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 heap in the THR zone.
Hydraulic values have been fairly evaluated : mean transmissivity is
T = 5,0 . 10 "4 m2/s and storage coefficient S = 3,0 . 10~2
These values have been used to test in a mathematical model three
hypotheses of decontamination by pumping/injection : it has been decided
to decontaminate putting into operation two pumping boreholes (P2 and
S5) at 8 rn3/h and four injecting boreholes (S2, S4, S6 and S7) at 4 m3/h.
Tests carried out during the spring season in 1989 have led to the
idea that this solution was feasible and efficient. In addition, leaching has
been carried out by injecting in three points and pumping in two others.
Nowadays, cadmium content as controlled in two boreholes under
1213
-------
permanent pumping at 2 m3/h does not exceed 50 mg/1, same value as in
the rest of the plant. '
3) Genera] balance at "Laboratoru" point
A permanent control located on the surface network (Riou Mort) on a
point considered as perfectly representative ("Laboratory" point) is used
in order to follow the impact of the installation.
Although hydrological conditions have been exceptionally dry for
two years, il is confirmed that cadmium emission has been drawn down
and that, even when flow rates are high due to important rainfall, the
maximum cadmium content approved by the authorities (500 q/h) is
respected.
Thus, in general, actions carried out on the site have led to a
cadmium lowering of ?0%. •
But this control leads the Vieille Montagne company to spend each
year about 5 MF (1 M$) to maintain in order the station in which pumped
waters are decontaminated. •
The Company staff is of course eager to reduce, or even to cut off
this operating costs, leading to a special study which has been launched in
THR zone (Traitement humide des residus = waste processing bu humid
way). .
Dewatering of the THR znns
Based on 10 boreholes, cadmium content analysis of water and of
driling cuttings, an interpreted log has been drawn up, which confirms the
interpretations which had been previously done: ;•
- phreatic ground water does not flow in a normal way due
to building fondations
- there exists a continuous and homogeneous clay layer
- ground water in porous medium flows under the buildings and
goes to the P2 zone.
1214
-------
Generally speaking, from a zone in borehole F1 (1 mg/1 Cd), there is a
progressive enrichment of waters up to a mean value of 50 mg/1.
Two options can be taken to solve this problem:
- a "methodologic" way aimed at understand all the mecanisms
involved,specially the role played by the clay layer (resorption)
- a pragmatig "field" way aimed at dewatering the downstream part
of the heap (THR) where the clay layer - which is supposed to
enrich the upper part of the aquifer in Cd - is located.
This second option has been chosen and includes three steps :
.1) construction, of a derivation system of surface waters coming
from the upstream part of an old river bed covered by the heap
2) evaluation phase : it is supposed that waters infiltrated in the
heap do not dissolv Cadmium, due to the high permeability of the
"thermic" part of the wastes and to the protective impervious
layer of the basins. If this hypothesis is confirmed, waters can
directly be evacuated in the environment without flowing through
the THR zone
Controls will be achieved through water analysis on Ft on a
complete hydrological cycle.
3) Water recovery : in this case, it has been decided to dig a
collecting system including a trench, an acces gallery and a final
drain.
Calendar : decision to be taken in 1991 (if the hypothesis n°2
appears not to be ascertained,waters will have to be
decontaminated); works to start in 1992.
Estimated cost: 2 MF.
1Z15
-------
PLAN dc, SlTUATlO M
1216
-------
O)
trs
O)
'
1217
-------
u
1218
-------
• •
tli
o
t-
O-
g
O
1219
-------
1220
-------
1221
-------
Vt'ou
ctM
A
*JL«>
cA
1222
-------
V1EILLE-MONTAGNE
Usine do Viviez - 12110 AUOIN
ETUDE UYDnOGEOLOGlQUE ot HYDHOCIIIMIQUE
de la PLAINE AI.LUVIALE du RIOU VIOU
LEGENDS GENERALS
S5
POMPAGE DU 26/07/08 sur S5 : REPARTITION DES CONCENTRATIONS
EN CADMIUM DE L'EAU DE LA NAPPE DANS LE SECTEUR DU P2
--S Concentration en mg/1
-------
IcOHCl-HrRATlON MOYEHNE EN CADMIUM DES EAUX AU POINT "
ANNEGS
89
10
It
12
90
01
02
03
04
05
06
07
08
09
10
11
x(mg/l)
0.17
0.1G
_
0.14
-
0.17
0.14
0.10
0.10
0.11
0.14
Qm3/h
445
683
—
515
—
1 000
1. 120
2 243
720
270
180
74
104
-
72
114
123
204
24
28
30
REHARQUES
(polnte & 2.260 m3/h -> 0.10 mg/h)
(pojnte 6 1 850 m3/h -> 0.08 mg/h)
(polnte a 3 230 m3/h -> 0.10 mg/h)
(polnte a 1 130 m3/h -> 0.15 mg/h)
1224
-------
CN
UJ
-z.
O
<£
H
z:
O
~^>
•**••
LJJ
1
_J
LJJ
c^~
»*•*
"
N
UJ
^
* —
>*--->
*"*^
CM
*— .
si
of
H
D
0
t:
VI
v/l
oC
O
A
^
C!
3
O
ul
\b
0
ci
O
X
111
O
7.
a.
v-\
~
2
2
O
u
ct
0?
,^
•v!
cc
u
0
A
H
..
c^
3:
•
.
;?'
2
10
*^>
2
a;
Vll
O.
D
C
O
J
fill I
^ t & it
•c- -<- -v- -- -r v -r —
I I I I I I I I
U -!, «) -^ u -J- wv^^
1225
-------
&TX*gTgiqwM v^^' //~\^~\
PIS^
AW t'/p;
LEGENDE
t ' • f. If -
.n.£. t-
-------
Cap
1227
-------
•lit11 p;:i'i'!i;;»";
d^r^Ef:
-•: ••;. .:•• ,.;!»••.'• ;. .& •
i-:.!-:-!i!iii'!ilj'-'.: •' -Sf/ i-'-'/i"?
-------
DEFENSES EMGAGEES POUR LE TRAITEHENT DES CRASSIERS
(exprirroSes en Francs fran
-------
-------
Appendix 6-D
Ground Water Pump and Treat Technology Case Studies
Separation Pumping, Skrydstrup, Denmark
1231
-------
The NATO/CCMS Pilot Study on
Demonstration of Remedial Action
Technologies for Contaminated
Land and Groundwater
Fourth International Conference
5-9 November 1990
Angers
France
THE SEPARATION PUMPING TECHNIQUE
by
Bertel Nilsson
Rasmus Jakobsen
Geological Survey of Denmark
8, Thoravej
DK-2400 Copenhagen NV
Denmark
1232
-------
0. Abstract
Demonstration tests of the separation pumping tetinnlque has been
performed at several sites in Denmark. The isejbafiltion pumping
technique relies on simultaneous pumping f rdift tcSfJ and bottom of
a fully penetrating well, through the polluted section of the
aquifer. Separation pumping is developed frolti tHe concept of
scavenge pumping. The method can optimize remedial fiction pumping
by minimizing the volume of polluted tfatae tB be treated.
•Furthermore the inflow distribution arid trSiiMmissivity of
selected sections of a well can be determined. Moreover the
method can be itsed for collection of level-accurate samples.
Finally, changes in the vertical distribution of pollutants in
the aquifer as a function of pumping time can be monitored.
Results from two different sites, a sand and gravel aquifer, and
a limestone aquifer show that separation of polluted groundwater
from slightly polluted or non-polluted groundwater was possible,
and for the sand' and gravel aquifer the amount of water to be
treated was initially reduced to 40%. Tests with separation
pumping as. an investigation tool at a third site, showed that the
influx distribution could be determined with high precision, and
very high vertical concentration gradients could be discerned
with separation pumping used as a level-accurate sampling method.
1233
-------
3. Methodology
3.1. Separation pumping
The water well is simultaneously pumped by two pumps at the top
and bottom, respectively, Fig. 1. A water divide is created in
the borehole between the two pumps. Water from above the water
divide will move towards the top pump and water below the water
divide towards the bottom pump.
By adjusting the pumping rates of the two pumps the water divide
can be established at any position in the borehole according to
the ratio between the pump rates and the hydraulic conductivity
above and below the divide. These position are unknown but can be
found indirectly by a separation injection test (SIT)/ see
section 3.2, by conventional flow-logs or by direct measurement
with a Heat-Pulse-Flow-Meter (HPFM), see section 3.3.
FLOWMETER
FLOWMETER
TOP PUMP
GROUND SURFACE
L
QROUNDWATER LEVEL
•DIVIDE
BOTTOM PUMP
DGW
Fig. 1.: Schematic illustration of separation pumping by
simultaneous pumping from top and bottom of a well.
From Andersen (1990).
1236
-------
The following parameters are measured during separation pumping
for every "steady state" situation (step), the total capacity is
kept constant for all steps.
a) The pumping rates for the top and bottom pumps.
b) The chemical composition of the water from the top and
bottom pumps.
c) The water level in the borehole.
The water samples from the top and bottom pumps are taken when
these parameters are stabilized at each pumping ratio (step). The
vertical distribution of a pollutant within the penetrated part
of the aquifer can be calculated.
The principles of a 9 step separation pumping is shown in Fig. 2.
Nine different combinations of pumping rates (step 1-9) divides
the screened part of the well in 10 inflow intervals.
The flow pattern shown on the sketch corresponds to step 3, where
intervals 11-13 supply the top pump and intervals 14-110 the
bottom pump.
For steady state conditions the total flux of each pollutant
leaving the well ((Q«C)t + (Q'C)b) should be constant.
The influx of the pollutants from the ten intervals: qi'Ci
(j.=l,2t ,10) can be calculated from the following equations:
q?
qs
(Qi «
(Q2 •
(Qs •
((Qs
((Qe
((Q7
(Q7 •
(Qs -
GZ)* - (Qi •
Ca)' - (Q2 '
• C4 - Qs •
• C5 - Q« -
• C6 - Q5 -
• C7 - Q6 •
C7)b - (Qs
C8)b - (Q9
C9)b
(Qs
(Q<
(Qs
(Q6
• C3 -
- c, -
• c5 -
• c6 -
QA. '
Qs '
Qe '
Q7 -
• C«)b)/2
• C5)b)/2
' C6)b)/2
• C7)b)/2
1237
-------
Ground surface
— st 1
-c' L L
1—No (low—I
| ___ 1 ___
. ,
---- ----- -------
j *~—— step 6
Water divide
Step 7
<5.,c« I I 18
c b
(Q-C)
T~~i~r~
I Bottom ra I
1 pumpgj I
19
110
Step 8
Step 9
Fig. 2.: Sketch of a 9 step separation pumping. Q == measured
pumping rate, C = measured concentration, b = bottom,
t = top, c = calculated concentration and q = calcu-
lated inflow from the aquifer.
The q-c-fluxes for the top and the bottom intervals are given
directly from the measurements, the fluxes for interval 12, 13,
18 and 19 are calculated using the top and bottom data respect-
ively while the middle intervals 14-17 are calculated using
average values from the top and bottom pumps.
1238
-------
All Q-values are measured and q-values are easily calculated.
q2 =
q6
q7
q8
q9
(Qi)t
((Q2 -
((Q3 - Q2)c + (Q2
((Q« - Q3)c + (Q3
((Qs - CM* + (OA
((Qs - Q5)c + (Qs
((Qr - Qe)* + (Qe
((Q8 - Q?)* + (Qy
((Q9 - QsT + (Qs
(Qio)b
Q2)b)/2
Q3)b)/2
Q5)b)/2
Q6)b)/2
Q7)b)/2
Q8)b)/2
Q9)b)/2
The principles of calculation and assumptions are described in
Andersen et al. (1989) and Gosk & Bishop (1989).
3.2. Hydraulic conductivity log
3.2.1. Separation pumping test
The object of the separation pumping test (SPT) is to determine
the inflow distribution in a well. The principle is seen in Fig.
2 for a 9 step SPT. The location of the water divides is done
with a Heat-Pulse-Flowmeter, described in section 3.3. The yield
of each section can be calculated as shown in section 3.1.'
3.2.2. Separation injection test
The separation injection test (SIT) is an inverted (SPT). Water
is injected to the top and the bottom of a well and of the
measured flow rates are the recharge rates of the intervals above
and below the convergence zone, the opposite of a water divide.
The location of the convergence zone between top-water and
bottom-water is determined by using an in-hole-detectable tracer
mixed in the top- or bottom injection water. For example
chloride, in a concentration sufficient enough to separate it
from fresh water, can be used. The interface between the salty
1239
-------
and the fresh water is detected by an electric conductivity
probe.
3.3. Level-accurate water sampling
The well is simultaneously pumped from top and bottom,,, Fig. 3.
The position of the water divide is found by a Heat-Pulse-Flow-
Meter (HPFM) . A HPFM-probe is a directional low-velocity
flowmeter. An electric energy pulse is fired through a wire grid
creating a small volume of warm water and its direction and
velocity, corresponding to the flow of the fluid in the borehole,
is determined by the monitoring of two thermistors, one above and
one below the wire grid.
A water sample is collected at the water divide by a third pump,
pumping with a yield of a few percents of the total capacity with
which the well is pumped.
By varying the pumping rates of the two pumps, the water divide
can be placed at any depth so that a chemical profile of the
penetrated sequence of the aquifer, can be made.
WATER SAMPLE
TOP POMP -
SAMPLING PUMP •
I E
i/ FLOWMETER
,— GROUND SURFACE
- HEAT • PULSE. FLOW. METER
• WATER DEVICE
- BOTTOM PUMP
B6W
Fig. 3.: Schematic illustration for level-accurate groundwater
sampling by separation pumping. From Andersen (1990).
1240
-------
3.4. Separation pumping as a remedial action
From the concentration profile determined in section 3.1, it is
known at which yield ratio between the top and bottom pumps one
of the pumps only recover water from the polluted layer and the
other water from the unpolluted layer, Fig. 4. In this way the
volume of polluted water is restricted to a minimum, the polluted
part of the inflow to a well. The efficiency of the method is
limited if the partial transmissivity of the unpolluted layer is
much smaller than the one of the polluted layer.
FLOWMETER FLOWMETER
POLLUTED WATER -*-~sQsSSav X=Q]=-»- UNPOLLUTED WATER
GROUND SURFACE
PUMPING WATER LEVEL
NON-POLLUTED ZONE
GROUNDWATER LEVEL
D6W
Fig. 4.: Schematic illustration of remedial action pumping by
, use of separation pumping -technique. From Andersen
(1990).
1241
-------
4. Results and discussion
The separation pumping technique has been tested at different
sites in Denmark and has been used at a site in United Kingdom.
Experiments from three sites in Denmark are presented.
4.1. Experiments of remedial action pumping in a sandy aquifer
4.1.1. Site description
The first tests have been performed at the chemical waste
disposal site registered under No. 543-01, Skrydstrup, Municipa-
lity of Vojens, County of S0nderjylland. The site is located
close to Skrydstrup Airport, Vojens City and approximately 2,5 km
up-stream one of the water works of the Municipality of Vojens
(Pig. 5).
Frederiksberg
Karfstrup.ifl
Fig. 5.: Locations of test sites in Denmark.
In the period 1963-1974 chemical waste has been dumped in an old
gravel pit near Skrydstrup by a,refrigerator factory. The waste
caused a considerable contamination of the groundwater by
1242
-------
chlorinated solvents, mainly 1,1,1-trichloroethane and organic
phosphorus compounds.
The waste was excavated in 1986, the drums with chlorinated
solvents were sent to destruction, and the contaminated soil was
placed in a special waste disposal site with aerobic/anaerobic
in-situ biodegradation of the chlorinated compounds. The latter
enter into the NATO Pilot Studies. -
The pollution plume has been found up to 1,5 km down-stream from
the dump site. The remedial action pumping is from four wells
screened to a depth of about 25 m.b.s. The pumped, water is
treated in a on-site groundwater cleaning plant by air-stripping,
and active carbon filtration. In November 1988, at the beginning
of the remedial action pumping, wells PB2, PBS, PB9 and PB10 were
used, (Fig. 6). In June 1989 well PBS was replaced by well PB16,
in which the demonstration experiments with separation pumping
were performed.
Chemical waste
disposal site
Pollution plume
Groundwater
Fig. 6.: Skrydstrup chemical waste disposal site and the
pollution plume. Location of the remedial action
pumping wells PB2, PBS, PB9, • PB10, PB16 and the
direction of groundwater flow are shown.
1243
-------
4.1.2. Hydrocreologv
The aquifer material consists of meltwater sand and gravel tinder
water table conditions. The thickness of the aquifer is about 60
m, just below the waste disposal site, decreasing down-stream,
and the water table is approximately 4-10 m.b.s. The thickness of
polluted groundwater was 7-10 m at the beginning of the experi-
ment, determined by a well drilled close to well PB16, section
4.1.4.
4.1.3. Experimental set-up
Well PB16 is 200 mm in inner diameter, 25 meter deep, and
screened from 6-25 m.b.s. The set-up consists of two submersible
pumps (max. capacity 11 mVhr) in the borehole, valves to adjust
the pumping rates of the two pumps, and electromagnetic flow-
meters to measure the actual flow from each pump.
4.1.4. Results and discuss!
on
First a 9 step separation pumping was performed, dividing the
inflow interval in 10 sections each contributing a flow partial
of approximately a 1/10 of the total discharge of 10 mVhr. (Fig.
2). The positions of the water divides have been estimated
indirectly from an impeller flow log, run in well PBS. Fig. 7
shows a calculated -concentration profile of three chlorinated
solvents (TCA, TCE and TeCE) and the electrical conductivity
(EC). It is seen that the concentration of the pollutants is
highest in the upper most intervals 11-12 and the concentration
decreases strongly in interval 13 and the underlying intervals.
Generally 1,1,1-trichloroethane occur in a concentration about 10
times higher than trichloroethylene while tetrachloroethylene
only occur in the upper most intervals 11-12 in a concentration
below 1 ng/1. The electrical conductivity is correlated to the
chlorinated solvents. Furthermore it appears from Fig. 7, that
approx. 30% of the total yield is coming from the inflow interval
7/5 - 17,0 m.b.s., while 40% comes from 17,0 -.20,5 m.b.s. and
the last 30% from a depth of 20,5 - 25 m.b.s.
1244
-------
500
-t—
200
300
O—O TCA, pg /1
•—• TCE,jig/I
A—A TeCE, jig /1
•—a EC, jiS / cm
Interval 1
O Interval 2
Interval 3
Interval 4
Interval 5
Interval 6
25
Bottom of well
Fig. 7.: Calculated concentration profile in well PB16, Skryd-
strup of the chlorinated solvents (1,1,1-trichlo-
roethane (TCA), trichloroethylene (TCE), tetrachioro-
ethylene (TeCE)) and the electrical conductivity (EC)
compared with the interval depths.
