0,9
:7,6 5,5 1,1
16,0 9,3 1,3
21,7 11,5 1,9
12,1 6,3 0,8
61,4 37,4 6,5
1£lo. 2flo.
4,2 4,5
0, .- — _
f * 
-------
PAHCt

Napht aline
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo ( a ) anthracene
Chrysene
Benzo (b+k)£luo-
ranthene
Benao(e)pyrene
Benzo(a)pyrene
Perylen
Indeno(l,2,3-cd)-
pyrene
Benzo(ghi)perylen
0,3

4*, 2
1,0
6,1
3,$
3,?

7^1
3,3
48?
1,5
1,4
2,3
 0,3
 0,3
 5,5
 1,4
10,3
 4,7

 *I?
 f,0
 3,9
 6,1
 1,2
 i,S
 2,6
0,4
0,4
4S3
1,1
4,3
6,4
4,0
3.7
4,3
0,7
2,0
     0,4
 0,7

 ©',4
 9,7
 0,4
 0,4
€0,1
 0,2
 0,2
0,1
§,1
1,2
1,4
§,?
0,7
1.2
0,3
§,?

oU
0,4
     0,4

     oU
     0,7
     0,4
     6,2
 6
17
17
32
17

?S
If
56
64
31
41
 4
19
21
 40
 33
 69

145
12?
 60
 54

 44
 60
 12
 24
 26
 33
117
 94
160
111
290
269
131
 99i
203
 94
134
 27
 45
 35
 Sum of  all P&HCfis     45.3 63.6 48.8  5.2  9.1  5.3  508  850 1866
 Comparing the  concentrations of  «,g.  lead  in  contaminated  soil
 (26 mo/kg),  In cleaned  soil  (10 fag/Kg),  and  in  the  flotation
 concentrate (250-2600 mg/kg), it is obvious  that the accumulation
 of pollutants in the concentrate has been achieved.


 The same result holds  for  the PAHCs.  Be for®, washing, the  samples
 contain  52  mg/kgj  cleaned  soil  contains 6  mg/kg,  and  the
 flotation  concentrate 1100  mg/kg  of  the  15 analyzed  PAHCs  on
 average. In contrast  to  the easily volatile PAHCs which probably
 vaporize already when  the  contaminated soil is moved, these less
 volatile PAHCs have accumulated in  the  flotation concentrate.
 The  overall   result   for   the  assessment   of  the
 concentrate is  similar to that for  the sample of June  10,  1987.
 There  is  an accumulation of pollutants  but not -of  all  of  them
 because some are filtered from the  exhaust air by the  activated
 carbon filters.  The flotation concentrate ranks 'Class II'.
                                1091

-------
  Sample 6

              oin   thf?S  -ft«P1«  of  contaminated soil and three
              cleaned soil were taken  from  the plant's  iimuf

                                                   ^*
  The following  concentrations of  heavy metals  were
             Lead Copper Chrome Cobalt Zinc Nickel Vanadium Cadmium
con. g}
Average
W aj
Average
3con. a)
»)
Average
1_ .
cl. *)
Average
2
cl* 2)
Average

3ol. a)
b>
Average

ni'vftrt MAmi»i'
123
128
160
135
US
91
101
96
67
71
69
98
106
102

62
63
A *f
158
163
161
106
41
74
34
39
37
46
84
& &
66
91
81
86

37
32
33
waj
17
11
14
12
13
13
9
12
11
9
8
9
10
9
10

8
6
8
ber
4
4
4
4
4
4
,5 4
4
4





3

4
4
4
content (
263
287
273
241
234
238
204
210
207
1813
185
187
134
135
135

140
141
141
»>
20
26
23
26
24
25
20
24
22
23
23
24
22
21
22

22
22
22

9
9
9
8
8
8
8
8
8
8
8
8
8
7
8

7
7
7
cyanide
2
2
2
1,5
1,*
1*5

j
1
1
1
1
1
i
.

1
1
1
(ng/kg);
contaminated soil


mixed  sample of

cleaned  soil
 9.1




15.8
     0.7
MOHC
                        110
     2con.    3con.
               65
33
                                                        cl.     cl.
                                                               31
                              1092

-------
                       "•con,
                                       3
           con.
                  I
                                                cl.
                                     3el.
PAHC8,

Naphtaline
Ac^naphten©
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysen*
Benzo(b+k)fluo-
ranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylen
Indeno(l,2,3-cd)-
pyrene
Benzo(ghi)perylen
Sum of  all P&HCsi
0,7
6,2
2,1
0,5
3,9
3,7
2,3
2,4
3,5
1,3
 1,*

 i'.4
 1,3
25.4
                                 on
                                0,1
3,2
2,6
3,0
3,8

2*,0
0,3
1,7
1,1
24.9
                                              illi  ?f§&*ft ki@ine? «*s
If"
0,2
1,3
0,9

2*. 2
1,2
1,5
2,2
** t
1,3
0,3
0,8
0,6
3,4
•€9,1
0,7
9,1
1,0
1,0
0,7
0,3
1,0
0,2
0,3
•£0,1
0,3
0,2
0,1
0,1
1,t
0,2
1,1
1,5
0,8
0,7
1»3
0,1
0,7
0,2
0,5
0,4
@,1kl,
0,1
1,2
0,4
1,4
1,3
0,8
o,e
1,2
0,3
0?£
kl. 0,1
0,3
0,4
                                        16.5
6.6   .10.1
                                  9.3
From the  three  samples of contaminated  and  the three samples  of
cleaned soil, mixed samples were taken and analyzed for PCBS. The
following concentrations were found*
PCB (in mg/kg)
contaminated
                                                    cleaned
No.
No.
No.
NO.
HO.
NO.
Sum
28
52
101
153
138
180
of all PCBs:
0,014 4
0,020 7
0,021 7
0,025 4
0,034 1
0,01* 8
0,133
0,004 3
0,005 3
0,012 2
0,011 4
0,@1P 1
0,010 1
0,062 5
 In  general,  the soil had not been very heavily polluted prior to
 washing.  Only referring to its content of mineral  oil,  lead,  and
 copper  they  belong to 'Class II'.  The samples of cleaned soil ar©
 ranking 'Class  I'.  Therefore the  cleaning  has  been  successful.
 With  respect  to  the  soil's  low  initial  contamination,
 calculation  of a cleaning ratio would not be meaningful.
                               1093

-------
Sample 7
Lead Copper Chz
1con. a) 24~9~ ""68
b) 251 77
Average 250 73
2con. *) 188 76
Average 1M 6B
18? 72
3eon. a) 201 86
b) 201 71
Average 20l ??
lrl a) W 26
CJ" b) 8t 25
Average 81 26
2cl. «) 65
/ 66
Average -r
vO
3cl. a) 78
b)
Average 7$

MOHC
Naphtaline
Acenaphten©
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo ( a ) anthracene
Chryaene
Benzo (b+k)£luo-
ranthene
Benzo (e)pyrene
Benzo ( a ) pyrene
Perylen
Indeno(l,2,3-cd)~
pyrene
Benzo (ghi)perylen
Bum of all PAHCsr
24 1
29
27
50
SO
Icon.
140
n.n.
2,0
2,0
o!?
3,6
3,6
2,0
1,7
3,3
2*, 3
0,6
1,7
1,6
28.3
ome Cobalt Zinc
u 7 ToT
14 7 214
U 7 211
U 6 204
U 6 205
15 6 205
15 8 211
15 8 213
15 8 212
7 3 108
6 3 108
7 3 108
1
6
9
9
9
'con.
210
n.n.
0,7 '
0,2
1,7
0,5
3,3
2^2
1,7
3,5
1,1
1,8
0,6
1,1
1,1
22.8
2 110
2 114
2 112
3 115
3 113
3con.
214
n.n. <
1,0 <
0,6 <
2,7
0,7
4,8
4,6
3,0
2,7
4,1
1,8
2,1
0,7
2,0
1,.4
32.3
Nickel Vanadium
tV""""""u"""
14 14
14 14
16 14
16 1S
14 15
S 12
9 12
? 12
4 $
4 6
4 6
9
7
8
8
8
iol.
i !• mi — ~m
o.
w,
o,
o,
o,
o,
It
0,8
0,'5
0,5
w ^
0,8
9f f w
0,4
0,4
0,1
0,5
0,2
6.3




2cl.
31

n.n,
ia,i
1,2
1»7
1,3
0,8
•^g P v
11 2
II B Bfc
0,5
o!i
0,9
0,4
9.3
6
6
6
€
6
3cl.
31

o!
Q,
o,'
1,
1,
o,
^ F
0.
,
1 .
1 ,
o,
o,
o,
o,
0,2
8.1
                             1094

-------
Although  the  initial  pollution of  the  contaminated  soil  is
relatively low,  it  has to be classified  'Class  li<  referring to
its content  of  lead and MOHCs.  The cleaned soil  is  of  'Class I'
quality. The target has been met with a  cleaning ratio  of 70-80
Saaple 8

Three samples  each of contaminated and cleaned soil from the Air
France/Garbage Dump No.  1  Marienfelde were taken  before and 20
minutes after  the  washing,  respectively, on October 08, 1387.
                                'eon,
'con.
                                                1
cl.
2cl'.   3cl,
MOHC
PAHC:
Naphtaline
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo ( a ) anthracene
Chryeene
Benso(b+k)fluo-
ranthene
Benzo(e)pyrene
Benzo ( a )pyrene
Perylen
Indeno(l,2,3-ed)-
pyrene
Benzo ( ghi ) perylen
fiura of all P&HCat
123

0,2
0,3
0,3
2,9
0.7
v f *
4*, 4
3,S
2,6
4,1
1,4

0,4
» s
1,7
1,4
33,0
84

0,1
0,3
0,2
1,3
0,6
V J V
ijs
1,8
2,3
1,1
1 ,4
0,3
t
0,8
0,7
18,4
110

0,6
1,0
0,6
4,8
1»5

6B,3
3,7
.. 3,5
4,8
t|1
2,4
0,4
1,5
1,7
41,5
12

^f Oil
*£Q ^
0,1
0,6
0,1
0,8
0,7
0,3
0,5
O*
,6
0,2
0,2
,•
<0, '
o,'
1l<
0,2
1,2
1,5
o,<
o,:
0»
,!
o,:
o,:
n
0,!
o,:
6.S
 These   samples   are  also   relatively  low   in   their
 contamination. Except  for the mineral Oil  content they could be
 claesified  'Claee I'. The cleaned soil definitely  is ©£ 'Class I'
 quality. A  cleaning ratio of about 70 % has been reached.
 Three  samples  each  of  contaminated  and  cleaned soil  -
 -Bisstadion Wilmersdorf/HKW Moabit" (containing  light oil) -
 taken  before  and 20 minutes  after  the washing,  respectively, on
 November  10,  1987.
                               1095

-------
 Mixed samples, one of contaminated  and one of  cleaned soil,  were
 created and analyzed for their content of water and cyanides*

Water content
Total content of cyanides
contaminated
6.7 %
53.4 mg/kg
cleaned
13,9 t
6 . 5 mg/kg
                        "con,
        'con,
        'con,
Sum of all PAHCat
        Lcl.
MOHC
PftHCi
Naphtaline
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo { a ) anthracene
CHvyaene
Ben20(b-f-k)fluo-
JLCUIUUOUB
Benzo(e)pyrene
Benzo(a)pyrene
Perylen
Indeno(l,2,3-cd)-
pycene
Benzo(ghi)perylen
21
0
0
1
3
1
7
8
*
3
o
3
3
0
3
2
4
i
,
,
,
»
,
,
,
,
,
•
,
»
,
»
I
3
2
3
5
4
7
7
$
3
jt
0 ,
5
9
0
2
1
0
0
0
3
0
5
3
s
3
^
2
2
1
2
1
4
|
i
»
i
,
»
,
»
»
1
V
t
i
,
,
,
7 17
3 0,
»
3 0,
9 1,
4 2,
8 0,
0 3,
7 3,1
S 9."
« 3,:

3 2,!
•- <• 1 •
6 2,i
0 0,2
2 2,]
8 2,1
9 47
0,2
V ^ «
0,3
s!o
1,2
6,
-------
Sample 10

On November  27,  1987,  three samples  of  contaminated and cleaned
soil originating from Ktfrtestrafle, Berlin 61, were taken from tha
running conveyor belt and analyzed.
Mixed samples

contaminated soil

cleaned soil
Water content

    6.4 %

   14.5 %
Total cyanides

  0,5  mg/kg

  0.3  mg/kg
Phenol©

 < 0,1
                                                        cl.
MOHC  (mg/kg)
  1,090   2,390    3,850     121     112     113
Referring  to  their mineral oil  content the contaminated  as-well
as the  cleaned  soil are ranking  'Class II' although  the cleaning
ratio for  MOHCs  is  80-90  %. In  this  case  the  target ratio has not
been met.
 3.3 Summary of the  Results

 »amp±es  or contaminated sou. prior to tne wasnxng  process ana 01
 cleaned  soil  after  the washing process in the plant on Treidelweg
 were  chemically  analyzed for  their content  of pollutants.  The
 soil  originated  from several locations within the city of Berlin.

 The  results  show that the  goal of  receiving 'Class  I'  soil  has
 been  achieved in most cases. The cleaning ratios reached from 70
 up to 95 %, depending on the initial degree  of contamination.

 In very few  cases,  the total  content  of  cyanides could  not
 puiJTAwitsitLl^  L/o  j-srauwsd. Tlio  UwLal  wwiiLaiiL  w£ u^a-iiiUoo ^w
 some  relatively  innocuous cyanides, though. A. threat from
 cyanide  could not be stated.

 The   pollutants  accumulate  in the  residual  concentrate  of  the
 plant as it  is  intended. This residue usually  belongs to 'Class
 II',  i.e. it has  to  be disposed  of  but can be deposited in a
 toxic waste dump without problem.
                               1097

-------
 This  shows  that the  easily volatile  substances  must have  - at
   2 S ti ?iair?ly^ ~ vaPori*ed, and  that  the water-solvent pollutants
 most likely have been carried away by the waste water, As long as
 th©  exhaust  air and  the  waste  water  are duly  treated/  the
 high-pressure soil washing process is  appropriate for soil which
 IB mainly contaminated  with mineral  oil  hydrocarbons ,  organic
 solvents, polycyclic  aromates,  and  -  with reservation  -  heavy
 l'                                           '
 Finally, the main  advantage of the method is  the possibility to
 adjust its parameters  according to the soil's  structure  and the
 kind and degree  of contamination given.  If  there is only  a low
 degree of initial contamination, an  experienced operator  will be
 able to run the plant in a way that the actual cleaning ratios do
                                                        efficient
 4 i  Conclusion and Prospects

 Tha  expertises  by  order  of  AFU  as  well  as  those  by  the
 independent experts show  that  the  pilot project on  the cleaning
      ntiminted                 -eMur* soil  washing plant  on
                                    almo8t a11  cases-
               of
       <    - -              pollution  in air and water have  always
      significantly lower than the permitted limits.


       ^Pril  19<89' ?  third-9eneration high-pressure soil washing
    nnn V  run"in^ in  Dtlsseldorf-Lierenfeldf Within  six months;
 70,000  tons  of contaminated soil  are  going to be cleaned  on  the
           f°3er  tu     ollinS «i".  Compared  with tha previous
aggregates. In

          - an enclosed coarse screening machine,
          - a steam injection,
          - two additional stages in the jet pipe,
          - large-scale scavenging technology for the soil,
          - process water treatment by flotation and separation,
          - regenerable activated carbon filters with attached
            solvent regeneration

have been added to the plant.

At  the  moment,   the   project in  Dttsseldorf-Lierenfeld  is  the
largest  environmental rehabilitation  project  carried out  with
this innovative technology in the Federal Republic of Germany.
                               1098

-------
fi?    &•    ®>
«p    s    i
—    &    a
            *
      S
                                    I

                                    E
                                           §
                                                3
                                                *i
                                                «e
                                                I
                                                •t
                                                j*
                                                         o
                                                    «


                                                     V
                                                         d

                                                         v
                                               1099

-------

-------
                                      Appendix 5-B



Physical/Chemical Extraction Technology Case Studies
                      Vibration (Harbauer), Germany
                1101

-------
EXPERIENCE GAINED WITH  A  SOIL-DECONTAMINATION  SYSTEM  IN BERLIN
H.-D. Sonnen, S. Klingebiel
1. Summary

What is being presented is an extractive soil-decontamination system de-
veloped in Berlin, the HARBAUER PB 2, which has been working on" the con-
taminated grounds of Pintsch-01 GmbH (in liquidation) since the summer of
1987 and has already cleaned more than 11,000 tons of contaminated soil
at various locations. The experience and knowledge gained with the opera-
tion of this system are described, the state of the art explained and fu-
ture perspectives outlined.

2. Introduction

Between 1925 and 1984 used oil was processed on the grounds of the
Pitsch-01 GmbH, now in liquidation. A distillation facility and CP facil-
ity, amongst others, were useed for this purpose.  Some of the production
residue leaked into the soil. Leaky tanks,  production breakdowns  and
careless work led to considerable contamination of buildings, soil and
ground water (Woltmann, 1985).

3. Location

The building of Pitsch-01  GmbH, now in liquidation, is located in  a typi-
cal Berlin industrial area in the south of  the city,  in the district of
Neukolln. It is approx. 16,000 sq.m.  in size.  The  soil consists of filled
and faulted layers of soil that contained sandy, clayish silts (drift
clay) to a depth of about  6 m followed by fine- to medium-grained  sand
down to a depth of 12.0 m. Between 12.0 and 14.8 m there are very   sandy,
gravelly and clayish silts (drift marl).  Underlying this drift marl  are
medium-grained and coarse  sands with  gravelly supplements.  These very
pervious sands are the gound-water carrier  proper. Layers of soil  and the
ground water are thoroughly contaminated with mineral  oils, phenols, CHC,
PCBs and PAHs

Table 1 provides a survey  of the values determined at the time the
clean-up began.

Since 1924 Pintsch-01 Berlin GmbH had been  collecting and processing used
oil from all over Berlin on the aforementioned property. The management
of the company allowed all the residue and  waste products resulting from
the processing of the used oil leak to away in unsealed pits in the
southern part of the premises. Since  used oil  containing PCBs was  also
treated during the reraffination process, PCBs made their way into the
production cycle, and thus into the soil  as well.  -Today's contamination
                                    1102

-------
of the soil, ground water and buildings was caused by this type of waste
disposal, by leaks in tanks and storage receptacles as well as by several
fires in the production buildings.

In March of 1982 the Senator for Urban Development and Environmental Pro-
tection commissioned a report to determine the soil and ground-water con-
tamination on the company's premises. The report was submitted in March
1983.
From this report it can be seen that 12,000 sq.m., i.e. some 75 % of the
property, are badly contaminated with mineral-oil products, chlorinated
hydrocarbons. Oil in phase was discovered on the ground water at a depth
of approx. 8.5 m under the ground over an area of 6,000 sq.m. As later
measurements showed, the oil floating on the ground water had local lay-
ers up to 3 m thick and was contaminated with biphenyls (PCB), polycyclic
aromatic hydrocarbons (PAH), chlorinated hydrocarbons CHC) as well as
phenols. Moreover, polychlorinated dibenzodioxins and dibenzofuranes'were
detected at several places.
      Contaminant
Ground water
   mg/1
 Soil
mg/kg
       Non-polar aliphates  (oils)

       Non-polar aromates.

       Misc.  aromates

       Benzene
       Phenols

       PCB

       Clophen  A 60

       CHC
       Dichloromethane
       1.1.1  trichloroethylene
       Trichloroethane
       Tetrach1oroethy1ene

       Inorganic compounds

       Cyanides
       Sulphides
       Lead
       Cadmium
       Chromi urn
    1400
     226
      40.5
       7.2
       4.3
      22.8
         0.135

         0.44
         0.0325
         0.022
 37,843

  5,620
    133
     80
                            270
  1,370
    210
  5,209
       0.5
      58.5
  14,418.6
       0.229
      70.48
 Table 1: Main groups of pollutants on the Pintsch grounds (Jan.  1986)
                                      1103

-------
 4«  General  Description of the Decontamination  Technology

 The original  clean-up concept was  limited  to decontamination of the
 ground water  and  directly related  measures, in keeping with the legal
 latitude. The buildings were  to  be demolished  only to the extent neces-
 sary to  sink  wells  with which to pump  the  ground water.

 A total  of  nine wells were planned, of which wells Br 1, Br 4 and Br 7
 are in the  area of  former building 6.  This building, which housed the
 distilling  furnaces,  therefore had to  be torn  down. Since the masonry of
 the building  and, especially,  the  plaster were contaminated with highly
 toxic substances, the building was  surrounded  with a protective hall be-
 fore its demolition to  prevent hazardous dust  from escaping into the sur-
 rounding area. After  all  the  production installations and the building
 had  been torn down the  protective'hall was used to accommodate the soil
 decontamination system  and  related research equipment. A biological  in-
 situ decontamination  of the soil was not possible due to the heterogenous
 subterranean conditions and the  sometimes unknown chemical  compounds. The
 use  of thermal methods was  ruled out because they were not adequately de-
 veloped and because of the  permit  procedures to be expected.  Thus,  the
 use  of an extractive  soil-washing!process based on the HARBAUER PB  1  sys-
 tem was decided on.

 5. The HARBAUER Soil  Decontamination System

 For the simultaneous  cleaning of contaminated  ground water  and  soil  as
well as the contaminated  soil  from other locations in Berlin  (utilization
 of the system's capacity) the HARBAUER Engineering Office for Environ-
mental Technology, Berlin has developed a system with which  it  Is possi-
 ble to clean soil  with a grain size of 15 urn to 15 mm so that the soil
can be refilled again and the ground and process water can be fed to  the
 drainage ditch with drinking-water quality.
At the centre of the HARBAUER solution is the  already tested HARBAUER PB
 2 soil-washing system with which contaminated  soil  can be treated and
cleaned by washing and/or extraction with chemicals,  depending  on the
type and quantity of the pollution and structure of the initial material
 (grain-size distribution).
The efficency with which harmful  substances are separated from  the soil
 particles is largely determined by the introduction of energy.  In the
 case of the HARBAUER process,  which was described by SONNEN and QRTWEIN
 (1986), mechanical kinetic and vibrational  energy is  introduced with  the
 help of mixing, stirring, vibrating and hydrocyclone equipment.
The introduction of energy neutralizes the various bonding forces between
the harmful  substances and the soil particle and is supported by the  ex-
tractive effect of the washing or extraction agent.  The energy  density
required depends mainly on the type of contaminated material, less on the
harmful substance.
Thus, gravelly soil requires the introduction  of only a small amount  of
energy to wash the harmful substances off the  surface of the material.  In
the case of very silty soil with a high percentage of clay, loam or marl
                                      1104

-------
as well as fine building rubbish (e.g. contaminated plaster) the amount
of energy introduced must be increased to 1 kW/t.
In the list analysis, however,, it is the individual combination of energy
and extraction agents that results in the cleaning success aimed for. In
the selection of the extraction agent, however, there are ecological and
economic restrictions, for the agent used must be either absolutely harm-
less and biologically degradable or completely regeneratable so that the
cleaning of the process water$ Which is obligatory with washing or ex-
tractive processes, is technically and economicaly feasible.
Therefore, efforts are always made to work with clean water or aqueous
solutions as long as the chemical and physical properties of the contam-
inants permit.
In individual cases it may be necessary to use organic solvents, complex-
ing agents or other reactive substances. For reasons of emission laws and
accident prevention, however, these systems belong more in the category
of chemical facilities than cleaning systems for contaminated soil.
However, in the case of extractive methods a system solution does not in-
volve only the decontamination system proper, and the related treatment
of the exhaust process air, but must also deal with the question of how
to treat the substances left over.
With the HARBAUER process all soil particles greater than or equal to 15
urn in size are cleaned in such a way that they can be used again as
building material (screening bottom) at the original location. Fine sub-
stances less than 15 urn in size are to be found in various types of wash-
ing and rinsing water. This water undergoes preliminary dessication in
settling facilities, is separated from remaining sludge on a belt screen
and cleaned to nearly drinking-water quality during the treatment of the
process water.
The remaining sludge is still polluted to a low to medium extent and re-
quires further treatment  (thermal treatment, disposal, compaction, pyrol-
ysis). This waste product is still being disposed of in the GDR for the
time being.
Further degradation products are concentrated mixtures of solvents and
separated  oil fractions, which can be disposed of by burning at high tem-
peratures. A mobile plasma facility is presently being tested in a joint
project by WESTINGHOUSE and KEMMER/HARBAUER to determine whether it is
basically  suited for such products.
This leaves, finally, only the substances adsorbed by the activated char-
coal of the  process-water conditioning system. These substances are de-
composed  during the thermal/chemical  regeneration of the activated char-
coal at the  manufacturer's and are turned into ash or acids.
The complete process"flow chart  for the  soil-washing system is  shown in
Fig.  1.