1245
-------
The calculated concentration profile of the chlorinated solvents
has been confirmed immediately after by drilling a BOTESAM-well
(Larsen & Andersen, 1988) a few meters from the testing well,
sampling groundwater for every meter (Fig. 8).
BOTESAM - well No.151.1072, Skrydstrup (Denmark)
Concentration of Chlorinated hydrocarbons in pg /1
(scale logarithmic)
0.1
,1,-Tnchloroethane =
Trichloroethylene = TCE
Tetrachloroethylene = TeCE
20
Fig. 8.: Chemical profile from the BOTESAM-well, Skrydstrup.
Based on the results of the separation pumping test it is
possible to optimize the remedial action pumping by separation
pumping, (section 3.4). The efficiency of the optimization has
been examined as a function of the pumping time (Fig. 9). Further
the optimized remedial action pumping with two pumps has been
compared to the situation where, traditionally, only a single
pump is used so, that water from polluted layers is mixed with
water from non-polluted layers. In the period July - October 1989
the content of chlorinated solvents in the bottom water was below
1 ug/1 while the concentration in the top water was 900 - 1450
ng/1 when the contribution from the top pump varied from, 35% -
50% of the total discharge and the bottom pump from 65% - 50%. In
the subsequent period November 1989 - March 1990 the pump ratios
are changed in order to keep the non-polluted bottom water free
of solvents. This means that less than 10% of the total discharge
is non-polluted water. This must be related to a change in
vertical distribution of chlorinated solvents within the
aquifer. The change is probably due to reinfiltration of the
cleaned water from the treatment plant, into a pit at the
1246
-------
disposal site. This must cause a- downwards movement in the
groundwater, which appears to have forced the lower boundary of
the plume, downwards by 10 meters in 9 months, at the position of
the test well.
Concentration of chlorinated hydrocarbons(yg/l)
2000-
'
1SOO-
1000-
•
soo-
0-
:|:
yt
V
[>x] Cone, in top water
E5B Cone. In bottom water
irea Cone. In mixed water
C
tc
BW
037
'
Mo
lal
^
rinati
ofT
BW
0.11
I
-
A
;d hydro
DA, TCE
1
55
-
S
carbons
TeCE
frn
:£
X
'f:
eu
0
M
O
1
M
"3
^
;
=
^
|
;|
[3
I
|
.;•:
1
N
D
X
i
1
!
I
s
i
j
TO
F 1
SB
:r
1
1
m i
M
The fraction of nonpolluted water
1JO-
os-
00-
;:
i
A
i
S O
1 1
N D
n
J F M
1990
Fig. 9.: Optimized remedial action pumping by separation pumping
in well PB16, Skrydstrup. Test period July 1989 - March
1990. A: The concentration of chlorinated hydrocarbons,
' a total of TCA, TCE and TeCE, in top, bottom and mixed
pump water. B: Fraction of non-polluted water of total
yield.
4.2. Experiments of remedial action pumping in a limestone
aquifer
4 .2.1. Site description
Tests have been performed in a Water Supply well at P. Andersens
Vej Waterwork, Municipality of Frederiksberg, situated in the
1247
-------
Chlorinated solvents, mainly trichloroethylene .(TCE),' have
polluted the limestone aquifer. One of the water supply wells at
the test site is slightly contaminated with up to 15-20 ng/1 TCE
in the production water. This was chosen as the test well.
4.2.2. Hydrogeoloorv
The confined limestone aquifer with fissure flow occurs from-
about 20 m.b.s. and downwards. The test well has a diameter of
300 mm, is about 71 m deep, and has an open-hole interval from
50-71 m.b.s. All the water supply wells in the municipality of'
Frederiksberg are situated in a 500 m wide, high yielding fault
zone, named the Carlsberg Fault.
4.2.3. Experimental set-up
The experimental arrangement consists of two submersible pumps
placed in the .borehole, one just above the bottom of the casing
with the intake approx. 47 m.b.s. and the other pump near the
bottom of the well with the intake approx. 68 m.b.s. The sub-
mersible pumps each have a max. capacity of approx. 50 mVhr.
4.2.4. Results and discussion
Modifying the equations presented in section 3.1., and calculat-
ing, gives the distribution of concentrations of TCE and the
electrical conductivity (EC) for a 7 step separation pumping
presented in Fig. 10. The calculated TCE values shows higher
concentrations in the upper inflow intervals 11-13 than .in the
lower ones, while the EC-values are highest in the bottom of the
well, inflow intervals 17-18. '
1248
-------
10 20 30 40 (19/I
—H .—' 1 "-T 1
500 1000 1500 2000 2500 uS/crr
- Top of limestone surface
, Pumping water • level
L—Bottom ol welt
-Botl<
LEGEND:
• TCE, pg /1 (calculated)
• EC, pS / cm (calculated)
7 TCE, pg /1 (level-accurate)
Fig. 10: Calculated concentration profile, and inflow distribu-
/ , tion in a water supply well at P. Andersens Vej
Waterwork, and a chemical, profile based on level-
accurate water samples collected at the water divides.
TCE = trichloroethylene and EC = electrical conductiv-
ity. I = inflow interval.
The boundaries of the inflow intervals have been located by a
Heat-Pulse-Flow-Meter, and level-accurate samples were taken at
the boundaries (see section 3.3). The results are shown in Fig.
10. The inflow from interval 11-18 has been determined as
fractions of total yield:
q: = 0,20, q2 = 0,10, q3 = 0,12, qA = 0,08, q5 = 0,11, q6 = 0,09,
q7 = 0,09 and q8 = 0,21.
60% of the total yield comes from a fissure system 50 - 57 m.b.s.
and 3.0% from an other fissure system at a depth of approx. 67,5 -
71 m.b.s. 1249
-------
As in Skrydstrup (section 4,1.4), the optimized remedial action
pumping have been examined as a function of the pump time. Pig.
11 shows that in the period 20. February 1990 - 22. May 1990, the
TCE concentration has stabilized at a value of 25 ug/1 TCE in the
top water and 5-6 ug/1 TCE in the bottom water, compared to
approx. 15 ug/1 TCE if only one pump was used. The pump ratio has
been held constant, so the fraction of slightly polluted bottom
water has been approximately 40% of the total yield through the
whole period. In the period 30. April to 22. May 1990 the total
discharge has been adjusted from about 45 mVhr to about 70 mVhr.
This change did not influence the concentrations of TCE in top
and bottom water.
a
3.
Ill
O
30-
25-
20-
15-
10-
5-
0 -
A
*!•
;
T
I
i!
m
1
1
I
1
B'M1
20/2-90
1
^
T'B'M
30/4-90
R
1
%&
T'B'M'
22/5-90
sz LOO-
• S 0.75-
° 'S0.50-
0.00-
20/2-90
LEGEND:
-60*|
30/4-90
22/5-90
Fraction of slightly
polluted water
Total yield (Q To,,,)
35$S Cone. In topwater (I)
£°?ec,: ln bollom (B)
Cone. In mixed
water
Fig. 11.: Optimized remedial action pumping by separation
pumping in a Water Supply Well at P. Andersens Vej
Waterwork, Frederiksberg in the period 20. February
90 - 22. May 90. A: The concentration of TCE in top
(T), bottom (B) and mixed (M) pump water, B: The
fraction of slightly polluted water of total dis-
charge, and the total discharge in mVhr.
1250
-------
The ratio of yields between the top and bottom pumps must
carefully be balanced, so the TCE-concentration is as high as
possible in the top water. However, in this case, the situation
is complicated by the high chloride content from intruding
saltwater from the bottom of the well which has the low content
of TCE.
This reflects an enormous water supply problem in the Copenhagen
area. Chlorinated solvents of anthropogene origin infiltrates the
limestone aquifer from above and saltwater intrudes the same
aquifer from below, because of the large demand for water.
4.3. Localization of a pollution plume in a chalk aquifer
4.3.1. Site description
A former chalk quarry has been filled with mixed industrial and
domestic wastes. The chalk is 1-4 m.b.s. and the groundwater
level is approximately 4-5 m.b.s. The chalk has been quarried
below the natural groundwater level, which means that the waste
has been dumped below the water table.
4.3.2. Hydrogeology
The aquifer is a fractured chalk, with closely spaced fractures,
with rather random orientation. The fracture porosity amounts to
around 1,5% while the total porosity varies from 20-35%. Around
the former quarry there are water table conditions in the
aquifer, but just south of the dump site the aquifer becomes
confined. The chalk is overlain by glacial till, with minor sand-
lenses .
1251
-------
4.3.3. SIT
Two open boreholes 150 mm in diameter, 25 m and 29 m deep,
downstream of the disposal site were tested with the separation
injection test (SIT), as described in section 3.2.2. Salt water
was injected at the bottom and fresh water at the top, and the
position of the interface, at the various injection rates, was
determined with an EC-probe.
4.3.4. Level-accurate sampling
Knowing the flow distribution, and the positions of the bound-
aries between the ten, 10% flow partials, level accurate sampling
at the boundaries, was just, a matter of pumping with the same
ratios from top and bottom as used for the SIT, and sampling at
the corresponding boundaries, The pumping of the sample was done
with a flow rate of approx. 2% of the total flow, and a limited
volume of approx. 3 litres. It should be noted, that the position
of the boundaries can only be considered equivalent, when pumping
and injection takes place through the same cross-section of the
aquifer. In other words the water level should in both cases be
above the highest filter interval. This was the case for the
well, for which the results are shown, but in the other well
there was a small difference of 30 cm in the two cross-sections,
because the casing did not reach the water table.
4-3.5. Results and discussion
The result of the SIT is shown in Fig. 12. It is evident that the
upper part of the aquifer has a much larger permeability, as 90%
of the flow takes place within the upper 25% of the tested
interval. Quite narrow flow zones can be discerned, down to
approx. 30 cm.
1252
-------
TERRAIN
0
GWL.
GWL:
CASING-
7,2m 7i7r.
o
111
cc
111
u.
ui
cc
O
111
m
«
cc
ui
i
a.
ui
a
BGfcJ
Fig. 12. : The relative influx distribution of ten 10% intervals
in well 207.2802, Karlstrup, determined by a. separ-
ation injection test (SIT).
The results from the sampling are seen in Fig. 13. The sampling
method is clearly capable of showing very high concentration
gradients. High contents of K+, Na+, Cl", NHA+/ and relatively low
contents of NCV and S0«~, show the existence of a pollution plume
at the top of the aquifer. This, together with the found flow
distribution, makes it unfeasible to use separation pumping as a
tool for optimizing remedial actions, in this case.
These investigation methods have a high potential for use in
similar cases, because they can be applied in existing wells,
even in the production well, where the pollution may have been
discovered.
1253
-------
SO}- „„
NOJ
Fig. 13.: Concentration prof lies from well 207.2802. The level-
accurate sampling was done using separation pumping.
1254
-------
5. Costs
Because the technique has only been tested at pilot scale, no
cost-benefit analyse has been made yet. Still comparing tradi-
tional remedial action pumping by use of. one pump, to simulta-
neous pumping with 2 or more pumps, the installation costs are
increased by using the separation pumping technique, but
contemporaneously the remediation time would probably decrease
and the costs for treatment decrease because the -volume of
polluted water is reduced. .
1255
-------
6. Concluding remarks and perspectives
The separation pumping technique is a method applicable for
locating and recovering the polluted water, from a vertically
partially polluted aquifer by simultaneous pumping from top and
bottom with two or more pumps in a fully penetrating well without
installation of packers.
The method can be used for determination of the inflow distribu-
tion and transmissivity of selected sections of a well. The
vertical distribution of pollutants can furthermore be determined
and a concentration profile calculated.
Moreover the, method is applicable for collecting level-accurate
water samples from any levels of a fully penetrating well only by
pumping from the well itself.
The method has been tested at two different sites, a polluted
water table aquifer in meltwater sand and gravel with pollution
down to 25 m.b.s., and a confined limestone, aquifer with fissure
flow. In both cases the effect of the separation of polluted
groundwater -from slightly or non-polluted groundwater were
successful. The separation pumping can be provided in existing
wells without disturbances of the pollution distribution with the
aquifer. Remedial action pumping can be provided in a production
well in a way, so the well continues the production of drinking
water at a waterwork using the non-polluted part of the inflow to
the production well.
1256
-------
7. References
Andersen, L.J., R. Jakobsen, F.L. Nielsen og B. Nilsson (1989) s
Separationspumpnings- og separationsinjektionstest. (SPT) og
SIT) i forbindelse med grundvandsforurening (Separation
pumping and Separation injection test, (SPT) and. (SIT) in
relation to groundwater pollution) i ATV-komiteen vedr0rende
grundvand, Vinterm0de, Vingsted 1989. pp. 303-318. (Text in
Danish).
Andersen, L.J. (1990): BOTESAM, separation pumping and capillary
barrier a remedial-action concept applicable to point-pollu-
•' tion. First USA/USSR Joint Conference oh environmental
hydrology and hydrogeology, June 18-21, 1990, Leningrad.
Gosk, E. & P.K. Bishop (1989): Coventry groundwater investi-
gation: Sources and movement of chlorinated solvents in dual
.. . - porosity rocks. EEC contacts no. EV4V-0101-c(BA) . DGU-intern
report. Joint venture between DGU and The University of
' Birmingham. . - . ;
Larsen, F. & L.J. Andersen (1988): BOTESAM-erfaringer med nye
: BOre-TEst og SAMpling udstyr. (BOTESAM-experiences with.a
new, combined borehole, testing and 'Sampling equipment).
Forurening fra punktkilder. ATV-komiteen vedr0rende grund-
vandsforurening, Vingstedcentret 2-3 marts 1988. (Text in
Danish) .• • '
Nilsson, B., L.J. Andersen, R. Jakobsen, E. Clausen & F.L.
Nielsen (1990): Remedial-action pumping by separation-pumping
technique, phase 1: Demonstration model. Lossepladspro jektet,
-Report R3-1, Dec. 1990. (Text in Danish, Summary in English).
Nilsson, B., E. Wille & L.J. Andersen (1990): Remedial Action
pumping by Separation Pumping Technique. Phase 2: Skrydstrup
Waste Disposal Site. DGU-intern report. Preliminary report.
(Text in Danish, Summary in English). , ,
1257
-------
Tate, T.K. & A.S. Robertson (1971): Investigations into high
salinity groundwater at the Woodfield pump station Welling-
ton, Shropshire. (Water Supply Paper Institution of Geol.
Sciences Research Report No. 6).
1258
-------
Appendix 7-A
Case Studies on Chemical Treatment of Contaminated Soils: APEG
Supplementary Information on the APEG Process,
Wide Beach, United States
1259
-------
APPENDIX 7-A
SUPPLEMENTARY INFORMATION ON THE APEG PROCESS
AND THE WIDE BEACH (UNITED STATES) DEMONSTRATION STUDY
Michael A. Smith, Clayton Environmental Consultants,, Ltd.,
68 Bridgewater Road, Berkhamstead, Hertfordshire HP4 IJB,
United Kingdom
The reagent, APEG, dehalogenates pollutants to form a glycol
ether and/or a hydroxylated compound and an alkali metal chloride
which are water soluble. These products are regarded as nontoxic,
although they may have an undesirable environmental impact if
they are released in an uncontrolled manner.
SCIENTIFIC BASIS OP PROCESS
The most widely used variant of the process employs potassium
hydroxide (KOH) in conjunction with polyethylene glycol (PEG) to
form a polymeric alkoxide anion, referred to as KPEG. This acts
as an effective nucleophile and as a phase-transfer catalyst
(Kernel & Rogers 1985). Other reagents can be used (see below).
The presence of oxygen is an essential requirement of the
process. The polyethylene glycol typically has a molecular weight
in the range 300 - 600.
The reagent, APEG, dehalogenates the pollutant to form a glycol
ether and/or a hydroxylated compound and an alkali metal chloride
which are water soluble. These products are regarded as nontoxic
(but see below), although they may have an undesirable
environmental impact if they are released in an uncontrolled
manner.
The reactions involved are shown schematically in Appendix
Figures. 7-1 and 7-2.
In some KPEG reagent formulations to treat chlorinated aromatics,
dimethyl sulfoxide (DMSO) is added to enhance the reaction
kinetics, presumably by improving rates of extraction of the
haloaromatic compounds and/or overcoming inhibitory effects
attributed to the formation of hydroxide anions. Sofolane (SFLN
- tetrahydrothiophene) may also be used to accelerate the
reaction.
Although potassium hydroxide has been the most widely used
reagent, sodium hydroxide has been used in the past and may find
increasing application, in part at least because of lower costs.
Other possibilities include the use of combinations of potassium
or sodium hydroxides with tetraethylene glycol (ATEG).
1260
-------
When KTEG reagent is used (i.e., potassium hydroxide with
tetraethylene glycol) to treat aliphatic halogenated
hydrocarbons, a dehydrohalogenation reaction occurs. Halogenated
compounds with one carbon atom react to form carbon dioxide and
potassium halide. Compounds containing more than one carbon react
to form acetylene (Rogers and Kernel 1987, Harden & Ramsay 1986).
THE WIDE BEACH CASE STUDY (UNITED STATES)
Only one case study was included in the NATO/CCMS Pilot Study:
the United States Environmental Protection Agency's (U.S. EPA's)
SITE (Superfund Innovative Technology Evaluation) Demonstration
at Wide Beach, New York (EPA 1989B, EPA 1991A, EPA 1991B). This
involved application of a KPEG process to the treatment of PCB
contaminated soils in combination with a complex four stage
"thermal processor". This is a continuous process in contrast to
the batch processes demonstrated or used at a number of other
sites. The basic processor has wider applications where thermal,
desorption or pyrolysis is required (Taciuk 1991).
Background. The Wide Beach site is located in a residential area
of the town of Brant approximately 48 km (30 miles) south of
Buffalo, New York.
From 1968-1978 PCB (Aroclor 1264) contaminated waste oil was used
for dust control on local roads at Wide Beach. About 155,000
liters of oil were applied. During installation of a sewer in
1980' highly contaminated soil was excavated from the roadways and
surrounding areas. After completion of the installation surplus
soil was used as fill in residential gardens (yards) and a
community recreation area. An odor complaint in 1981 led to
discovery of 19 drums, some containing PCB-contaminated oil.
Subsequent sampling revealed widespread PCB contamination in air,
homes, private water supplies, ground water, drainage ditches and
soil. About 274 residences were affected in Wide Beach community
and surrounding areas. There are 60 residences in the Wide Beach
community itself accommodating about 120 people in summer and 45
on a permanent basis. A detailed description of the site is
available (EPA 1991B).
Pending long term remedial action, ameliorative measures were
taken to limit public health and environmental impacts. Roadways,
drainage areas and driveways were paved to prevent public
exposure to dust and runoff. Measures to decontaminate homes
included vacuuming, cleaning of carpets, etc., and replacement of
air conditioner and furnace filters. Particulate filters were
installed to protect individual private wells from sporadic PCB
contamination.
Soil contamination ranged from 0.18 to 1026 mg/kg in samples
collected from residential driveways, roadways and drainage
ditches. PCB concentrations in excess of 10 mg/kg were detected
at depths of up to 1 meter in drainage ditches, although in
general'' these concentrations reached only to about 0.15 to 0.3
1261
-------
meters in depth. It was estimated that 21,700 tonnes (15-16,000
m3) of soil were contaminated. In practice, ho
tonnes of soil were processed (see below).
however, about 43,000
PCB concentrations up to 126 mg/kg were found in surface waters ••
from adjacent wetlands.
After a review of possible remediation options, KPEG soil
treatment was chosen as the most cost effective. Preliminary
pilot scale studies showed that initial PCB concentrations
ranging from 30 to 260 mg/kg could be reduced to less than 10
mg/kg. Final concentrations ranged from 0.7 to 1.7 mg/Jcg.
Accordingly, a clean up target of not more than 2 mg/kg was been
set (EPA 1989B, AOSTRA 1990). • . .. ,
The Demonstration Pro-ject. The remedial action at Wide Beach
consisted of.excavation of soils with PCB concentrations greater
than 10 mg/kg from all areas; excavation of contaminated asphalt
material from roadways for disposal off-site, with uncontaminated
material being retained for reuse; treatment of the PCB-
contaminated soils using the an Anaerobic Thermal Processor (ATP)
in association with a KPEG process; use of the cleaned soil as
fill in excavated areas; repavement of the roadways and
driveways; and treatment of perched water in the sewer trench
(this water has been found to contain up to 10 j«g/L PCB's).
On completion of the work the residences were subjected to a -
further thorough cleaning to remove any stray dust, etc.
About 43,000 tonnes (23,000 m3)contaminated soil were treated in
a continuous process. An AOSTRA (Alberta Oil Sands Technology'and
Research Authority) ATP (anaerobic thermal processor) unit was
used under license by SoilTech Inc. to thermally desorb the
organic contaminants from the soil for subsequent dechlorination
(EPA 1991A, EPA 1991B). The AOSTRA "Taciuk" processor, conceived
in about 1975 (Taciuk 1991) was developed in about 1985 for the
continuous extraction and primary upgrading of bitumen and oil
from oil sand and shale feedstocks but has been adapted to the
removal of oily materials from soils (AOSTRA 1990, Taciuk I99f£
Taciuk S^Ritcey 1991). Development work leading to its '""•
application to treat PCB-contaminated soil has been described by
Taciuk (1991).
The ATP system is designed to heat and mix contaminated soils,
sludges and liquids and is heated indirectly. It has four
separate, internal heating zones: the preheat, retort,
combustion, and cooling zones. , :
The process is designed (EPA 1991A, EPA 1991B) to work as follows
(see also Appendix Figures 7-3 through 7-5): ;
Before entering the preheat zone the contaminated soils are • r; :
-
1262
-------
sprayed with a diesel fuel and oil mixture containing alkaline
polyethylene glycol reagents (the oil mixture acts as a carrier
for the dehalogenation reagents and is also necessary to achieve
optimum operation of the various stages of the process). In the
preheat zone, water and volatile organic compounds (VOC's)
vaporize. At the same time, the reagents dehalogeriate or ' ' ,
chemically break down chlorinated compounds including PCB's. The
vaporized contaminants and water are removed at a temperature of
200-315 °C under a slight vacuum to a preheat vapor cooling
system consisting of a cyclone, condenser and 3-phase preheat
system. Uncondensed light organic vapors are fed by a gas blower
directly into the combustion chamber of the processor. Condensed
liquids are separated: the oil fraction is recycled to a reagent
blending tank, and recovered water is pretreated and sent off-
site for disposal.
From the preheat zone, the hot, granular solids pass through a
sand seal to the retort zone (approximately 600 °C) . Here heavy
oils vaporize and thermal cracking of hydrocarbons forms coke and
lower molecular weight gases. The vaporized contaminants are
removed under vacuum to a retort gas handling system. Entrained
particles are removed by a two-stage pair of cyclones (the
collected dusts and fines are blended with the treated soil).
After passing through the cyclones the vapor is cooled by oil
circulating in two packed columns, acting as a two-stage direct
contact condenser for the high boiling point compounds. Remaining
uncondensed vapors are cooled in a water-cooled non-contact
condenser and passed to the combustion chamber of the process .-
The oil phase is combined with the condensate from the packed
columns. This oil condensate is then sent to the blending unit to
mix with the APEG reagents. The blend is pumped at a measured
rate and is applied to the untreated soils in the feed chute of
the processor. Condensed water is pumped directly to the on-site
treatment system.
The coked soils pass through a second sand seal into the
combustion zone (650-815 °C), where they are combusted and either
recycled to the retort zone or sent to be cooled in the cooling
zone. Exhaust gas from the combustion zone is treated in a system
consisting of a cyclone and baghouse that removes particulates; a
scrubber to remove acid gases; and an absorption bed to remove
trace organics. The treated gas is then discharged to atmosphere.
The combusted soils that enter the cooling zone are cooled in the
annular space between the outside of the preheat and retort zones
and the outer shell of the kiln. Treated soils leaving the
cooling zone are quenched with scrubber water and then
transported to a storage pile. l
Residence time in the preheat and retort zones is approximately
15 to 30 minutes each, sufficiently long to ensure complete
volatilization of the hydrocarbons. The heat requirements are
provided by natural gas burned in the precombustion chamber. A
fraction of the coke in the combustion chamber may also burn.
1263
-------
The optimal moisture content of the waste is 5 to 10 weight
percent. Although wastes with a moisture content greater than 20
percent may be treated the high moisture content will affect net
throughput rates. The ATP system is designed to treat wastes with
a_nominal hydrocarbon concentration of 10 percent. However, heavy
oil contaminants have been reduced from as high as 60 percent in
the feed to near detection limits in the treated solids.
The rate of contaminant desorption and dechlorination from soils
and sediment is influenced by the contaminant concentration.
SoilTech report that the contaminant concentration in the feed,
waste is generally independent of the contaminant concentration
in the feed waste and will be near the detection limit of,the
contaminant. The processor treats wastes containing.low boiling
contaminants more effectively than wastes containing high boiling
contaminants. However, high boiling organics such a PCBs and
polycyclic hydrocarbons can be removed to concentrations below
detection limits of 1 mg/kg.
Demonstration Results. The ATP technology was demonstrated under
the U.S. EPA SITE program in May 1991. Three test runs were
conducted, each consisting of five and one half hours of solids
and liquids sampling and five hours of stack sampling. Key
findings of the demonstration were:
The SoilTech ATP unit removed PCB's in the contaminated soil to
levels below the desired cleanup concentration of 2 mg/kg. PCB
concentrations were reduced from an average concentration of 28.2
mg/kg in the contaminated feed soil to an average concentration
of 0.043 mg/kg in the treated soil.
The SoilTech ATP does not appear to create dioxins and/or furans.
No volatile or semivolatile organic degradation products were
detected in the treated soil. There were also no leachable VQC/s
or SVOCs detected in the treated soil. ..
No operational problems affecting the ATP's ability to treat the
contaminated soil were observed.
HEALTH AND SAFETY ISSUES
Polyethylene glycol and dimethyl sulfoxide are essentially
nontoxic materials. The addition of the glycol to the aromatic
PCB or dioxin ring produces a water soluble, low toxicity
product. Testing by the US EPA indicates that replacement by
polyethylene glycol of a single chlorine on a PCB or dioxin ,;
molecule produces a material of low toxicity (LD50 greaiter than '
5000 mg/kg) which does not appear to bioaccumulate or cause
mutagenic effects. However, the long term stability of these
compounds in a soil environment where they may be subject to
microbial and other degradative reactions does not appear to have
been determined. -
1264
-------
Treatment of certain chlorinated aliphatics in
boncentrations ttfith APEG may produce compounds' th&t are
potentially explosive (e.g., chloroacetlylene§) SHd/or cause a
fire hazard, ihe USe of DMSO or similar reagettti may result in
the formation of highly flammable volatile organibs (e.g., methyl
sulphide). Alkaline reactive materials such as aluminium can
result in formation of hydrogen. Vapors from, heating oily
soils, which are often the matrix in which PCBs are found, can
also create such potential problems as fires and noxious fumes.
These potential problems can all be overcome by appropriate
process design.
The process must also be conducted with care because of the
elevated temperatures and production of steam, the use of
caustics in the process, and the acids used in neutralization.
If DMSO, which is a powerful solvent and skin penetrant, is used
care must be taken to prevent contact with skin as it enhances
transport of PCB's through the skin, thus increasing the risk of
exposure.
KTEG processes are exothermic and can produce vinyl halides as
intermediate compounds. Special considerations will therefore
apply in the design of plants to operate these processes (Harden
and Ramsay 1986).
REFERENCE:
See References following Chapter 7, Vol. 1.
1265
-------
HO-PEG + KOH > KO-PEG + H2O (1)
Aryl-ci + KO-PEG > Aryl-o-PEG + KCI ...,,(2)
Aryl-o-PEG > Aryl-OH + Vinyl-PEG .......... (3)
Note: In reaction (1) the polyethylene glycol (PEG) is reacted
with potassium hydroxide to form the reactive KPEG species. This
preparative step may be performed directly in the contaminated
matrix or externally. Reaction (2) takes place takes place over a
wide temperature range from ambient to about HOC. Reaction (3)
represents the conversion of the ether linked PEG/Aryl moeity to
a phenolic with consequent release of a vinyl terminal
polyethylene glycol.
Equation (4) below represents this overall reaction.
R-(C1)..+ A-PEG
R-(Cl)y-OR' + AC1 +R-(Cl)y-OH ...(4)
Pig 6.1 APEG REACTIONS
1266
-------
HO-PEG 4- KOH > KO-PEG + H2O (1)
Aryl-Cl + KO-PEG -> Aryl-O-PEG 4- KC1 ....(2)
Aryl-O-PEG —• > Aryl-OH + Vinyl-PEG (3)
Note: In reaction (1) the polyethylene glycol (PEG) is reacted
with potassium hydroxide to form the reactive KPEG species. This
preparative step may be performed directly in the contaminated
matrix or externally. Reaction (2) takes place takes place over a
wide temperature range from ambient to about HOC. Reaction (3)
represents the conversion of the ether linked PEG/Aryl moeity to
a phenolic with consequent release of a vinyl terminal
polyethylene glycol.
Equation (4) below represents this overall reaction.
R-(C1)X+ A-PEG
•> R-CC1) -OR' + AC1 +R-(C1)-OH ...(4)
Appendix Figure 7-1. APEG reactions.
1267
-------
ROK+HOH
Appendix Figure 7-2. Chemical reactions during the KPEG process.
VM*
Alt
I.
OilPtodua
Mulmum 5,500 ACFM
170F. «% Wal« Vigor
In Caita OH
Reagent
Makeup
Unit
Rtagenl
fiaidiing
Uiiil
ToOn-Sile
liialmenl
Appendix Figure 7-3. Simplified process flow diagram.
1268
-------
|
.2
Q>
%
•5
CO
1»
s
c
i
s
o
o
CO
2
i
O)
x
TJ
o>
Q.
a.
1269
-------
OBOvwce
V
MS r——
ace —•—
\ J
COOUNC ZONE
COUBUSnON ZONE
P. STEAM
MO HYDROCARBON
ROW
RUE (MS
PREHEAT ZONE |
"•••» •[
•V SWM SEH. (TW.j -^
K/j RETORT ZONE
w
v/\
^/A
/
/
/
/
/
\
fHEO
STOCKS
{PENT SOUOS
WU.NCS
HYDROCARBON
ANO STEAM
VAPORS FU3W
— COMBUSnON
AM FLOW
=f
Appendix Figure 7-5. Simplified sectional diagram showing the four internal zones.
1270
-------
Appendix 7-B
Case Studies on Chemical Treatment of Contaminated Soils: APEG
The AOSTRA-Taciuk Thermal
Pyrolysis/Desorption Process, Canada
1271
-------
THE AOSTRA-TACIUK THERMAL
PYROLYSIS/DESORPTION PROCESS
AUTHOR: William Taciuk, P.Eng.
ADDRESS: UMATAC Industrial Processes,
210-2880 Glenmore Trail, S.E.,
Calgary, Alberta, Canada.'
Postal Code: T2C 2E7
TELEPHONE: 403-279-8080
FAX NO.: 403-236-0595
PRESENTED TO: NATO/CCMS Conference
Washington, D.C.
November 18-22, 1991
1272
-------
ABSTRACT
The AOSTRA-Taciuk Thermal Pyrolysis an4 Desorption Process, for use in
contaminated waste clean-up applications, has been under development since 1986.
The basic AOSTRA-Taciuk Processor machine was conceived in 1975 and was actively
researched and developed for oil sands and oil shale treatment applications for oil
production purposes. The patents for this technology are held by the Alberta Oil Sands
Technology and Research Authority (AOSTRA) which is located in this oil-producing
province in Western Canada.
UMATAC Industrial Processes, located in Calgary, Alberta, Canada, are responsible
for the research, development, design and manufacturing of equipment and plants
which utilize this technology. UMATAC is actively involved in oil sands and oil shale
prototype projects and, in 1989, constructed a 10 ton/hour transportable waste
treatment plant which has recently successfully completed at 42,000 ton clean-up
project where chemical dechlorination and desorption of PCB-contaminants was
combined within the unique features of the Processor machine.
Unlike incineration, the use of the Processor system provides options for recovery and
reuse of separated by-products.
This paper provides a brief view of the development history, description of the Process
and carries on to describe current development status and potential future development
for treating other specific contaminated waste streams.
1273
-------
DESORPTION PYROLYSIS UTILIZING THE AOSTRA-TACIUK PROCESSOR
SYSTEM
The basic design and concepts for the AOSTRA-Taciuk Processor were conceived in
1974-75 for application in the Athabasca Oil Sands located near Fort McMurray,
northern Alberta. The Processor is a horizontal, rotating unit which has internal
components to permit staged processing, removal and separation of product streams
depending on thermal sensitivity. The internal reactor design permits thermal
processing of preheated feed materials in an anaerobic atmosphere at temperatures in
the range of 950° to 1100°F. Vaporized and/or thermally cracked hydrocarbons from
the reactor are removed and condensed so there is potential for recycling or reuse of
this concentrated energy stream. The Processor can accept a wide range of feed
material sizes, water content and hydrocarbon content. Since the Processor is a basic
machine consisting of a heat exchange zone, reaction zone and combustion, zone, it can
be used to treat any type of contaminated feed material that responds to temperature
phase changes within the reactor operating range.
DEVELOPMENT HISTORY
The UMA Group, a Canadian emplqyee-owned consulting engineering company with
the head office in Vancouver, B.C., formed UMATAC Industrial Processes in 1976.
UMATAC's primary function was to research, test, develop, design and market the
AOSTRA-Taciuk Processor Technology. This work was carried out under funding
agreements with the Alberta Oil Sands Technology and Research Authority (AOSTRA)
which owns and licenses use of the technology. UMATAC retains the engineering,
development, design and manufacturing rights for all proprietary equipment associated
with this technology.
A 5 ton/hour Pilot Processor was constructed in Calgary in 1977 and has been operated
on a wide range of feed materials to provide test data for use by UMATAC engineering
involved in design, modeling, scale up and assessing suitability of the Process for
handling these materials.
During the period 1976 through 1984, development of the Technology was
1274
-------
concentrated on Athabasca oil sands processing capability. In excess of 12,000 tons of
various grade oil sands were processed in this development which verified the predicted
stability, ease of operation and uniformity of Oil product yield that could be obtained
from use of the ATP.
In 1985, with the extreme decline in world oil prices, plans for oil sands
demonstration/commercial prototype projects had to be delayed and UMATAC shifted
its research efforts towards Processor applications in processing of oil shales, treatment
of various hydrocarbon-contaminated wastes and separation of hazardous waste
constituents from contaminated soils.
During the period 1985-1987, UMATAC successfully completed various phases of test
work relating to Australian oil shales processing. These efforts included pilot testing of
1,400 tonnes of Australian oil shales in 1987, and UMATAC is currently working with
the Australian lease operator towards a large-scale (6,000 tonne per day) demonstration
plant.
In 1986, UMATAC began a program of research and testing for developing the
AOSTRA Taciuk Process Technology for clean-up and separation of waste materials
containing various hydrocarbon contaminants. A series of batch pyrolysis tests and
pilot plant tests has been successfully completed since that time.
Programs relating to the waste treatment application are summarized as follows:
1. Oil Refinery Wastes such as API Separator Sludges, Emulsions and Oil-
Contaminated Solids.
a) In early 1987, a series of ATP batch tests was carried out on the typical oil
refinery waste streams of API separator sludges, slop oil emulsions and
DAF float sludges. Samples were received from a refinery located in
Montreal. The results of these tests demonstrated that the Process could
treat these materials and produce a solids reject that meets acceptable
leachate criteria, produce an oil product that could be recycled, and produce
a water product that could be treated to acceptable environmental standards.
1275
-------
This program was conducted for inclusion in the evaluation by the American
Petroleum Institute (API) of available process technology applicable to
clean-up treatment of these particular wastes of oil refinery operations. The
UMATAC results were submitted to API and were included in the
subsequent API report of its Task Force Study (API Report #4465, May,
1988).
b) A series of batch tests was carried out on samples of oil-contaminated solids
from a decommissioned refinery located in Montreal. These tests were
successful and were followed in December, 1986, by a pilot plant run with
the Processor on a 30 ton bulk sample of oil-contaminated solids from this
site. Results of this test run substantiated the environmental acceptability of
the ATP System treatment.
c) In mid-1987, a series of batch tests was carried out on oily sludge and oil
samples from a refinery site in the U.S.A. Results of these tests were
similar to those obtained from the a) and b) programs discussed above, and
confirmed Processor applicability for treatment of these wastes.
2. Heavy Oil Production Wastes
A major test program was completed in 1988 on samples of oily wastes from
production of heavy oil obtained from 17 different heavy oil operations located in
Alberta and Saskatchewan. This program included pilot plant test runs during
which 400 tons of bulk waste samples from four representative sites were tested
in the pilot plant. This program was sponsored and funded by Environment
Canada, DSS, CANMET, and AOSTRA.
3. PCB Contamination Removal
In 1988, UMATAC carried out a series of batch tests on PCB material (Aroclor
1242 and 1260) to examine and demonstrate the ability of the Taciuk Processor
to vaporize and extract these materials from host solids and condense them into
concentrated liquid form. The concentrated PCB condensates could then be
disposed by destruction in remotely located incinerators, or could be burned in a
pre-combustion chamber connected to the ATP Processor combustion zone.