 6. Results

 After  the soil-washing  system had been  put  to several months  of success-
 ful use treating  the contaminated soil  of  the Pintsch grounds,  cleaning
 tests  were begun in October  1987 with contaminated soil  excavated from
 two Berlin gas-works locations in Mariendorf and Wilmersdorf.
 The following cleaning results with the material  from Mariendorf are ex-
 emplary of the two soils;
 The grain-size distribution ('Fig* 2)  shows that  the soil  is extremely
 fine-grained with 37 % of the particles being less than 100 urn in size.
                                       1105

-------
1106

-------
          V. -
                                O   O   O
                                3   !S   oo
        5
•§
s
                                                                        o
                                                                        en
                                                                       o
                                                                       T3
                                                                       01
                                                                       IO
                                                                        0)
                                                                       CO
                                                                        o>
                                                                        o
                                                                        to
                                                                        O>
                                                                        O


                                                                        C
                                                                        o
                                                                        S-
                                                                       .«->
                                                                        in

                                                                       •5

                                                                        O)
                                                                        CO
                                                                       evi


                                                                        en
                                         1107

-------
Such  soil  cannot  be  processed  by  conventional  soil-washing systems since,
on the one hand,  the high  percentage of  extremely fine grains can no
longer be  effectively separated from the harmful substance and, on the
other, they lead  to  complete si 1ting-up  of the system.
In comparison, during some eight  weeks of tests with this material it was
seen  that  the system is capable of effectively cleaning soil with a prob-
lematically high  content of si It land clay without significant modifica-
tions or conversions.
The results  of the analysis (Table 2} substantiate the high cleaning ef-
ficiency in  relation to all the relevant parameters of the harmful sub-
stances. The pollution remaining  in the decontaminated soil was nearly
always lower than reference category A of the guideline for soil decon-
tamination,  but in general clearly beneath the B values.
The cut of the system, as  related to the grain size to be processed, was
determined to be approx. 15 urn within the scope of a balance sheet drawn
up for the complete  system.
On the average, the  amount of residue to be disposed of during the test
period lasting more than 6 months amounted to approximately 2 % of the
initial material.

In general these statements also apply to the soil  from Wilmersdorf.  But
since the balance sheet for the system was not yet finished for this  ma-
terial by the copy deadline,  reference is made in this connexion to  the
authors'  paper.
                            Initial
                            values
                 Remaining
                 pollution
              Washing
              success (%}
Petroleum ether extract:
(index for mineral-oil
contents)
PAH's

Phenol index:

Total cyanide:
(as per DIN 38 405)
476,000 ug/kg    67,000 ug/kg
752,000 ug/kg

 60,500 ug/kg

  5,300 ug/kg
2,000 ug/kg

    n.n .

   59 ug/kg
       86



       99.7

appr. 100

       98.9
Table 2: Cleaning efficiency, soil from gas works in Berlin-Mariendorf


7. Conclusions

With the system solution presented it is possible to clean contaminated
soil with a complex matrix of harmful substances and problematical  grain
sizes by means of large-scale technology and then refill  the soil.  The
entire system can be considered by now to have undergone  large-scale
testing, and the suitability of the system components for a process-en-
gineering solution to the problem has been demonstrated.  The process
costs less than 200 DM/t on the average. The present potential  for  devel-
opment of the system presented is, for one, a reduction of the  cut  to ar-
eas less than 10 urn in size and thus to further minimization of the resi-
                                       1108

-------
due to be disposed of and, for another, further optimization of the sep-
aration of harmful substances from the soil. These two areas are the fo-
cal points of the R&D work being done by HARBAUER at the present and,
among others, the subject of a research project sponsored by the Federal
Ministry of Research and Technology and the Senator for Science and Re-
search.
8. Literatur

Ortwein, H.
Sonnen, H.D.


Woltmann, M
Bodenreinigung auf dem Gelande der Pintsch-01  GmbH i.L.,
Vorstellung einer Bodenwaschanlage, Abfallwirtschaft 18
(1987), p. 117-126
Die Sanierung des Pint sen-Gel a'ndes, "Sanierung kontaminier-
ter Standorte 1985", FGU-Seminar, Wiesbaden 1985
                                       1109

-------

-------
                                     Appendix 5-C




Jet Cutting Followed by Oxidation (Keller), Germany
                                 No final text available.
                1111

-------

-------
                                      Appendix 5-D




Physical/Chemical Extraction technology Case Studies






    Electro-reclamation (Geokinetics), The Netherlands
                  1113

-------
           NATO/CCMS Pilot Study

Demonstration of Remedial Action Technologies
    for Contaminated Land and Groundwater
            Montreal,  Canada

           6-9 November 1989



Electro-Red amaM. on : State-of-the-Art
 drs. Reinout Lageman
 Geokinetics
 Delft, Groningen
 the Netherlands
                                 co-authors  :
                                 drs. W.Pool    ,.
                                 drs. G.A.Seffinga
                                 Geokineticsj
             1114

-------
                Electro-Reclamation : State-of-the-Art
Electrokinetical phenomena

During the  last A  years Geokinetics  has been  developing a method to
remove heavy metals and other contaminants  from soil  and groundwater.
The method  is based on electrokinettcal phenomena, which in one way or
another have  been made  use of  since''the  end of  last century. These
phenomena  occur  when  the  soil  is, electrically charged with direct
current by means of one or several electrode arrays :

1. Electro-osmosis : Movement of soil moisture or groundwater from the
2. Electrophoresis

3. Electrolysis
anode to the kathode.
Movement of soil particles within the soil
moisture or groundwater.
Movement of ions and ioncomplexes within the soil
moisture or groundwater.
Electro-osmosis

the electro-osmotic transport depends on the following factors  :

  -  the mobility  of the   ions and  charged particles  within  the soil
     moisture or groundwater;
  -  the hydratation of  the ions and the charged particles;
  -  the charge  and direction of the ions and charged particles, which
     cause a net water movement;
  -  the ion concentration;
  -  the viscosity of the pore  solution,  depending  on   the capillary
     size;
  -  the dielectrical  constant, depending on the amount of  organic  and
     inorganic particles in the pore solution;
  -  the temperature.

From existing literature and own experiments  the average  electro-osmo-
 tic mobility  has been   calculated  to be  in the order of 5.10-' mz/U.s,
where  U = potential drop (V).

To drain 1 m3 of  soil   by   electro-osmosis,  the   following parameters
 should be known  :

  -  the porosity;
  -  the moisture content of the soil to be treated;
  -  the conductivity of the pore solution;

Apart  from   these, other factors like  the desired  time period, the  use
 of  the soil  after  treatment and safety  requirements  regards maximum
voltage and  current are  also of  importance.

 Electrophores i s

 Electrophoresis  (cataphoresis)  involves the movement  of particles under
 the influence of an  electrical   field.  This   definition  includes  all
 electrically charged  particles  like  colloids,  clay  particles  floating
 in the pore  solution,  organic  particles,  droplets  etc.
                               1115

-------
                      circulation  system



                      current  supply



                      boundary of  electrokinetical  treatment
                                                                 GENERATOR

                                                                 OR MA»I
                                          '>• < •••contaminated
                                         --., *..'•• •/,••;.• . •
                                         vv'} •.•..'Soil.;;••..:
Fig. 1   :  SchemaHc representation of ER-field unit and

          electrokinetical  transport in the  soil
                                1116

-------
The mobility of these particles corresponds  with that  of ions. Within
the pore  solution these  particles transfer the electrical charges and
affect the electrical  conductivity  and  the  electro-osmotic current.
Clay  minerals  as  such  have .2 electrical polarity possibilities. One
consists of the stucture based  dipole  moment,  which  depends  oh the
atomic masses  and has  an orientation  parallel to the longest axis of
the clay particle. The second polarity  stands at  right angles  to the
first and is caused by the external electrical field. It depends on the
way of  polarization of  the electrical  double layer.  The mobility of
clay, particles  is an interplay between these two moments and is, ther-
fore, less than the electro-osmotic mobility. It varies between
1.10"10 and 3.10" m'/U.s.

Electrolysis

In the case of electro-osmosis and  electrophoresis one  considers only
water transport  or particle   transport respectively; with electrolysis
only the movement of ions and  ioncomplexes is taken into consideration.
The average  mobility of  ions lies  around 5.10" mz/U.s, which is ten
times greater than that of the electro-osmotic mobility. Therefore, the
energy  necessary  to  move  all  ions  over an average distance of 1 m
through a crosssectional area of 1 m2 of  soil is  ten times  less than
with electro-osmosis.

To calculate the energy necessary to dispose of the contaminants within
1 m3 of soil, the following factors are of importance  :

  -  chemical form of the contaminants;
  -  concentration of the contaminants;
  -  required concentrations of the contaminants;
  -  behaviour of the contaminants at different pH levels; (
  -  pH control around the electrodes within the soil;
  -  removal of the  contaminants and particles at the  respective  elec-
     trodes ;
  -  supply of  a conditioning solution to replace the removed  contami-
     nants and other particles at the electrodes;
  -  processing of the contaminated solution removed at the electrodes.
 The application of electrokinetical  phenomena  in practice

 The effectiveness of electro-reclamation  is largely  determined by  the
 chemical  composition  of soil  and groundwater.  In marly soils for exam-
 ple it depends mostly on the kind of clay minerals and the  calcium  and
 magnesium  bearing  minerals  like  carbonate   (e.g. lime) and sulphate
 (e.g.  gypsum).
 Another important element is  iron,  whose  concentration in groundwater
 depends a.o.  upon pH.
 Generally  speaking,  the  concentrations  of   the different metal ions
 depend on the C032- content and pH,  while the  following  chemical equi-
 libria are of importance :
           Me" (Clay-mineral)
           Me(OH)n
           Me«(CO,).
           HCOS-
Men* + Clay-mineral1
Me"* + n(OH)'
n(Me)
H* +
                                1117

-------
                                                                 vj& :»>%\ *•'«&./, ' •-•• j AW! y>"
 A)   Remediation  of residential areas
 B)   Remediation  of industrial areas
 Cl   Remediation/fencing of hazardous waste sites
 PJ   Preventive electrokinetical fence around potentially hazardous industrial  complexes
Fig. 2  : Some applications of in  situ Electro-Reclamation
                                     1118

-------
 The element  Me can be any metal  element like sodium,  potassium, calci-
 um, magnesium,   iron, etc,   but also  heavy metals'  like copper, nickel,
 lead,  chromium, cadmium etc.

 The concentration of the different metal ions in the groundwater depend
 on the solubility-products  of  the  hydoxides  and  carbonates  of the
 metals  concerned  and  the  pH  (H* concentration) of the solution. At
 lower pH  levels (higher  H* concentrations)  metal concentrations will
 increase.

 When groundwater  is contaminated with salts of heavy metals like lead,
 copper, nickel, chromium,  zinc etc., the metal ions  will influence the
 original chemical  equilibrium. At the newly formed chemical equilibria
 part of the heavy metal ions will exchange with the original metal ions
 in the mineral phase (like carbonate, hydroxide, clay mineral), whereby
 one- heavy metal ion will exchange more easily  with the  original metal
 ions than  the other. The mobility or the displacement of a heavy metal
 ion in the groundwater  or the  soil depends  on its  exchange capacity
. with the original exchangable ions.
 When soil
 observed :
is electrically charged,  the following processes can a.p.  be
 At or near the anode

   -  The positively charged particles  move into  the direction  of the
      negatively charged  cathode. As a consequence the concentration of
      the metal ions in the liquid phase (soil  moisture or groundwater)
      will  decrease.  The  decrease  in  concentration of the displaced
      ions in the groundwater will  be  restored  by  exchange  with the
      solid phase (mineral phase). This ion displacement and  ion-exchan-
      ge will continue as  long as  the electrical  field is.maintained.
      The final  concentration of  a certain heavy metal thus depends on
      the initial concentrations in  the  liquid  and  solid  phase, the
      electrokinetic mobility  and the mutual exchange capacity with the
      other metal ions.                                       *      .,

   -  At the anode moreover, H* ions are being formed through electroly-
      sis of water. These positively charged ions move via soil moisture
      or groundwater into the direction of the cathode.  As the  H* ions
      exchange rather  easily with the (heavy) metal ions of  the mineral
      phase and lower  pH  of  the  groundwater,  the  concentrations of
       (heavy) metals  will increase,  thus accelerating the processes of,
      exchange and displacement.

 At or near the cathode

   -  At a certain point near the cathode, the   (heavy) metal   ions move
       into the  direction of the cathode, but total concentration of the
       (heavy) metal  ions will stay the same, because an equal   amount qf
       ions  is  being supplied from the point situated at the  anode side.
      Total concentration  of (heavy)  metals will  decrease  when supply
       from the anode side  is less than transport  to the cathode side.
                                 1119

-------
   0.
   a.
       1000




        800
  =    600 -
   re
   2    400-
   c
   o
   VI


  3    200-
\
                 20    40    60    80    100    120   '140   160


                               time (hours)
 Fig.  3a   :  Decrease of Copper  during etecfrokineh'cal  treatment

            of  contaminated  pottery clay
 Q.
 Ol
 o
 u
•a
250
200-
150-

100-
50-
0-

\
\
•x^-
^^^.^^
^^^.

0 5 10 15
time (days)
Fig. 3b   : Decrease of Cadmium during electrokinetical treafmenf

          of confaminafed fine argillaceous sand
                                        1120

-------
The efficiency  of electro-reclamation  is less  for soils  with a high
cation or metal ion  exchange capacity  and an  acid buffering capacity
(like marls).  If feasible the efficiency of in situ remediation can be
increased by  irrigating the  soil with  slightly acidified  (pH : 3-4)
water .

When treating  the soil  on or off site, the soil material can be mixed
with a  slightly acidified  solution after  it is  being excavated. The
amount of  acid to be added depends mainly on the cation exchange capa-
city of the soil.


Electrokinetical Installation

The core of an electrokinetical installation  (fig. 1)  consists of the
electrode series and their housing. These can be installed in principle
at any depth, either vertically or horizontally. The  cathode and anode
housings are   interconnected but  form  two seperate circulation systems
(one for the cathode, one for the anode), filled with  different chemi-
cal  solutions.  In  these  solutions the contaminants are captured and
brought to a   connected  treatment  system,  installed  in   a container
together with  the solution tanks and measuring and monitoring devices.
The energy is  supplied by a generating  set or taken from the main.

As explained before, electro-reclamation  can  be  applied  both   for  in
situ remediation  of contaminated  soil (fig. 2) and for on  or off site
remediation  of excavated polluted soil  or river and/or   industrial and
sewage sludges.  For the latter application a  (semi)permanent  installa-
tion would be  most effective.
 Laboratory experiments

 The method of electro-reclamtion has been tested on the basis  of nume-
 rous laboratory experiments.  They focussed on important parameters like
 kind of current, strength of  current, voltage,  moisture content,  chemi-
 cal additives and the like.  Besides, the effectiviness of the method as
 regards certain soil and heavy metal types has  been examined  with the
 help of  several simulation  experiments (clay, peat, fine argillaceous
 sand polluted with As, Cd, Co, Cr, Cu, Hg, Mn,  Mo, Ni, Pb, Sb, Zn).

 These experiments also provided good  insight  into  the  energy demand
 and the  time duration. Some results are presented in table 1 and figu-
 res 3a and 3b.
                                 1121

-------
 4a.
          ••  3m   ••
           cathode-series
           anode-series
|       Cu>500ppm
H  1005000ppm
    600
-------
Soil type
Peat

Pottery clay
Fine argil-
laceous '.sand
Clay
Fine argil-
laceous sand






River sludge







"
Metal
Pb
Cu
Cu
Cd

• As
' .Cd
..Cr
Ni
Pb
Hg
Cu
Zn

Cd
Cu
Pb
Ni
Zn
Cr
Hg
As

Cone, before
(ppm)
9000
500
1000
275

300
319
221
227
638; .-
334
570
937

10
143
173
56
901
72
0.5
13

Cone, after Decrease
(ppm) (per cent)
2400 ",
200 . ..
; 100 ' "',. ;
40

30 '. •
20 : ;
34
230
110
50
180
Average :
5
41
80
5
54
26
• 0.2
4.4
Average :
73
60
90
85

89
99
91
85
64
67
91
81
83
50
71
54
91
94
64
60
66
69
     Table 1 :  Some results of laboratory electro-reclamation.
Field experiments
Site 1  .

The first  field experiment took place alongside part of a waterbearing
ditch, on one side bordered by a former paint factory and'on : the other
side by  open grassland. , The bank on the latter part was heightened by
sediment dredged from the ditch.  This  sediment  was  heavily polluted
with metals  in the form of paint residuary. The raised sediment layer,
height 20 - 50  cm, length  70 m  and width  3 m,  contained Pb  and Cu
concentrations up  to 10,000 ppm and 5,000 ppm respectively. The origi-
nal peat soil underneath was contaminated by leaching of this overlying
layer with  Pb concentrations  ranging from  300 ppm to more than 5,000
ppm, while Cu concentrations were in the order of 500 to 1,000 ppm.

A preceding electrokinetic laboratory test with  a sample  of the sedi-
ment reduced  the  concentration  of Pb  from » 9,000 ppm to « 5,000 ppm
and that of Cu  from »  4,500 ppm  to «  1,600 ppm,  all within  a time
period  of 320 hours.
For the field experiment one cathode and one anode array were installed
both with a length of 70 m and a mutual distance of 3 m.

                                1123

-------
  The cathode was installed horizontally, while the anodes  were implaced
  vertically into  the soil  about 2  m apart. On the basis of the energy
  consumption during the laboratory test the field experiment  was confi-
  ned to 430 hours.
  The changes  in Pb  and Cu concentrations were monitored at 26 sampling
  locations, sampled at regular [intervals at  several depths (10,20,30,40
  and 50  cm below ground surface). The following table lists part of the
  results for Cu and Pb within the peat at a depth of  30 to  40 cm below
  ground  surface.  The  spatial:  distribution  of  the pollutants at the
  beginning and at the end of the test are shown in figs. 4a and 4b).
Sample point
(30-40 cm)
1

2

3

4

5

6

7

8

9


Metal
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu

Cone, before
(ppm)
440
185
3900
540
> 5000
1150
> 5000
475
> 5000
1170
> 5000
580
3780
410
380
35
340
50 ,

Cone, after
(ppm)
110
35
700
220
560
580
2450
250
610
230
300
45
285
30
180
15
90
15
Average :
Decrease
(per cent
75
81
82
59
89
50
51
47
88
80
94
92
92
93
53
57
74
70
74
      Table 2. Electro-reclamation,  field results,  site 1
 Site 2

 The second field experiment  was carried out  on   the  site  of  a  galvani-
 zing plant.  According to preceding investigations  the  soil (sandy clay)
 around the plant was  contaminated with   Zn to  a depth  of 40  cm below
 groundsurface.  In  the upper  10 cm  Zn concentrations were reported  to
 have a maximum  of 3,000 ppm. At greater depths Zn   concentrations were
 indicated as being in the order of 500  ppm.

 For the experiment an area was  selected with dimensions  of 15 m x  6 m x
 1 m.  Two  cathode drains were installed  at a depth of 50  cm below groun-
 dsurface,  while  33 anodes, divided along 3 rows were  implaced  in  holes
 of  1  m depth with a mutual distance of  1.5 m. The  distance between the
 cathode and  anode series was also  1.5 m.

 Energy was supplied by a  100 kVA generating set.  The resistivity of the
 soil was 5 ftm. The  installation  was calculated  for a  DC supply  of 8
Amps/m2 of soil, which  should result in a potential drop of 40'y/m.
This potential drop could not be maintained during the whole period.
As^a  result of some material problems it was neither possible to main-
tain a 24 hour energy supply to the soil. '
                           1124
126

-------
Within 2 weeks temperature rose from 12 'C to an average of 40  C. As a
consequence soil resistivity decreased to 2.5 ftm and the potential drop
to 20 V/m with an average strength of current of 8 Amps/m2.  The effec-
tive energysupply  per 1000 kg of soil amounted to 160 kWh during the 8
week period.

Changes in Zn concentration  were monitored  at 12  sampling locations,
which were  sampled at  3 different depth intervals (10, 30 and 50 cm).
Changes in  groundwater concentration  were monitored  in 2 observation
wells. In table 5 and fig. 5 the results are given for the depth  inter-
val of 30 cm.
Sample point Metal Cone, before
(30 cm detpth) (ppm)
1 	 5120
2
3
4
5
2030
1600
2320
2450
6 > Zn 4390
7
8
9
10
11

1960
3250
2400
70
150

Cone, after
(ppm)
4470
1960
800
2320
2450
2360
940
1960
2000
30
120
Average
decrease
(per cent)
13
3
50
0
0
48
52
40
17
57
20
: 20
      Table 5 :  Electro-reclamation,  field results site 2

 The energy demand for this test amounted to 160 kWh/ton.  At the begin-
 ning of  the test  the highest  Zn concentration  amounted to 7,010 ppm
 with an average of 2,410 ppm  over the  whole area.  At the  end of the
 test the  highest Zn  concentration was  5,300 ppm  and the average had
 been decreased to 1,620.  The concentrations  of Zn,  Pb and  Cd in the
 groundwater and the filtercake are presented in the tables 6 to 8.

 A total  of some  1000 kg of filtercake was produced with an average Zn
 content of  117 g/kg. This comes  to a  total removal  of some  50 kg of
 zinc, assuming  an average'moisture content of the filtercake of 60 % .

 A rough mass balans can be summarized as follows  :

   -  treated volume of soil  : 15 x 6 x 0.5 x 3/4  = 34  m3 (1/4  of the
      area did not show  increased Zn concentrations).
   -  weight : 34 ms x 1.8 =  61 tons.
   -  weight of filtercake :  1000 kg.
   -  average moisture content : 60 %
   -  total  dry matter : 400  kg.
   -  average Zn concentration : 117 g/kg.
   -  amount of zinc removed  : 47 kg
   -  removed per  1000 kg of  soil : 47 x  10V61 x  103 =  770 ppm.

 The  last  value is in the same order of magnitude, as the average decrea-
 se  in Zn  concentration  (2410 ppm -1620 ppm - 790  ppm).

                               1125

-------
      An important  outcome  of   the  test  was   the  relatively
      saaple from   the area  It was  found that the  energy necessary to
        '                                                     *  'o =00
                                                             WA thl3  Hould
              6m
   Zn-concentrafions 30 cm below  qs
   (24/10/88)
                                            Zn-concentrah'ons 30 cm  below as
                                            (16/12/88)
Zn>4000ppm
                            2000
-------
Metal r Zn
sample treatment : not
date
24-10-88
01-11-88
09-11-88
17-11-88
25-11-88
30-11-88
24-10-88
01-11-88
09-11-88
17-11-88
25-11-88
30-11-88
obs. well
1
1
1
1
1
1
2
2
2
2
2
2
Pb
Cd
acidified '
Zn
Pb
Cd
acidified
ppm . . .
200
120
130
172
130
120
10
1.5
2.5
6
2.8
4
0.09
0.07
0.06
0.07
0.17
0.13
0.17
0.09
0.06
0.03
0.03
0.09
0.06
0.00
0.02
0.03
0.03
0.03
0.02
0.09
0
0
0.01
0
270
140
160
198
180
150
40
2
3
8
5.8
5.5
1.4
0.09
.0.17
0.11
0.22
0.16
0.34
0.15
0.15
0.07
0.18
0.14'
0.07
0
0.02
0.04
0.03
0.03
0.02
0
0
0.01
0.01
0
Table 6. Zn-content (ppm) of the groundwater in obser-
         vation wells 1 and 2.
Metal : Zn
date
24-10-88
09-11-88
17-11-88
25-11-88
30-11-88

136.9
199
99
89
61
Pb
g/kg
1.9
1.1
2
1.5
0.58
Cd moisture en t
in %
0.34
0.18
0.12
0.16 78
0.11
Table 7.  Zn-content (g/kg) of the filtercake during
          Electro-Reclamation.
Metal : Zn
date
30-11-88
Pb
Cd
ppm
30
0.6
0
       Table 8.  Zn-content  (ppm) of the solution  in
                 the anode-circulation  system.
                          1127

-------
   Remedial  action

   The  first  'official' electro-reclamation project started at  the end of
   January 1989.  It  involved a site of a former timber  impregnation plant
   containing arsenic  levels up  to 400  - 500 ppm. After a fire in 198*'
   which destroyed a large part of the plant,  it was   decided not  to re-
   build the plant. After dismantling the same a 'statement of  unpolluted
   soil  was needed in  order to  allocate the  land to  building plots  A
   following investigation  established the  presence of As concentrations
   up to several 100 ppm in part of the heavy clay soil to a maximum depth
   n-m? H?*n  ?SSm °f  ?-  P011^1011 was attributed to 'Superwolmansalt
   D (Na3HAs04.7HaO), used for impregnation.
  In April 1988 Geokinetics was requested to investigate  the possibility
  of remediating  the soil by Electro-Reclamation.  A following laboratory
  test with a soil sample reduced; the As concentration of  300 ppm  to 30
  ppm against  an energy  consumption of 115 kWh/ton.  An additional field
  investigation delineated the pollution to an  area  of  10  m  x  10 m
  contaminated to  a depth  of 2 '.m and  an adjoining  area of 10 m x 5 m!