1276
-------
Two test runs on PCB-spiked soil samples were carried out in the pilot plant
whereby approximately 300 Kg of Aroclor 1242 was injected into the 5 TPH
Processor unit along with feed sand to simulate a contaminated sand feed. The
test results verified that the PCB's could be removed and condensed with no
production of dioxins and with minimal production of non-toxic furans. The
tailings solids contained no trace of PCB's, dioxins or furans. In addition,
dioxins or furans were not produced in the oxygen-deficient vaporizing and
condensing phase of processing.
In 1989 and 1990, batch tests on PCB removal using a combination of chemical
dechlorination and desorption, were successfully completed and was first utilized
for commercial clean-up operation at Wide Beach in New York State, U.S.A.
Other Materials Tested
Following is a listing of other materials tested in UMATAC's batch pyrolysis
equipment with potentially successful results:
a) Scrap Rubber Tires
b) Coal Tar Residues and Emulsions
c) Plastic Containers
d) Compacted, Shredded Garbage
e) Emulsions from Heavy Oil in the United States
f) Athabasca Bitumen/Solids/Water Mixtures
Processor Flexibility
The AOSTRA-Taciuk Process Technology provides a rugged, simple unit that
has excellent operating stability and flexibility to deal with feedstock variations
and process variations. It can produce liquid products, which have been
separated, at temperatures near their vaporizing points, and oil by-products that
can be recycled or transported to other locations for final treatment or reuse.
These functions are achieved in an oxygen-free atmosphere in the Processor,
which also incorporates heat exchange to minimize fuel requirements and flue
•"gas treatment. The internal design of the combustion annulus provides for long
1277
-------
residence time and intimate mixing of gases and solids to ensure complete
combustion.
During the past 10 years, the pilot plant has been operated in excess of 10,000
hours and has processed in excess of 15,000 tons of various feed materials.
This operation has demonstrated the ability to provide safe, stable and consistent
operation while testing a range of feed materials. These features of the Process
and equipment are now being confirmed in the commercial operation of the
SoilTech 10 TPH ATP waste treatment plant.
BASIC PROCESS DESCRIPTION
The AOSTRA-Taciuk Processor is a single, horizontal, rotating unit that contains
several zones where specific functions are performed. Below is a diagram illustrating
the major zones and flows which cooperate to achieve the following:
AOSTRA Thciuk Processor
FLUE GAS
STEAM
WASTE FEED
(LIQUIDS t SLUDGES)
CLEAN SOLIDS
AUXILIARY BURNER
HYDROCARBON VAPOR
COMBUSTION AIR
SPENT SAND
I.
2.
High thermal efficiency by use of an internal heat exchange which cools hot flue
gases and combusted solids, while preheating the incoming feed.
Anaerobic vaporizing of water and light hydrocarbons, which are extracted from
the preheat zone as a separate low-temperature vapor stream.
1278
-------
3. Anaerobic vaporizing and/or thermal cracking of hydrocarbons, which are
extracted from the reaction zone as a high-temperature vapor stream.
4. Oxidization of carbon residues deposited on inert solids during thermal cracking.
This is carried out in a combustion zone which envelopes the reaction zone.
This combustion provides a portion of the Processor heat requirements.
5. Thermal stability and zone separation by use of patented, unique, sand seals and
a sand recycle system, which provide heat to the reactor via a "locked-in"
recycle sand charge.
Auxiliary treatment and recovery systems are used to contain and process the various
Processor effluents. Below is a block diagram illustrating the auxiliary systems, which
can be easily altered
AQSTRA Taciuk Process System
STEAM
CONDENSER
uamo WASTES
WATER
WATER AND Oil. PRODUCTS TO STORAGE.
TREATMENT. RECYCU ANQOH
TRANSPORT TO DISPOSAL
STEAM
CONOEHSWQ
to suit the requirements of a specific feed material and site conditions.
SOILTECH COMMERCIAL TRANSPORTABLE PLANT
In 1989, UMATAC designed and constructed a 10 ton/hour transportable waste
treatment plant for use by SoilTech, Inc. of Porter, Indiana. This plant was designed
1279
-------
with parallel, full capacity, flue gas treatment trains and, where possible, standby
pumps, blowers, burners, compressors, etc., were provided to maximize plant
reliability. This plant has recently successfully completed treatment of 43,000 tons of
PCB and oil-contaminated soils located at Wide Beach, south of Buffalo in New York
state. The plant is being relocated to Waukegan Harbor, north of Chicago, where it
will be used to treat 20,000 tons of dredged sediments and drainage ditch soils
contaminated by PCB's and various cutting and lubricating oils. SoilTech has gained
commercial acceptance for the AOSTRA-Taciuk Processor in the United States and is
actively involved with UMATAC in batch testing and evaluating use of the ATP for
other projects and industry problem areas in wastes treatment.
AOSTRA RESEARCH AND DEMONSTRATION PORTABLE PLANT
In early 1991, AOSTRA commissioned UMATAC to design, construct and operate a 2
to 5 ton-per-hour portable AOSTRA-Taciuk Processor plant. This plant will be
completed in early 1992 and will be initially used at UMATAC's pilot plant site for
developmental testing of several internal Processor configurations leading towards use
of the ATP for processing of municipal solid wastes, and investigate potential for use
on sulfur or phosphorous contaminants. This unit will be available for use as a
demonstration facility for oil sands development and for clean-up operations in. Alberta.
On-site demonstration at locations outside Alberta will be available, dependent on
operating schedules. .
AOSTRA PORTABLE
5 TON/HR ATP TREATMENT PLANT
ISOMETRIC LAYOUT
1280
-------
PYROLYSIS - AN OPTION TO INCINERATION
The use of staged heating and separation of vapor phase products at the lowest possible
temperature provides an efficient and flexible way of dealing with complex waste
mixtures. Certain incinerator manufacturers are evolving options to their base
equipment tending towards staged separation and partial pyrolysis, which is then
followed by second-stage combustion.
The AOSTRA-Taciuk Processor System can provide this flexibility if a benefit can be
derived as a result of its use. These benefits are accentuated if:
a) The feed contains a high water content that can be removed by vaporizing at low
temperature.
b) The feed contains a high hydrocarbon content where product oils in excess to
that required for Process heating can be stored and used for other purposes.
c) The feed contains other impurities, metals, etc., that concentrate in the fine
solids fractions that are removed in the cyclones and baghouses.
d) The flue gas quantity remains at a low level since combustion is determined by
Process heat requirements, not by total product combustion.
e) The combustion process is carried out at 1300° to 1400°F instead of 1800° to
2300°F so this reduces potential for slagging and partial oxidation of heavy
metals, etc. Most contaminants are never exposed to combustion since they are
stripped off in the preheat zone and the reaction zone.
f) The feed can contain minor amounts of coarse rocks and debris since the ATP
unit has an internal screening and rejecting system for handling difficult
materials.
The ATP system can be more complex to construct and operate than a straight
incinerator so its use must be closely evaluated to determine its costs/benefits for a
particular feed material.
1281
-------
COMPARISON OF INCINERATION VERSUS PYROLYSIS FOR HYDRO-
CARBON SLUDGES
a
UMATAC carried out a series of tests and comparisons of Process parameters for
sludge feed using the ATP Process versus a portable incinerator. The key results from
this comparison are as follows:
Feed Characterisfir.s
Water Content
Hydrocarbon Content
Solids Content
Sulfur Content
Feed Rate (assumed)
Process Parameters
Process Combustion Heat Release
(MM BIU/Hour)
Flue Gas Flow (Ibs./hour)
Inert Gases
"Water Vapor
Light Oil Recovered (Ibs./hour)
Reagent Consumption (Ibs./hour)
Solids Residence Time (minutes)
35%
50%
10%
5%
15 tons/hour
Pyrolysis Incineration
28.3
30,000
5,000
11,000
250
15 to 20
270.0
450,000
175,000
-0-
7,000
.02 to .04
The much reduced flue gas flow requirements, cooling water requirements and the
recovery of oil condensate, indicate the basic advantages of pyrolysis/desorption as
compared to straight incineration.
DESORPTION/DECHLORINATION OF PCB'S
PCB's are chemically produced compounds which are very stable, hence their
advantages for industrial products such as dielectrics and high temperature lubricants.
1282
-------
This characteristic also makes them difficult to destroy chemically so that, after the
toxic characteristics of the compound became know, destruction by high temperature
oxidation was, for some time, the only effective way to break the chlorine-to-biphenyl
bond. PCB's vaporize at temperatures of 325°C to 425°C so they can be stripped away
from the host solids at the appropriate temperature.
In the late 1970's, there were several discoveries that reagent mixtures of an alkali and
glycols could detoxify chlorinated aromatic compounds, which include the PCB's.
Most of these findings addressed liquid-state PCB's. Since these discoveries, there
have been many efforts to develop industrial processes which could be used efficiently
and economically to treat, using the alkali-glycol chemistry, the great quantities of
PCB's and PCB-contaminated materials which have been collected and identified for
destruction.
One of the difficulties encountered is the variability of the materials in which PCB's are
dispersed, the worse being soils and sludges where PCB concentrations can vary
upwards from zero, and the host solids can include almost the whole range of natural
and man-made inorganic materials. For process detoxification, the difficulty these
materials presents is largely the need for a process vessel which can condition and
present the contaminated solids in a manner in which the alkaline reagents can address
the PCB's so that the dechlorinating reaction can occur uniformly, economically and
reliably in a continuous flow.
The AOSTRA-Taciuk Processor was designed to process soils and rock materials which
can vary in consist and particle size up to 75mm. Although it functions best in
processing coarse materials, it can also handle the fine silts and clays which are so
commonly found as the host for drainage-collected contaminants like PCB's. The
Processor internal zones, Preheat and Reaction, are designed to vigorously and
simultaneously mix and heat the feed solids in the turbulent, tumbling action of the
rotating vessel. This process condition was considered desirable for enabling reagents
such as NaOH and PEG400 to address the PCB's in the aroclors and askarels coating
the solids particles of the feed soil.
In the summer of 1989, UMATAC, in cooperation with the U.S. EPA Risk Reduction
Engineering Laboratory (RREL) in Cincinnati, carried out an initial series of batch tests
using various ratios of NaOH and PEG400 to chemically dechlorinate PCB's. These
1283
-------
studies continued in early 1990, based on the previous experience, including desorbing
the PCB's from the soils. Communication was continued with the EPA RREL, which
was also actively developing the dechlorinating technology. Extensive batch tests were
conducted at laboratory and bench scale. UMATAC uses a small, rotating batch
reactor which simulates the reactions of the Processor Preheat and Reaction zones.
Numerous test runs were made with various alkalis, concentrations, etc. The test work
concentrated on studying methods of mixing, reagent preparation, sensitivity of
reactions to time and temperature, and ability to recycle (reuse) reagents not consumed
by the reaction.
The basis of the UMATAC testing was to introduce alkaline metals into the water-free
and oxygen-free environment of the Preheat Zone to react with the chlorine of the PCS
molecules to produce salts, i.e., the alkali strips the chlorine from the PCB's leaving
biphenyl and salt. The polyethylene glycol is used as a carrier for the alkaline metals
and probably acts as a catalyst for the reaction between chlorine and the alkali. Sodium
hydroxide (NaOH) was used as the source of the alkali. The results were positive and
showed that the dechlorinating reaction, known as the phase transfer catalysis effect,
occurs in the temperature range and residence time of the Processor Preheat Zone.
The fraction of PCB's reduced in the tests was consistently high, but was rarely 100%.
Thus, there would be some carry-over of the PCB's from the Preheat to the Reaction
Zone where the higher temperature would ensure vaporization and/or thermal cracking.
This was considered desirable in that these PCB molecules would be recovered in the
condensates from the vapor product streams, directed to and recycled with the PEG for
treatment again in the Preheat Zone.
The results were similar with Na and K, and Na was selected for the confirming tests
and further development because its Use is more practical. The overall results are
shown graphically in Figure 1 as a plot of PCB's remaining (wt%) versus contact time
in the reactor.
These UMATAC R&D programs resulted in the conclusion that the ATP System is
commercially capable to undertake remediation treatment of soils and sludges, using
one of the two processes; a) desorption, or b) dechlorination. In effect, as explained
above, the dechlorination process is used followed by desorption in the Reaction Zone,
wherein the latter is used to ensure removal of the PCB's from the treated soil. This is
important since the consumption of chemical can be minimized by utilizing the reactor
1284
-------
for final and absolute removal from the solids of any remaining PCB's and PEG400
reagents. However, overall dechlorination of the PCB's is accomplished by way of the
recycling for retreatment of any carry-over PCB's with the PEG reagent charge.
On the basis of the experience, developmental and commercial, which UMATAC has
with the technology, the factors and/or limitations to the applicability of the ATP for
PCS decontamination of soils and sludges relate mostly to the physical characteristics
of the feed material, the water content, and the size and variability of the particles.
Higher water content reduces the throughput capacity of a given size of Processor
because of the higher heat loading versus heat transfer capacity. Throughput is also
reduced with higher fines content because heat transfer capability is less than with
coarser solids. Conversely, throughput increases with coarse feeds. Feed stocks, with
materials larger than can be processed, can be pre-separated or sized (screened or
crushed) with ancillary equipment to a size which the Processor can handle. The
conditions and requirements of each site or feed material must be carefully examined to
determine the overall configuration and plant in which the basic ATP Processor and
plant can perform most effectively.
The following tabulations of "ATP Proof-of-Process Emission Results" provide a
summary of test results obtained during performance testing of the 10 ton/hour plant at
the Wide Beach Superfund Site.
TABLE 1
SUMMARY OF THERMAL DESORPTION OF PCBs FROM SOILS
MATERIAL
Feed Soil
Product Soil
Oil Product
PCS CONCENTRATION, ppm
TEST 1 TEST 2
7,000 16,000
ND(0.1) ND(0.1)
65,431 157,725
TABLE2
SOIL CHARACTERIZATION DATA
WIDE BEACH SUPERFUND SITE
Total Volume to Remediate, cu.yds.
Soil Particle Size Distribution (weight percent):
Clay
Silt
Sand
Gravel
Moisture Content (weight percent)
PCB Concentrations:
Discrete Samples from the Site Characterization
Studies - lest Samples, ppm (1 was 5,300) 0 to 1,000
Plant Operations -12 hour Feed Composites, ppm
Treated Soil Residual (detection level 70 ppb)
ESTIMATED
19,000
0
8
65
27
8 to 15
ACTUAL
30,000
16.0
26.3
32.3
25.4
12 to 30
Oto46
Not Detected
1285
-------
TABLE 3J
WIDE BEACH SUPERFUND SITE
ATP PROOF-OF-PROCESS EMISSIONS RESULTS
Test Date
Soil Feed to Plant (TPH)
PCBs in untreated soaw - ppm
PCBs in treated soil0'' - ppb
Stack Gas Flow (dscfm)
Stack Temperature - °C
PCBs in stack gas (gm./hr)
Polyethylene Glycol (gm/hr)
C0(kg/hr)
CO2 (volume %)
Qj (volume 96)
Paniculate gm/m3
SO2 (volume 56) .
NO2 (gm/hr)
2,3,7,8 PCDD/PCDF(ng/dscm)
EPA equivalents
New York State equivalents
90-09-07
8.1
0-46
<63
3600
63
0.0048
(a)
1.6
6.4
11.4
0.76
0.03
613
< 0.062
<0.015
TEST 2 TEST l
90-09-07
8.1
6-46
<63
3500
64
0.0027
<0.019
2.25
6.6
11.2
0.69
0.02
618
< 0.076
<0.019
90-10-04
7.1
_
3160
58
H
*•
—
7.2
10.8
0.023
_•
—
_
_
90-10-05
7.1
2964
59
7.3
10.7
0.09
90-10-05
7.1
3013
50
4.7
13.9
0.07
JARGET
25 (avg)
2000
3233
57
0.015
0.019
5.59
0.11
3.3
1200
0.2
dscfm - dry standard cubic feet per minute
gm/m3 - grams per cubic meter
TPH - tons per hour, average
- J^dffm - nanograms per dry standard cubic meter, at 7% residual oxygen
detected "* ""^ ***** * ** deteCti°n limlt« Such ^ ^ *"*»* enriuion. limit for PEG could not be
Much more complete data and results of various UMATAC's test programs, as well as
batch testing of any candidate feedstocks, is available at UMATAC's facilities in
Calgary.
1286
-------
ACKNOWLEDGEMENT
The author expresses his appreciation of the financial and technical support received
from the Alberta Oil Sands Technology and Research Authority. Their representatives
have visited many countries of the world and completed "memoranda of understanding"
agreements with these countries. UMATAC has carried out test and evaluation
programs resulting from these government-to-government agreements and these joint
efforts are continuing with the purpose of developing new technology and applications
for the ATP as well as other AOSTRA-funded technologies.
1287
-------
-------
Appendix 7-C
Case Studies on Chemical Treatment of Contaminated Soils: APEG
AOSTRA-SoilTech Anaerobic Thermal Processor
Wide Beach, United States
1289
-------
United States
Environmental Protection
Agency
EPA/540/MR-92/008
March 1992
&EPA
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
Demonstration Bulletin
AOSTRA-SoilTech Anaerobic Thermal Processor
Wide Beach Development Site
SoilTech ATP Systems, Inc.
Technology Description: The anaerobic thermal processor
(ATP) was developed by UMATAC Industrial Processes^
tho sponsorship of the Alberta Oil Sands Technology and Re-
Iv^f10^,^03™' and is licensed bv SoHTech ATP
Systems, nc.. a U.S. corporation. The ATP technology involves
a physical saparatfon process that thermally desorbs organlcs
such as polychlorlnated biohenyls (PCBs) from soil and sludge
Tho ATP process was used in conjunction with optional
uhe,mically treat over 42'000 '°ns °f
6 WWe Beaoh Development site in
4 • demonstration, the contaminated soils
o rh a,d£?muel and Oil mixlure staining alkaline
B glycol (APEG) reagents before entering the preheat
reagent?9 " 3C'S 3S * °arrier for the dehalogenation
(40°-650° F), water and volatile organic
c , .P0"29 [Rgure 1>" At the same «me, the
reagents dahalogenate or chemically break down chlorinated
compounds (including PCBs). The vaporized contaminants and
water are removed via a vacuum to a preheat vapor cooling
system consisting of a cyclone, condenser, and 3-phase preheat
h!Pa WoJ; T?6 "
-------
separator. The final noncondensable gases are returned to the
combustion chamber of the process. The oil phase is combined
with the condensate from the packed columns. This oil conden-
sate is then sent to the reagent blending unit to mix with the1
APEG reagents. The blend is pumped at a measured rate and is
applied to the untreated soils in the feed chute of the processor.