                                                f polluted soil  :  2SO m
 The project   started  in  January 1989. Along the  length  of  the polluted
 area  4 x  2 cathode drains were  installed  : one at a  depth of  1.5 m and
 the other at  0.5 m. The cathode ;arrays had a mutual  distance of 3 m
 In between 36 anodes were implaced  in the soil, divided  along 2 rows of
 14 and 1  row of 8 pieces. Within the area  of 10  m  x  10 m the anodes
 were  installed  to a  depth of  2 m  below ground surface.  In the other
 area of 10 m x 5 m the depth of the anodes  was limited  to 1  m depth
 All anodes were placed at a mutual distance of 1.5 m.

 On the  basis of  both the  laboratory test and the field investigation
 cn/^!   ? °f the Electro-Reclamation period  was calculated  to last.-
 50 (24 hour)  days,  using an energy supply of 200 kVA (= 44 kW effective
 into the soil.

 At the beginning the resistivity of the  clay was 10  ftm and  soil  tempe-
 rature at  a depth of 0.5  m was 7 "C. After 3 to  4  weeks temperature  had
 risen to an average  of 50  °C,  while the  resistivity  decreased to   5 Ora
 The original   potential drop  of 40 V/m  decreased  accordingly to  20 V/m
 with  an average  current strength of 4  Amps/m'   (total  crosssectional
 area  being 110 ma).

 Changes in As concentrations were monitored at 10 fixed  sampling  loca-
 tions  and  numerous randomly  distributed  sampling   points. Of the  fixed
 locations, 2  were sampled   at 0, 0.5, 1, 1.5 and  2 m depth. The others
 at 0,  0.5  and  1 m depth. The analysis results from samples taken at the

                      Tgu6at  a depth of ' m belou
When starting the project, the average As concentration  over the whole
area amounted to 115 ppm, which comes to a total As content of on ample
«?U K§*


During the remediation process it was  observed that  at one particular
               338 in
     h   .                 concentration proceeded much more slowly than
at other locations. After April 30*" there was almost no  reduction ob-
served anymore.

                          1128

-------
                            Project Loppersum  x-
                            O cathodes

                            • anodes
-          15 m

As-concentrations 1 m below g.s.
   |As>2SOppm
pllOO As SO
6
7
8
9
10 —
75
40
175
40
60
Concentr a t i on Energy
250 -,
< 20
< 20
190
< 20 > 150
30
< 20
< 20
< 20
< 20 -J
       Table  9  :  Results of remedial action project.

                                    1129
                                                                                131

-------
   A total  of some 40  m>  of soil  had to  be excavated.  Periodical  treatment
   of the electrode solutions  resulted in some  800  kg  of  filtrate.  A
   mass  balans can be  summarized  as  follows :
    -   treated volume  of soil
    -   weight of soil  .
    -   average As concentration
    -   amount of As before remediation
    -   weight of excavated' soil
    -   average As concentration of excavated soil
    -   amount of As in excavated Isoil
    -   amount of filtrate
    -   average moisture content
    -   total dry matter
    -^   total amount of As removed: by electro-reclamation
    -"  total amount of As removed; after remediation
    -  amount of As remaining in the soil
    -  average As concentration of remediated soil


  Other applications / future developments

  Electrokinetical fencing

  The electrokinetical  phenomena occurring when the  soil  is electrically
  charged  can  also be used for  fencing purposes. These  so-called elec-
  trokinetic fences can be installed either at refuse sites/factory
  complexes, where soil pollution  has already been ascertained,  or where
  soil pollution is likely to occur. Depending on the local
 250  m3
 450  tons
 115  ppm
 5.2 kg
 71 tons
 200  ppm
 14 kg
 800  kg
 70 %
240 kg
 34 kg
48 kg
4 kg
10 ppm
   -  the elctrokinetical  transport is  directed towards  the source of
      the pollution (fig. 7a).  The cathode series is situated nearest' to
      ™rmS£TCe -f *?"ution-   Such a set-"P should be applied in less
      permable soils without substantial groundwaterf low (  < l a/year).

   -  the contaminants,  which are  carried  along with   the groundwater
      flow are diverted,  collected around the  electrodes and periodical-
      ly removed.  In this case  the cathode series is  farthest  away from
      the source of pollution and cathode and  anode  series  are installed
      perpendicular to the direction of  groundwater flow (fig.  7b)   Such
      a  set-up should be  applied  when the soil  and/or subsoil is relati-
      vely permeable (groundwaterf low velocity  >  1 m/year).

Desalination  of arable land
                          is a cor°n Pr°*lem in those countries, where
Mon          ,      KK  10W Hd evapotranspiration high. In combina-
tion with relatively high groundwater levels  (coastal areas  and river
fS'irS'   iS°J    ?^ 10W  P6™6^11*/ ^d irrigation water with high
total dissolved  solids, the  accumulation of  salts in  the top laye?s
prohibits further agriculture.
                           1130

-------

                               source'; :
-------
  The most  common technique for landreclamation consists of the lowering
  ?nrf£  f o^ndwatertable and/or drainage of the soil  by means  of we if
  and/or horizontal  drains.  The  soil is  then frequently ir^ea?ed with
  relatively fresh water, thus leaching the salts f?om  the lof  However

  latio-nMo?e^eaiiliHy °f"the S0il hamperS ffi°re Often tha» ™  the pe?co-
  lation of the leachate to the deeper layers..                      v^-f-o

  By applying electrokinetical processes these problems  can be overcome
  redaction    a^illaceous  *>"» «e specifically suited Tor ete'rol


  The As remediation project  mentioned earlier  showed,  that  after elec-
  trokinetica   treatment,   the  permeability  of the heavy clay soifhad
  increased significantly.  When furthermore gypsum is added at  the anode
  "£j *XM '  the1f°i1 S5ucture wil1  ^ improved even  more and higher crop
  production will  be obtained (Collopy,  1958).

  Removal of organic contaminants
An R t D project investigating the possibilities of removal

                                  -**
                                                              of
 The second research objective is to examine the possibilities of combi-
 ning electrokinetical techniques with biodegradation,  more specif leal-
 *y •

   -  can micro-organisms be distributed more  evenly  in  the  soil   or
      can micro-organisms be added to the soil

   "  ?* f 6 f 7,S°aii iTS^S1™.1" -111"1-11 "  l««--«.l  levels
   -  can addtional oxygen and nutrients be brought into the soil  ?
 Cost  estimates

 Electro-Recalamation

 Fig.  B shows a set of graphs depicting remediation cost per ton of soil
 as a  function of  remedial action  time  (fig. 8a) and as a function of
 the measure of contamination (fig. 8b),  assuming a  polluted area with
 dimensions of  500 m x  100 m x 1 m. From the graphs it is evident that
 short remediation  periods and  highly polluted  soil (low resistivity)
 require  a  high  amount  of  energy, having the greatest effect on the
 costs.

 In practice, however, there is a limit to the electrical  current which
^ £! **?   f0 ^e 30i1' F°r 6Very 3Pecific caae,  therefore,  an opti-
mum must be calaculated for energy supply and time duration.
                         1132

-------
a.
                                           b.
    ISCh
                       T
      10    30    60   90   150   365

          remedial action time (days)


 Assumptions:
 - Area  dimensions: 500 m x 100 m x 1 m
 - Weigth of soil: 90,000 ton
 - Distance  between C-  and A-series:  10 m
 - Mutual distance anodes: 10 m

 - Soil resistivity: 10  Ohmm
                                             I/)
                                             3
                                             Ol
                                             0.
                                             I/I
                                             ^»

                                             O
                                                1501
                                                100-
                                                50-
          10   20  30  40  50  75   100

             soil resistivity (Ohmm)
- Duration of ER-proces: 60 days
Fig. 8  : Cost  estimate of electro-reclamation
        a. costs  per .time-period
        b. costs  versus grade of pollution
      Electrokinetical fencing

      In fig. 9 the  energy costs per  year are given  as  a   function of  groun-
      dwater flow  velocity (fig.  9a) and as a function  of the rate of  pollu-
      tion (fig. 9b),  assuming an  electrokinetical fence of 500 m  length and
      10 m depth.

      In areas of  low groundwater  flow velocity (clay, argillaceous sand) and
      low soil pollution, the  yearly  energy  costs   of  an electrokinetical
      fence  are   insignificant.   This  changes rather quickly, when  the soil
      becomes more permeable  (sandy   formations)  and  the  groundwater flow
      velocity increases together  with the concentration of the contaminants.

      For relatively high groundwater flow velocities  a  combination of  hydro-
      logical measures and electrokinetical  techniques  will  render  the most
      economic results.
      Desalination of arable  land

      Preliminary cost estimations amount to US $  1000  to 2000 per ha.
                                       1133

-------
 a.
 o
 C  300-
 in
 •B  250
 s
200

150-1
 g.
 g   100-i
 >»   f.
 01   50-
 O)
        0   10   20   30   40
         groundwafer  flow velocity (m/year)

 Assumptions:
 - Length of fence: 500 m
 - Depth of fence: 10  m
 - Average  ion mobility: 1.56  mW.year

 - Soil resistivity: 50 Ohmm
                                                         20  CO   60   SO  100  120
                                                         soil resistivity (Ohmm)
                                            - Groundwafer  flow velocity: 10 m/year
Fig. 9  :  Cost estimate of electrokinetical fencing
         a. energy costs versus groundwater flow velocity
         b. energy costs versus grade of pollution
     October  1989

     Geokinetics
     Poortweg  A
     2612  PA Delft
     the Netherlands
                                      1134

-------
                                       Appendix 5-E



  Physical/Chemical Extraction Technology Case Studies
In Situ Acid Extraction (TAUW/Mourik), The Netherlands
                  1135

-------
In situ cadmium  removal
- full-scale remedial  action of contaminated soil  -

L.G.C.M. Urlings*, V.P. Ackermann**, J.C. v.  Woudenberg*,
P.P. v.d. Pijl*, J.J.  Gaastra***                        '      .
*   TAUW Infra Consult B.V.,  P.O. Box 479, 7400 AL DEVENTER  (NL)
**  Provinciale Waterstaat Utrecht, P.O. Box 80300, 3508 TH UTRECHT (NL)
*** MOURIK B.V., P.O.  Box 2,  2964 ZG  GROOT  AMMERS (NL)
1.   INTRODUCTION

     A photopaper  producing plant discharged  Cd containing waste.water
     into  two  infiltration  ponds  in  the  years  1935-1955.  Periodical
     flooding  of  these ponds  caused a  soil  pollution with Cd  in the
     adjacent dune plot. In the vicinity of the polluted area a ground--
     water -pump ing station is siuated.                                ,
     This paper deals with the set up and the results of the in situ Cd
     "remedial action".  There are  three contributors  to  the  project:
     province  of  Utrecht as principal,  Mourik as contractor  and TAUW
     Infra Consult as consultant. The  calculated total  costs  of  the in
     situ  remedial action  amounts  to  only  80%  of the  conventional
     sanitation costs. The costs of the whole project are approximately
     4 million DM.

     Figure 1  gives  an outline  of  the  cadmium pollution in  the dune
     plot (horizontal profile)  and the plant  grounds  (vertical  profi-
     le) .  The total Cd content of the soil  was estimated 725 kg.
     Otflli 5-50t«.
                     Otpth Z50-300e«.
                        f i> » ,»«
    figure 1. Cadmium pollution in horizontal en vertical profile1

    In table 1 a few soil characteristics of the contaminated soil are
    presented. The analyses are carried  out  on more or less represen-
    ative samples.
                                    1136

-------
     Table 1. Soil-characteristics  for three soil-layers
sieve size

< 16 urn
16 - 63
63 - 90
90 - 125
125 - 180
180 - 250
250 - 355
355 - 500
500 -1000
1000-2000
> 2000 urn
pH •
CEC meq/100 g
org. carbon %
CAC03- content
Fe(oxalate
exac tractable)
mmol/kg dm
Al(oxalate
extractable)
nunol/kg dm
dry matter %
top layer

0.6
1.3
3.0
10.1
29.1
27.4
18.9
6.8
2.6
0.3
<0.1
4.9
3
0.3
% 0.4


5


10
98.1
layer above
groundwatertable
3.7
1.7
7.0
14.6
26.3
19.7
13.6
6.1
5.0
1.5
0.7
4.7
2.5
<0.1
0.6


6


7
96.7
layer below
groundwatertable
2.4
1.6
7.6
16.1
28.1
20.3
13.0
5.1
3.9
1.1
0.7
7.4
1.8
<0.1
0.6


3


5
90.4
     The  soil  can be  characterised as  middle fine  to middle  coarse
     sandy soil. The adsorption capacity of the soil in general is very
     low.

     The whole  in situ treatment  involves  approximately 30,000 nr soil
     within an  area of  6,000  m^. For  the authorities a priority for
     remedial  action was  given  because  of  the  expected  additional
     pollution of the ground water and the possibility of direct contact
     for recreants with  contaminated topsoil. The  groundwater is used
     for the preparation of drinkingwater.

2.    DEVELOPMENT AND INSTALLATION OF THE REMEDIAL ACTION TECHNIQUE  ,

     Removal of the  cadmium pollution out off the environment  can be
     obtained by remedial  action  on the solid soil on location of the
     former ponds and the dune plot and by remedial actions  of the deep
     groundwater downstreams of the plant (upstrearas of the  groundwater
     resource). In  this  presentation,  only the remedial action of the
     solid soil is outlined.

     Concerning  in   situ  remedial action  three  aspects  need  special
     attention, because there is no or very little experience available
     in -this matter:
     - the Cd-desorption of the polluted soil (soil chemistry)
     - the hydrological system of infiltration and withdrawal of water
       (hydrology)                                                     .
     - the purification of Cd-containing groundwater (watertreatment).

                                      1137

-------
     According  to the three topics  mentioned above, laboratory experi-
     ments, designs  and/or -installations are .outlined underneath.

2.1. Desorption  of cadmium    '              •
     A desorption liquid was selected by batch-experiments with polluted
     soil.  By  calculation  of  the   Cd-distribution . coefficients  ,(Kd-
      soil/cwater) from the batch-experiments  a selection was made for
     column-leaching experiments.
     Hydrochloric acid (1CT3 mol)  turned out  to be useful; in figure 2
     the results  of  the  column-leaching experiments are compiled.
Cd :
20mg/kg.ds
          cadmium cone.
          in percolafe
          mg Cd/1.
cummulah've
percolated Cd.
                                r  100%
                                 - 50%
Cd
: 70mg/kg.ds
cadmium cone.
in percolate
mg Cd/t.
(	1
                                  cummulative
                                  percolafed Cd.
                                    (	j
                                     r  10(1%
100  140 pore Yolumes
    20
                                                              - 50%
                                                  60   100  140
                                                                  volumes
     figure 2.  Leaching of Cd-polluted soil bv  10  -1 mol HCH fpH .. 3.51

     The  Kd-value for  the batch experiments  range between  1.6  and 7.8
     dm3/kg and for the  column experiments between  approximately 2 and
     3 dm3/kg.
     Results of the batch experiments can  indicate an overestimation of
     the  Kd-value due to  buffer capacity of the sandy soil.  The column
     experiments give  a  more  accurate  value  for the  Kd because  the
     Liquid Solid ratio is much higher.
     The  expected quality of the percolate,  calculated for  a Revalue
     of 4 (dm3As) with  the  assumption of lineair desorption behaviour
     of Cd,  is  shown  in figure 3.  The hydraulic permeability  of the
     soil used  for the calculation is 1 m/day.
                                     1138

-------
     figure 3. Calculated cadmium concentration in the percolate of five
               separate compartments
2.2. Hvdrologie  system
     The total area for remedial action amounts to 6,000  m^ (see figure
     1).  The capacity of  the  groundwater treatment  installation was
     limited  to  approximately 250 m^/h,  so the  at  one  time  treatable
     polluted area is dependent on the hydraulic  properties of the  soil.
     Horizontal  drains are preferable to vertical deepwells in order to
     get straight  groundwaterflow lines. A  cross  section  of infiltration
     and withdrawal systems is outlined in  figure 4.
                   pH control
 pH control
                                                  •m**-*  drain » 10cm.
                                                       extension of the drains
        0—TCd removal Treatment!
iCd removal  treatmentr~0
       in            iv
      figure 4.  Cross section of  the  infiltration and withdrawal  systems
                                        1139

-------
 Design and  dimension  of the  infiltration/withdrawal  system  was
 carried out by a two-dimensional  (vertical)  computer model.  The
 calculated results are converted into the three-dimensional  space.
 Full recirculation of the infiltrated water can be guaranteed when
 only a slight  quantitity of aquifer water is discharged additional-
 ly.                            :
 For  one compartment the groundwater flow pattern  is represented in
 figure 5.
figure  5. Two-dimensional  flow pattern of  the infiltration water

The  end of  the remedial action will also be determined by the Cd-
concentration  in the pumped  percolate  for each separate compa.rt-
ment:  in principal,  infiltration  with acified  water  should be
continued untill the Cd-concentration reaches a constant low value
(approximately 20 ug/1) . The  installed infiltration and withdrawal-
system  for  the whole contaminated  area is outlined  in figure 6.
The  compartmentsize is bases on practical  experience during the
proceedings of the  remedial action.
                       • ai
                           = •----19
               	=.-,, _ jn --H-*
figure 6. Overview of  the horizontal  drains  in the four compart
          ments
                                1140

-------
2.3. Watertreatment system •
     A literature survey was conducted on the removal of Cd from waste
     water.
     The three most important treatment techniques are:
     .  precipitation
     .  biosorption
     . .ion-exchange
     Because  the treated  acid  percolate  is  infiltrated again  in the
     polluted soil and the waterflow is quite high, viz 250 m^, sorption
     on resins is preferred.

     First  in the laboratory batch  experiments were  carried out. The
     main goal was to select a resin with a high specific Cd-adsorption
     in presence of  a high iron-,  aluminium- and calcium-concentration
     and a pH of 3.5.
     The IMAC GT-73 turned out to be the most selective for Cd.
     After the batch experiments a column experiment was set up to test
     the IMAC GT-73  resin  of Rohm  and Haas.  The experimental  set up is
     shown in figure 7.
               oH tantrel
              ca !~~~
      HCI sol.9%

Contaminated
Sind 95kq.
Glass-
i
'•'•£&:•':•
'££$&
gsssnd
INFILTRATE :  pH - 3.5-4
HYDRAULIC
LOAD RESIN :  20 bedvolu-
             mes/h
Cd-SOIL    :  20-30 mgAg
                            £ump_
     figure 7. Laboratory  testunit for the resin  column
     The  load of the first resin  column is 6.7 g Cd/1 resin while -the .
     effluent of  the  second resin  column  contains  no  Cd (<1  ug/1).
     Loads as high as 34  g Cd/1 resin were  found in previous experiments
     with high Cd-influent concentrations.
     Regeneration of the  Cd-loaded resin is more or less  complete when
     flushed  with 4 bedvolumes of  5% HCI.

     Based on the  laboratory results the  dimensions  of the full scale
     installation are  designed.  Due to the  relative  short  operation
     time (approximately one year)  there  were made some  modifications
     in   the  original  design  so  the  contractor  was  able to  instale
     smaller  resin filters.

     The  installed water  treatment system  is represented  in figure 8.
                                          -.
                                      1141

-------
  The  ion-exchange system  consists of 5 filters with 3 ra3 resin IMAC
  GT  73 each.  Two  parallel  streets of  each  two filters  have been
  installed with one filter  stand-by used during  regeneration.
                  HAIH FILTERS
mfilrration/rtcirculation
90 J «SnVn
Backwash facilities

Regeneration facilities : air-hold down systeit

               HCl 55C (2.3BVI
Storage
 figure 8.  Scheme of the water treatment plant

 The  regeneration  is  upflow by an  air-hold  down system  to reduce
 the production of eluate. The step-by-step regeneration is perfor-
 med  with  two  times  3 bedvolumes; hydrochloric  acid of  5% upflow
 flushing and  proceeded by  2.5 bedvolmes fresh  water  upflow-  and
 2.5 bedvolumes fresh water  downflow flushing. The  first 3 bedvolu-
 mes  hydrochloric  acid are discharged to  an  industrial  cleaner
 (precipitation and  ion-exchange);  the  sludge is  conditioned  and
 discharged to  a controlled  waste  dump.
                                                                           A discharge
                                                                           v 10 a 25mVh
                                 1142

-------
3.   THE RESULTS OF THE REMEDIAL ACTION OF COMPARTMENT I

     In .August  1987 'the  in situ remedial action star'ted with  the. less
     contaminated  compartment  I.  Compartment  I  is  the  most  distant
     compartment from  the  formal discharge pounds  on the  plantground.
     The compartment I remedial action forms a testcase and the results
     will be applied for the treating device for the rest of the conta-
     minated soil.

3.1..Hydrologic svstem
     First  infiltration was  carried  out with neutral water  to test
     hydrological permeability of the soil. The permeability of the 6 m
     soil  layer above the horizontal withdrawal drains was much lower
     than  was  measured in the topsoil.  Extrapolation of  the granulair
     composition of the different soil .layers as pointed out in table,1
     and  application  of  the  results of cone penetration tests indicate
     also higher permeability.
     The  inclusion of air in the  pores was expected  but no complete
     evidence  could be gained.  The  delivery of the drainage system was
                           7  -  for  compartment I  during  the  remedial
 3.2.
between
action.
50 and  55
m3/h
The  flow  direction of the groundwater  was observed by  6  shallow
piezometers around the pound and  1  deep piezometers  in the -p^und.
The  deep  piezometers  have  filters at the  depth of 5,  10 and 15 m
below surface.

During the remedial action of compartment I only slight changes in'
flow direction have  taken place.  The  system was  adjustable  to
change the flow in the good direction.

Desorption of cadmium
The  Cd-concentration  of the influent  (percolate)  and effluent of
the  first resin filter are represented  in  figure 9.
In figure 10  the  cumulative  quantity of Cd removed is outlined as
well as the pH of  the percolate:
In the begin of August after the  soil was  saturated the percolated
water beared higher Cd-concentration than was measured during
the  normal  unsaturated circumstances in 1985. Probably the little
pores of  the  soil,  which are not flushed under unsaturated condi-
tions, are responsable for the additional  Cd  release. The course
of  the measured  Cd concentrations in  the influent  is comparable
with the  calculated concentrations as represented in figure 4. In
order to  give a more  complete impression of the percolate composi-
tion some data  from chemical  analyses of  September 1  1987 are
outlined  in table  2.
                                      1143

-------
 C» 'n/l
  ill
                                          onosca
1 H
KOVEnats
 figure 9. Cd-concentration of:influent  (percolate)  and effluent
           of  the  first resinfilter                         '    ~
          >ticusr
                                                           nwvlnul M
figure  10.  Cumulative Quantity  of Cd removed and pH of  the per
            colate
Table  2.  Influent
analysis. September 1
water from fi
9th September 1
pH
E.G.
Cl"
HCO 3
C03"
Cd
Ca
Fe
Al
Mn
us/cm
mg/1
meq/1
meq/1
ug/1
ug/1
ug/1
ug/1
ug/1
5.0
276
66
0.09
<0.01
440
26000
<15
160
145
Iters of 2 rne^n
.987

L KiUUIlU.-
meters on

Piezometer
11 12


filter • Cd
1m
2m
3m
4m
5m
(ug/1)
16
80
383
1150


pH
4.1
4.6
4.75
5.9

,
Cd
(ug/1)
8
48
2370 '
7580
70 ,

pH
3.8
4.3
5.3
5.7
6.6
                                 1144

-------
                          •             10

     The  piezometers  II  and  12  (see  figure  6) with filters on  1,  2,  3,
     4 and 5  m were of great value to  controll and predict the process
     of the remedial  action.  Tn table  3  a moment survey of the Cd.
     concentration and pH is given for the  piezometers II and  12.
     The  effectiveness of the, remedial action is controlled by soil
     analyses.  The 12th  October the  Cd .concentration  in the percolate
     was  less than 10 ug/1 so acidification of the infiltrate  was
     stopped. The 26th October  neutralization of the  acid soil was
     started  with NAOH pH 8.5.
     The  neutralization  stopped when the percolate Cd concentration  of
     every seperate drain was not detectable (< 10 ug/1) anymore.

3.3  Water treatment  svstem
     The  performance  of  two  resin filters in series is excellent. By
     the  end  of September the first  filter  was totally loaded  (6.7 g/1,
     resin),  while the second filter operated correctly. Due to  low
     influent concentration the Cd  load  was relatively low.

     The  first resin regeneration pointed out a recovery of more than
     98%  of the adsorbed cadmium. The main metals  found  in  this  eluate
     are  pointed out in table 4.
     Table 4: Eluate composition after resin regeneration
       cadmium
       calcium
       aluminium
2600 mg/1
 940 mg/1
 170 mg/1
zinc
silver
iron
23 mg/1
15 mg/1
12 mg/1
     The conclusion can be drawn that the resin IMAC GT 73 is very
     selective for cadmium.