Condensed water is pumped directly to the onsite treatment
system.. , . '
The coked soils pass through a second sand seal into the
combustion zone (1,200-1,450° F). Here the coked soils are
combusted and either recycled to the retort zone or sent to be
cooled in the cooling zone. Flue gas from the combustion zone
is treated in a system consisting of a cyclone and baghouse that
remove particulates; a scrubber that removes acid gases; and a
carbon adsorption bed that removes trace organics. The treated
flue gas is then discharged to the atmosphere through a stack.
Treated soils exiting the cooling zone (500-800° F) are quenched
with water and are then transported by conveyor to an outside
storage pile. ' '
Waste Applicability: SoilTech reports the following specifica-
tions of the ATP system. The optimal moisture content of the
waste to be treated is between 5 and 10 weight percent. Wastes
with a moisture content up to 20 percent can be treated, but will
impact the net throughput rates. Wastes with a moisture content
greater than 20 percent may need to be dewatered to optimize
process economics. The ATP system is also designed to treat
wastes with a nominal hydrocarbon concentration of 10 percent.
Heavy oil contaminants have been reduced from as high as 60
percent in the feed to near detection limits in the treated solids.
The rate of contaminant desorption and dechlorination from soils
and sediment is influenced by the contaminant concentration.
SoilTech reports that the contaminant concentration in the treated
solids is generally independent of the contaminant concentration
in the feed waste and will be near the detection limit for the
contaminant. The processor treats wastes containing contami-
nants with low boiling points more effectively than wastes con-
taining contaminants with high boiling points. However, high boil-
ing point organics such as PCBs and polycyclic aromatic hydro-
carbons can be removed to concentrations below detection limits
of 1 part per million (ppm).
Demonstration Results: The ATP technology was demon-
strated at the Wide Beach Development Superfund site in Brant,
New York, in May 1991. Three test runs were conducted during
the SITE demonstration, each 5V£ hours. The solid and liquid
locations that were sampled during each run were contaminated
feed soil, treated soil, combined flue gas cyclone fines and
baghouse dust, preheat vapor cyclone fines, scrubber liquor,
condensed water before and after treatment, vapor scrubber oil,
'and preheat oil. The noncondensed preheat and retort off-gases
were also sampled during each run.
Laboratory analyses included analyses of the solids and liquids
for PCBs, dioxins/furans, VOCs, and semivolatile organics
(SVOCs) to determine the PCB removal efficiency of the proces-
sor, the potential degradation products of the PCBs, and the
potential formation of dioxins and furans. Total chlorides and
total organic halogens (TOX) were also analyzed in an attempt to
trace the fate of chlorine throughout the system. In addition, a
variety of other parameters were analyzed to characterize the
feed and treated soils.
Key findings from the Wide Beach site demonstration are sum-
marized below:
• The SoilTech ATP unit removed PCBs in the contaminated
soil to levels below the desired cleanup concentration of 2
ppm. PCB concentrations were reduced from an average
concentration of 28.2 ppm in the contaminated feed soil to
an average concentration of 0.043 ppm in the treated soil.
• The SoilTech ATP does not appear to create dioxins and/or
furans.
• No volatile or semivolatile organic degradation products
were detected in the treated soil. There were also no
leachable VOCs or SVOCs detected in the treated soil.
• No operational problems affecting the ATP's ability to treat
the contaminated soil were observed.
For Further Information:
EPA Project Manager:
Paul R. de Percin
U.S. EPA Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, OH 45268
(513) 569-7797 (FTS: 684-7797)
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/MR-92/008
1291
-------
-------
Appendix 7-D
Case Studies on Chemical Treatment of Contaminated Soils: APEG
Site Demonstration of the SoilTech "Taciuk" Thermal
Desorber, Waukegan Harbor, United States
1293
-------
SITE DEMONSTRATION OF THE SOILTECH "TACIUK"
THERMAL DESORBER AT WAUKEGAN HARBOR
Paul R. de Percin
Demonstration Section
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 West M.L. King Dr.
Cincinnati, Ohio 45268
INTRODUCTION
The Outboard Marine Corporation (OMC) operates a
recreational marine products manufacturing plant on the western
shore of Lake Michigan in Waukegan Harbor, Illinois. From 1961
Un? ^?he.early 1970s, OMC used a hydraulic fluid containing
polychlonnated biphenyls (PCBs) in its operations. in the
process of plant cleaning operations, large quantities of PCBs
escaped into Waukegan Harbor and onto OMC property. Over one
million pounds of PCBs were estimated to be contaminating the
harbor sediments and property.
In 1984, the U.S. Environmental Protection Agency (USEPA)
signed a record of decision (ROD) authorizing site cleanup;
however, litigation between OMC and the USEPA suspended •
implementation of this Decision. In 1988, OMC and the USEPA
signed a consent decree, which specified the final terms of the
cleanup. Generally, OMC is required to excavate or dredge all
PCB contaminated soils and sediments, treat some soils and
sediments, and place all treated and untreated solids in
containment cells.
1 i'
Beginning in 1991, OMC initiated -the .cleanup of the PCB
contaminated sediments and soils. Based on the 1988 Consent
Decree the following actions were taken (Figure 1): '
1) A new ship slip (#4) was constructed.
1294
-------
2) Ship slip #3 was isolated from the upper harbor by
construction of a double-walled sheet-pile cutoff wall
and slurry wall.
3) Two other (East and West) containment cells were
constructed.
4) Ship slip #3 is now being dredged and the sediments
dewatered.
Those solids with concentrations greater than 500 ppm PCBs
are to be treated by the SoilTech "Taciuk" thermal desorption
process. The treatment requirements are 85% removal or 500 ppm
whichever results in the lower final PCB concentration. The
treated solids and any solids with an initial concentration of
less than 500 ppm will be placed in one of the three isolation
cells (i.e., ship slip #3, and East and West containment cells).
The SoilTech "Taciuk" process is being set up in the West
Containment cell. Contaminated sediments will be pumped to this
cell and dewatered before treatment. Contaminated soils will be
trucked to the cell.
\_y ,
EAST
CONTAINMENT
CELL
..RELOCATED..
•-' PARKING —
:: FACILITY ::
LAKE
MICHIGAN
Cleanup of
OMC/Waukegan
Harbor
Superfund
Site
FIGURE I
• n
SILT ]
CURTAIN
1295
-------
SITE DEMONSTRATION PROGRAM
Before final approval is given by the USEPA and the State of
Illinois a "proof-of-process" test of the Taciuk process is being
required. This test is a full-scale demonstration of both -the
treatment effectiveness of the process, and the gas treatment and
air_pollution control systems. Treatment effectiveness is '.
defined^as the PCB residual level, percent removal of PCBs, and
identification of any contaminants created by the thermal process
and left in the soil or sediments. Contaminants removed from the
soils and sediments are supposed to be recovered in the gas
treatment and air pollution control systems. Effectiveness of
the gas treatment and air pollution control systems is defined by
the mass and emission rate of PCBs, organics and dioxins/furans
emitted from the stack.
The test of the Taciuk process is tentatively scheduled for
January/February 1992. A SITE Demonstration will be performed
during this "proof-of-process" test to provide supplemental
information to the USEPA Region V (Chicago) office and to develop
the detailed information necessary for other regulatory offices
to consider use of thermal desorption technologies. The SITE
Demonstration includes sampling to meet the objectives of the
"proof-of-process" test previously described, and additional
evaluations required by the SITE program.
All SITE Demonstrations are designed to determine the
technology treatment effectiveness, field (mechanical)
reliability and operating cost. How these objectives are
measured depend on the technology. Table 1 lists sampling
required to meet these general SITE objectives for the five
thermal desorption SITE Demonstrations to be performed in 1991
and 1992.
WIDE BEACH SITE DEMONSTRATION
In May 1991 a SITE Demonstration was performed usjing the
SoilTech "Taciuk" process at the Wide Beach Superfund site in
Brant, New York. Dirt roads of this community were contaminated
with PCBs when used oil was applied to control dust problems. A
combined process of thermal desorption (Taciuk) and
dechlorination (APEG) was used to treat the contaminated soils.
Preliminary results from thi£ demonstration are:
1) Greater than 99 percent PCBs were removed from the soil
and residual levels were less than 2 ppm,
2) No dioxins or furans appeared to be created, and
3) No volatile or semivolatile organic degradation
products were detected in the treated soil.
1296
-------
TABLE 1
SITE PROGRAM
THERMAL DESORPTION SAMPLING
Treatment Effectiveness
1) Measure the4 contaminant removal efficiency aM residual
levels,
2) Measure the air emission rates,
3) Determination if dioxins, furans or other toxics are being
created by the thermal process, and the fate of these
compounds,
4) Determine if the physical properties of the soils and
sediments are being changed (e.g. increased leaching),
5) Evaluate and compare the toxicological properties of
contaminated and treated soils,
6) Evaluate the physical (stability) condition of the treated
material and its possible future use, and
7) Determine the treatment efficiency of the condensed water.
Field Reliability
1) Document operating conditions of process and identify
potential operating problems, and
2)
Determine percent on-line during operating period.
Operating Costs
1) Document the capital and operating costs of the remediation.
1297
-------
-------
Appendix 8-A
Microbial Treatment Technology Case Studies
Aerobic/Anaerobic In Situ Degradation of Soil and
Ground Water, Skrydstrup, Denmark
1299
-------
Biodegradation - Chlorinated Aliphates
A SUMMARY OF
Research and Demonstrations at The Skrydstrup Site.
Denmark
1300
-------
0 100 200 300 U)0 SCO m.
SI GKRTURFORKLARING ;
• FBI Undersegelsesboring nr.
5 og koncentration
\ Privat boring med nr.
/gj} Kemikalieaffaldsdepot
^rijSI Forureningsfane
Isokoncentrationslinie
med koncentra'tion i ug/1
Fig. 1.1 The Skrydstrup Site and the location of the pollution plume.
1301
-------
Biodegradation of Chlorinated Aliphates.
A Summary of Research and Pilot Tests at The Skrydstrup Site.
The Municipality of Vejens, The County of S0nderjylland, Denmark.
(ed. Neel Str0ba5k, M.Sc., Danish EPA).
1. The Skrydstrup site.
The Skrydstrup Site is a former landfill, where chemical waste from a refrigerator manufactory
was dumped during the period from 1963 to 1974. A gravel pit was used a dumping site, and
the waste was left without any form, for protection against percolation. This has caused a
considerable contamination of the groundwater by chlorinated solvents and organic phosphoruos
compounds.
Investigations and clean up's have taken place, and today, the waste is removed and a full scale
plant for a pump and treat remedy to clean up to groundwater have been in operation since '89.
^ progress of the activities in connection with the Skrydstrup site was phased as follows from
a wish to place the decision-making at the lowest possible level:
Phase 1
Phase 2
Phase 3
Phase 4
Phase 5
Orientation survey
Boring inspection
1. Draft project for excavation of the disposal site
2. Investigation of spread of contaminant
3. Testing methods for cleaning of groundwater
1. Excavation of the waste disposal site
2. Draft project for cleaning of groundwater contamination
3. Supplementary investigations of spread of contaminants
Establishment of groundwater cleaning plant.
= The waste disposal site was excavated in 1986, and by that time, no treatment plant for the
additional excavated soil was in operation in Denmark. The soil was, therefore, re-deposited in
a new controlled depot on the site.
The treatment plant for contaminated groundwater consists of 3 treatment units: An airstripper (3
plates in series) for removal of volatile alip'hates, quartz-filtration (pre- and after filtration in
columns) to remove iron and manganese and active carbon filtration (3 columns in series) for
removal of the organic phosphorous compounds. The cleaned groundwater is discharged to a
stream app. 1,5 km from the site, even though, it can meet the Danish drinking water standards.
1302
-------
p-
p-
p-
Rfivandsanlceg
38m3/h
Skyllevands
beholder
Inka-beluftning
3 trin- serie
Mellem- I T
pumper P -Jf
/ Kvarts \
(for) hi
\ filter J
Bortpumpning til fi
Grundvand i Skyllevand.kvarts
Slamvand ; Skyllevand.AC
Fig. 1.2 Process diagram of full-scale groundwater treatment plant.
1303
-------
As the remedy at the site contains unsolved problems in forms of the diposited contaminated soil,
and the airstripping, which only remove the contaminants from the water to the air, research and
development projects have been carried out in order to look into biological methods as alternative
remedies.
This paper summarizes the results from the projects concerning biodegradation of the
contaminated soil in the on-site depot and from the on-site pilot plant for biodegradation of
chlorinated aliphates in groundwater. As an introduction, a brief summay of the results of the
laboratory experiments is presented in order to give a review of the framework for the work done
The Danish EPA has laid down the funding for the laboratory experiments as well as for the pilot
demonstrations. The work is primary carried out at The Technical University of Denmark,
Department of Environment Sciences, supported by the Department of Microbiology at the
University of Copenhagen together with the Danish Environmental Survey and Kemp & Lauritzen
A/S, Consulting Eng.
The following chapters are a translation of parts of the report: "Biologisk nedbrydning af
klorerede opl0sningsmidler. Projekter gennemf0rt ved Skrydstrup Specialdepot. Rapport Al 1991"
(ed. Jens Aamand og Kim Broholm). , .
2. A summary of Laboratory Experiments.
Some laboratory batch and column experiments have been carried out to examine the possibilities-
of carrying out in-situ biodegradation of chlorinated aliphates in the unsaturated soil The
experiments showed, that it was possible to build up a methane-oxidizing biomass which was able
to degrade tri-chloroethylene (TCE) and 1,1,1-tri-chloroethane (1,1,1-TCA) but not tetra-
chloroethylene (PCE). The column experiments revealed that stripping due to the addition of
methane and air was a significant removal process in columns without recirculation of the injected
air. In practices it is difficult to avoid stripping, which means that the biological processes have
to be optimized before the technique will be useful. '
Laboratory batch experiments have been carried out to examine the degradation of chlorinated
aliphates in groundwater.
Aerobic degradation experiments showed that the degradation of TCE and methane in the
groundwater at Skrydstrup was slow compared to the degradation of TCE and methane in a
mineral medium. In opposition to that, the degradation rate of TCE was increased 88 times when
bacteria grew in absence of copper. It is unknown whether the natural concentration of copper
in the groundwater at Skrydstrup results in copper limited growth or not.
The groundwater at Skrydstrup contains high concentrations of chlorinated aliphates. Therefore
the toxicity of 1,1,1-TCA and TCE towards a mixed culture of methane-oxidizing bacteria was
examined in batch experiments. The consumption of methane was inhibited even by small
concentrations of 1,1,1-TCA and TCE. A total inhibition of the methane consumption was
observed at a TCE concentration of 13 mg/1, whereas a total inhibition was not observed at a
1304
-------
\ ««M4MWMa •*
0 10 .«) JO 40
Fig. 2.1 The Skrydstrup Site with the controlled soil depot.
1305
-------
1,1,1-TCA concentration of 103 mg/1. Experiments also showed that the presence of methane
inhibited the degradation of TCE. A. mathematical model describing the growth of bacteria and
the degradation of methane and chlorinated aliphates is proposed. The model was able to simulate
the observed degradation curves for methane and TCE but not for methane and 1,1,1-TCA.
The use of methanol instead of methane as primary substrate for the methane-oxidizing bacteria
was examined. The experiments show that methanol did not stimulate the degradation of TCE
probably because methanol does not induce methane monooxygenase.
Anaerobic .biodegradation of chlorinated aliphates have been investigated with groundwater from
Skrydstrup, groundwater sediment from the site M0rkh0j Bygade, and methanogenic sludge from
different wastewater treatment plants. No degradation of PCE was observed in the groundwater
from Skrydstrup, probably because of the high redoxpotential in this reservoir. The reservoir
contains high concentrations of oxygen, nitrate and sulphate. The anaerobic degradation of
chlorinated aliphates depended on the presence of a primary substrate.
Laboratory experiments show that the primary substrates resulting in formation of hydrogen
sustained the fastest degradation of PCE.
3. Biodegradation in The Soil Depot
In connection with the remedies at The Skrydstrup. Site, a separat soil depot was establish in
1986. In the depot excavated contaminated soil and slag from the original waste site were
deposited. The depot was established with an anaerobic compartment (section 1), and an aerobic
compartment (section 2). Tests for biological degradation was carried out.
3.1 The Soil Depot.
The anaerobic compartment contains app. 3.000 m3 of contaminated soil. The aerobic
compartment contains app. 500 m3. The soil was placed on a HPDE liner, and the anaerobic
compartment was covered with a LPDE liner to secure anaerobic conditions in the compartment
In the two sections, waste water sludge was build in, to secure an anaerobic and aerobic activity
respectively. In each section, separate leachate collection systems and recirculation-facilities was
established. The leachate from the two sections is separately collected. In the anaerobic section
the leachate has been recirculating since the beginning of 1987 by a rate of 100 % and without
contact to the surrounding environment. j
The leachate from the aerobic section is recirculating through a storage tank, with contact to the
surrounding environment Dependent of the llocal precipitation and evaporation from the section
the storage tank is emptied.
1306
-------
Methaner
Ethaner
Ethener
diklormethan
triklormethan
tetraklormethan
klorethan*
1, l-diklorethan*(l, 1-DCA)
1,1,1-triklorethan (1,1,1-TCA)
vinylklorid*
1, l-diklorethen*(l, 1-DCE)
cis-l,2-diklorethen*(cisl,2-DCE)
trans-l,2-diklorethen*(trans-l,2-DCE)
triklorethen (TCE)
tetraklorethen (PCE)
Table 3.1 Scheme of analysis for chlorinated aliphates in the soil depot.
* means that the compound may be a degradation product of 1,1,1-TCA, TCE or PCE.
3500
3000 --
A 2000 4-
0
1-7-86
1-7-87 30-6-88 30-6-89
Dato
30-6-90
30-6-91
Fig. 3.2 A time dependend describtion of the occurrence of contaminants in section 1.
1,1,1-TCA(8), TCE(a) and PCE(f).
1307
-------
From Dec.'86 the concentrations of chlorinated aliphates and their degradation products are
monitored by sampling and analysis of the leachate. From 1987 exist several sampling reports,
while in the later years 3-4 sampling schemes have been carried out each year. All leachate
samples are analysed in a scheme including inorganic parameters, chloro-alkyl-phosphates and
chlorinated aliphates, as shown in table 3.1.
The methods of analysis are head space GC with a electron capture detector (GC/ECD), and for
vinyl chloride by GC with a mass spectrometer as a detector (GC/MS).
Besides the chemical analysis monitoring, a microbiological investigation in both sections was
carried out in 1990. The aim of the investigation was to determine the microbial activity in the
contaminated soil.
Soil samples from different depths in the two sections were taken to the laboratory, and the
number of bacteria was determined by direct counting. The rate of growth was determined by
incorporating titrated thymidine in bacterial DNA under anaerobic and aerobic laboratory
conditions. Furthermore testing of the ability of the bacteria to transform PCE and TCE under
laboratory anaerobic and aerobic conditions was carried out.