3.4  Conclusions and recommendations
     The experiences with fche full scale in situ remedial action of the
     first compartment showed that:
     -the desorption of cadmium is good comparable with the laboratory
      studies. However neutralization takes a long time;
     -the permeability of the whole soil layer is rather different
      from what was measured in the.topsoil;
     -the watertreatment system operates according to the design crite-
      ria .                                        9                 .
     Hence the remaining contaminated area, 5000 mz, would be remedia-
     ted  in  situ. Due to the low end values of Cd in soil for compart-
     ment I,  the remedial action limit was lowered from 5 rag Cd/kg  d.m.
     to 2,5  rngAg d*11-

     The  experiences of the remedial action of compartment I resulted
     in the  .following recommendations:
          enlargements of the compartment size;
          3  m instead of 4,5 m distance between the deep horizontal
          withdrawal drains  (5,5 m below surface);
          instale additional drains on a depth of 2.25 m below  surface
          to accelerate the remedial action.                          [
                                      1145

-------
A.   THE RESULTS OF THE REMEDIAL
     AND IV
                                        11
              'ACTION OF COMPARTMENTS IT.  TTT
4.2  Desorption of cadmium                     •              :

     The compartment configuration is outlined in figure 6.  In October
     and November the infiltration/withdrawal of compartment IV (2000'
     DV"), III (1500 m/2), and II (1000 m2) started.
     From the beginning of December acid water has  been supplied to the
     compartments.                                           .
     The Cd-concentrations in the withdrawed percolate form  the com-
     partments II and III is given in the figures 11  and 13.
     Concerning compartment IV two withdrawal pumps were used;  at
     the eastside pump 2 and at the westside pump 3.
     The Cd-concentrations of the pumped percolate  are given in figure
     15 (pump 2)  and figure 17 (pump 3).  The corresponding cumulative
     quantity Cd pumped from each compartment are shown in the  figures
     12,  14,  16  and 18.                                           £'

     Compartment  IV,  figures 15 and 17  indicate  lower  Cd concentrations
     than calculated (see  figure 3   part  I and II). Especially  for  the
     eastside of  compartment IV the initial  Cd concentrations  without
     acid infiltration was  quite high (figure  15).

     The  cumulative  quantity Cd pumped  from  each  compartment is given
     in table 5. While the  Cd removal by  the watertreatmentinstallation
     for  the  whole remedial  action  period was not compete, see chapter
     4.3   the Cd quantity removed from  the soil is less  than stated in
     table 5.

     Table 5  .Cumulative Quantity Cd pumned per compartment
    Compartment
    Compartment
    Compartment
    Compartment
    Compartment
    Total
II
III
IV east
IV west
I
 77 kg Cd
154 kg Cd
143 kg Cd
 45 kg Cd
 24 kg Cd
443 kg Cd
                                    1146

-------
                                   12
   01
   c
   o
   c
   o
   i
   •o
                                    1	1	1	1	1	1	1	1	r

          09-Dec  18-Jan  27-Feb  07-Apr  17-May  E6-Jun  05-Aug  14—Sap
figure 11.  Cd-concentration in pumped  percolate of compartment II
         10
         o  -¥•—i	1	1	1	1	1	1    r

          09-Dec  18-Jon  27-Feb  07-Apr  17-May  26-Jun  05-Aug  14-Sep 24—Oct
 figure 12. Cumulative  quantity Cd removed  and pH of the percolate

            (compartment II)
                                   1147

-------
                                       13
   .£

   c
   Q


   TO
   t_

   "£
   I
  •o
         19-Nor 29-Dac 07-Peb  18-Mar  E7-Apr OB-Juu  16-Jul  25-Aug ' 04- ct
  figure 13. Cd-concentration in;pumped percolate of

             /eastside^
        19-Nov  29'-Dec 07-Feb  18-Mar 27-Apr 06-Jun 16-Jul  25-Aug  04-Oct
figure 14.  Cumulative quantity  Cd removed and pH of  th

            (compartment TIT)
                                  1148

-------
                                   14
   en
          30-S
                        08-Jan
17-Apr
28-Jul
03-Nov
figure 15. Cd  -concentration in pumped  percolate of compartment
           easts icie
                            TV
      cd (kg) .
        140 -

        130 -

        120 -

        110 -

        100  -

         90  -

         80  -

         70  -

         60  -

         SO  -

         40  -

         30  -

         20  -

         10  -
           30-Sep
                        	1	
                         08—Jan
 	1	
 17-Apr
 	1	
  26-Jul
 03-Nov
 figure 16. Cumulative quantity Cd removed and pH  of the percolate
            (compartment  IV)
                                  1149

-------
                                      15
     en
     a.
    .£
     §
    c
    o
         450  -r-
         400
         350
300  -
         E50  -
200  -
         150  -
         100  -
         50  -
  30-Sep
                         08-Jan
                                        17-Apr
                                                                     03-Nov
 figure  17.  Cd-concentration in pumped percolate  of compartment: TV,
             westside
      cd (kg)
         30-Sep
               08-Jan
                                       17-Apr
                                                     26-Jul
                                                                  •  03-Nov
figure 18.  Cumniulative quantity Cd removed and pH  of the percolate
            (compartment IV. westside)
                                  1150

-------
                                      16
    The Cd-concentrations in the percolate water showed large  dif-
    ferences (see for example table 3 difference of 5000 ug/1  .within 1
    m). Figure 19 gives an impression how Cd scattered in the  deep
    horizontal drains of compartment IV.  Drain length is about 20 m.
       X
       at
                     EAST
    Figure 19.
           Cd  concentrations  in  deep horizontal withdrawal drains
           compartment  IV  (draindistance  3 m)
    The progress of  the  in  situ  remedial action is judged on the Cd
    concentration  of soil samples.  In  table  6 the end values of Cd
    content  in soil  samples are  given.

    Table  6.   The  distribution of Cd-concentrations  (me Cd/kg d.m") in
               soil analyses

Compartment II
Compartment III
Compartment IV
1 and <2.5
4
1
6
>2.5 and <5
2
1
2
>5
0
0
3*
4.2
     The origin of the Cd contamination is probably different.  The
     natural soil layers are disturbed by antropogenic activities;
     incineration stags are observed.

The latest samples of table 6 were taken in the beginning of
October 1988. Hence at the end of October when infiltration stop-
ped the remedial action goal of 2,5 mg Cd/kg dm was reached, with
the exception of the 3 samples taken in this middle off compart-
ment IV west. Further investigation are carried out to determine
the origin of this immobile Cd contamination. Additional excava-
tion of the small area is probably the most suitable remediation
technique.

Hvdrologic system
The  start up procedure took, as in compartment I quite a long
time. In all compartments the escape of air could be observed. The
maintenance of the desired discharge flow rate was difficult,
mostly the discharge was too low.
                                      1151

-------
                                    17

  The low permeability of the soil is  due to  the high resistance of
  a rather thin soil layer.  The infiltration  increased by making'
  use of the infiltration drains on 2.25  .below  surface. Especially
  in the heavily contaminated area of  compartment IV, the shallow
  drains are used succesfully.
  The withdrawal capacities'  for the compartments were-
  compartment I           25-50 m3/h
  compartment III         20-40 m3/h
  compartment IV east     45-65 m3/h
  compartment IV west     35-40 m3/h

 The maximal withdrawal flow was approximately 200 m3/h,  and tb»
 average flow  was about 145 m3/h. The discharge flow was  only 8~3%
 of  the infiltration water flow.         '.

 In table 7  the calculated infiltration capacity of the different
 compartments  is stressed. During the first period discharge took
 place in compartment I, and for the second period the sewer was
 used.

 Table 7. Infiltration capacity during 2
 compartment
period 1   period 2
dec.-may      may-aug.

II
III
IV
(m/d)
0.81
0.77
0.60
(m/d)
1.02
0.81
0.91 (in
                                    (influence shallow drain)
In figures  20  and 21  computed infiltration/groundwater stream line
patterns are presented for respectively discharge of surplus water
onto compartment  I or the sewer.
During discharge  on compartment I (figure 20) there is a stream
off of infiltrated water from compartment II. The streamline
pattern showed in figure 21 indicate that discharge on the sewer
is preferable. Still  there is a little stream off, this must be'
related to  the discharge quantity of 8.3% instead of 10% of the
infiltrated water.
For the two periods it was calculated that 70% of the compartment
areas are flushed more than 75 times.
                                1152

-------
                                  18
M
155. _
130. .
105.
80. .
55.

0

~ r^ ""N
c"< -$ — -
c-jyii.jii i i tjflj!) '. [ 1 IIIU^J3/Q^^^\

I ^-* 1 t 1 i i t i i i t i 	 i t r f f r ,•
{ l/U^T1 ' '' tmt-t-L- •••'•»'''• L w '
/ 1/ if K i 1 1 f i li u \iliU\\\y 'i,.f, ! ! \ \ v ' j
i ,, ,| , 1 III , , | | 1 1 1 1 M 1 1 1 | 1 1 1 1 M 1 1 1 | 1 1111 M II pi 11 1 II II 1
30. 55. 80. 105. 130. 155. 180. (m)
PROGRAM TRIFLO
Chance So*itdu!n*n lit p*rio4
/•infill wai/jite 0.0
palnl wUrdlvU* 0-0 OX)
aiqli vatcrdlv. z-*xli d*?r. OJ)
inHornf lo« m/«! 0.011
thlckntt* «f aquifer m ' 125.0
•urtar flf vtrt. f«l Ucltena '1
aurtxr at •«!!• 0
»urrti«r •! dr«i(it 33
Mirtxr of eonparlrwnti <
|44J TU U tU 09 :
I4U TU U IU *Jfl
!+*j tu u iu oa
1*4J wj u tu oa
!14J UJ U IU 12]
tU IU U IU 131
iu TU u iu in
iiu TU u iu in
nu iu u tu in
IIU tU U IU 171
nu ITJ u iu in
itu iu u tu m
nu »u u iu in
tu tu u iu xn
iu *u u tu xn
iu tu u iu xn
7U 71* u TU xn
7u wj u iu xn
74-i MJ u 3u xn
TU iu u iu xn
7u tu u MJ xn
IJi ,*Si tS Si IS
7IJ IIU U IU Xll
TiJ IHJ u tu xn
TfJ 11X1 U tU XII
TU iiu u iu xn
I44J IU tU IU -T14.
IIU IU 1IJ IU -1147.
iu txa iu iu -«!.-
ni nu tu tu -m.
f'g. 20 ^
figure 20. Streamline pattern, discharge surplus water  on
           compartment I
(m)
155. _
130. .



105. .
BO.
55.



o
\/
- (, 	 > "1' l( ll 1' 	 .
~ C^* "y
_ c!rS~s j
" ^,^3' 'I*' ~yji\ n /]i i i 1 1 | i I i A . . . . .
^f-1 1 , III ' ' ' III'11 I1 I1 1 1 1 f f
gj 1 1 1 1 1 1 1 1 ' ' 1 1 1 1 1 1 I 1 1 1 ' ' ' ' ' >-] —
: ^— 	 	 ' 	 '• '• : : '• • • • ' ' 1 1 'i i n^
- (W^\ !!',',!!" !'!!!!!!!!" ' ' ' ' ' eg
-
;,,,,,,,,, i mi [i.iiiiiinmiiiii.ii-LLllllllllllHIIIII 1
30. 55. 80. 105. 130. 155. 180. (m)
PROGRAM TRIFLO
Oi*nxo Sceitdulotrt 2n4 p«ri*d
fdnlcll rnn/ytar 0.0
Mlnt Mt«dl*U* ' _ 0.0 0.0
aflota Mttrdiv. i-*iit 4«(f. 0.0
•nllarmllevm/d 0.011
thickntu *! tquf(«r m 122.0
•w6«r •( v«rt. fatlicltoni t
•urt>*r •( ««IU ' 0
•i*rti«r of draini 33
I4IJ tU U IU XJl
1,15 ili iS !u til
ISi £i i3 !u S

11 U tU U IU 111
ttO TU U IU Ul
1HJ TU 1* tU Ul
i::j, s.i a IH ii:
II4J «TJ U IU Xll
1I4J IU U tU Ut
\\\A IU U IU Ul
iu iu u iu xir
iu tu u tu in
7U TU U IOJ UT
TU I4J U IU U)
7U tU U »J XiT
7U «-• U 744 UT
74J tU U MJ 117
TU IMJ U IOJ Ul
TU t»4J U IU 1IT
79J 160 U IU XII
7t,» tlU U tU U7
71,8 IIU U IU XIT
TU IHJ U IU UT
144J TU IU IU -HI
'iu nj iu tu -*»4.
TU UU tU IU -440.
f '9- 21 VVV,
ti/u/ttu, *fi VV
 figure 21. Streamline pattern, discharge surplus water in. sever-

                                  1153
•••>?-" vr

-------
                                        19

     Downstreams of the compartment (east) an additional groundwater-
     withdrawal system is installed to minimize the stream off  In
     September the system started with groundwater extraction at a
     flowrate of approximately 60 m3/h, and probably it will be con-
     tinued till the end of November 1988.  The Cd concentration was
     approximately 120 ug/1 in the beginning of November 1988.

4.3  Water treatment system

     -Flow
      The average hydraulic loading of a resin tank was  almost  24
      Bedvolumes/h.  The design flow was 33  Bedvolumes/h.  The  average

          "                                 the «*•»  T-d ^o
     -pH- influent
      The  percolates from the  separate compartments differed from
      pH and ionic composition (Al, Ca) . An accurate pH adjustment was
      necessary to prevent flocculation in the resin tank at high pH
      (>5.5). While at  low pH  «5.0) the Cd-adsorption of the resin
      decreased.

     -Cd loading  of the resin
      The  Cd loading of the resin varied between 7 to 0 5 g Cd/1 resin
      and  the disign value was 6 g Cd/1 resin. No effect of hydraulic
      loading rate  could be found on the Cd loading of the resin  To
      the  resin only slight mechanical demage was observed,  the resis-
      tance  for a clean resin bed increased from 0.3 till 0  5 bar
     Although humic acid substances are  adsorbed to the resin no"
     significant effect on Cd loading of the resin was found durinr
     normal operation.                                             b

    -Removal efficiency
     Under  "normal" circumstances the Cd-concentration of the  disehar-
     Stfi?V?n    /fX ^ f°r the recirculati°" water (infiltration
     sS  Lr-    ?B/   The average removal  efficiency  can be  estimated
     96%.  During two periods  the performance  of the watertreatmenf
     ^ooo      10n WaS P°°r- From ?5   of December 1987 till January
     i!Sr^ ?? °f ?e infJU6nt was  4 instead  of 5 *s shown on  the pH
     meter display  Hence the  necessary pH adjustment  of  the influent
     did not take place.  High  Cd influent concentrations  caused  short
     hold  out times for the resin tanks in that  period
     In the period 4»  of Augustus  till 6th of  September  1988 the
     steady increase of Al- and  Ca-concentrations  in the  influent
     together with low  Cd- concentration resulted in low Cd- loading of
     the resin.  In table  8  four  influent -analyses  are  shown
                                  1154

-------
                                   20
Table 8 . . Cation. concentrations of. the Influent: (percolate)
Cation

Cd (ug/1)
Cu (ug/1)
Ni (ug/1)
Zn (ug/1)
Ca,(ug/l)
Ag (ug/1)
Al (ug/1)
Fe (ug/1)
pH (ug/1)

5 jan '88
1550
3
10
. .720
135000 :
1450
15
5
• . date
21 June '88
125
3
• •. 10 ;.
360 • ,
96000 ;
840
15
5.2

9 aug. '88
310
20
40
1400
, 160000
'20
7600


 At Start, of the .neutralization of compartment II  and III .Al -and Ca
 concentrations dropped strongly.due to flocculation.

 Taking the two above mentioned periods of poor removal efficiency
 in account the average removal efficiency drops from 96%. to',88%.
 The total amount of Cd removed from the influent is 385 kg. The
 pumped.cumulative quantity Cd should be around 440 kg, as mentio-
 ned in 4.2.:                               •            '     '  •

 -Regeneration of the resin     .           •  • '•    : •    •    ••'•  -
  The regeneration caused less trouble. The applied air hold down
  procedure worked out well. Additional regeneration in laboratory
  experiments with a NaOH/NACl mixture and with cyanide had only
  slight influence on the Cd adsorption.

 Conclusions           • .•        .   •• •'  •            •            .:

 Desorption of cadmium      •   .   •" •    :
 *    the, Cd desorption from the .soil was almost complete;
      90%-of .the soil samples • gave Cd values of   Img/kg d.m;'
 *    the desorption of Cd  is  good comparable with  the laboratory
      studies especially the column leaching experiments;-     '
 *  •• approximately 400 kg  Cd  is  removed from the soil;
 *    change of soil pH- took quite a long time,.,

 Hydrologic system
 *.   start up took a long  time due to air inclusion  in  the ;soil;
 *    the permeability of the  soil- layer was rather low, not with-
      standing the greater  part middle to course sand present  in
      the soil;
 *    the hydrological system  reacts slowly  to  flowrate  changes.
      The discharge  flow was mostly too low  according to the  design
      criteria.
 Water  treatment  system
 *    the watertreatment  system operated according to  the  design
      criteria;
 *    for proper  operation special attention has  to be given to
      influent pH,  and Ca-,  Al-concentration in the influent.

                                  1155

-------
                         •           21

 Application
 *    The  remedial  action  continued the whole winterperiod; 1987-
      1988.  Temperatures o'f less than minus 10 C cause no difficul-
      ties.  The water  treatment' plant was roofed and kept free of
      frost.

 Costs
 *    The  total costs  of the in situ remedial action are comparable
      with the estimated costs.
 *    The  resin will be  sold after this remedial action. The com-
      mercial value is quite high.

 Recommendation
 *    Apply more in situ remedial actions for the extraction of
      contaminants from  the soil;
 *    laboratory experiments can reasonably predict the full scale
      soil desorption performance of the contaminant;
 *    additional attention should be given to the hydrological
      charactaristics of the contaminated site.

 From the carried out in situ remedial action it is clear that
 several specialists are required viz.  hydrologists,  environmental
 engineers, watertreatment engineers and soil scientists. To make
 in situ treatmenfa success they have to co-operate  very close.
november 1988
                                 1156

-------
                                             Appendix 5-F

    Physical/Chemical Extraction Technology Case Studies
                            Debris Washing, United States
 The attached article has been used by permission of Air and Waste
Management Association, Pittsburgh, PA. A full report of this study
         has been issued by U.S. Environmental Protection Agency:
   EPA 540/5-91/006a "Technology Evaluation Report:  Design and
     Development of a Pilot-Scale Debris Decontamination System."
                      1157

-------
             CONTROL  TECHNOLOGY


           Development  and Demonstration of a
             Pilot-Scale  Debris  Washing System
                          Michael L. Taylor
                       IT Environmental Programs, Inc.
                       (formerly PEI Associates, Inc.)
                            Cincinnati, Ohio
                            Naomi P. Barktey
                       U.S. Environmental Protection Agency
                       Risk Reduction Engineering Laboratory
                              Cincinnati, Ohio
  Metallic, masonry, and other solid debris that may be contaminated with
  hazardous chemicals litter numerous hazardous waste sites in the United
  States. Polychlorinated biphenyls (PCBs), pesticides, lead or other metals
  tretome of the contaminants of concern, in some cases, cleanup standards
  h*vc been established (e.g., 10 fig PCBs/100 cm1 for surfaces to which
  humans may be frequently exposed). Decontaminated debris could be either
  returned to the site as "clean" fill or, in the case of metallic debris, sold to a
  metal smelter.
  This project involves the development and demonstration of a technology
  specifically for performing on-site decontamination of debris. Both bench-
  tcale and pilot-scale versions of a debris washing system (DWS) have been
  designed, constructed, and demonstrated. The DWS entails the application
  of an aqueous solution during a high-pressure spray cycle, followed by
  turbulent wash and rinse cycles. The aqueous cleaning solution is recovered
  *nd reconditioned for reuse concurrently with the debris-cleaning process,
  which minimises the quantity of process water required to clean the debris.
 More than 1,200 sites are included on
 the UJ5. EPA's National Priority List
 (NPL), and numerous other sites have
 been proposed for inclusion. Many haz-
 ardous waste sites contain toxic organic
 and/or inorganic chemical residues
 which are intermingled with remnants
 of razed structures (wood, steel, con-
 crete block, bricks) as well as contami-
 nated soil, gravel, concrete, and metal-
 lic debris (e^., machinery and equip-
 ment, transformer casings, and
 miscellaneous scrap metal). Decon-
 tamination of these materials is impor-
 tant in preventing contamination off-
 tite and in facilitating debris disposal
 in an  environmentally safe manner.
 Since the majority of contaminated de-

 C«P}T||1>1 IWl—A»4 Wuie Mintienxm Auoeution

Jftft
 bris at Superfund and other hazardous
 waste sites has no potential for reuse,
 the purpose of a debris decontamina-
 tion system would be to decontaminate
 the material sufficiently to permit its
 return to the site as "clean" fill or to
 allow its disposal as a non-hazardous
 rather than a hazardous waste.
  After hazardous waste sites suitable
 for demonstration of the on-site DWS
 technology were located and evaluated,
 two versions were developed; the Ex-
 perimental Debris Decontamination
 Module (EDDM) and the  DWS. The
EDDM was demonstrated at a Region
V Superfund site and the DWS at two
Region IV Superfund sites, under the
Innovative portion of the EPA SITE
Program which includes EPA-devel-
oped technologies.

             1158
  Study Objectives

   The design goal for the DWS was to
  produce a transportable, hydrome-
  chanical, self-contained cleaning sys-
  tem consisting of an enclosure for
  washing debris (with an aqueous clean-
  ing solution)  and a closed-looped
  cleaning solution purification system.
  An  aqueous cleaning solution was de-
  liberately sought rather than a cleaning
  system which utilizes an organic sol-
  vent. Organic solvents tend to be more
  costly, more difficult to handle, and are
  often perceived (if not actually so) to be
  more toxic than aqueous cleaning solu-
  tions. Field-testing of the system at ac-
  tual hazardous waste sites and with
  various contaminants was the primary
  objective.
   Development arid demonstration of
 a system that achieves these objectives
 has proceeded as follows:
 (1)  Bench-scale testing of a 10 gallon
     "off-the-shelf debris washing
     system.
 (2)  Design, development, and field
     trial of a pilot-scale, 300 gallon,
     EDDM system at the Carter In-
     dustrial Site, Detroit, Michigan.
 (3)  Additional bench-scale testing of
     a 20 gallon version of the EDDM,
     to identify the most effective
     wash solution!).
 (4)  Design, fabrication, and prelimi-
    nary testing of a transportable
    DWS demonstration unit that
    could be field-ready in all weather
    conditions.
(5)  Demonstration of the DWS to de-
    contaminate PCB-contaminated
    transformers at the Gray Super-
    fund site, Hopkinsville, Ken-
    tucky.
(6)  Treatment of process water used
    during  cleaning operation to al-
   lowable discharge levels.


        J. Air Waste Manage. Assoc.

-------
                          TABLE t. SUMMARY OF RESULTS FOR OIL/GREASE AND TOTAL
                              :           SUSPENDED SOLIDS ANALYSIS

Experimental
	 Run Ho. 	
1

2

3

1

2

3

1

t"
2
2
3

3
Total o(1. 2. 1 3


Sample Type*
Cleaning solution

Cleaning solution

Cleaning solution

Cleaning solution

Cleaning solution

Cleaning solution

Wp.No. 1

Wipe No. 2
W^eNo. 1
WfceNo.2
WipaNo. 1

Wipe No. 2
Oil from skimmer


Anelyalt
OH and grease.
mgfliter
Oil and grease.
mgMer
OH and grease.
mg/liter
Total suspended
solids, mg/liter
Total suspended
solids, mg/Mer
Total suspended
solids. mg/Mer
Oil and grease,
me/cm2
Oil and grease.
Oil and grease,
Oil and grease,
Oil and grease,
mg/tm*
Oil and grease,
Oil and grease.
mgXiler
Cleaning Solution .
Water
42

tst

241

S

7

15

1.77

142
10.S4
4.40
2.43

NA»
NA

Sulfuric Add Cone.
~ 161

143

138

128

255

148

Tie

1.75
0.7
4.8
3.81

3.27
1540

BB-100
Cone. 15%, v/y
7

182

319

600

904

1000

oS

,0.25
0.15
0.42
0.26

0.33
3380

Power Ctean
Cone. 1:4 Ratio
1670

1470

2440

ioi

£76

434

030

0.49
0.43
0.48
0.34

0.61
3900

        * Al timptti are postueslmem samples.
        b Not Analyzed.
(7)  Demonstration with herbicide
     (Dicamba) and benzonitrile con-
     taminated drums at the Shaver's
     Farm Superfund site,  Chicka-
     mauga, Georgia.
(8)  Design,  fabrication  and shake-
     down of the full-scale DWS fol-
     lowed by field tests at selected Su-
     perfund sites.

Bench-Scale Tests

  The bench-scale version of a Turbo-
Washer (Bowden Industries)  served as
the debris washer for initial studies.
Experiments  were performed in  a  10
gallon hydromechanical cleaning unit
which incorporated an axial flow
pump; propeller shaft, propeller, pres-
sure chamber (all housed in a heated
tank with a rotating disc for  removing
oil), and a cleaning solution  reservoir.
Measured (the same weight was added
to each piece) quantities of used motor
oil, grease, and soil were applied  to
rusted iron parts to simulate the kind
of grime that is likely to be encountered
on oily, PCB-contaminated metal parts
and debris in the field.
   Four cleaning solutions—(1) tap wa-
ter, (2)  10 percent sulfuric acid solu-
tion, and (3 & 4) aqueous dilutions of
two proprietary detergents  [BB-100
(Bowden Industries) and Power Clean
(Penetone Corporation)] were evaluat-
ed. Three tests  were  performed  with
each cleaning solution. The oil and
grease cotitaminated metal parts were
used for each test. For consistency,
each set of contaminated parts was
matched closely in size, shape, and type
of metal. The parts were arranged simi-
larly in the washer basket and lowered
into the heated tank where they were
exposed to the turbulent cleaning solu-
tion for 30 minutes.
  At the completion of each test, the
cleaning solution was analyzed for oil
and grease and total suspended solids.
Two surface wipe samples from select-
ed  metal  parts were  analyzed  for oil
and grease to determine the level re-
maining on  the metal  surfaces after
treatment in the parts washer.
  Results of the oil/grease  and total
suspended solids analyses are summa-
rized in Table 1. Analytical results ob-
tained for the wipe samples indicated
that relatively large amounts of oil and
grease remained on the metal surfaces
after cleaning  with water or  sulfuric
acid. Clearly more oil/grease  was re-
moved after cleaning with BB-100 and
Power Clean.  Moreover, handling 10
percent sulfuric acid was difficult, and
had a corroding effect on the hydrome-
chanical  cleaning  equipment. Water
and  sulfuric acid were  eliminated as
potential  cleaning  solutions  for  oily
PCB-contaminated debris. Based on
the results of surface wipe testing listed
in Table  1, the  BB-100 solution  was
judged to be a better cleaning solution
than the Power Clean solution.

Pilot-Scale Field Tests at Carter
Industrial Site

  A 300 gallon-capacity pilot scale sys-
tem, the EDDM, was designed, assent
bled, installed (on a 48-foot semitrail-
er) and tested at the Carter Industrial
Superfund Site  in Detroit, Michigan.
The process flow diagram for the sys-
TABLE 2. CONCENTRATION OF PCBs FOUND IN SURFACE WIPES AND BLANKS
                               (ng/tOOcm2)                 .
Belch Number
1





2





Sample Number
1
2
3
4
5

1
2
3- • '•
4
5

Pretraatment
134
490
1280
73
203
F ..Id Blank <1 0
8.0
6090
374
96
1690
Field Blank: 1.0
Posttreatment
SO
178
•56
43
23

13.0
1800
126
10
18

* Reduction
63
64
33
41
87
Avg, 58
, :63 '•
70
66
90
99
Avg 81
 April 1991     Volume 41, No. 4
                                                     1159
                                                                       489

-------
   CONTROL TECHNOLOGY
   tern is illustrated in Figure I. Two 200-
   Ib. batches  of metallic  debris were
   cleaned with the system.  BB-100 sur-
   factant solution was used  as the clean-
   ing agent. Before cleaning, five individ-
   ual pieces of metal from each batch
   were sampled for PCBs by a surface
   wipe technique  in  which hexane-
   •oaked cotton gauze pads were used to
   wipe a 100 cm* area on the surface of
   the object being sampled.1 The metal
   items in Batch 1 were then placed in a
   basket, transferred to the EDDM, and
   cleaned for two hours. In Batch 2, the
   metallic debris was rearranged after
   one hour, then cleaned for an addition-
   id hour. A portion of the cleaning solu-
   tion in the Turbo-Washer was pumped
   through a closed-looped paniculate fil-
   ter into the  oil/water separator. The
   effluent from the  oil/water separator
   was recycled into the module. After the
  cleaning process, five additional wipe
  samples were taken from the same
  pieces of metallic debris, at a location
  directly adjacent to that  of the pre-
  treatment  samples, to determine the
  post decontamination PCB levels.
    The quantity of PCBs on each metal
  surface before and after  cleaning  is
  summarized in Table  2. The average
  percentage reduction of PCBs was 58
  percent for Batch 1 with a  range from
  33-87 percent. Batch  2  had a  range
  from 66-99 percent and an  average re-
  duction of 81  percent. Better cleaning
  results for Batch 2 may be the result of
  removing the  debris basket from the
  EDDM after  one hour and manually
  rearranging the debris so that all sides
  of the debris were exposed to the clean-
  ing solution with the same force of the
  Turbo-Washer. It was then placed
  back into the washer for the second
  hour. In Batch 1 the cleaning process
  was for two hours without  disturbing
  debris placement.
   The surfactant solution in the Tur-
 bo-Washer  was sampled  once during
 each of the two cleaning operations.
  PCB values were 420 and 928 ^g/L. Af-
  ter the debris experiment, the cleaning
  solution was pumped through a series
  Qf particulate  filters and activated
  carbon. The PCB concentration of the
  surfactant solution was reduced to 5.4
  Mg/L  following this treatment.  The
  cleaning solution was passed through
  the activated carbon a second time and,
  the PCB level was reduced to less than
  1.0 ftg/L. Subsequently, under the di-
  rect supervision of EPA's  On-Scene
  Coordinator for the site, the water was
  discharged to the PCB-contaminated
  soil located on site.

  Bench-Seal* Testing of Surfactant

   Based on experience gained during
  the Carter site field test, a bench-scale
  (20 gallon surfactant solution capacity)
  hydromechanical cleaning apparatus
  was designed, constructed, and assem-
  bled. This system consisted of a spray
  tank, ash tank, oil-water separator, and
  ancillary equipment (i.e., heater,
 pumps, strainers, metal tray, etc.). De-
 velopment  of a  bench-scale  DWS al-
 lowed assessment of the system's abili-
 ty to remove contaminants from debris
 and to facilitate  selection of the most
 efficient surfactant solution.
   A survey of surfactant products
 identified five  nonionic surfactants
 (BG-5, MC-2000, LF-330, BB-100, and
 L-422) for an experimental evaluation
 to determine their capacity to solubi-
 lize and remove contaminants from the
 debris surface. Unlike anionic and cat-
 ionic surfactants, nonionic surfactants
 perform adequately  during moderate
 pH changes and in the presence of elec-
 trolytes.*
  Prior to  each  bench-scale  experi-.
 ment, six pieces  of  debris, including
 three rusted metal plates, a  brick, a
concrete block and a piece of plastic,
were "contaminated" by dipping each
• Manufacture™ of thwe lurfacunti «r«r BB-100. Bowien
InduitriM, Hunuvillt. AL; BG-S. Modern Chemical, Jack-
•onville. AR; MC-2000, AlcoUc, Baltimore. MD; LF-330,
GAF Chemicali Corporation, Wayne, NJ; and L-422. Du-
Boi» Chemical*. Cincinnati, OH.
                       piece into a spiking material consisting
                       of a known amount of used motor oil,
                       grease, topsoil, and sand. The pieces of
                       contaminated debris  were then  ar-
                       ranged on a metal tray, inserted in the
                       spray tank and subjected  to a high-
                       pressure spray of  surfactant solution
                       for 15 minutes. At the end of the spray
                       cycle, the tray was transferred to the
                       high turbulence, wash tank, where the
                       debris was washed for 30 minutes with
                       a solution of th« same surfactant as
                       that in the spray tank.  After the wash
                       cycle was completed, the tray was re-
                       moved from the wash tank and the de-
                       bris was allowed to air-dry.
                         Before and after treatment with sur-
                       factant solutions, surface wipe samples
                       were obtained from the six pieces of
                       debris. These wipe  samples Were ana-
                       lyzed for oil and grease. Tjhe results are
                       summarized in Table 3. Based on the
                       results of the  wipe  testing, L-422 was
                       selected  as the surfactant best suited
                       for cleaning oily metal  parts and de-
                       bris.
                        To evaluate the ability of the bench-
                      scale  system to remove specific con-
                      taminants from debris, DDT, lindane,
                      PCB and lead sulfate were mixed into
                      the spiking material. The six pieces of
                      debris were spiked with this mixture,
                      then washed using L-422. Three trials
                      were performed. Surface wipe samples
                      from debris from the first two trials
                      were analyzed for PCB, lindane, and
                      DDT. The surface wipe  samples from
                      the third trial  were analyzed for total
                      lead.
                        The average overall percentage re-
                      duction  of PCBs  and pesticides
                      achieved  during Trials 1 and 2 were
                      greater than 99 and 98 percent, respec-
                      tively. The overall percentage  reduc-
                      tion of lead was greater  than 98 per-
                      cent. The results  are summarized in
                      Tables  4  through 6. It is recognized
                      that actual debris, especially porous
                      material such  as concrete blocks or
                      bricks, is often permeated by PCB-con-
                     taining oils. Testing of these surrogate
                     materials, while important, is not nec-
                     essarily representative of a worst case
                     scenario, although testing did provide
                 OIL        DEBRIS
             COLLECTION    WASHER
  CHIP
REMOVAL
OIL/WATER
SEPARATOR
             Flfluro 1.  Plktt-scate process flow diagram.
                                                                                            DISPOSAL
 SOLUTION
TREATMENT
4QO
                                                    1160
                                               J. Air Waste Manage. Assoc.

-------
       O
       a
       5<
       81
       Sg
       ffiu.
       U.O
       OUI
         Q
Mnw
OOetn'
838 3
  333

"•00
o « « «i
  to o
-. 3 K

                     3


                   o' «
                   | o o

                   IM' S S
                            88 8 8
                            sill
  333

^«po
o n w  » w o>
                 S S O O» O) O
                              o to *- •- f^ eo
                              h. CM 10 —. r) cj
                               oi o> en
                                                             en o
                                                             O) o
                                      S; 5 g g S;
                                      a g s s s
  8^ ^ S o --
  o o o o o
o o o o' ei o
                                      r» \f> w f** n* 5
                                      *- *n cvi w> <£ 4
                                      000800
                                S Z
                                s s
                                               S £
                                               s s
                         R S 2 S fe fc
                         S 8 2 S I S
                             r*. « »- n in en
                             § i i i i §
                      8
                           ui uS tri -«* in
                    ^
                  •«  CM o r^
                  ^  o o r»
                                        S S 9 fc 2 R
                                        S $5 S § 2
                                        in' in tri «c V V
                           8 R S S
                           s s a g
                           •«• in •• n
                                                                              in u> o 9
                                                                              •D CM 0 O
                                                                              en « IA uj vi V

                                      CM CM CM N CV CM

                                      5 S3 3 S S
                                                     in tn in
                                            
-------
   CONTROL  TECHNOLOGY

   data that indicated the cleaning pro-
   cess was efficacious for removing con-
   taminants from the surface of the de-
   bris. Whether or not removal of surface
   contamination is sufficient for a partic-
   ular situation would depend on clean-
   up guidelines.
     After the completion of the'bench-
   scale debris washing experiments, the
   cleaning solution was neutralized to a
   pH of 8 and then pumped through a
   teries of particulate filters and finally
   through activated carbon. During this
   treatment, the PCB, lindane, and DDT
   concentrations were reduced to <2.0,
   0.03, and 0.33 /ig/L, respectively. The
   concentration of lead was  reduced to
   0.2 rag/L after treatment During this
   process, it was noticed that a gel-like
  precipitate was formed when the L-422
  cleaning solution was  neutralized,
  which quickly plugged the particulate
  filters and had the potential of clogging
  the activated-carbon drums. As a re-
  lult, BG-5, which performed almost as
  well «s L-422 in removing oil and grease
  and  formed only a fine precipitate
  when neutralized, was selected  as the
  surfactant to be  used  for the subse-
  quent pilot-scale DWS study.

  Design, Fabrication, and Initial Testing
  of the Flirt-Scale DWS

    Based  on results obtained from
  bench-scale studies, a  transportable,
  pilot-scale debris  washing system was
  designed and constructed. The pilot-
  scale system consists of a 300-gallon
  •pray tank; a 300-gallon wash tank; a
  surfactant holding tank; a rinse water
  holding tank; an  oil/water separator;
  and a solution-treatment system with a
  diatomaceous earth filter, an activated
  carbon column, and an  ion-exchange
  column. Ancillary equipment includes
  a spray tank heater, pumps, particulate
  filters, a metal basket, and a stirrer mo-
  tor. The process flow diagram for the
  DWS is illustrated in Figure 2.
   The pilot-scale  system was assem-
  bled in a Cincinnati, Ohio warehouse.
 Several  tests were  conducted using
 pieces of oil/grease-coated objects
 found in the warehouse. Surface wipe
 samples were obtained before and after
 washing in the pilot-scale system and
 analyzed for oil and  grease. In three
 trials using five objects in each trial, oil
 and grease removed ranged from 19 to
 91 percent end  averaged  76  percent
 The warehouse testing also involved
 optimization of several  test parame-
 ters, including spray  and wash cycle
 durations,  and cleaning solution tem-
 perature. Based on these results and a
 visual inspection of the washed debris,
 the system was judged to be effective in
 removing oil and grease from the sur-
 face of these objects.
          TABLE 5.  SUMMARY OF BENCH-SCALE RESULTS OF CONTROLLED
               DEBRIS ANALYZED FOR PCBs AND PESTICIDES (TRIAL 2)
1 Contaminant
Lindane
! 4.4- DDT
, PCB- 1260
Lindane
4,4' DDT
' PCB-1260
Lindane
4. *• DDT
PCB-1260
Undane
4, 4- DDT
PCB-1260
Lindane
4. 41 DDT
PCB-1260
Lindane
4. 4' DDT
PCB-1260
Pretreatment
Oig'IOOcm')
11.800
8320
1770
6160
7540
1760
6150
5640
1450
5610
5660'
1220
6440
6610
1390
10.300
8400
1620
Posllreatmenl*
(MB'IOOcm'J
0.13 U
2.32
2.0 U
0.31 U
4.6
2.79
0.41
2.61
2.0 U
3.49
10.5
• 4.1
397*
369
66.1
52
223
35
Porcent
Redwnion
-'•
100
99.07
299.69
101)
99.SI4
99.64
99.99
99.95
299.1)6
99.94
99.81
99.66
93.63
94.11
95.24
99.4!) '
97.34
97.8'l
Average Overall Performance
Average
Performance
•™ ™™"^^«»»™
299.91
99.80
94.39
98.22
S98.08
            "U indicates that the target compound was not detected at this level.
Demonstration at the Gray PCB Site

  The Ned Gray Superfund site was
selected for the first demonstration of
the upgraded pilot-scale debris wash-
ing system. The twenty-five acre site is
located in Hopkinsville, Kentucky.
From 1968 to 1987, a metal reclaiming
facility which conducted open burning
of electrical transformers to recover
copper for resale was operated at the
site. Soil, where the transformers were
 burned, was contaminated with lead
 and PCBs. The  demonstration took
 place during December 1989. Ambient
 temperatures wen> at or below freezing
 during the entire operation.
  Subsequent to  the warehouse test-
 ing, the system was disassembled and
 loaded onto a 48-foot semitrailer. The
 system and the ancillary equipment
 was transported to the Gray PCB site.
The entire system was reassembled on
a 25 ft X 25 ft concrete pad that was
     TABLE 6.  SUMMARY OF BENCH-SCALE RESULTS OF CONTROLLED DEBRIS
                        ANALYZED FOR LEAD (TRIAL 3)
Concrete
 : Block
Contaminant
Lead
Lead
Lead
Lead
Lead
Lead
Pretreatment
(MOlOOcm*)
676
414
450
508
414
446
Posttrealmem
(pg/IOOcm1)
6.0
6.0
<3.0
<3.0
<3.0
<3.0
Percent
Reduction
. 99.31
98.55
>9SI.33
>9941
>9S'.27
>9S.33
Average Performance on All Materials Tested
Average
Performance
>99.06
>99.34
>99.20
492
                                                   1162
                                                                                      J. Air Waiste Manage. Assoc.

-------
poured on site prior to the arrival of the
equipment. A temporary enclosure was
also built on the concrete pad (approxi-
mately 25 ft. high) to enclose the sys-
tem and to protect it and the surfactant
solution from rain and cold weather.
  Before initiation of the cleaning pro-
cess, 75 transformer casings, with sizes
ranging from 5 to 100 gallons, were cut
in half with a partner  saw. Pretreat-
ment samples were obtained from one
half of each of the transformer casings
by  using the surface wipe technique
previously described.1
  The transformer halves were placed
in the wash basket and lowered into the
spray tank, equipped with multiple wa-
ter jets that blast loosely adhered con-
taminants and dirt from the debris. Af-
ter the spray cycle, the basket was re-
moved and  transferred to  the  wash
tank, where the debris was washed with
a high-turbulence wash. Each batch of
debris was  cleaned for a period  of 1
hour in the spray tank and 1 hour in the
wash tank. During both the spray and
wash cycles, a portion  of the cleaning
solution was cycled through a closed-
loop  system where the oil/PCB-con-
taminated  cleaning  solution  was
passed through an oil/water separator,
and the cleaned solution was then recy-
cled. After the wash cycle, the debris
  TABLE 7.    RESULTS OBTAINED DURING FIELD DEMONSTRATION OF
                         DWS  AT GRAY PCB SITE

Batch Number

1
2
3
4
s
6
7
a
9
10 .
11
12
13
14
15
IB
17
18
19
20
21
22
23
24
Average PCB Concentration on Surfaces (jig'flOO cm2)
Before Cleaning • Atltr Claming
Average
19.7 (N« . 10)
9.9 (N. 6)
6.6 (N - 4)
4.1 (N.6)
4.0 (N . 8)
2.0 (N. 4)
2.8 (N . 2)
23.5 (N . 5)
8.3 (N . 4)
5.2 (N . 4)
94(N.4)
48.6 (N. 4)
12.3(N.2)
16.7 (N . 2)
18.5 (N. 4)
11.3 (N. 2)
24.8 (N . 4)
6 4^N . 5)
8.3 (N . 4)
24.0 (N. 3)
18.6 (N . 8)
25.0 (N . 4)
8.6 (N . 4)
6 8 (N . 8)
Ring*
<0. 1-94.0
4.8-17.0
5.0-9.9
«0.1 - 12.0
<0. 1-28.0
<0. 1-7.8
1.4-4.3
<0.1-70.0
2.9-23.0
<0 1-9.7
<0.1 -17.0
2.3 - 98.0
9.6- ISO
8.7-250
8.1-27.0
8.6-14.0
1.1-80.0
<0.1-19.0
<0.1-18.0
13.0-45.0
<0.1-44.0
12.0-35.0
1.5-18.0
<0.1-31.0
Average
1.5 (N. 10)
1.S (N . 6)
1.4 (N- 4)
0.8 (N.6)
<0.1 (N . 8)
2.9 (N . 4)
3.9 (N . 2)
1.3 (N. 5)
3.1 (N. 4)
1,9 (N. 4)
3.0 (N . 4)
1.1 (N. 4)
5.1 (N . 2)
<0.1(N.2)
<01 (N.4)
2.0 (N . 2)
2.2 (N. 4)
3.4 (N. 5) ,
3.2 (N . 4)
3.3 (N. 3)
0.4 (N . 8]
<0.1(N.4)
<0.1 (N . 4)
0.3 (N . 8)
Rang?
<0.1-9.7
<0.1-4.7
<0.1 - 3.3
<0.1-4.1
<0.1-«0.1
<0.1 -10.0

<0.1-3.8
1.5-4.9
<0.1-2.8
<0.1 - 9.5
<0.1-3.2
<0.1 - 10.0
<0,1-<0.1
<0.1-<0.1
1.5-2.5
<0.1-8.4
<0.1 - 7.4. •. ,
<0.1-5.3 .
<0.1-9.8
<0.1-2.1
<0.1 -<0,1
<0.1 - <0.1
<0.1-1.4
Average
Percentage
Removed
92 .
85
79
80
>98 , '
b

94
63
63 .
67
98
59
>99
>99
82 '
91
60
61
86
. 98
>99 ,
>99
96
* N indicates the number of samples.
b The distribution of PCB contaminatcn on the surfaces of these transformers is obviously not uniform and therefore in
some uses a meaningful comparison of post-treatment PCB levels with pre-treatment levels cannot be achieved.
                                                                                             Siep 1 • Spray Cycte
                                                                                             siep 2 • Wash Cycle
                                                                                             Siep 3 • Rinse Cycle
                                                                                             DE Filler
                                                                                             water Treatment Step

                                                                                             Pump

                                                                                             Activated Carbon
                                                                                               * e
          Flgur* 2.  Schematic of pilot-scale debris washing system.
  April 1991     Volume 41, No. 4
                                                         1163
                                                                                                                    493

-------
   CONTROL TECHNOLOGY
                     TABLE 8. RESULTS OF SURFACE WIPE SAMPLES ANALYZED FOR BENZONITRILE
                         2,4-DICHLOROPHENOL, 2.6-DICHLOROPHENOL, 1,2,4-TRICHLOROBENZE.NE
                                                   (ug/IOOcm1)
Btleh NumN
1
2
3
4
&
t
7
9
9
10
Simpl* Numbtr
1
2
1
2
1
2
1
2
1
2
1
2
2
1
1
2
1
2
BenzonKrlle

U0» (50)
1301 (50)
125
SO
43
28
4400
2700
47000
22000
10»(5)
8«<5)
200
320

3500
22»(5)
1400
Pctniumwn
NO6
ND
117 '
7.8«<5)
NO
ND
NO
ND
10»(5)
7.9* (S)
ND
ND
ND
10»(5)
28
ND
7»(5)
ND
ND
2,4-Dlchtorophenol
Pnnimwn
NO (SO)
ND(SO)
34
43
ND
ND
MAC
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Pocttrunwn
ND
ND
ND
ND
16* (5)
14»(5)
NA
NA
ND
' ND
ND
ND
NO
ND
ND
ND
ND
ND
ND
2,6-Dfchlorophenol
Pmnauiwnt
ND(SO)
ND(SO)
ND
ND
ND
NO
NA
NA
ND
ND
NO
ND
ND
ND
ND
ND
ND
ND
ND
PocnrMtiwn
ND
ND
ND
ND
ND
NO
NA
NA
ND
ND
NO
ND
ND
NO '
ND
ND
NO
NO
ND
1,2,4-Trlchlijrobeiuene
Pranaimtm
ND(SO)
ND(SO)
ND
ND
ND
NO
NA
NA .
NO
ND
NO
ND
ND
ND
ND
NO
ND
NO
ND

ND
ND
ND
ND
ND
NO
NA
NA
ND
ND
ND
ND
ND
ND
ND
NO
ND
ND
ND
                Now dttKttd In nent ot lh« minimum dcttclabb conc»ntr«ifcn of 5 |i8/lOOcm2 urttu olhciwiu ipcdtM In ptremhMit.
  basket was returned to the spray tank,
  where It was rinsed with water.
    After completion of the cleaning pro-
  cess, posttreatment wipe samples were
  obtained from each of the transformer
  pieces to assess the post-decontamina-
  tion levels of PCBs.s All Superfund site
  activities described in this document
  were governed by EPA approved
  Health and Safety and Quality Assur-
  ance Plans.3-4 In the case of the metallic
  debris sampled in this study, the post-
  treatment wipe sample was obtained
  from a location adjacent to the location
  of the pretreatment sample. This was
 necessary because wiping the surface
 removes the contamination and if one
 were to wipe the same surface after
 cleanup, the results obtained would be
 biased low.  !
  The average concentrations of PCBs
 on the internal surfaces of the trans-
 former casings before and after clean-
ing are summarized in Table 7. Con-
 centrations before  treatment ranged
 from 0.1 to 98 ng/100 cm2. Posttreat-
 ment analyses showed that all but sev-
en of the transformers that were
cleaned had a PCB concentration less
than the acceptable level of 10 Mg/100
 cm2. The seven transformers with  a
 concentration greater than the accept-
 able level were rewashed in the DWS,
 and posttreatment samples were ob-
 tained and analyzed. The concentra-
 tion of PCBs in these seven samples
 after the second wash was below the
 detection limit of 0.1 jig/100 cm2.