3.2 Anaerobic Compartment (Section 1)
At the time of establishing, it was estimated, that section 1 in total contained 109 ke of 1 1 1-
TCA, 31 kg of TCE and 0,6 kg of PCE. '
The first leachate was sampled Dec. 22, 1986, while the latest sets of samples dates to Nov. 20,
1990. In fig. 3.2. the connection between nieassured concentrations in leachate samples and the
dates of sampling is illustrated.
In the evaluation of the concentrations of the contaminants in the anaerobic section, the closed
conditions in the section, and the recirculation of the leachate must be taken into account.
Assuming the leachate is recirculated 100 % anaerobic, the very small concentrations of
contaminants in the beginning of the monitoring period, probably can be explained by the lack
of equilibrium concerning the partition between the soil and the leachate. Later in the first year
(1987) a substantial increase of especially i;i,l-TCA and TCE is seen. The increase is followed
by a decrease, which for 1,1,1-TCA constitutes app. 90 % of the highest value in 1987. For TCE
the decrease is app. 80 % of the highest value in 1987. The concentrations of these two
contaminants have in the later years been app. 50-250 microgram/1.
Looking at PCE, the concentrations have been low and at changing levels in the monotoring
period.
In fig. 3.3 the connection between meassured concentrations of degradation products and the dates
1308
-------
2500
1-7-86
1-7-87 30-6-88 30-6-89 30-6-90 30-6-91
Dato
a 1 3 A time dependent! describtion of the occurrence of pressumed degradation products in section 1.
' and 1,1-DCE(*).
0 -
. 0
50
100 150
Dybde (cm)
200
250
Fig. 3.4 Microbial activity determined by incorporating radioactive marked thymidine and re-calculated to cells produced per gram soil
dry matter per day. Section l(n) and Section 2(*).
1309
-------
of sampling is shown. At an early stage in the monitoring period, high concentrations of 1,1-
DCA are seen. This level of 1,1-DCA is re-found in general in the monitoring period; Some
months later in the same year (1987) as well 1,1-DCE as cis-l,2-DCE are detected The
concentration of the latter is increasing by time. Twice-in 1988, vinyl chloride is found in
concentrations approximately at the detection limit of 1 microgram/1.
In comparing the decrease of the concentrations of the original contaminants 1,1,1-TCA, TCE
and PCE with the formation of the pressumed degradation products 1,1-DCA, 1,1-DCE cis-1 2-
Dra and vinyl chloride, there is a reason to believe, that microbial anaerobic transformation'of
chlorinated aliphates have taken place in.the section.
The results of the microbial investigations in the anaerobic section show, that in that section a
relatively large and active bacteria population is present, especially in greater depths (190 cm
beneath the top liner), ref. to fig. 3.4. Laboratory tests have shown, that the microorganisms are
able to transform PCE under anaerobic conditions. The transformation velocity is dependent of
the surrounding conditions, f.ex. the presence of primary substrates and essential nutrient.
There is concordance between the occurrence of degradation products and the presence of
anaerobic active bacteria in the anaerobic section of the soil depot.
3.3 Aerobic Compartment (Section 2)
When die depot was established, it was estimated, that the aerobic section in total contained 02
kg of 1,1,1-TCA, 0,1 kg of TCE and less than 0,01 kg of PCE. Afterwards, there has been
doubt of the correctness of this estimation, while later measurements of concentrations in leachate
from the section indicates, that the start concentrations probably have been higher. As mentioned
beforehand, the section has no top liner, and is therefore flushed by precipitation, and evaporation
from the top soil takes place.
In fig. 3.5 the connection between meassured concentrations in leachate samples and dates of
samphng is illustrated. ;,
As mentioned, the leachate is recirculating in the aerobic section via a storage tank in contact
with the atmospheric air. .-.>- ^
Taken into account the probable flushing and evaporation, it is difficult to evaluate the changes
in concentrations of the contaminants in the section. '
For 1,1,1-TCA it is seen, that the concentrations in the first year (1987) are of the same order
of magnitude as for the anaerobic, section, with a maximum of app. 3.000 microgramft In the
later years the concentration has been app. 100 microgram/1, which is about hatf'of the
concentration in the anaerobic section.
Concentrations of TCE have in the monitoring period been lower than in. the anaerobic section.
1310
-------
3000
o
I
, 1500-•
1000--
500--
1-7-86
,iM *
\SL ^**«r
^*=^>
1-7-87
30-6-88 30-6-89
Dato
30-6-91
Fig. 3.5 A time dependend describtion of the occurrence of contaminants in section 2
, TCE(») and PCE(«J).
0
^7-86
1-7-87 30-6-88 30-6-89
Dato
30-6-90
30-6-91
Fig 3.6 A time dependend describtion of the occurrende of pressumed degradation products in section 2.
l,l,-DCAg», cis-U-DCEW and 1,1-DCE»
1311
-------
C0ncentration of app' 50 % of ** hi8hest val"e is seen. The same tendency applies
In fig. 3.6 the connection between meassured concentrations of degradations products and dates
or sampling is shown. .
in
m
n- H c beginning °f 1987' ^ concentration amounts
period only app. 15 % of what is found in the anaerobic section. Approximately
at the same time, 1,1-DCE is found in concentrations in the same order of magnitude as in the
is f°Und' but in verv low concentrations compared to the
tvK- de
-------
pH
Ledningsevne
Calcium
Magnesium
Natrium
Kalium
Ammonium
Jem
Mangan
Kobber pgCu/1
Bicaarbonat
Klorid
Fluorid
Sulfat
Nitrat
Nitrit
Fosfor, total
Optest ilt
1,1-DCA
1,1,1-TCA
1,1-DCE
TCE
PCE
mS/m
mgCa/1
mgMg/1
mgNa/1
mgK/1
mgNH«/l
mgFe/1
mgMh/1 . ;
mgHCOj/1
mgCl/1
mgF/1
mgSO«/l
mgNOj/1
mgNOj/1
mgP/1
mg
-------
In the aerobic section, a large portion of the contaminants are presumeably removed physically
from the soil medium to water arid air. Furthermore, different processes have been difficult to
control, f.ex. the partly anaerobic degradation of the contaminants.
In comparison with other remedy technologies, the biological anaerobic degradation is evaluated
as functioning, however, the technique is slow, and problems with toxic degradation products
can occur. ,-: . ••
It will be of great interest to maintain the monitoring program in both of the sections
supplemented by further microbial investigations, which should be related to the occurance of
organic and inorganic parameters and heavy metals.
4. On-site biodegradation - Groundwater >
4.1 Introduction
In connection to the established full-scale treatment plant (ahr-stripping, quartz- and AC-filtration)
a demonstration on biological on-site treatment was carried out.
The objectives were to investigate the aerobic biodegradation of TGE in a pilot-plant, secondary
to optimize the process. «»"<«>
The demonstrations have been divided into 2 phases. In phase 1,'tests were made with
downstream filtration in a column with antracite carbon as filter material, In phase 2 tests were
made with upstream filtration in a column with quartz as filter material. The following
presentation includes the results from phase 2. .
The perspectives of application of these processes are expected to be an alternative to the current
used air stopping, where the contaminants are removed from water to air.
4.2 Grundwater Quality
Table 4.1 shows the overall water quality at The Skrydstrup Site. The figures are based on
samples from 4 wells, which are used in the full-scale plant. It must be noted that the
concentrations of chlorinated aliphates in the raw water varies from well to well, dependent On
their location in the pollution plume.
The inflowing water to the pilot plant was taken after the first of the three airering units of the
full scale plant. In table 4.1 the water quality data are shown. The concentrations of 1 1 1-TCA
and TCE in the inflowing water is shown in! fig. 4.1.
1314
-------
Udtedreng til det fri
r 0.02- 0.2 tf/h
Tilledning fra \
hovedanlaeg >px
SIGNATURES:
(Hy Slangepumpt
^4 Rowrcgulering
(\) Hanomcter
^^S Rowmiltr
Vand»(cd(ftig
—— Vandaflednsts '
Lumiltdnmg
—— , Gai/luf Ufiedninj)
2 V2 V.
Hethan
iaha.
tuft
V^^x
Afro.luft
Fig. 4.2 Process diagram of Pilot Plant
1315
-------
4.3 On-site Pilot Plant
4.3.1. Configuration
Fig. 4.2 shows a process diagram of the pilot plant. The contaminated (oxidized) water was led
to a pre-filter (column) with quartz as filter material for removal of iron and manganese. From
the pre-filtration unit equivalent dicharge runs through the two test-filter columns, in the following
named biofilter and controlfilter, respectively. The.plant was primary made of glass and stainless
steel. Packers and tubes of viton. Methane is added to the biofilter as primary substrate for
methane-oxidizing bacteria. The methane (2,5 % (vol/vol)) was mixed with synthetic air (21 %
(yol/vol) oxygen; 79 % (vol/vol) nitrogen). To the controlfilter an equivalent amount of synthetic
air was added. In the tests, upstream filtration was used, where gas (methane/air and clean air,
respectively) were added the test filters with the waterstream, ref. to fig. 4.2).
4.3.2 Filter and inoculation material.
Quartz with a grain size of 2-3 mm was used as filter material. The porosity was 0,4 and the
surface of the material was less than 1 m2/g.
As inoculum filtersand (quartz) from a pre-filtration unit from a public waterwork was chosen.
This waterwork treats groundwater from an aquifer which contains methane. Test made at The
University of Copenhagen showed 'that .this material contained a large population of methane-
oxidizing bacteria. The degradation potential of TCE was tested on several other media (lake- and
aquifer sediments) and the mentioned filter material was chosen due to a large potential and a
relatively small methane consumption in the degradation process.
In the demonstration, the filters contained only quartz from the waterwork filtration unit. The
material was thereby used as filter material as well as inoculum.
4.4 Operation
4.4.1 Running-in
The running-in of the process take place the first month of operation. In mat periode the
methane-oxidizing bacteria was multiplied. During the period the methane consumption in the
biofilter was meassured. The consumption was used as an expression of the biological activity
in the filter. • •
4.4.2 The Demonstrations
To investigate the influence of the residence time on the biological activity and on the
degradation of chlorinated aliphates, tests was made with different residence times (Th). In these
1316
-------
BIOF1LTRETS METHANFORBRUG
too
Alternerende drift-
Th
H—
•o
^c
H_
O
!!
o>
l_
l_
o
.C
:>
90-
80 -
70 -1
60 -
50 -i
40 -
30 ^
20 -
10-
Th <* Hime ,Th « Ztimer , Th <* 3.5timer ,
/^^ th
/ \ -»-"&~|
7 *
^.'.^-
^,x'*
,
: '. ^ ,..,.., : ,
i v i
•a
o
Fie 4 3 Methane consumption in biofilter(ti>, fixed relatively to the added methane. Changes in inflow of water(+) and the thereby
resulting hydraulic resistence time in the biofilter (Th=l, Th=2,Th=3 and Th=4). Gas to water rate is app. 1.
METHANFORBRUG OG OPHOLDSTID
S
•a
E
o
£
V
100-
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
0 -
0
-"•tr"°
. n^.--'
"<<""
a »-*****"
V*
a '•
1 1 \ i • ' ! ' ~ln
.0 1.0 2.0 3.0 4.0
Opholdstid i timer
Fig. 4.4 Methane consumption in biofilterffiD as a function of the hydraulic residence time (Th). The methane consumption is fixed
in relation to the added methane. Furthermore a regression line is shown.
1317
-------
tests, the gas to water rate was nearly constant at app. 1.
Furthermore tests were made with alternating addition of methane, by means of excluding and
including methane in the airsfream to the biofilter. The waterflow through the filter was kept
unchanged. The aim of these tests was to investigate the degradation of chlorinated aliphates by
the methane-oxidizing bacteria without methane addition. Methane is a competitative inhibitor for
the degradation of chlorinated aliphates. It was therefore expected, that the degradation of
chlorinated aliphates would increase immediate after methane stop at the biofilter.
4.4.3 Methods of Analysis
The concentrations of methane in the air disharge from the biofilter were meassured on the site
OR-sensor URAS 10E from Hartmann & Braun). The concentrations of methane in the water
disharge from the biofilter was analyzed by a GC with a flame ionization detector (GC/FID)
Concentrations of chlorinated aliphates in samples from inflowing and outflowing water from the
biofilter and controlfilter respectively, were extracted by pentane and analysed by GC with a
electron capture detector (GC/ECD). Concentrations of chlorinated aliphates in the outflowing air
from the biofilter and the controlfilter were collected on activated carbon in glastubes. The carbon
was eluated by sulphur/carbon and analysed by GC with a flame ionization detector (GC/FID).
4.5 Methane Consumption
In fig. 4.3 the methane consumption in the biofilter is shown. The consumption is shown
relatively to the ingoing quantity of methane to the biofilter. In the figure the quantity of
inflowing water to the biofilter is shown as well as the order of magnitude of the hereby resulting
hydraulic residence times (Th) in the biofilter. The hydraulic residence time is calculated as the
residence time of water in the test columns without taken into account, that the gas "takes up
space , and thereby reduces the real hydraulic residence time. From the figure it can be seen that
the methane consumption increases, as expected, when the hydraulic residence time increases.
In fig. 4.4 is the methane consumption in per cent drawn up as a function of the hydraulic
residence^ time. It must be noted, that the first 3 measurements, of the curve of methane
consumption in fig. 4.3 dates back to the period of tunning-in, ref. to sect 441 These
meassurements are not included in fig. 4.4. A linear regression a made on the per cent'data of
the methane consumption, and the regression line is included in fig. 4.4.
The-gradient of the regression line differs significantly from 0 on all levels greater than 0 1 %
m an inteval for the residence time on 1-4 hours. There is thereby stated a connection between
the residence time and the methane consumption. Li the latter part of the tests, the methane
consumption decreases to 0 % at two times (Two periods where methane are excluded the
inflowing aur). In this case, as shown in fig. 4.3, a trial is made on alternating operation on
methane, ref. sect. 4.4.2). By re-adding methane, a methane consumption on the same level as
before the addtion was stopped, is seen. It is therefore evaluated, that there is no effects on the
biological activity in the biofilter, when excluding methane in shorter periods of time
1318
-------
TRICHLORETHYLiN (TCE)
110
100
_c
~o
c
'E
a.
o>
o
•o
l_
.Q
e>
c
<0
o
40
30
20
10 -
i
°2&09 09*. 10 23'. 10 06ll1 20111 04112 18I12
Date 1990 - 1991
OS'.Ol'
Fig. 4.5 Break-thiough(a) and Break-through+Stripping@) of TCE in the biofilter, Break-through(+) and Break-through+StrippingW
of TCE in the controlfflter in test with the residence time(Th). The Break-through is calculated as the amount of TCE,
removed from the filters by the outflowing water. Stripping is calculated as the amount of TCE, removed from the filters by
the outgoing air. The residence time is changed according to fig. 4.3.
o>
•o
50
40-
6 30-
20-
10-
-10.-
-20
0.0
NEDBRYDNINGSPROCENTER OG OPHOLDSTID
n
D
—T
1,0
—I 1
2.0
Opholdstid i timer
3.0
4.0
Fig. 4.6 Relatively degradation of l,U-TCA(b) and TCE« in fte biofilter as a function of the hydraulic residence time(Th).
Fuithermoore regression lines are shown.
1319
-------
4.6 Results from Tests with Residence Times.
In fig. 4.5, die break-through and stripping of TCE in the biofilter and controlfilter is shown for
different residence times. Break-through is calculated as the amount of TCE leaving the biofilter
and the controtfilter, respectively. Stripping is calculated as the amount of TCE in the outgoing
air from the biofilter and control filter respectively. Break-through and Break-through plus
Stopping is related to the ingoing amount of TCE to the filters. The curve of Break-through plus
Stopping is at the level at 100 %, which shows, that a valid mass balance for the filter can be
^nt6*! Vlf dlfferenc.e ? *e values for Break-through plus Stopping for the biofilter and the
control filter, respectively in evaluated as an expression of the degradation of TCE in the biofiUer
Degradation m per cent). For 1,1,1-TCA the per cent of degradation is calculated in the same
manner.
I[ 8' thercalcu! for M.1-TCA differs not significantiy from 0 on levels less than
11 f'™ennore'ls *e lme aro^d 0 %. It is evaluated, that no significant degradation of
iorfflJjrVi f-t^™ Ilacenm *?, bi°filter in ^ test Period For TCE *» regression line is
sigmficandy different from 0 on all levels greater than 1 % in an interval between 1 an 4 hours
inere is thereby stated a relationship between the residence time and the degradation of TCE.'
In fig. 4.7 the degradation per cents, re-calculated to absolute values (in microgram/hour) are
shown as a function of the methane consumption, also re-calculated to absolute values (in
mg/hour). A linear regression is made on the absolute values for the relationship between the
absolute values of degradation and the absolute values of the methane consumption In fig 47
are the resulting regression lines shown. Furthermore, a linear regression is made on the relation
between the degradation per cents and the methane consumption per cents. For these analyses
of regression similar results are obtained, as for the hydraulic residence time. This is however
not supnsmg, as a linear relationship between methane consumption and hydraulic residence time
is stated, ref. to sect. 4.5.
By the regression analysis, the following linear relationship, between the absolute values of the
methane consumption and the degradation of TCE is found:
Degradation of TCE in microgram/hour = 0,7 x methane consumption in mg/hour.
The linearity is found in an interval of app. 50-200 mg/hour for the methane consumption.
Looking at the results of the tests, it is not possible to evaluate, if an increased degradation of
TCE primary can be achieved by an increase in the residence time (time of contact) or by a
arger methane consumption (larger biomass). It must be expected, that as well a longer residence
time as well as a larger methane consumption will give an increased degradation of TCE
1320
-------
140
: J 80 -.
NEDBRYDNING ,OG METHANFORBRUG
180
200
Methonforbrug i mg/time
Fig 4 7 Degradation of 1,1,1-TCAfe) and TCE« in the biofdter as a function of the methane consumption. The methane consumption
g' Ranging by changing the hydraulic residence time(Th). Furthermore regression lines are shown.
NEDBRYDNINGSPROCENTER - ALTERNERENDE DRIFT, l.PERIODE
Drift uden
jnethantilledning
Normaldrift med methantiliedning
-20.
03.01
08.01 09.01
10.01
Dato 1991
Fig. 4.8" Relatively degradation of l,l,l-TCA(p) and TCE(+) i the biofflter in alternating operation. Methane addition is stopped on
1991-01-08.
1321
-------
The results from the tests of the residence time will to a certain extent be applicable when
dimensioning a full-scale plant It must, however, be noted, that for instance the results in fig.