 Treatment of the Procau Water

  After all transformers at the site
were decontaminated, the surfactant
solution (BG-5) find the rinse water
were neutralized to a pH of about 8,
                     TABLE 9. RESULTS OF METAL STRIP SAMPLES ANALYZED FOR BENZONITRILE,
                       2,4-DICHLOROPHENOL, 2,6-DICHLOROPHENOL, 1A4-TRICHLOROBENZENE
Bitch Numb*
1
2
3
4
S
6
7
S
9
10
Sunpl* Number
1
1
1
1
1
1
1
1
1
1
BeruonNrlte

4.9
S.4»(12)
NO (0.15)
NO
190
1.0
140 ,
16
61
NO (1.2)
PostnMnw*
0.89
0.31« (0.12|
ND(O.IO)
ND
0.61 (0.22)
ND (0.08)
5281 (0.16)
0.41* (0.15)
O.i1»(0.17)
NO (0.11)
2,4-Dtehlorophenol
Pneunrnx
6 J3» (0.33]
S.0«(1i)
26
MAC
NO (028)
ND (0.085)
NO (0.19)
NO (0.14)
NO (0.10)
NO (0.12)
PoUBMtTIM
ND* (0.15)
NO (0.12)
1.9
NA
NO (022)
NO (0.08)
NO (0.16)
ND (0.15)
ND (0.17)
NO (0.11)
2^-Diehlorophenot
PrMvjMrrwnt
NO (0.15)
ND(12)
ND (0.15)
NA
NO (028)
NO (O.OS5)
ND (0.19)
ND (0.14)
NO (0.10)
NO (0.12)

ND (0.15)
ND (0.12)
NO (0.10)
NA
NO (022)
ND (0.08)
NO (0.16)
NO (0.15)
ND (0.17)
NO (Oil)
• Ellm»:.d it*t*l Ini than 5 tnwt dttaction lim». Daucton bnti arc indicated in partnthMh
6 Men* dttKKd at sp*cr*dd«»c1i[>n limit.
C Mgj •n«ltflad
1,2,4-Trlehlorabenzent
Pnnalnwil
NO (0.15)
ND(1.2)
NO (0.15)
NA
ND (028)
0.13* (0.085
NO. (0.19)
NO (0.14)
ND(0.10)
ND (0.12)


NO (0.1S)
MO (0 15)
DO (0.10)
NA
»IO(022)
ND (0.080)
MO (0.16)
MO (0.15)
ND (0.17)
NO (0.11)

494
                                                  1164
                                             J. Air Waste Manage. Assoc.

-------
                                TABLE 10.  RESULTS OF SURFACE WIPE SAMPLES ANALYZED
                                                 FOR DICAMBA, 2,4-D, 2,4,5-T
                                                          (lig/IOOcm2)
Batch Number
4
6
6
7
•
9
10
Sample Number
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Dtettnba
Pnmum**
1.9
3.4
NO
NO
NO (2.7)
NO (2.7)
7 J» (2.7)
IS
£5
13
1.7
NO (2.7)
41
180
Pa.nrn.trwi!
0.63* (0.27)
NO
NO
2.6
NO
NO (2.7)
1.1
23
5.7* (2.7)
0.62* (0.27)
0.63* (0.27)
NO
0.30* (0 .27)
0.34*10.27)
2,4-0
Pivtrtilmcnt
ND°
NAe
NO
NO
NO (12)
NO (12)
NO
NO (12)
NO (12)
NO
NO
NO (12)
NO (12)
NO (12)
PotQnMQrwnt
NO
NA
NO
NO
NO
NO (12)
NO
NO
NO (12)
NO
NO
NO
NO
NO
2,4,5-T
Pretvatment
NO
NA
NO
NO
NO (2.0)
NO (2.0)
NO
NO (2.0)
NO (2.0)
NO
NO
NO (2.0)
NO (2.0)
NO (2.0)
Posttreament
NO
NA
NO
NO
NO
NO (2.0)
NO
NO
NO (2.0)
NO
NO
NO
NO
NO
                      •  Estimated result lest thin S times (Mcaion Urn*. Detection limits are indicated in perenlhetit.
                      0  None detected In excess of minimum detectable concentration oMJieamba it 0.27; 2,4-D at 12: and 2.4.S-T at 0.20
                        unless otherwise specified.
                      e  Not analyzed.
                                  TABLE 11.  RESULTS OF METAL STRIP SAMPLES ANALYZED
                                                FOR DICAMBA, 2,4-D, AND 2,4,5-T
Batch Number
2
3
4
6-
7
9
10
Semple Number






1
Dicamba
PretresBnent
0.37
82
0.057
25
2.3
2.3
0.40
PMtnnrwnt
0.030* (0,0097)
13
0.0201 (0.0065)
NO (0.059)
0.023* (0.062)
0.11
0.017*(0.0062)
2,4-D
PretrMknent
NDb (0.34)
NO (0.36)
NO (0.029)
NO (0.26)
NO (0.50)
NO (0.32)
NO (0.24)
POt BTV Ul 1 • lit
NO (0.043)
NO (0.38)
NO (0.029)
NO (0.26)
NO (0.28)
NO (0.040)
NO (0.028)
2.4M
Pretrai&nent
NO (0.0056)
NO (0.060)
NO (0.0048)
NO (0.044)
NO (0.084)
NO (0.054)
NO (0.040)
PeettMtment
NO (0.0072)
ND (0.064)
NO (0.0048)
ND (0.044)
NO (0.046)
ND (0.0066)
NO (0.0046)
             * Estimated result less then 5 times detection limit. Detection limits ve indicated in parenthesis.
             0 None detected at specified detection limit.
                                TABLE 12.  RESULTS OF METAL STRIP SAMPLES ANALYZED
                                                 FOR DIOXINS AND FURANS
                                                            (ng/g)
Analyses
HpCDD
HpCDF
HiCDD
HiCDF
OCDD
OCDF
PeCDD
PeCDF
TCOD
2.3.7.8-TCDD
TCDF
Semple Number 1
Pretreatment
1.2
NO (0.16)
2.8
ND (0.12)
4.1
NO (01 9)
NO (0.11)
ND (0.066)
NO (0.10)
NO (0.10)
NO (0.062)
Posttreatment
ND° (0.30)
ND (0.23)
ND (0.24)
NDI0.11)
ND (0.66)
ND (0.36)
NO(O.U)
NO (O.OBS)
ND (0.13)
NO (0.13)
ND (0.079)
Sample Number 2
Pretreatment
ND (0,22)
ND(0.17)
ND(0.18)
NO (0.11)
1.6* (0.45)
ND (0.22)
NO (0 10)
ND (0.065)
NO (0.10)
ND (0.10)
ND (0.065)
Posttreatment
ND (0.31)
NO (0.23)
0.71* (0.27)
NO (0.17)
5.5
NO (0.34)
ND (0.16)
NO (0.092)
ND (0.15)
ND (0.15)
ND (0.089)
                             * Estimated result less than 5 times detection limit.
                             e Not detected at specified detection limn. Detection limits are indicated in parenthesis.
April 1991     Volume 41. No. 4
                                                           1165
                                                                                                                         495

-------
  CONTROL TECHNOLOGY
  using concentrated sulfuric acid. The
  neutralized surfactant solution and
  rinse water was treated by passing it
  through a series of particulate filters,
  an activated-carbon drum, and finally
  through an ion-exchange column. The
  treated water was stored in a 1000-gal-
  lon polyethylene tank pending analy-
  sis. The before- and after-treatment
  water samples were collected and ana-
  lyzed for PCBs and  selected metals
  (cadmium, copper, chromium,  lead,
  nickel, and arsenic).
   The PCB concentration in the water
  was reduced to below the detection lim-
  it of 0.1 /ig/L. Concentrations of five of
  the six metals were reduced to the al-
  lowable discharge levels set by the City
  of  Hopkinsville. Arsenic remained
  above the allowable level. After receipt
  of the analytical results for the water,
  the treated water, which was stored in
  the holding tank, was pumped into a
  plastic-covered, 10,000 cubic yard pile
  of contaminated soil at the site. An in-
  cision of about % inch was made into
  the plastic covering at the  top of the
  •oil pile, and a rubber hose was inserted
  into the incision. After all of the water
  was pumped into the contaminated
 soil, the hose was pulled  out and the
 incision was covered with tape.
   Equipment was  decontaminated
 with a high-pressure wash. Wash water
 generated during the decontamination
 process was collected and treated in the
 water treatment system. The DWS and
 the enclosure was disassembled, loaded
 into the semitrailer and transported to
 Cincinnati
   During this site cleanup, 75 PCB-
 contaminated transformer casings (ap-
 proximately 5000 Ib) were cleaned. A
 total of 1000 gallons of process water
 (surfactant solution and rinse water)
 was utilized in this demonstration. Ad-
 ditional waste generated included
 three Activated-carbon drums and  four
 miscellaneous drums containing safety
 gear, solids,  and plastic. All of the
 transformers are now considered clean
 and could be sold to a scrap metal deal-
 er or to a smelter for reuse.

 Demonstration at Ihe Shaver's Farm
 Drum Disposal SHe

  A second demonstration of the DWS
 was  conducted at the Shaver's Farm
 drum disposal site in Walker County,
 Georgia. Fifty-five gallon drums con-
 taining varying amounts of a herbicide,
 Dicamba (2-methoxy-3,6-dichloroben-
 zoic ncid), and benzonitrile, a precursor
 in the manufacture of Dicamba, were
 buried on this 5-acre site. An estimated
 12,000 drums containing solid and  liq-
uid residues from the manufacture of
  Dicamba are thought to have been bur-
  ied here from August 1973 to January
  1974. EPA Region IV  had excavated
  more than 4000 drums  from one loca-
  tion  on the site when this demonstra-
  tion occurred in August 1990.
    The  pilot-scale system  was trans-
  ported to this site on a 48-foot semi-
  trailer and assembled on a 24 ft. X 24 ft.
  concrete  pad. The temporary enclo-
  sure, used previously at the Gray site,
  was reassembled to protect the equip-
  ment from rain. Ambient temperature
  at the site during the demonstration
  ranged from 75 to 105 degrees Fahren-
  heit.
    Fifty-five gallon, pesticide-contami-
  nated drums were cut  into four sec-
  tions. Pretreatment surface-wipe sam-
  ples were obtained from each section
  and two metal strips (approximately 6
  cm by 3 cm) were taken from one of the
  sections with a nibble. The latter were
  analyzed in an attempt  to obtain data
-  which would corroborate surface-wipe
  test res.ults and to determine if contam-
  inants were imbedded in the metallic
  surfaces. The drum pieces were placed
  in the spray tank of the DWS for one
  hour of surfactant spraying,  then
  placed in the wash tank for an addi-
  tional hour of surfactant washing, fol-
  lowed by 30 minutes of water rinsing in
  the spray tank. The drum pieces were
  then allowed'to air-dry prior to obtain-
  ing the posttreatment  surface-wipe
  and metal strip samples. Ten batches
  of one to two drums per batch were
  treated during this demonstration.
   Data in Tables 8 to 12 provide an
  indication of the effectiveness of the
 DWS technology for removing  pesti-
 cide and related contaminants (other
 than PCBs) from the surfaces of exca-
 vated  drums. Pretreatment concentra-
 tions  of benzonitrile in  surface wipe
 samples ranged from 8 to 47,000 /ig/100
 cm2 (average: 4556 pg/100 cm2)  while
 post treatment samples  averaged 10
 fig/100 cm2 with a range from below the
 detection limit to 117 >«g/100 cm2. Pre-
 treatment Dicamba values averaged 23
 >ig/100 cm2 (range: below  detection
 limit to 180 /ig/100 cm2) while  post-
 treatment concentrations ranged from
 below  detection limit to 5.2 fig/100 cm2
 and averaged 1 /ig/100 cm2.

 Conclusions

  Field-test results obtained using the
 upgraded pilot-scale DWS in demon-
 strations at two Region IV Superfund
 sites showed the unit to be both trans-
 portable and rugged. Extreme high and
 low  temperatures had little effect on
 the operation of the equipment. The
system was successfully used to remove
PCBs from transformer casing surfaces
and certain pesticides, dioxins and fu-
ran residues from drum surfaces. While
the system has not been proven effec-
  tive for removal! of all types of organic
  contaminants from surfaces of debris,
  results obtained to date are considered
  promising.
    The cleaning isolation was recovered,
  reconditioned and reused during the
  actual debris-cleaning process  which
  minimized the quantity of process wa-
  ter  required for the decontamination
  procedure. The water  treatment sys-
  tem was effective in reducing contami-
  nant concentrations, with the excep-
  tion of arsenic a.nd Dicamba, to below
  the  detection limit.
    Planned progression of this U.S.
  EPA-developed technology includes
  design, development, and demonstra-
  tion of a full-sc«je, transportable ver-
  sion of the DWS unit. Commercializa-
  tion of the technology will be handled
  by IT Environmental Programs, Inc.
  (formerly PEI Associates, Inc.)

  Acknowledgments

    The authors acknowledge the efforts
  of IT Environmental Programs, Inc.'s
  key  engineers, Majid  Dosani,  John
,.,.Wentz, and Ayinash., Patk.ar. who .per-
  formed a portion of this work under
  Contract No. 68-03-3413 with the En-
  vironmental Protection  Agency's Risk
  Reduction Engineering Laboratory.

  References

  1.  Field Manual for Grid Sampling of PCB
    Spill Sites to Verify Cleanup U.S. Envi-
    ronmental Protection Agency, EPA 560/
    5-86/017, May 1986.
 2.  Test Methods for  Evaluating  Solid
    Waste, Volume ]C, 3rd ed., Office of Sol-
    id  Waste and Emergency Response,
    Washington, D.C., November 1986.
 3.  Standard Operating Safety Guides, Of-
    fice of Emergency and Remedial Re-
    sponse,  Hazardous Response  Support
    Division, Edison, NJ, November 1984.
 4.  Quality Assurance Procedures for
   RREL, Risk Reduction Engineering Lab-
   oratory, Cincinnati, OH, June 1989.
     N. P. Berkley is an Environmental
   Scientist with the U.S. EPA's Risk
   Reduction  Engineering  Laboratory
   in Cincinnati, OH 45268. M. L. Tay-
   lor is Director of Research and Devel-
   opment with IT Environmental Pro-
   grams, Inc.,  1M99 Chester Road,
   Cincinnati, OH 45246. This manu-
   script was submitted for peer review
   on January 3, 1991. The revised
   manuscript was received March 1,
   1991.
486
                                                     1166
                                                 J. Air Waste Manage. Assoc.

-------
                                          Appendix 6-A

   Ground Water Pump and Treat Technology Case Studies
Decontamination of Ville Mercier Aquifer for Toxic Organics,
                           Ville Mercier, Quebec, Canada
                     1167

-------
ASSESSMENT OF CONTAMINATED GROUNDWATER

    TREATMENT AT VILLE MERCIER, QUEBEC
                        by

   M. Halevy, R.M. Booth, J.W. Schmidt and S.A. Zaidi
     Environment Canada, Conservation & Protection
            Wastewater Technology Centre
                 Burlington, Ontario
                    January 1992
                       1168

-------
1.0    INTRODUCTION

       In Canada, the first large, full scale aquifer restoration effort was initiated in Ville Mercier,
Quebec, by the Quebec Ministry of the Environment (MENVIQ), in July 1984.  Due to the nature of
this large scale remediation project, a unique opportunity existed to collect and analyze operational
data in order to assess the performance of the remediation effort.

       Consequently, a Canadian consortium consisting of The SNC Group, the University of
Sherbrooke and Laval University, was  awarded, under the scientific authority of the Wastewater
Technology Centre (WTC), in July 1988, a research, development and pilot demonstration contract,
as a result of an unsolicited project proposal submitted to the Department of Supply and Services
Canada (DSS) in June 1988, entitled "Aquifer Decontamination for Toxic Organics: The Case Study
of Ville Mercier, Quebec".

       The aim of the study (1988-89) was to determine the impact of the existing restoration program
on the rehabilitation of the aquifer and to establish the best strategy to complete the requirements of
the remediation program.

       The project focused on the following two (2) areas of research:

a)     assessment of the hydrogeological impact  of the  existing restoration  program on  the
       contaminated aquifer.
b)     performance evaluation of the groundwater treatment plant, and the conduct of bench and
       pilot scale treatability studies to improve the plant's performance. The treatment study dealt
       exclusively  with physical chemical technologies similar to  those used in the full scale
       groundwater treatment plant of Ville Mercier.
                               1234
 1.1    Historical Background

       The Ville Mercier site is located on the south shore of the St. Lawrence River, 20 km southwest
 of Montreal, Quebec (Figure 1).

       In  1968, the "R6gie des Eaux du Quebec", a Quebec government agency, authorized an oil
 carrier to temporarily store oily liquid wastes, mainly from petro-chemical industries, in a former sand
 and gravel pit located in an esker formation near Ville Mercier, with the intention of recovering the
 hydrocarbons at a later date.

       From 1968 to  1972, an estimated 40 000 m3 of organic liquid wastes were  stored in an
 abandoned pit (lagoon) located on a ridge composed mainly of sand and gravel overlying an extensive
 fractured dolomitic sandstone aquifer.  Contaminants were soon noticed in nearby farm wells and
 disposal of liquid waste was discontinued. Shortly after, clean-up of the lagoon was conducted and
 incineration of the organic waste was performed at a liquid waste incinerator built nearby for the
 purpose.

       Ten years after the disposal of liquid wastes had ceased, a plume of contaminated groundwater
 (Table 1) extending over 30 km2 was mapped south and west of the lagoon affecting local municipal
 water supplies.

        In 1982, studies to understand the migration of the contaminated groundwater, suggested that
 a pumping and treatment system, which would consist of three (3) purge wells with an operating
 hydraulic capacity of 0.076 m3/s, would create a hydraulic barrier sufficient to control the progression
 of the contaminated plume, and restore the aquifer within a five-year period.
                                             1169

-------
                                                                   VHle Mercler site
                Otto wo
                  Toronto
                                                VILLE MERCIER SITE
FIGURE 1   LOCATION MAP OF THE VTTJJK MERCIER SITE (REF 2)
                                       1170

-------
TABLE 1     DETECTED GROUNDWATER CONTAMINANT'S
              AND PLANT EFFLUENT QUALITY OBJECTIVES
Concentration (jig/L)

Contaminant
Total Phenol
1,2 Dichloroethane (1,2 DCA)
1,1,1 Trichloroethane (1,1,1 TCA)
1,1,2 Trichloroethane (1,1,2 TCA)
Trichloroethene (TCE)
Chloroform (CCls)
Chlorobenzene (-)
Trans 1,2 Dichloroethene
(Trans 1,2 DCE)

Minimum
167
76
23
16
19
1
2
6


Maximum
1000
517
326
326
349
165
16
139


Mean
507
187
127
-
114
-
-
-

Effluent
Objective
2.0
50
33
-
4.5
-
-
-

     PCBs:
     Arochlor 1242, 1254, 1260

     Iron (mg/L)
     Manganese (mg/L)
0.01

0.84
0.06
 0.08

12.00
 0.82
0.05

8.19
0.18
0.01

0.30
0.05
       Based on this information, the first large scale Canadian program to remediate an aquifer was
 commissioned in 1983 by the MENVIQ in Ville Mercier, Quebec.

       The MENVIQ allocated, at the time, $5.7 million to restore the site, as follows: $3.0 million
 to build the treatment facility, and $2.7 million to cover its 5-year operation by The SNC Group/Hydro-
 Mecanique Inc. (SNC/HMI). In 1986, an additional $1.7 million supplemented the original operating
 cost estimate in order to continue the remediation work until July 1989.  Since then (1989),  the
 MENVIQ has assumed the responsibility for the operation of the groundwater treatment plant.

       The treatment train incorporated the following physical chemical unit operations:

 a)     air stripping, to partially remove the volatile organics (VOCs) of concern, and to aerate the
       groundwater for iron and manganese oxygenation; :-•-
 b)     alum and polymer additions;
 c)     clarification and rapid sand filtration; and
 d)     granular activated carbon adsorption, to remove the remaining VOCs and other organics of
       concern.


 1.2    Operational Difficulties and Modifications Implemented

       The original treatment train incorporated the following elements: three groundwater purge
                                           1171

-------
  wells,  an air stripper, a flash-mixer, a pulsating clarifier, two rapid sand filters, three activated
  carbon absorbers, and a sludge dewatering chamber.

         During the first year of operation, serious operational difficulties were encountered. These

        biological fouling throughout the plant;
        a surge of Dense Non-Aqueous Phase Liquids (DNAPLs) in the pumped groundwater, and
        ensuing fouling of the treatment train;
        excessive iron precipitation in the air stripper;
        incomplete removal of iron;
        substantial granular activated  carbon loss during the frequent backwashes required to
        minimize plugging of the GAG adsorbers;
        reduced capacity of the wells due to fouling of the strainers;
        incompatibility of flocculating and dewatering polymers; and
        increased load of organic contaminants in the pumped groundwater.
were:
a)
b)

0
d)
e)

f)
g)
h)
 1.3    Existing Treatment Train
           i  section briefly Presents the treatment train since its last modification in January 1987
 It should also be noted that the remediation effort to decontaminate the aquifer has been extended
 by the MENVIQ beyond its 5-year program, originally scheduled to terminate in July 1989. The plant
 is still in operation today (January 1992).

        The presently operating treatment train incorporates the following process elements- three
 groundwater purge wells, hydrogen peroxide addition, chlorine addition, an air stripper, a flash-mixer
 chlorine dioxide addition, a pulsating clarifier, two rapid sand filters, three activated carbon adsorbers
 and a sludge dewatering chamber.                                                          '

   , „  The existing treatment plant's process diagram and characteristics are summarized in Fieure 2
 and Table 2, respectively.


 2.0     FULL SCALE PLANT PERFORMANCE ASSESSMENT

        SNC/HMI, was mandated by the  MENVIQ to sample, on a semi-monthly basis, raw and
 treated groundwater.  Starting in the second year of operation, sampling was extended to include
 selected locations along the treatment train, and in the fifth year of operation, the original list of
 orgamcs to monitor was  expanded.  The  data were collected and analyzed, and are presented in
 Section 2.1, entitled Historical Performance".
 * A •            ;          sampling monitoring campaigns (1989) were undertaken in the context
ot this study. The intent of these campaigns was to evaluate the impact that each unit process had
on the contaminants found in the groundwater.  The list of targeted contaminants was expanded
solely for the purpose of this performance evaluation. The data were subsequently analyzed and are
presented in Section 2.2, entitled "Full Scale Plant Sampling Monitoring Campaigns".


2.1     Historical Performance

       In order to gain a better understanding of the operation of the treatment train, and to better
assess the impact of the modifications implemented to the train, the historical performsmce data were
sub-divided into three (3) populations:
                                          1172

-------
                                                               i o
                                                           UI
                                                           o

                                                                   CO
                                                                   OL
                                                                   W
                                                            CO   O O3
                                                           	< a:
                                                              . . o o

                                                            fco    S
                                                            Qi ^    ^
                                                            CZ 03
                                                           ^
                                                                CO
                                                            gee
                                                              cc
                                                              tt   ^^
                                                              9- Q co tu
                                                              9: Z < X
                                                              CO	
                                                              oj   td co
                                                              Q   z z
                                                              xaE9
                                                              o z o t:
                                                              fc <=!Q
                                                              UJ   X Q
                                                              Q-   o"
                                                               ST13M
FIGURE
VUJLE MERCEER GROUNDWATER TREATMENT PLANT PROCESS DIAGRAM
                                               1173

-------
                 E
 CO
 cc
 Ul
 CO
 «x
 o
 CO
o
CO
C£
•<
o

a
Ul

I

B
          E -  0
         7fc  E10

         O «o
                              ?  e
                                   CM
a
o
                 E
•° ° r? T*

EV f o


5lJ."?
•  ^"K c ¥

=  S E^ I
(r)  O U = M
       o
       o.
                                  E    «
                                                                  !J2    -B
                                                         p •=" Q.   k *
                                                         E * *-* "O ^ UI i

                                                        z Q   <3 C
                           ot

                           E
                      a

                      E
              «o
              to
              o
                                                                  E E

                                                                  «0 
-------
1.      July 1984 to April 1985: this period corresponds to an evaluation of the original design of the
       treatment train, prior to any modifications.

2.      May 1985 to December 1986: this period corresponds to an evaluation of the partially modified
       treatment train, following hydrogen peroxide and chlorine additions.

3.      January 1987 to August 1989:  this period corresponds to an evaluation of the presently
       operating treatment train, following the additions of hydrogen peroxide, chlorine and chlorine
       dioxide.


2.1.1   Data Analysis: Period 1 (1984-1985)

       Due to the operational problems experienced during the plant's first operating year, a limited
number of samplings (4) were conducted during this time period.

       Table 3 presents the performance evaluation for the groundwater treatment plant for this
period.

       Figures 3 and 4 summarize the information presented, and display influent and effluent mean
concentrations for iron and manganese,  and total phenol and total  chlorinated hydrocarbons
monitored, respectively.


2.1.2   Data Analysis:  Period 2 (1985-1986)

       A more extensive  and detailed data base, consisting  of 35 samplings, and allowing the
performance evaluation of the two following segments of the treatment train, is available for this time
period:

a)     hydrogen peroxide  and chlorine additions, air stripper, pulsating clarifier, sand, filters; and
b)     sacrificial and polishing GAG adsorbers.

       Table  4 presents the performance evaluation for the groundwater treatment plant for this
period.

       Figures 5 and 6 summarize the information presented, and display treatment profiles for iron
and manganese, and total  phenol and total chlorinated hydrocarbons monitored, respectively.
 2.1.3   Data Analysis: Period 3 (1987-1989)

        An even more extensive and detailed data base, consisting of 65 samplings, and allowing the
 performance evaluation of the following segments of the treatment train, is available for this time
 period:

 a)     hydrogen peroxide and chlorine additions; air stripper;
 b)     pulsating clarifier; sand filters;
 c)     sacrificial GAG adsorber (GAC1); and
 d)     polishing GAG adsorbers (GAC2).
 period.
        Table 5 presents the performance evaluation for the groundwater treatment plant for this
                                            1175

-------
       10

       8
           as?
       o
          il
           § S
           O O
           8 ii
          a A*
          Cj C ^^

          iJl
as?


§sg
                O
                r-t

                0
 5
 0

 0
10   on
o   o
0   ^
      I
03
                     I
     S
     o
     
-------
     4.50 -
     4.00-
     3.50-
z    3.00-
-  5 2.50-1
_J  0>
<  e 2.00-
 o
      1.50-
      1.00-
      0.50-
      0.00
                                 TOTAL MANGANESE
               TOTAL(ROM
                      RAW
                                              O.01 mg/L
                                              0.20 mg/L
                                        POST-GAC2
FIGURES   HISTORICAL DATA 1984-85: INORGANICS
                                                        1-0.10
      UJ
      CO
      UJ
      z
      S ;S
0.05  ^.   d
      ^   E
                                                                o
                                                                I-
                                                         •0.00
   o
   ca
   oc
   <
   o
   o
   Q
   :n
    o
    _i
    31
    O
 3000 •

 2500

-2000
i
, 1500

 1000 -

   500 -

     0
                                TOTAL CHLORINATED HYDROCARBONS
                  TOTAL PHENOL
                                                   1479 og/L
                                                    65 ug/L
                         RAW                POST-GAC2
 FIGURE 4   HISTORICAL DATA 1984-85: ORGANICS
   500
  -450
  -400
  -350 .
  -300
  -250
  -200
  -150
   -100
   -50.
    0
                                                                    O-  en
                                                                    t-
                                                                    O
                                                                    t-
                                      1177

-------
  CO

 8
 PW
 5
R
O
O
a
      c
      en
      a o ^
§ §1
cj n


s§3
9 §§
CJ *•*	.
     Hi
     §§
             S   8
            a
            O
            s
            
-------
 O
 cc
 O
 t-
     3.00 -



     2.50 -



     2.00-



     1.50-



     1.00-



     0.50-
      0.00;
                                  TOTAL MANGANESE
              TOTAL IRON
                                                    O.O7 mg/L
                                                    O.O9 mg/L
                                                           -0.16


                                                           -0.14


                                                           -0.12  •«
                                                                   LU

                                                           -0.10   <   .

                                                                   O  I?

                                                           -0-08   5  ^


                                                           -0.06   _j  —


                                                           -0.04   £

                                                                   f-

                                                           -0.02


                                                            0.00
                  RAW
                              POST-FILT     POST-GAC2
FIGURES    HISTORICAL,DATA 1985-86:  INORGANICS
o
OQ
OC
      1400 -i
      1200 -
O
o
g    1000H
 ui
 cc.
 o
 o
      800 -


      600 -



      400 -


      200 -



         0
             1374 ug/L


              90 ug/L
                            TOTAL CHLORINATED HYDROCARBONS
                TOTAL PHENOL
                                                           r100
                                                            -80
                                                   3 879 ug/L
                                                     4 ug/L
                                                                 O
                                                              60
                                                                 UJ
                                                                    -«
                                                            r-40
                                                            -20
                                                                 O
                                                                 t-
                                                              0
                   RAW         POST-FILT     POST-GAC2


FIGURE 6   HISTORICAL DATA 1985-86: ORGANICS
                                       1179

-------
iO
TAL
NOL
P
           o
           8
          PW
                    s  §s
                     S2
                     £3
s§
                              cn
                              •^
                              O
                                      CD     O3
                                      «0     10
                  t-
                  «-l
                  O
                                    CO


                                    CQ
                                    0
                                            00
                                            55
                         T}<
                         cS
                                     »-<     
                     CO
                     o
                                             n
                                             §
                                             
-------
       Figures 7 and 8 summarize the information presented, and display treatment profiles for iron
and manganese, and total phenol and total chlorinated hydrocarbons monitored, respectively.

       During the last year of operation (1989), a series of 18 samples of influent, post sand filters
and  effluent waters  were  collected  and  analyzed for the  following  additional  parameters:
1,1,2 Trichloroethane (1,1,2 TCA); 1,2 Dichloroethene (1,1 DCE);  1,1 Dichloroethane (1,1 DCA);
Tetrachloroethene (PCE); Chloroform (CClg); Toluene; and Benzene.

       Table 6 presents the performance evaluation for the groundwater  treatment plant for this
period.

       Figure 9 summarizes the information presented, and displays  treatment profiles for the
aromatics and for some of the additional chlorinated hydrocarbons monitored.


2.1.4   Observations

       The following observations can be  made, based on the analyses  of the data collected by
SNC/HMI:

1.     1,2 DCA and phenolic compounds did not meet their respective effluent objectives during the
       first period (1984-1985).

2.     Manganese, 1,2 DCA and phenolic compounds did not meet their respective effluent objectives
       during the second period (1985-1986).

3.     Manganese, 1,2 DCA, phenolic compounds and TCE did not meet their respective effluent
       objectives during the third time period (1987-1989).

4.     Although a combination of chemical oxidants were added to  the train during the second and
       third period, manganese was more effectively removed during the first period.

5.     The addition of chemical oxidants during the second and third period allowed the groundwater
       treatment plant to operate more effectively than  during the first period, by minimizing all
       operational problems associated with the presence of iron and manganese, and with the growth
       of microorganisms,  e.g. Iron bacteria, Pseudomonads.   However,  the addition of chemical
       oxidants had little impact in abating the concentration of the organic contaminants monitored.
       Removal efficiencies for each targeted organic contaminant were  comparable  during all
       three (3) periods.

6.     1,2  DCA and 1,1,2  TCA amounted to approximately 90% (w/w) of the concentration of
       chlorinated organics monitored in the groundwater (1989), with 1,1,2 TCA accounting for 28%.
       The concentration profile for this contaminant was monitored solely in 1989. Also of interest,
       is that a large percentage of the removal observed took place prior to the GAG adsorption step,
       i.e. 89% of the 1,2 DCA removed by the train was eliminated prior to the GAG adsorbers.


2.2    Full Scale Plant Sampling Monitoring Campaigns

       The data presented  in Section 2.1 has confirmed the existence of problems encountered in
 achieving an effective treatment  of Ville Mender's groundwater. Performance data were  provided
 solely on selected segments of the treatment train, and only for selected parameters.  For these
 reasons, a study program was initiated in order to carry  out an in-depth performance evaluation of
 the treatment train.
                                            1181

-------
                                                           1-0.20
                                                        -0.15
                                                            -0.10
                                                            -O.'OS
                                                                   tu
                                                                   O3
                                                                   UJ.
                                                                   2:
                                                                   <

                                                                   a
                                                                   2:
                                                               t~
                                                               o
               RAW   POST-STRIP POST-HIT  POST-GAC1 POST-GAC2




 KCGTJRE7    HISTORICAL DATA 1987-89: INORGANICS
                                                            0.00
                                                                      at

                                                                      £
 CO
 z:
 o
 CO
 cc
 <
 o
 o
 cc
 a

 a:

 a
 ui
cc
o

a:
o
_j
<:

o
  2000


  1800


  1600


  1400


^1200
O)

=,1000 -


   800 -


   600 -


   400 -


   200 -


     0
     1937 ug/L
390Q/L \  TOTAL CHLORINATED HYOROCARBOKS
                                     1069 ug/L
                                                          rso
             TOTAL PHENOL
                                                    4 ug/L
                                             -40
                                             -30
                                                           -20
                                                                _j

                                                                O


                                                                ui
                                                                X
                                                                CL
                                                  f-
                                                  o
                                                  h-
                                                                    DJ
              RAW   POST-STRIP POST-RLT  POST-GAC1  POST-GAC2
                                                            0
5IGTJBE8    HISTORICAL DATA 1987-89: ORGANICS
                                      1182

-------
CO
     ffl
      8
      cu
      3
      a
       a,
          « ^ *
          Ul
          «3 o 2"
          OOW)
          SO a
            It
e 6
Of 5
III


See
S35
                     SCO
                     w
                     O
                     
-------
   SQNnodwoo
                                   ivioi
SNoaavooyaAH
                                      ivioi
FIGURE 9   HISTORICAL DATA 1989: ADDITIONAL ORGANICS
                          1184

-------
       Scans for organic and inorganic contaminants were carried out separately by the MENVIQ and
by "WTC  Good correlation was obtained. The data showed that 61 organic compounds were detected
in the groundwater, amounting to a total concentration of 2 500 ug/L, 97% (w/w) of which were
volatiles  The 3% remainder consisted of non-volatile organic compounds. Of these volatiles, 1,2 DCA
and 11 2 TCA accounted for more than 60% (w/w) of the total VOC count. The inorganics scan showed
iron  and manganese above the drinking water objectives  set by the MENVIQ.  The organic and
inorganic  contaminants, shown in Table 7, were selected as target compounds.   The organic
parameters selected, amounted to 84% (w/w) and 81% (w/w) of the total organic count, determined by
the MENVIQ and WTC laboratories' analyses (GC-MS), respectively.


TABLE 7     LIST OF ORGANIC AND INORGANIC PARAMETERS MONITORED
       Organic Contaminants

       Cis- 1,2 Dichloroethene (1,2 DCE)
            1,2 Dichloroethane (1,2 DCA)
            1,1,1 Trichloroethane (1,1,1 TCA)
            1,1,2 Trichloroethane (1,1,2 TCA)
            Trichloroethene (TCE)
            Tetrachloroethene (PCE)

            Benzene
            Toluene
            meta-Xylene
            ortho-Xylene
            para-Xylene
Inorganic/Parameters

Iron (Total/Dissolved)
Alkalinity
Hardness
Dissolved Oxygen/Temperature
pH/Temperature
Turbidity
Suspended Solids
       Two (2)  monitoring campaigns  were carried out on the plant.  The first campaign was
 conducted four (4) months after the last GAG replacement, in March 1989, in order to gather a
 performance profile of the adsorbers, once in operation for some time. Samples of raw and treated
 groundwater along the treatment train were drawn daily for a period of two (2) weeks. The second
 campaign began immediately following a GAG replacement, in July-August 1989, in order to
 specifically assess the performance of the GAG adsorbers. Samples of raw and treated groundwater
 along the treatment train were drawn regularly for a period of six (6) weeks in order to compile the
 necessary data required to assess the performance of the GAG adsorbers (breakthrough),  and of the
 train in removing the organic contaminants of concern.

       Furthermore,  the concentration of the following parameters were occasionally  monitored
 throughout the treatment train by an outside laboratory, i.e. Montreal-based Lab Elite Lt£e and/or
 WTC  since the field  groundwater laboratory trailer was not equipped to conduct these analytical
 determinations: Manganese; Metal Scan: Ag, Al, As, B, Ba, Ca, Cd, Co, Cr*3, Cr*6, Cu, Hg, K, Mg, Mn,
 Na, Ni, Pb, Sb, Se, Si, V and Zn; and Total Organic Carbon (TOG).

       These data are presented in the body of the text under the sub-heading "Other  Analyses".
 Tabulated data are presented together with the data for the full scale sampling monitoring campaigns.


 2.2.1   First Full Scale Plant Sampling Monitoring Campaign: Data and Results

        Table 8 presents the performance evaluation for the groundwater treatment plant during this
 sampling monitoring  campaign.
                                           1185

-------
            S  g
       u
       g

 J5

 I
 0
 o
 I
 §
 C3
O
CO
s
a
I
CO
TABLE
           aj?


           :sa j

           s|g
          Ill
          38
      I
          s s«
          ^a;
                        23
                        *"  *
                     asssssgg
                        ass
                                         s   i
                    S
                                     S  S
                       ui  o
                                     §   §

                              i
                                           I
                                                 s    K
s    s
QO
                            1186

-------
       Figures 10 and 11 summarize the information presented and display treatment profiles for
iron, the aromatics and for the chlorinated hydrocarbons monitored, respectively.

Other Inorganic Parameters

       The pH of the groundwater remained fairly constant throughout the train. Mean influent and
effluent pH values were 7.0 and 7.2, respectively.

       The hardness of the groundwater remained fairly constant throughout the train.  Mean
influent and effluent hardness values were 576 mg CaCOg/L and 588 mg CaCO/L, respectively.

       The alkalinity of the  groundwater showed a slight decrease throughout the train.  Mean
influent and effluent alkalinity values were 306 mg CaCOg/L and 252 mg CaCO/L, respectively.

       The dissolved oxygen content of the groundwater increased to saturation level following air
stripping, and remained saturated throughout the train. Mean influent and effluent dissolved oxygen
values were 2.7 mg/L and 8.4 mg/L, respectively.

       Suspended solids  values increased immediately after TL£)Z addition and decreased following
clarification and sand filtration.  Mean  influent and effluent values were 0,3 mg/L and 0.9 mg/L,
respectively.

Other Analyses

Manganese:  Manganese  was poorly removed by the train (52%). Removal was essentially observed
in the pulsating clarifier, i.e. 25% of that unit's influent concentration, and irt the GAC2 adsorbers,
i.e. 35% of that unit's influent concentration. Influent and effluent concentrations for manganese were
0.23 mg/L and 0.11 mg/L, respectively, and therefore the effluent objective of 0.05 mg/L was not met.

Metals: The following metals were below the detection limit of the analytical method used: Al, Ag,
Ba, Cd, Cr*3, Cr*6, Hg, Ni, Pb, Se, Sb and Si.  The concentrations of the following metals remained
essentially constant throughout the train: As, Ca, Cu, K, Mg and Na.

Indicator Organic Parameters

        TOG  values  were generally low. Influent and effluent TOG values were 15.4 mg/L and
 11.6 mg/L.


2.2.2   First Full Scale Plant Sampling Monitoring Campaign:  Discussion

        Throughout this  sampling monitoring campaign, both 1,2 DCA and TCE did not meet their
 respective effluent objectives.  The same conclusions were drawn during the historical performance
 data analyses.

        Aside from benzene which consistently showed signs of desorption from the polishing GAC2
 adsorbers, all ring compounds and double-bond compounds monitored were still adsorbing effectively.
 In addition,  it should be noted that benzene was the only compound desorbing from the GAC2
 adsorbers. Furthermore, it also desorbed from the sludge bed of the pulsating clarifier. Analysis of
 the historical data indicated that benzene would still be adsorbing on the GAG when the chlorinated
 hydrocarbons would start desorbing (Section 2.1.3). A possible explanation for this observation might
 reside in a higher concentration of benzene in the raw groundwater prior to the start of the sampling
 monitoring campaign. This  would explain the  desorption observed for this contaminant from the
                                            1187

-------













o
o



I



^f.
OJ
E



I






| n

-------
                         (I/fin)
                  SO1JLVWOUV 1V1OJL
             o    o    o    o    o
             O    U)    O    00    O    O
             CO    CM    CM    T-    T-    U)
             i      i     i	I	I	I
o

                                               OL
                                               S
                                               CO
                                               g
                                               05
                                               O
co A

-------
        pn -V^ g    i  *   W dSy?' followinS ** start of ^e campaign, and the subsequent shift to
            ^?3!        °V relatively «*«»ted GAG bed,  resulting in  the observed steady
  nn,M ; ^  T^ T^^ sensitivily to such s^ngs in incoming benzene concentration
  only adds to the reality that the GAC2 adsorbers are evidently close to exhaustion,,
  hvrfrnp                Temoval efficiency of each process unit with  respect to the chlorinated
  hydrocarbons and aromatic compounds monitored, the following is noted:

                    ,wei$ted Ration rates for the chlorinated hydrocarbons and aromatics were 7%
                    1^ One dwdd note that the oxidation of all unsaturated molecules was higher
  than for saturated ones, e.g TOE (22%), PCE (17%). Hence, in addition to oxidizing iron, HA
  addition contributed to a partial removal of the organic contaminants monitored.


  SteffilS" ^HnT^t^ 'f ^ f°?°^n? C]* addition' f°r the ^drocarbons studied was nil.
  S*i S2i w  ™- ??   * ? 10ad °f chlorinated organics to be air stripped, by a weighted
  ™™SJ •   iV™'S haloSfatlon was more pronounced for some of the chlorinated organics
  monitored, i.e.  18% increase for 1,1,2 TCA,  than for others, i.e. no increase for TCE. The sole
  beneficial impact, therefore, of C12 addition is to control biological growth in the air stripper.

              The weighted _ stripping rate for the hydrocarbons studied was 21%. Of particular

                                          ^ DCAand w TCA> which
 Pulsating Clarifier/Sand Filters/Storage  ReservmV-  Some removal, i.e.  3%, of the hydrocarbons
                                               removal of *•
 SS ^nf'n ' S? n  adso* a^y Conger onto the media. Aromatics still adsorbed well, however, with
 the exception of benzene which showed signs of desorption.                           ««=vci, wim


 each  mPSJ^ rf°S ^ fi-St ^.P11.11? monitoring campaign was to assess the: performance of
 S,?Sk P«         !       ?mce this information was lacking from the historical data collected
 during the five years of operation of the plant. The behaviour of a series of inorganic and organic

 SSSS?KEK ?TS<5'      ^ folf d to C0rrelate V6iy wel1 ^^ the "overa11 P^ture" presented
 dunng the histoncal performance analyses of the plant.
2<2'3   Second Full Scale Plant Sampling Monitoring Campaign: Data. Results and Discussion


rAnlMiiArSTpf??11 s,cale,samPhng monitoring campaign commenced immediately following the
SfiffEST of flie adsorbers with fresh GAG (July 1989). The focus of that campaign was put on

rtson S°ong2 U                                                                      "
on GAG adsorption
                                                                   GAG adsorrs. Fo  tha
                                               the presentation of the analytical data collected
seement
segment
              train
                                                  monitored, and on the organics from the first
                                   sand filtration) are presented together below.
                                         (91%) f°llowing H^ addition-  CWorine addition did
rot   rp -           u        remaiPin& iron COX The shut-down of the plant, required to
replace the GAG in the adsorbers, allowed the plant's operators to clean all unit processes of the
                                         1190

-------
treatment train, including the air stripper packing. Throughout the second monitoring campaign, an
extensive re-deposition of iron hydroxides took place on the stripper's packing: twenty-three percent
(23%) of the incoming iron precipitated in the stripper.  Iron was mainly removed (97%) by the
pulsating clarifier, with the remaining 3% removed in the sand filters. Overall, 100% of the incoming
iron was removed by the train. The mean influent and effluent concentrations were 8.64 mg/L and
0.01 mg/L, respectively, and therefore the 0.30 mg/L effluent objective was met (Figure 15).

Other Parameters: Similar results were observed during both sampling monitoring campaigns for the
following parameters: pH, hardness, alkalinity, dissolved oxygen content, suspended solids, turbidity,
manganese and all other metals monitored, COD and TOG.

Organic Contaminants:  Generally, the experimental findings for the overall removal of organic
contaminants from the first process element (H,jO2 addition) to the seventh unit process (sand
filtration) matched that observed during the first full scale plant sampling monitoring campaign and
during the historical data analyses.

       Table 9 presents the performance evaluation for the ground water treatment plant during this
sampling monitoring campaign.                                      ,

       Figures 12 and 13  summarize the information presented, and display treatment profiles for
iron, the aromatics and chlorinated hydrocarbons monitored, respectively.


2.2.4   Second Full Scale Plant Sampling Monitoring Campaign:  GAG Adsorption  Study

       The removal performance of the organic contaminants monitored by the GAG adsorbers was
studied for a period of forty (40) days.

       Breakthrough of a contaminant from a GAG adsorber was defined as the detectable appearance
of that contaminant in the unit's effluent stream. The information collected has been summarized in
Table 10.


2.2.5   Full Scale Plant Performance Assessment:  Conclusions

       Based on the information presented, the following is concluded:

a)     The process elements of the treatment train, responsible for organics removal, do not perform
        as well as originally anticipated.
b)     The capacity of the GAG adsorbers to remove the ground water contaminants of concern to
       their respective effluent objectives, is exhausted only three  (3) days following fresh GAG
        replenishment, for 1,2 DCA.
c)     The process elements  of the  treatment train, responsible for inorganics removal, perform
        satisfactorily for iron but not for manganese.
d)      Given the presence and the significant concentrations of the contaminants of concern still
        observed in the groundwater, comparable to that observed prior to the start of the remediation
        exercise, it can be  concluded that the remediation of the aquifer is not complete after five (5)
        years of Pump-and-Treat.
                                             1191

-------
  I
 o
 o
 1
 o
 a<


 o
 co

 3
CO
o
S
I
             2,2?
              S g?
             sal


              Sg£
|aE|




l<2?


2S?
             ss;
        w
        n
       I
              |g
                               E   S   S

                           *•>   Oi   to   C«|   o
                           T   &   eo   c3   S
                              0    C-    C4
                          CO   03
                      en   C3   
-------
                                        en

                                        e


                                        o

                                        "o
_J
CD
e


-* n

-------
              soiivwoyv IVIOJL
         o
         o
         CO
         L_
       o
       to
       CM
     o
     o
     CM
       o
       10
         o
         o
                 o
                 to
                                o
         I-
        o
        o
        o
    o
    o
    CD
T~
 o
 o
 CO
o
o
o
o
CO
~i—r~
 o  o
 o  o
 IO  ^f
-i—r~
o  o
o  o
CO  CM
                                         I
o
o
SNoaavooaaAH
                                  ivioi
FIGUEJE 13 SECOND MONITORING CAMPAIGN: ORGANICS
                            1194

-------
CO
  a>

    -4J
                 Ol  03

                 0  0

                 P*  P-i
                                   -StJ
                                   13 10
                                   Is
                                     1
                                   2
                                    1
                                               g
                                               t-
                                          o
                                          •s
                                          I
                                          o

                                          I
                                               i
                                    -s
                        *  0 j-J

                    li lit
                     3"
                     o
«
O
Q,


II
pel
B JJ
OPQ

                                   o «
uen
E
                                     1195

-------
  3,0     BENCH AND PILOT SCALE TREATABILITY STUDIES

         A continuous flow pilot plant designed to integrate various combinations of physical chemical
  unit operations was installed in the vicinity of the full scale treatment facility. The entire pilot plant
  was fitted into two trailers. A third trailer was used to conduct on-site analytical determinations and
  preliminary bench scale treatability studies.

         Figure 14 shows a schematic of the pilot plant. The pilot plant included the following unit
  operations:

         a diffused aeration vessel - for iron (II) and possibly manganese (II) oxygenation

         three (3) chemical oxidation reservoirs designed for use with hydrogen peroxide, sodium
         hypochlorite and chlorine dioxide - for iron (II) and manganese  (II) oxidation and also, some
         oxidation of organics.

         a rapid mix chamber and a flocculation tank - to prepare groundwater for clarification/
         sedimentation.

         a sedimentation tank - to remove suspended solids.

         a granular media filter - to remove suspended solids.

        an air stripping column - to remove volatile organic compounds  (VOCs).

        three (3) activated carbon columns  - to remove organics (volatile and others),


 S.I    Pilot Scale Treatability Studies: Summarized Data and Conclusions

        This work has illustrated the need, through performance data analyses of the full scale
 groundwater treatment plant and  experimental results from the bench and pilot scale treatment
 studies, to remove inorganics prone to precipitation at the earliest possible stage in the physical
 chemical treatment of groundwater contaminated by toxic organics.  This allows for the proper
 operation of subsequent unit processes, responsible for the removal of the organics of concern from the
 groundwater, by minimizing the potential for physical fouling and the ensuing biological fouling.

        One of the objectives of the experimental project was the development and demonstration of
 an effective iron removal process train, as  a pre-treatment prior to the removal of the organics of
 concern from the groundwater. In order to fulfil this objective, several  iron removal process trains
 were evaluated at pilot scale for their ability to remove iron effectively (Table 11).  In addition,
 attempts were made to minimize the number of unit processes involved, eliminate the requirement
 for coagulants), flocculant(s), chemical oxidant(s), aqueous pH conditioners), and also minimize sludge
 production.