4.6 can not be extrapolated the given interval of the residence time, if a higher per cent of
degradation must be achieved. This is due to the assumption of linearity, which may not apply
outside the interval. It is likewise relevant to note, that the tests are carried out in a given
interval of concentrations of chlorinates aliphates in the inflowing water, ref. to fig. 4.1 and table
4.1. The tests have not reveiled, to which extent the degradation of TCE is dependent on the
concentrations of TCE and other chlorinated aliphates in the inflowing water. Similar remarks
must be noted on the overall groundwater quality.
4.7 Results from Test with Alternating Operation.
In fig. 4.8 the calculated degradation per cents for 1,1,1-TCA and TCE are shown from a test
with alternating operation. The hydraulic residence time was kept constant at app. 3 hours. The
first two calculation points on the curve represent the degradation of 1,1,1-TCA and TCE,
respectively under testoperation with addition of methane immediate before a period without
methane addition.
The figure shows', that alternating operation does not have any significant effect on the
degradation of 1,1,1-TCA. Furthermore, it is seen, that the degradation of TCE all in all is
unchanged right after the stop in methane addition, and that the degradation is decreasing after
app. 24 hours without methane addition. The alternating operation has by thai: not showed a
possitive effect on the degradation of TCE under the test conditions. Similar results are found
in tests carried out in the US (Canter et al., 1990).
The decreasing degradation of TCE after app. 24 hours may be caused by the lack of an energy
source for the methane-oxidizing bacteria, when the methane addition is stopped. An effect of
alternating operation can therefore be determined by the addition of another energy source in
the period without methane addition. As energy source, it could be revant to apply methanole og
format^ (Semprini et al.,1991). The lack of; effect of alternating operating can furthermore be
determined by the low competitative inhibation, when methane addition in the operation. The
effects in stopping the methane addition are thereby smaU. It must be noted, that the alternating
operation have had an effect in relation to the methane consumption, while the degradation of
TCE keeps on for a while after stopping the methane addition. Compared to operation with
continues methane addition, a similar degradation of TCE can be achieved, by a smaller methane
consumption.
4.8 Conclusions
In the on-site tests, a methane consumption .of between 40 and 90 % was meassured. A linear
relationship between the hydraulic residence time and the methane consumption is found in an
interval for the residence time of 1 - 4 hours.
1322
-------
Neither in the tests with the residence time, nor in the tests with alternating operation, any
significant degradation of 1,1,1-TCA was seen. In contradiction 19 that, a substantial degradation
of TCE between 20 to 40 %' of the inflow was seen, which is equivalente to an absolute
degradation of TCE between 40 and 130 microgram/hour. Furthermore linearity is found between:
hydraulic residence time and degradation of TCE in an interval of 1 - 4 hours
for the hydraulic residence time.
methane consumption (in per cent) and degradation (in per cent) for TCE in
an interval of 40 - 90 % for the methane consumption of the inflowing water.
absolute values of methane consumption and degradation of TCE in an interval
of the methane consumption of 50 - 200 mg/hour.
By alternating operation the same degradation of TCE was seen in periods with and without
methane addition. The alternating operation has by that not showed a positive effect on the
degradation of TCE under the test conditions.
4.9 Perspectives.
It is shown, that an increased biomass will give a relatively increase in the degradation of TCE.
In the tests, the biomass may have been limited by the accessibility of nutrients (nitrogen,
phosphor, iron etc.). An addition of nutrients will therefor be a possibility to optimize the
utilization of the added methane, and by that be able to develop an increased biomass. The nature
and structure and surface of the filter material have likewise importance for the size of the
biomass.
The largest methane consumption in the systeme will probably take place at the the location in
the filter, where methane is added. Under operation conditions, where the growth of the biomass
is unlimited witout lack in accessibility of nutrients, a larger biomass can be obtained by applying
methane at different locations in the filter. As an alternative, filters in series could be more
succesful.
In laboratory test, it is shown, that a potential of increasing the degradation of TCE will be in
alternating operation (Broholm et al.,1991). In the on-site tests this has not been proven, perhaps
because the lack of an available energy source for the bacteria in the periods without methane
addition. A possibility for increasing the degradation of TCE could therefor be in the alternating
operation to apply f.ex. methanol or format in the periods without methane addition (Semprini
et al.,1991). It must, however, be noted, that an addition of f.ex. methanol or format may give
raise to other problems, for instance an unnessesary growth of heteretrofe bacteria.
In laboratory tests it is furthermore shown, that methane-oxidizing bacteria can degrade a
relatively larger amount of TCE under conditions, where their growth have been limited by the
availability of copper (Oldenhuis et al.,1989; Tsien et al.,1989). A possibility to increase the
degradation of TCE can therfore. be to etablish operation conditions, where copper is removed
1323
-------
from the contaminated water, before inflow in the filter with methane-oxidizing bacteria. It must
here be noted, that in the contaminated water at the Skrydstrup site, the concentrations of copper
have been so low, that this effect have been obtained.
In the tests it is shown, that it is possible to degrade TCE .in an aerobic biological process. It is
furthermore probable, that other chlorinated aliphates with a lower rate of chlorine can be
degraded in a similar way - among others chloro ethylene (vinyl chloride) (Vogel et al.,1987) The
degradation is running without an increase of toxic end products (Little et al.,1988).
1324
-------
Appendix 8-B
Microbial Treatment Technology Case Studies
In Situ Biorestoration of Soil, Asten, The Netherlands
No final text available.
1325
-------
-------
Appendix 8-C
Microbial Treatment Technology Case Studies
In Situ Enhanced Aerobic Restoration of Soil and Ground
Water, Eglin Air Force Base (AFB), United States
1327
-------
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
Mr Douglas C. Downey
HQ AFESC/RDVW
Tyndall AFB, FL USA
1328
-------
BIOIOGICAL AND PHYSICAL TREAdMQtt
A JET FUEIr
-------
and reapplication system has been operating for over one year and a
portion of the fuel residuals have been removed through a ccirfcirijation of
biodegradation, hydraulic washing, and aboveground aeration. Observations
on system operation and soil and ground water data from th€» initial 12
months of this field project are reported in this -oaper. Final sioil and
ground water sampling win take place in Sepbianber, therefore only
preliminary results were available before the publication deadline,
INTKODCCTICN
Each year the U.S. Air Force stores and transfers millions of gallons
of ^ JP-4 jet fuel at over 200 Air Force installations. Fuel leaks and
spills are by far the most "frequent sources of soil and groundwater
contamination on these installations. The Environics Division of the Air
Force Engineering and Services Center (AFESC), located at Tyndall
AFB, Florida, is responsible for developing and testing improved methods
of soil and groundwater decontamination. Soil decontamination has been
emphasized, because in most soils the majority of the JP-4 will rtsnain in
the unsaturated zone and slowly leach into the groundwater (Davis et al,
1972). These soil residuals represent the long-term source of ground
water contamination and must be addressed in site remediation.
Enhanced in situ biodegradation is an innovative technology which has
received much attention and research during the past two decades -('Lee, et
al, 1988). While several commercial firms have reported sucJcessful
ground water remediations, published results have lacked sufficient data
to ^ determine the effectiveness of biodegradation in reducirg fuel
residuals in the vadose zone. For this and other reasons, AFESC has
conducted independent field tests of this technology prior to reconmending
it for widespread Air Force application.
In 1984, AFESC initiated a pilot-scale test of enhanced biodegradation.
at> a site on Kelly AFB, Texas. As this test progressed, problems with
soil permeability were encountered reducing the delivery of hydrogen
peroxide and nutrients through injection wells. This reduction in
permeability was attributed to both natural silt and clay soils and the
precipitation of calcium phosphates which formed as injected phosphates
reacted with calcium in the son (Wetzel, 1987). Barmeabnity problems
reduced the delivery of oxygen and consequently little biodegradation
occurred. Based on these results, a second site with more favorable son
permeabnity was selected at Eglin AFB, Florida.
In 1986 AFESC awarded a competitive contract to EA Engineering,
Science and Technology and their subcontractor FMC Aquifer Remediation
Systems (now a part of IT Corporation) to conduct an enhanced
biodegradation demonstration on the site. EA Engineering, Science and
Technology was responsible for overall system design, operation and site
monitoring whne IT Corporation has provided microbiological support and
operational expertise in nutrient and hydrogen peroxide application
systems.
1330
-------
SlTK
In April of 1984 a leak was discovered in an underground jet fuel
pipeline in the Bglin AFB petroleum storage area. A preliminary site
characterization estimated 30,000-45,000 gallons of JP-4 jet fuel had
contaminated approximately 4000 cubic yards of soil and shallow aquifer
material (Weston, 1984). Follow up sampling in August 1985 decreased that
estimate to 20,000 gallons. A series of shallow, gravel-filled trenches
were installed perpendicular to fuel movement and skimmer pumps recovered
over 7000 gallons of free product. By early 1986, free product had been
removed to non-recoverable levels and skimmer pumps were turned off.
Prior to system design a complete site characterization was
accomplished to better define hydraulic, contaminant, and microbiological,
conditions. The site is located in an area of unconsolidated and
relatively homogeneous coastal sands that extend from the surface down to
40 feet. Below the sand, a 400-foot confining clay layer protects the
deeper Floridian Aquifer. The contaminated shallow aquifer is only three
to five feet below the surface across the site and has an average
hydraulic conductivity of 7 X 10-2 cm/sec. The highly permeable **ni« and
the shallow depth to contamination make this an excellent site for testing
enhanced biodegradation.
One of the primary objectives of this test was to insure that a
sufficient number of ground water and soil samples were taken to assess
the performance of this technology. An initial soil vapor survey was
performed to delineate the extent of free product influence and to assist
in monitoring well locations and application system design. A Ehotovac
10S50 portable gas c±romatograph was used to analyze soil gas samples
taken at a constant depth of two feet on a radial grid across the site.
Monitoring wells and soil sampling locations were selected on the
basis of the soil vapor survey. Initially 22 monitoring wells were
constructed and soil samples taken at one-foot depth increments at 12
locations. During the course of the research, the number of ground water
monitoring wells has increased to 40 and soil sampling locations to 16. In
order to compare natural degradation rates to degradation rates under
enhanced nutrient and oxygen conditions, monitoring locations were
established in a control contaminated area which received no nutrients or
hydrogen peroxide. The "control area" was a unique feature of this test
which is normally unacceptable for commercial site restorations.
Total Organic Carbon (TOG) was selected as the general indicator of
fuel
-------
monitored in the ground water. The list was expanded to 43 compounds to
racnitor the specific degradation rates of a broad range of alij±atic and
arcnatic fuel components. A complete discussion of GC/MS results is
beyond the scope of this paper. This data will be included in an AFESC
Technical Report to be published in early 1989. Figure One illustrates
the areas of free product influence prior to bicdegradation qaerations,
and the location of initial monitoring wells in relation to the basic
operation system discussed in the next section.
Microbial enumerations were also performed on soil and ground water
samples at most mcnitoring locations. Microbial monitoring provided a
measure of microbial response to nutrient and oxygen enhancements over
time. Soil samples at each mcnitoring point were collected at two depth
intervals, one from the capillary fringe zone above the water table and
another sample one foot below the water table. The microbial peculations
at these two depths would be most likely influenced by ground waters
enhanced with nutrients and oxygen. A sutanary of initial site conditions
within the contaminated area is included in Table One.
ABOVEGROUNDj |
TREATMENT
FREE PRODUCT
INFLUENCE
i4
•6
' '
, ; '3 N. /
:.''\* ,'
0 20 49
feel
O RECOVERY WELL
• MONITORING WELL
A INJECTION WELL
Figure 1. Bglin AFB Site Profile
1332
-------
TOG
TABLE CKE
Initial Site Conditions in Contaminated Area
Soil (mg/kg) Ground Water fmg/LI
126 (18)
Fe(total)
pH
Dissolve O-
Temp
Total Bacteria (105 CPU)
HC Degraders(10 CPU)
2000 (20)
110
640
850
2
.9
6
3
8
12 (<.5)
5.4 units (6)
<1 (3)
20° C
15
1.6
1 Average of contaminated monitoring locations
( ) = Uncontaminated background level
2 Total benzene, toluene and xylenes
3 Total alkanes detected on GC/MS
laboratory microbial enumerations were completed using a modification
of spread plate-count method for heterctrophic bacteria (AEHA, 1985).
Ground water samples were prepared for enumeration by performing serial
decimal dilutions of the sample in a sterile mineral salts broth.
Subsurface soil samples were prepared using a modification of the method
reported by Balkwill and Ghiorse, 1985. The method involves homogeniza-
tion in a Waring blender of five grams of soil in 50 mis of a solution
consisting of 1% sodium pyrophosphate, 0.1% polyvinylpyrrolidone-360
(F7P-360). This homogenization step is designed to facilitate release of
bacteria attached to particles. The homogenate is then diluted and
plated in the same manner as for water samples.
The concentration of total heterotrophic bacteria is defined as the
number of colony forming units per milliliter of ground water or per gram
(dry weight) of soil, that can form iiacroscopically visible colonies on
0.23% nutrient agar after one week incubation at ambient temperature and
oxygen. Hydrocarbon-degrading bacteria are defined as those capable of
forming colonies on carbon-free mineral salts agar when incubated under a
hydrocarbon-vapor atmosphere at ambient temperature for one week
(Jamison, et al., 1976).
1333
-------
Laboratory 'Ri
studies
One of the important tasks undertaken in the laboratory pheise of this
project was to conduct a bench-scale degradation study, using
contaminated soils and ground water from the site, to debarmine the
effects of nutrient and oxygen supplementation on the rate of cxaTtaminant
biodegradation, and to estimate the extent of contaminant removal that is
achievable biologically under laboratory conditions. These studies have
historically been based on batch reactors, or "microcosms," which consist
of soil/ground water slurries in sealed glass vessels. The microcosms are
then treated with the appropriate nutrient and hydrogen peroxide
amendments, and one set of microcosms is sterilized and/or inhibited with
a biological poison to account for physical/chemical mechcjnisms for
contaminant removal from the system. Aqueous samples can then be
periodically withdrawn for microbial, nutrient and/or csontaminant
analyses, or the entire reactor can be sacrificed for analysis. While
studies such as these provide a limited model of the natural aquifer
environment, the resultant information is useful in determining the
relative effectiveness of different nutrient types and concentrations on
biodegradation rates, and for demonstrating the heterotrophic potential
for contaminant destruction.
For reasons of simplicity and cost, most microcosm degradation studies
monitor aqueous phase contaminant biodegradation. For the Eglin project,
however, we attempted to quantify the degradation of the total amount of
organic material within the reactors. Microcosms were prepared be
combining 20 gms of contaminated soil and 20 mis of site ground water in a
40 ml VGA vial. Vials were tightly sealed and various nutrient
concentrations injected through a septum to determine their effect.
Excess oxygen was maintained through hydrogen peroxide additions. The
microcosms were separately analyzed; the aqueous fraction was extracted
with pentane and analyzed by capillary GC-FID. The soil underwent Soxhlet
extraction according to Standard Methods 503D, followed by analysis of the
extracts using capillary GC-FID.
A GC/F3D analysis of the aqueous phase showed that biodegradation was
enhanced by 25 ppm concentrations of Restore 375^ and that the total
extractable organics were degraded from 35 ppn to <.2 ppm in nine days.
This data confirmed that indigenous microbes would degrade soluable fuel
components with minimal nutrient additions if adequate oxygen was
available. Restore 375^ is a patented, prepared nutrient mixture
consisting of 50% ammonium chloride and a blend of disodium •phosphate,
sodium tripolyphosphate and monosodium phosphate.
Analysis of the solid phase (soils) exhibited considerable noise and
due to extraction problems actually showed an increase in organics. These
analytical procedures cannot be reliably used to measure biodegradation of
low solubility compounds adsorbed or occluded in soils. (We are currently
repeating the microcosm study using a modified purge and trap procedure to
measure organics removal from both the aqueous and soil phase. The
results of this improved procedure will be published in the f:Lnal AFESC
Technical Report.)
1334
-------
SXSEEM DESIGN
Tests
The ability of the shallow aquifer to produca sufficient water for
recirculation of nutrients and oxygen was a critical design factor that was
first determined, through a series of pump tests on all recovery and
injection wells. Four recovery wells were installed on the down-gradient
perimeter of the contaminated area using the cable tool drilling method.
The six-inch wells were constructed with 5 feet of .02 slotted PVC screen
located 5 - 10 feet below the static water table to prevent free product
drawdown and movement into the wells (Figure 2). Injection well design is
discussed later in this section.
From Polishing
Filter
V Stick-up
—6" PVC Casing
with bottom plug
— Centra 8zer
Grade
- Cement Grout
I to the Surface
-1—6- PVC
! Schedule 40
i Casing •
Bentonite Seal
Gravel Pack
I
6- PVC Casing
with bottom plug
Cantralizef
INJECTION WELL
RECOVERY WELL
Figure 2. Recovery and Injection Well Design
1335
-------
After initial installation, the wells were air-developed. All wells
were tested at maximum pump yield for three hours; this was sufficient to
determine the hydraulic parameters of the highly permeable aquifer. Water
levels were measured during testing using an electronic water level
indicator or a MK oil-water interface probe. Well yields were measured
with a flow meter. Ifantoring wells located at varying distances from the
pumped well were used as observation wells, m general, the treatment area
can be divided into two zones based on the observed and calculated data:
'1. a northern zone of higher elevation (wells INI, IN2, Rl, ard R2) with
t?5?T^^LtY/«feaE^:od3nately 2°'000 gpd/ft <250
of 270 ft/day (9.5 10 *• cm/sec)
2. a southeastern zone of lower elevation with
10,000 gpd/ft (125 mVd) and permeability of 130
cm/sec)
transmissivities of
ft/day (4.7 x ID"2
Tbe> lower transmissivity of well R4 may be related to an area of finer-
grained sediments deposited by an old stream channel.
Discussion of 02 Supply Requirements
Natural aerobic degradation at this site has been limited by depleted
oxygen in the groundwater and capillary zone. Background oxygen levels of
.5 - 1.0 ppm were measured in monitoring wells in the contaminated area.
She primary< objective of system design was the delivery of adequate oxygen
for mineraliztion of fuel hydrocarbons. Bie oxygen required was calculated
based ^on an estimate of the mass of hydrocarbons present and the
stoichiometric requirement for- complete mineralization. ^ihe basic
mineraliztion equations for typical hydrocarbons follow :
—+• 6C02 + 3H20 for benzene
or
12^02
8C02 + 9H2O for octane
Complete mineralization requires approximately 3.1 pounds of oxygen per
pound of benzene or 3.5 pounds of oxygen per pound of octane for complete
mineralization. Eased on our site characterization, it was estinated that
the 16,000 Ibs of hydrocarbon remaining in the treatment area's soil and
groundwater would require approximately 40,000 Ibs of oxygen for
mineralization. Assuming 100% oxygen utilization efficiency, the necessary
oxygen and corresponding peroxide I concentrations can be calculated for a
variety of pumping rates:
1336
-------
Peroxide
cmxsitration
tacr/Ll
0
100
300
500
Effective Oxygen
Ccaxsntration
facr/L)
8
50
150
250
Required Water
Volume
(gallons) (qgn/16 mo.