        The study resulted in the selection and  subsequent optimization of a simple iron removal
 process train consisting of a fine pore diffused aeration vessel followed by rapid direct contact sand
 filtration. The hydraulic retention time of the groundwater in the aeration vessel was 22 min  and
 the air to water flow rates ratio was 5.5:1.  The oxygenation of iron (II) was shown to be virtually
 complete.^ Direct contact sand filtration, conducted in the subsequent unit process, was shown effective
 in removing the  majority of the iron, thereby consistently meeting the 0.30 mg/L  iron effluent
 objective. A sand bed depth of 110 cm, effective  sand diameter of 0.46 mm, and a filtration rate of
5.3 x 10 m/s resulted in a mass storage capacity of 2.8 kg iron/m3 of sand, expressed as Fe
                                            1196

-------
               o   X
               s
                                       JO
FIGURE 14  PILOT PLANT SCHEMATIC FLOW DIAGRAM
                                    1197

-------
 TABLE 11    IRON REMOVAL TRAINS (VILLE MERCD3R, QUEBEC)

 1.   Chemical Oxidation - Coagulation - Flocculation.- Sedimentation - Sand Filtration

 2.   Chemical Oxidation - Direct Contact Sand Filtration

 3.   Diffused Aeration - Coagulation - Flocculation - Sedimentation - Sand Filtration

 4.   Diffused Aeration - Chemical Oxidation - Direct Contact Sand Filtration

 5.   Diffused Aeration - On-Line Coagulant Addition - Direct Contact Sand Filtration

 6.   Diffused Aeration-- On-Line Flocculant Addition r Direct Sand Filtration

 7.   Diffused Aeration - Direct Contact ,Sand Filtration


        The organics of concern were removed from the groundwater by air stripping, due to their
 volatile nature. An air stripper was demonstrated .effective in removing 1,2 DCA and all other VOCs
 from the groundwater for an air to water flow rates ratio of 240:1.

        Figures 15 and 16 illustrate the performance, of the diffused aeration-direct contact sand
 filtration-air stripping process train for iron and organics removal, respectively.  This performance
 evaluation was conducted during a three (3) month period of continuous operation.

        Performance data analyses of the full scale plant, and performance data collected during the
 pilot scale treatment study have strongly indicated the poor removal performance offered by liquid-
 phase GAG adsorption for the volatile organics found in the groundwater. This was observed for high
 VOC concentrations (full scale plant) and for low VOC concentrations, i.e. in the evaluation of the
 diffused aeration-sand filtration-air stripping-GAC adsorption pilot scale process train. Hence, due
 to the poor removal performance' offered by GAG adsorption, it was not  recommended in order to
 remove the remaining VOCs from the groundwater.

        A biological fouling control technique, developed by F.J. Dart (Ontario Ministry  of  the
 Environment), incorporating zinc granules in the sand bed, to abate and control Pseiidomonads was
 shown effective.
4.0    HYDROGEOLOGICAL STUDY: DATA AND SUMMARIZED CONCLUSIONS5

       Boreholes and monitoring instruments  were installed to measure the  distribution  of
contaminants in the area located between the former lagoon (the "source") and the purge wells (the
"sink"). Monitoring procedures also included a program to sample groundwater at, fifty points  at
different depths and location with respect to the source and the sink. Measuring points are located
in unconsolidated sediments (sandy till formation) and bedrock. The depth of investigation reaches
45 m which is below the depth of the purge wells with intake -screens partly in loose sediments and
the fractured bedrock.

       Measurements of aquifer hydraulic properties, monitoring water levels and use of numerical
models were the tools used to assess  the performance of the hydraulic trap created by the high
performance purge wells.   :   .   •  ,    .  .     ..     ! '. ?:.  ; •'   :;,  'f«   J;;j.   -';  •;
                                          1198

-------
                     NOIJLVH1N33NOO NOd! 1VJL01

                o  o o
                ooooooooooooo
                      o
                         oooooooooo
1 f If 1



j

1

J
f f 1 1 1 1

-J -4 <0
<» ^ £
O
co -
CM <
H co Q
"^ _
^  *~~
" ^ oc
tu
_ CM CL
^~ o
^ 0 CO
o
                                              CO
                                              O
                                              CL.
                                                   -  co
                                                     CO
                                                     CM
O
O
                   i   i   i   i   i  i   i   i   i  i   i
                   o  o o o o  o  o o o o  o  o
                   OO O OO  QO O OO  Q  O
                                                     o
                    .NOIIVaiNHONOO NOHI 1VI01
FIGURE 15  IRON REMOVAL BY A-F TRAIN: PERFORMANCE EVALUATION 2
                              1199

-------
                          
-------
       Monitoring data show that a plume of highly contaminated ground-water is still present
between the lagoon and the Jrtlrge Wells both in unconsolidated sediments arid bedrock fractures. "Oil"
samples were taken in the former lagoon and in several observation wells tapping the rock fractures.
The characteristics of this bil residue are unusual.  The organic fluids have a density greater than
water and reach the bottom of the sand aquifer to infiltrate the fracture network. Their composition
shows the presence of aliphatic hydrocarbons, a proportion between 30 and 35% of polyaromatic
hydrocarbons (PAH), and the presence of chlorinated hydrocarbons, solvents and pesticides.

       The existence of these dense non-aqueous phase liquids (DNAPL) far from the source area is
unexpected and raises several problems.  Many constituents  of the DNAPL are slightly soluble in
water so that the movement of these fluids in the pore space of the unconsolidated sediments and in
the fracture  space of the rocks increases  considerably  their area of contact with groundwater.
Solubility equilibria between the organics and groundwater will persist for a very long period of time
(hundreds of years or more) if the DNAPL are not removed. The purge wells should continue to pump
highly contaminated groundwater for a comparable period of time. The problem has evolved from a
concentrated fixed source area to a moving and dispersing source with a large contact area with
groundwater.

       Poor vertical connectivity between the surface and bedrock aquifers has been documented.
High upward vertical gradient from the bedrock to the sand exists but flow is small because of the
presence of a low hydraulic conductivity layer at the sand/rock interface.  During an interval when
pumping was stopped to permit maintenance in the treatment plant, recovery of the water levels in
the bedrock was extremely slow. Concentrations of volatile organics in a few samples coming from the
bedrock aquifer are from 20 to 50 times higher than the water actually entering the purge wells. More
data is needed before concluding that the bedrock aquifer is less affected by the restoration scheme.

       Presence of DNAPL in the deepest bedrock piezometer may indicate that the hydraulic trap
is of limited efficiency in controlling the movement of the organic phase.  The higher density of this
fluid and the poor vertical connectivity between the two aquifers may create a situation whereby some
of the DNAPL can escape the trap by the overflow.

       Monitoring data from observation wells located outside the study area show that the hydraulic
gradients as far as 1.5 km downstream of the purge wells are oriented toward the sink. The hydraulic
trap is  therefore  considered effective to  control the  spreading of the contaminant plume in the
dissolved phase.

       Conclusions at this stage are that:

1.     the hydraulic trap must be maintained to prevent further migration and spreading of the
       plume of dissolved organics, and

2.     that any long-term management of the plume must consider the removal of the DNAPL by a
       method more effective than pumping the contaminated groundwater.


5.0    CONCLUSIONS

       Based on the analyses presented, the following is concluded:

       The unit operations of the groundwater treatment plant do not perform as well as originally
       anticipated.  Hence,  the treatment train of  the Ville Mercier plant will require further
       modifications in order for the treated effluent to meet the objectives set by the MENVIQ.

       Remediation of the aquifer is not complete after five (5) years of pump-and-treat.
                                             1201

-------
 REFERENCES
 1.



 2.





 3.



4.




5.
 Mattel, Richard. Document devaluation et d'Analyse de la Probl^matique de ITTsine de Ville
 Mercier. Document Interne. MENVIQ.                                          '•    •


 Simard, Georges and Jean-Paul Lanctot. Decontamination of Ville Mercier Aquifer for Toxic
 Orgamcs   First International Meeting of the NATO/CCMS Pilot Study Demonstration of
 Remedial Action Technologies for Contaminated Land and Groundwater, Washington D C
 November, 1987.                                                         6  '     '


 Lanct6t, Jean-Paul.  Usine de Traitement des Eaux Souterraines Contaiminees de Mercier.
 Sciences et Techniques de 1'Eau, Vol. 18, No. 2, Mai 1985.


 Poulin, Martin, Georges Simard and Marcel Sylvestre. Pollution des Eaux Souterraines par
 Sr8 • -S     Oreani(lues & Mercier, Quebec. Science et Techniques de 1'Eau, Vol.  18, No. 8
 JM.81  "
Aquifer Decontaniination for Toric Organics:  The Case Study of Ville Mercier, Quebec
                                                                   ai^.  Montr^
                                         1202

-------
Vieille-Montagne France S.A.                                                  30.01.91
Viviez-12110 Aubin

1. Principles of Operation of the Treatment Facility
   1.1    Sources of Water
       1,1.1  Old  Storage Basins for Hydrometallurgical Sludges
             Infiltration water crosses three basins.  The polluted water flows by gravity to the
             treatment plant.                                   .  ,   .,,
             Flowrate      6 to 30 m3/h (during wet weather)
             Composition  Zn = 0.2 to 1.5 g/L
                           Cd = 10to30mg/L

       1.1.2  Water pumped from the aquifer
             Comes from  the surroundings of the old facilities  for the hydrometallurgical
             extraction of zinc from the ore
              Flowrate      4to10m3/h
             Composition  Zn = 0.2 to 0.8 g/L
                           Cd = 30to50mg/L      ;
                           pH = ± 5

       1.1.3   Rinsewater from zinc metal finishing plant
              Flowrate      ± 8 m3/h
              Composition pH = 2 to 4

    1.2    Process
       1.2.1   The polluted waters are mixed and treated continuously in 5 reactors in series of
              100 m3 each.  PH adjustment is accomplished by adding hydrated quicklime at
              150 g/L (suspended solids) to 3  of the reactors.  In each reactor the amount of
              lime is controlled by a  pH feedback loop,  (combination electrodes  and on/off
              pneumatic valves)
                  pH reactor No. 1  =  6.5 to 7
                  pH reactor No. 2  =  8 to 8.5
                  pH reactor No. 4  =  8.8 to 9

       1.2.2   The free  acidity and the dissolved elements such as Zn and Cd form CaSO4,
              hydroxides and sulphates of zinc  and cadmium. This mixture of solids and liquids
              is separated in one gravity clarifier with a volume of 800m3. A flocculant dosage
              of ±15 mg/L is necessary to ensure clarification because of:
              •   the effluent quality required  (the  effluent  to the  river must  meet  the
                  requirements set by DRIR)
              •   the increase in sludge density (density  1  030 to 1 100).  The sludge is
                  dewatered by a filter press (FAURE) in order to minimize the moisture content
                  and then it is shipped to VMF AUBY.

       1.2.3   To  complement the lime addition and overcome the partial solubility of Cd which
              would precipitate as sulphates and hydroxides, a stoichiometric quantity of sodium
              sulphide  NagS is added to precipitate the cadmium as cadmium sulfide.  The
              concentration of Na-jS in solution is 200  g/L.
                                          1203

-------
       1.2.4  Operation
                 The facility has a design maximum fiowrate of 140m3/h.
                 Without exception this permits an operation through an 8 hour shift each day,
                 given the volumes to be treated and the storage capacity provided by the 800
                 m3 clarifiers.

                 The suspended solids produced are filtered and shipped to another site of the
                 company for recovery of zinc, the clean residue is then disposed of.

                 The purpose of this shipment is more to accelerate the clean up of old sites
                 of the Viviez plant than to reclaim the solids.

2.0    Principle of Operation of the Facility
                 Recycled material was used for the hydrometallurgical extraction of zinc, which
                 explains  the  volume of tankage  installed for treatment,  clarification and
                 storage.

       2.1     Preparation and distribution of lime
              -  1 silo of powdered quick  lime, ultra fine, 100T
              -  preparation in industrial water at 150 g/L
              -  distribution in a continuous loop

       2.2     Reactors
              -   5 reactors cascading (100 m3 per reactor)
              -   Agitator motor:  50 HP at 750 rpm
              -   Turbine mixers-0=1.20m -12 blades
              -   Construction - mild steel, rubber and bricks

       2.3     Clarifiers
              -  1 clarifier in use, 1 standby
              -  Volume 800 m3
              -  Construction:  mild steel, rubber and bricks
              -   Equipped with a rake with 4 arms; stainless steel
             -  Motor:  4KWat11 rph

      2.4     Filter Press
             -  TITAN model 219 with recessed plates of polypropylene 1 200 x 1 200 mm for
                127 chambers 30 mm
             -  Volume: 3 800  L
             -  Pressure: 15 bars
             -  Filter surface:  296 m2
             -  Closing, opening and cake discharge mechanically
             -  Cake characteristics
                •   Density 20 to 30% with:  Zn:    25 to 35%
                                              Cd:    0.2 to 0.5%
                                              Ca:    1.5 to 6%
                                         1204

-------
                 •   For 1  shift of 7 working hours:
                    •  dry solids 6T
                    •  cycle time  1 hour 40 minutes or 4 to 5 cycles per shift

       2.5    Storage capacity for the waters to be treated
             800 x 3 = 2 400 m3

3.0    Effluent Criteria set by a Regional decree in Aveyron as of February 1989
       -   PH: between 5 and 9
       -   Temperature: less than 30°C
       -   Suspended Solids:  less than 30 mg/L
       -   Phosphorus:  less than 10 mg/L
    ,   -   Total hydrocarbons: less than 5 mg/L
       -   Total metals: less then 15 mg/L
          (Zn + Cu + Ni +  Al + Fe + Cr + Cd + Pb + Sn)
       -   Zinc:  less than 5 mg/L

       (a) Specific to  Cadmium
              Monthly effluent concentration: less than 0.2 mg/L
              Daily effluent criteria: 0.4 mg/L

       (b)Self-Monitoring
              •   Continuous  effluent monitoring  for flowrate and  pH,  prior to  discharge.
                 Flowrate values are recorded in a notebook.
              •   Control monthly by a representative daily sample for suspended solids, COD
                 pH  and Zn.

                 Daily sample  for  Cd.   Results communicated monthly to the  inspection
                 authority.

       (c) Daily average effluent results (discharge to a river)
                 pH  =  8.5 to 9
                 Zn     =   0.5 to 3 mg/L
                 Cd  =  0.03 to 0.18 mg/L


 4.0    Principle of Operation of the Facility
       4.1    Staffing Requirements
              Treatment, clarification:  filtration for one shift, each weekday:
                        -    shift from 5:00 a.m. to 1:00 p.m. for 7 hours of effective work
                 From 1:00 p.m. to  5:00 a.m. the next day, the water to be treated  is stored in
                 800 m3 reservoirs
                 •   Shift 5:00  a.m. to 1:00 p.m.: 2 workers
                 •   Shift 1:00  p.m. to 9:00 p.m. and 9:00 p.m. to 5:00 a.m.: no staff
                 •   1 Foreman: (daytime hours)

       4.2    Cost breakdown
                 •   Material costs
                                         1205

-------
                     (lime, filter cloth, flocculant)
                     Handling
                     Energy
                     Analyses
                     Clean up, maintenance
                     Personnel Salaries
Original version in French: J.M. Granier, M. Darcy
Translation: J. Schmidt
                                                                 TOTAL
 2460  F/d
   600  F/d
 2190  F/d
   260  F/d
 1 890  F/d
 2830  F/d
10230  F/d
                                        1206

-------
                                                Appendix 6-B

        Ground Water Pump and Treat Technology Case Studies
  Evaluation of Photo-oxidation Technology (Ultrox International),
Lorentz Barrel and Drum Site, San Jose, California, United States
                           1207

-------
 vvEPA
                             United States
                             Environmental Protection
                             Agency
                                            EPA/540/M5-89/012
                                            November 1989
                            SUPERFUND INNOVATIVE
                            TECHNOLOGY EVALUATION
                            Demonstration  Bulletin

                             Ultraviolet Radiation and Oxidation
                                           Ultrox International
 TECHNOLOGY DESCRIPTION: The ultraviolet (UV)
 radiation/oxidation treatment technology developed by
 Ultrox International  uses  a combination of  UV
 radiation, ozone, and  hydrogen  peroxide  to oxidize
 organic compounds  in  water.  Various  operating
 parameters can be adjusted in the Ultrox® system to
 enhance the oxidation of organic contaminants. These
 parameters include hydraulic retention time, oxidant
 dose, UV radiation intensity, and influent pH level.
 A schematic of the Ultrox system is shown in Figure
 1. The treatment system is delivered  on  four skid-
 mounted modules, and includes  the following  major
 components:

 •  UV radiation/oxidation  reactor module
 •  Ozone generator module
 •  Hydrogen peroxide feed system
 •  Catalytic ozone decomposer  (Decompozon) unit
    for treating  reactor off-gas
                        Treated Off Gas
                 Catalytic
             Ozone Decomposer"'1
                     Reactor Off Gas
 Water Chill
                           Treated
                           Effluent
                          Ultrox
                        UV/Oxidation
                         Reactor

                    "Hydrogen Peroxide
 Cooling Water
   Return
sAir Compressor
Figure 1. Isometric view of Ultrox System.
                                      The commercial-size  reactor used  for the  SITE
                                      Demonstration is 3 feet long by 1.5 feet wide by 5.5
                                      feet high. The reactor is divided by five vertical baffles'
                                      into six chambers. Each chamber contains four UV
                                      lamps  as well  as  a diffuser which uniformly  bubbles
                                      and distributes ozone gas into the groundwater being,
                                      treated.
 WASTE APPLICABILITY: This treatment technology
 is intended to destroy dissolved organic contaminants,
 including  chlorinated  hydrocarbons  and aromatic
 compounds,  that are present  in  wastewater or
 groundwater with low levels of suspended solids, oils,
 and grease.


 DEMONSTRATION  RESULTS: The  SITE Demon-
 stration  was conducted at  a  farmer drum recycling
 facility in San Jose, California, over a 2-week period in
 February and  March  1989.  Approximately  13,000
 gallons  of groundwater contaminated with  volatile
 organic compounds  (VOC) from the site were treated
 in the Ultrox system during 13 test  runs. During the
 first 11  runs, the  5  operating  parameters were
 adjusted to evaluate  the system. The last 2 runs were
 conducted under the same conditions as Run 9 to
 verify the reproducibility of the system's performance.


 To evaluate  the performance  of  each  run,  the
 concentrations of indicator VOCs in the effluent were
 analyzed overnight. Three of the 44 VOCs identified in
 the groundwater at the site were selected as indicator
 VOCs. These indicator VOCs  were  trichloroethylene
 (TCE); 1,1 dichloroethane (1,1-DCA);  and   1,1,1-
trichloroethane  (1,1,1-TCA). TCE  was  selected
because it is a major volatile contaminant at the site,
and the latter two VOCs were selected because they
are relatively difficult  to oxidize.
                                             1208

-------
Key  findings from  the Ultrox  demonstration  are
summarized as follows:

•  The groundwater treated by the Ultrox system met
   the  applicable  National Pollutant Discharge
   Elimination System  (NPDES) standards  at the 95
   percent confidence  level.  Success was obtained
   by using a hydraulic retention time of 40 minutes;
   ozone dose of 110 mg/L; hydrogen peroxide dose
   of  13 mg/L; all  24 UV  lamps operating;  and
   influent pH at 7.2 (unadjusted).
•  There were no volatile  organics detected  in the
   exhaust from the Decompozon unit.
•  The  Decompozon  unit  destroyed  ozone in the
   reactor off-gas to levels less than 0.1  ppm (OSHA
   Standards). The  ozone  destruction  efficiencies
   were observed to be greater than 99.99  percent.
•  The Ultrox  system  achieved  removal efficiencies
   as high as 90 percent for the total VOCs present
   in  the groundwater at  the  site. The removal
   efficiencies for TCE were greater than 99 percent.
    However, the maximum removal efficiencies for
    1,1-DCA  and 1,1,1-TCA were  about 65 and 85
    percent,  respectively (Table 1).

Table 1. Performance Data During Reproducible Runs
Mean Influent
(P9/L)
Run Number: 9
TCE
1,1 -DCA
1,1,1-TCA
Total VOCs
Run Number: 12
TCE
1,1 -DCA
1,1,1-TCA
Total VOCs
Run Number: 13
TCE
1,1 -DCA
1,1,1-TCA
Total VOCs

65
11
4.3
170

52
11
3.3
150

49
10
3.2
120
Mean Effluent
(W/L)

1.2
5.3
0.75
16

0.55
3.8
0.43
12

0.63
4.2
0.49
20
Percent
Removal

98
54
83
91

99
65
87
92

99
60
85
83
                     •  Within the treatment system, the removals of 1,1-
                        DCA and 1,1,1-TCA appear  to be due to both
                        chemical  oxidation and stripping. Specifically,
                        stripping accounted for  12 to 75  percent  of the
                        total removals for 1,1,1-TCA, vinyl chloride, and
                        other VOCs.  .
                     •  No semivolatiles,  PCBs, or pesticides were found
                        in the groundwater at the site. Among the VOCs,
                        the contaminant present at  the  highest
                        concentration  range  (48 to 85 ug/L) was  TCE.
                        The groundwater also had  contaminants such as
                         1,1-DCA  and  1,1,1-TCA  in  the  concentration
                        ranges of  10 to 13 ug/L and  3  to  5  pg/L,
                        respectively.
                     •  The organics  analyzed  by Gas Chromatography
                         (GC) methods represent less than 2 percent  of
                        the total  organic  carbon  (TOC)  present  in the
                        water.  Very low  TOC  removal occurred, which
                         implies that partial oxidation of organics (and not
                         complete conversion to  carbon dioxide and water)
                         took place in the system.

                     A Technology Evaluation Report and  an Application
                     Analysis  Report  describing  the  complete
                     demonstration will be available in the Spring of 1990.
                                                    FOR FURTHER INFORMATION:
                                                    EPA Project Manager:
                                                    Norma M. Lewis
                                                    U.S. EPA
                                                    Office of Research and Development
                                                    Risk Reduction Engineering Laboratory
                                                    26 West Martin Luther King Drive
                                                    Cincinnati, OH 45268
                                                    (513) 569-7665 (FTS: 684-7665)
  United States
  Environmental Protection
  Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PAID
        EPA
   PERMIT No. G-35
  Official Business
  Penalty for Private Use $300

  EPA/600/M-89/014
                                                  1209

-------

-------
                                              Appendix 6-C




       Ground Water Pump and Treat Technology Case Studies
Zinc Smelting Wastes and the Lot River, Viviez, Averyon, France
                         1211

-------
                          VIEILLE MQNTAGNE
       The origin of zinc production in VIVIEZ goes back to the years  1870.
 Until 1930, the process was thermic, and all the wastes were thrown near
 •to the plant,  on the hill  slope. Such is the origin of this waste  heap.
 Afterwards, the Vieille Montagne  company shot hydrometallurgic refuse
 coming from the electrometallurgic process.

       The total  amount of wastes stored  in the heap is evaluated to
 700000 tons.

       Natural  lixiviation by rainfall takes metallic  ions  to the ground
 water located  at  the hill-foot  then  to surface  water  which  are
 contaminated.

      The amount of cadmium thrown  in the environment amounted  to a
 mean 36 kg/day up to 1987.

      When  this permanent pollution was proved,studies were carried out
 in order to  understand the circulation  phenomena of metallic  ions to the
 superficial  hydraulic system and, then decide  which means had to be put
 into operation so as to eliminate this pollution.

      1) Hudrooeoloaical studij •

      The aim of this study was to evaluate:

      - the hydrodynamic behaviour of the alluvial aquifer,
      - the hydraulic relations aquifer/heap,
      - the surface water regime.

      At the same time,the qualitative aspect,  based on water analysis
had to be taken into account.                                         '

      Analysis of boreholes  and water have led to collect a certain amount
of information which can be summed up as follows:

     - initial geological scheme showing various overlaid structures,

     - drawing  of a piezometric map on which a mathematical model was
       set up in order to localize flow repartition,

                               1212

-------
     - detailed analysis of  data  taken  from all the control  equipment,
       including cadmium content in surface and ground waters leading
       to a global balance.

     In this general context, the first actions, brought up in  1988, were
taken so  as to cover ponds containing muds and to relocate the smallest
storages.

     The impact  on  natural  environment .was in  a very  short  time
effective since the day cadmium content went down from a 36 kg/day
(1987) to a 2 kg/day early in 1989.
      2) P2 point

      P2 is a special point highly polluted  in cadmium (1250 mg/1) v/hich
was due to a temporary storage  of ore. P2 is an important element of
cadmium re-emission (about a third of the pollution).

      Tests carried out in July 1988 have led to a better knowledge of the
"P2 point"  which  had  been  located  before  and  to  locate cadmium
concentrations around this point.

      it was possible to locate a zone of cadmium dissolved content higher
than 300 mg/1 around boreholes P2, P3, S7  and S5. Maximal contents are
surrounded by concentric circles: a 200-300 mg/1 circle and another one
at 100-200 mg/1, both  10-20 m wide are limited by  boreholes S6, SI, S2
and S3 ; further around, cadmium contents  are similar to that measured
elsewhere, particularly in the waste