650,000,000
100,000,000
30,000,000
20,000,000
1100
170
55
33
Based on pump tests and O_ requirements > an initial peroxide cxncentration
of 500 mg/1 and a pumping^ range of 30-40 gpm were selected to deliver the
required O_ over the 16 month fixed contract length.
Iron Removal System
Iron concentrations in ground water were reported to be below 1.0 mg/1
(Weston, 1984). Our sampling found iron levels in contaminated ground
water in the range of 8 to 26 mg/1. The recovery well discharge water was
found to contain an average of 12 mg/1 of total iron. . JXje to highly
reducing conditions common in anaerobic, contaminated groundwaters, this
iron emerges from the wells in the ferrous (Fe+2) form. Oxidation to the
less soluble ferric (Fe+3) form dating aboveground treatment leads to
precipitation, and clogging of the injection systems. The (xncentration of
iron at injection points can also reduce the stability of the peroxide
(Britton, 1985).
In order to reduce iron concentrations in the injection systems a
sedimentation basin was added following the aerator. To further reduce
iron ccancentrations going into the injection wells a pressure filter was
installed. The combined sedimentation and filtration removed approximately
90 percent of the total iron entering the injection wells.
Three Application Ifethods
Three basic technologies are available for oxygen and nutrient
injection; injection wells, infiltration galleries and surface application;
each'with its own advantages and disadvantages. The injection well most
directly delivers the nutrients to ground water, however, nutrients are
poorly delivered to the unsaturated zone. The injection well has a
relatively small surface area and is therefore prone to clogging.
Infiltration galleries and surface application deliver the nutrients in a
more uniform pattern which allows more effective delivery to the
unsaturated zone. At many sites the drawback to infiltration galleries
and surface application techniques is the potential difficulty in
delivering nutrients to the saturated zone. If an impermeable or less
permeable strata exists above the groundwater it may prevent or limit
percolation. In the case of the infiltration gallery, this can be
overcome by placing the system beneath the less permeable strata. ^In
general, the injection well is the most expensive application method, with
the infiltration gallery being intermediate in cost and surface application
being least expensive, in difficult situations such as paved or developed
1337
-------
sites, or sites with a relatively deep, less permeable s-trata the
infiltration gallery can be substantially more expensive than an iniection
well.
As the ground water is qui|» shallow at the. Eglin site and no
significant stratification exists in the unsaturated zone or the shallow
saturated zone, surface application is the technique of choice. The site
however, is certainly amenable to all three technologies and due to the
research nature of this project all three were demonstrated. It was not
intended that three separate treatments be tested for remediation
efficiency, rather the relative effectiveness of each technology was
evaluated and each injection system modified as necessary to improve
nutrient and oxygen delivery. The application systems were <3esigned to
receive approximately 10 gpm each or approximately 6,000,000-7,000 000
gallons over the course of the study. '
Injection well cxrtstruction is shown in Figure 2. The wells were
screened from approximately one foot above ground water surfaoa to a depth
of nine feet below groundwater surface to anew uniform nutrient
distribution. The design diameter of six inches was considerably
oversized for the hydraulic needs of this site to provide a .safety margin
against plugging. The stainless steel screens anowecl vigorous
development, redevelopment and acid cleaning.
t Shanow, gravel-filled trenches were used as infiltration gzidleries as
inustrated in Figure 3. The large surface area of these ganeries
provided a broad distribution of nutrients and peroxide perperidicular to
flow lines. Surface application was the simplest form of nutrient and
peroxide distribution. The surface spray system provided an even
application of nutrient and oxygen over an 1800 sq ft area using a
sprinkler system. Figure 4 provides a schematic of the iaboveground
treatment and application system.
e
V
Native Material Backfill
Large Stona Backfill -
(1-mln.size)
4" PVC
riser -^^_^
—
18-
—
,— -
, •
... -
°°0
-G>°
J\
\J
sr^
'A
00
°o
12"min.
1" gate valve
,1-PVC 1
/ T 1- pun
-
vS I? '
6'
i Light-weight plastic sheeting
added in later installations
_^— 4" perforated PVC Installed
-—-"^ level and with perforations
in an 'UP1 position
Figure 3. Infiltration Gallery Design
1338
-------
Nutrismi
Injection
W«Ue
Aeration
Basin
2000 galena
Sedimentation
Basin
4000 gtltons
lOgpm
Figure 4. System Flow Diagram
Nutxient Addition
Nutrient addition was designed to be in excess of micriobal
requirements. laboroatory microcosm tests indicated that nutrient levels
as low as 25 ppm would support microbial degradation. 'Die decision to use
excess nutrient addition was based on three factors. First the phosphate
present in the nutrient solution could act to stabilize the peroxide
(Britton, 1985). Second, laboratory -column tests showed that
orthphosphate precipitation in site soils was inhibited in the presence of
sodium trdpolyphosphate at Restore 375^ injection concentrations of at
least 1000 ppm. Finally, an excess addition would insure that oxygen alone
was limiting biodegradation. One concern of excess nutrient addition is
the potential for precipitate formation such as calcium phosphates
resulting in reduced aquifer permeability. The potential for aquifer
plugging at the Eglin site was lessened due to the initial high
permeability.
Restore 3751* was a*M«ad on a batch basis of 150 Ibs/week and meterpd into
the recirculation flow to make a delivery concentration of 1000 ppm.
Nutrient solutions were added three times a week and each pulse was
approximately four hours. Typically, nutrients were added on Monday,
Wednesday, and Friday.
1339
-------
This application rate exceeded the
requirement for microbial
degradation of fuels. Assuming 16,000 Ibs of fuel hydrtcarbons and a
Carbon: Nitrogen: Ifcosphate ratio of 100:6:3, and that 33% of the
hydrocarbon was converted to biomass -(less than 33% would probably be
converted), nitrogen requirements would be 7.20 Ibs aid phosphorus
requirements 360 Ibs. Despite reduced oxygen delivery rates due to
peroxide stability problems, the design nutrient injection schedules were
maintained and to date 750 Ibs of nitrogen and 440 Ibs of available
phosphorus have been delivered to the site. Subsequent ground water
sampling across the site indicate that increased levels of P04 and m, were
well-distributed in the saturated zone.
In-Situ Peroxicfe Stability
Tests of in-situ peroxide stability and oxygen transport weare carried
out in the infiltration galleries. Due to the shallow ground water at this
site, mounding occurred in the vicinity of the galleries which resulted in
saturation to the ground surface. Shortly after peroxide injection was
initiated, gas bubbles were observed rising through the saturated sandy
soil immediately above .the galleries. ihis observation coupled with a
failure to observe peroxide in any down gradient wells (including EA 18
installed 1 ft down gradient of the injection galleries) lead to the
suspicion that peroxide decomposition was very rapid and resulted in off
wasteful gassing of oxygen. Initial laboratory batch studies rising a 3:1
ground water to soils ratio had indicated that a 500 ppm hydrogen peroxide
addition would have a half-life of approximately six hours in the Eqlin
aquifer.
A monitoring well (EA-22) was installed directly in one of the injection
galleries. While the system was operating under continuous peroxide
addition, peroxide concentrations observed in EA-22 were very close to, or
slightly below the 500 ppm measured in the feed lines to the galleries. To
evaluate in-situ peroxide stability, tests were conducted in whicii the flow
to the galleries was shut off and the disappearance of peroxide in EA 22
was measured over time. Peroxide half-lives, based on first-order decay
rates, were observed in the 30 to 90 minute range, far less than the six
hours predicted in the laboratory.
In order to determine the effect of phosphate pretreatment a new series
of galleries were constructed. These galleries received nutrient
solution, including phosphates, prior to initiating peroxide injection.
Similar peroxide stability tests were conducted in these "pretreated"
galleries, however, no significant increase in peroxide stability was
noted. Due to the significant off gassing and waste of oxygen at the 500
ppm peroxide addition rate, a decision was made to reduce the operating
concentration to 300 ppm.
Throughout the project several experiments have been conducted to
determine the cause of hydrogen peroxide dcomposition. Both biological and
inorganic reactions have been identified as contributing to this
instability. These findings are briefly discussed in the conclusions.
1340
-------
SYSHM EEHTCSMHKE
Withdrawal and Application Systems
During the initial weeks of operation, the four recovery wells produced
over 40 gpn of cxsntinuous flow. Over the past 15 months of operation, well
production has decreased to 32 gpra. Regular treatments with acid and
vigorous scrubbing have been required to clean both recovery and injection
well screens.
The total volume of water passing through each of the primary
application systems to date is:
Infiltration Galleries 6,500,000 gallons
Spray Irrigation 7,800,000 gallons
Injection Wells 1.000.000 gallons
Total 15,300,000 gallons
Ihe infiltration galleries and injection wells have shewn a tendency _to
clog and flow^tttcough has been reduced. In the case of the infiltration
galleries this problem was remedied by increasing their number. Initially,
4 galleries with 40 linear feet of injection area were installed and were
capable of taking the full 10 gpm. After surface flooding was noted three
additional galleries with an additional 30 feet of linear injection area
ware added. Finally four additional galleries were installed in an old
free product recovery trench which was approximately 40 feet long, 10 feet
wide, and was filled with approximately 4 feet of gravel (approximately 2
to 6 feet below land surface). Ihe galleries have received a continuous 10
gpm flow for over one year.
Flow through the injection wells was significantly reduced over time.
Efforts were made to acid wash and redevelop the wells by adding 1 to 2
gallons of 31% industrial HC1 and applying 100 ± gallons of water. The
wells were then allowed to sit for a 16 ± hour contact time. The wells
were then purged until the pH returned to 6.0 or greater. At times the
wells were also air surged and/or vigorously brushed. Despite the cleaning
and redevelopment the wells continued to clog and within a few months were
only able to accept only 1 ± gpm. For this reason the total volume
introduced through the injection wells was signficantly less than design
volumes.
The spray irrigation area performed well and consistently received 10±
gpm over an area of 1,800 ft2 without notable reduction in efficiency.
Occasional roto-tilling of the soil improved permeability and reduced
overland flow. Additionally, peroxide concentrations in water puddles on
the ground were not significantly less than those in the feed to the spray
irrigation area indicating little peroxide loss in the spray process.
"However, samples of infiltrate taken in the spray area vadose zone revealed
that peroxide rapidly decomposed in the first six inches of the soil. In
one test, dissolve oxygen in the infiltrate varied from 23 mg/1 at the
surface to 2 mg/1 at 2.5 feet below land surface indicating oxygen
utilization in the contaminated soils.
1341
-------
GCtJCEDSiaNS
An Oxygen. Siortfall
*° .f3^ Peroxide destabilization and oxygen loss at point of
^ •Was "^ P03311316 to deliver sufficient oxygen f br ccnblete
hydrocarbon mineralization. Ihe causes of hydrogen peroxide irBtability
at the Bjlin srte have been identified through a seriekof laboratorTand
field experiments (Spain et al, in press) ihey concluded that catolase
enzymes released Toy aerobic bacteria near the points of injection are the
manor cause of rapid peroxide decomposition and oxygen off gassing' '
To a lesser extent, humic materials and inorganics such as -Iron also
ircrease the rate of peroxide decomposition at the Eglin site. ' • • •*
Estimated
. ^ extent of bioremediation, or hydrocarbon biodegradation may be
estimated from the actual oxygen delivery rates. Peroxide was injected
into approximately 16,000,000 gallons of water. Based on peroxS
stability and oxygen measurements at the points of injection, it is
estimated that an average 25 mg/1 of oxygen was actually delivered in this
VS^' - ^J^ oxygen concentrations down gradient of the injection
points ieated utilization, ffe estate that sufficient oxygeTwS
1'2°° pamds of hydrocarbon? It is
'
some of this oxygen was incorporated into microbial biomass,
and therefore, greater than 1,200 pounds of hydrocarbons may have been
removed.
Estimated Volatilization
A significant quantity of fuels residuals were removed in the above
ground aeration basin. Based on TOO analysis of groundwater entearim and
^T™% Slaeratlon basin' a loss of volatile organics (measured as 15-
20 mg/1 TOC) occurred throughout the past 12 months, -mis means that the
nf^S;^?331?, haS removed at least 2000 pounds of hydrocarbons from the
16,000,000 gallons of water pumped through the site. Contributincf to this
process is an unknown quantity of fuel residuals which have been washed
from the^ soils into the groundwater due to natural leaching arxd soils
washing in the spray application system. A better understanding of the
hydrocarbon mass balance is anticipated after all the site data has been
collected and analyzed. ^^
lour More Years
In order to deliver sufficient oxygen with the observed peroxide half-
llJ\,5Pd4_.subse<3Pent oxygen delivery rates, and assuming continued
volatilization of 2000 pounds/yr, four more years would l^reqSredto
deliver the minimum stoichiometric oxygen requirements. Oxygen utilization
agency by microbes was difficult to estimate from field measurements?
However, it is likely that less than 100% could be used. Volatilization
rates would also decline over time as ground water quality contiued to
improve, therefore, more than four years would be required to remove and/or
degrade remaining fuel residuals from the soil. Switching the entire site
to the spray system would be a practical way of speeding up the clean up.
rihe estimated non-research cost asspciated with this effort is $600,000.
1342
-------
Soil Ocntandxiation Data
To date fuel residuals in soils above the water table in areas
influenced by injection wells and infiltration galleries have shown no
solid evidence of enhanced biodegradation. There is no evidence that
enhanced nutrient and oxygen levels in the ground water have influenced
biodegradation in the soils above the water table. Soil samples from the
spray application area have shown nriypd results. Figure 6 illustrates the
difficulty in assessing biodegradation in the unsaturated zone. Samples
taken near EA-2 indicated a definite decrease in TEH has occurred.
Unfortunately, samples taken at EA-3, which is also in the spray
application area, show no consistency and even show an apparent increase .in
TEH. The TEH values at the EA-5 control area have also been subject 'to
variations which have made data interpretation very difficult.
There has been no increase in free product in these areas that would
indicate a new fuel leak. However, we have observed small lenses of free
product on the site which have randomly moved through our monitoring areas.
Data
Microbial enumerations of soil and groundwater have suffered from
similar variability. Once again the most significant microbial activity
appears to be taking place in the spray application area. Although not an
impressive exponential growth, both total and hydrocarbon degrading
bacteria have multiplied their numbers 10 to 15 timpfl in the soils and
groundwater within the spray area. Bacteria counts in the EA-5 control
area have shown a ten-fold decrease in numbers over the past nine months.
Ihere baa been a general increase in the proportion of hydrocarbon
degrading to total bacteria suggesting an adaptation of microbial
populations to growth on fuel substrate.
mg TPH/kg soil (xlOOO)
Mar. 87
Jul 87
Sap 87 Jan 88
Sampling Event
Mar 88
-+- EA-2 -*- EA-3 -a- EA-5 CONTROL
Avg of 3 to 4 Depths at Each Location
Figure 6. TEH Date" From Spray Area
1343
-------
VcjlatJJ.es Ranoval in Aeration System
One aeration basins which were originally installed to oxidize iron were
upgraded to also remove the volatile organics. A series of tests to
evaluate volatile organic removal efficiency were conducted at air to
water ratios varying frcm 8-36:1. Based on these tests the aj to water
ratio was increased to approximately 25 to 1 in order to achieve 95%+
removal of volatile organics. Biis inexpensive method of VCC removal
must be factored into the overall mass balance on fuel removal anl
separated from biodegradation.
Ground ffeter Contamination Data
Data fron two monitoring wells in the active treatment zone and two
wells in contaminated control areas is provided in Figure 5. BEX was
selected as an indicator of soluable fuel contamination for this
ccnpariscn. EA-2 is located in the spray application area. EA-8 is in a
control area receiving a spray of ground water without nutrients or
peroxide. EA-EA-19 is located 12 feet downgradient of the infiltration
galleries while EA-5 is located in a contaminated control area not
influenced by nutrients or peroxide. Both EA-2 and EA-19 have shown
decreased levels of BOX. Some removal of BIX has also occurred at EA-8
possibly due to the continuous application of partially restorfisd/aerated
ground water at the surface. Ground water at EA-8 has ai»i been influenced
by nutrient additions through the injection wells which may account for
some BEX removal in the saturated zone. BEX levels in the EA-5 control
area decreased in July 87 for an unknown reason but have remained stable
over the past nine months.
mg/L BTX
Mar 87
Jul 87 Sep 87
Sampling Event
Jan 88
Mar 8-a.
EA-2 —»— EA-5 Control -*- EA-8 Control
Figure 5. Ranoval of wry Feat Ground Water
EA-19
1344
-------
OCNCEDSIONS
An Oxygen Shortfall
Due to rapid peroxide destabilization and oxygen loss at point of
injection it was not possible to deliver sufficient oxygen for complete
hydrocarbon mineralization. The causes of hydrogen peroxide instability
at the Bglin site have been identified through a series of laboratory and
field experiments (Spain et al, in press) They concluded that catalase
enzymes released by aerobic bacteria near the points of injection are the
major cause of rapid peroxide decomposition and oxygen off gassing.
To a lesser extent, humic materials and inorganics such as iron also
increase the rate of peroxide decomposition at the Eglin site.
Estimated Biodegradation
The extent of bioremediation, or hydrocarbon biodegradation ^ may be
estimated from the actual oxygen delivery rates. Peroxide was injected
into approximately 16,000,000 gallons of water. Based on peroxide
stability and oxygen measurements at the points of injection, it is
estimated that an average 25 mg/1 of oxygen was actually delivered in this
water. Very low oxygen concentrations down gradient of the injection
points indicated utilization. We estimate that sufficient oxygen was
delivered to mineralize approximately 1,200 pounds of hydrocarbon. ^ It is
possible that some of this oxygen was incorporated into microbial biomass,
and therefore, greater than 1,200 pounds of hydrocarbons may have been
removed.
Intimated Volatilization
A significant quantity of fuel residuals were removed in the above
ground aeration basin. Based on TOC analysis of groundwater entering and
leaving the aeration basin, a loss of volatile organics (measured as 15-
20 mg/1 TOC) occurred throughout the past 12 months. This means that the
aeration basin has removed at least 2000 pounds of hydrocarbons from the
16,000,000 gallons of water pumped through the site. Contributing to this
process is an unknown quantity of fuel residuals which have been washed
from the soils into the groundwater due to natural leaching and soils
washing in the spray application system. A better understanding of the
hydrocarbon mass balance is anticipated after all the site data has been.
collected and analyzed.
Four More Years
In order to deliver sufficient oxygen with the observed peroxide half-
life and subsequent oxygen delivery rates, and assuming continued
volatilization of 2000 pounds/yr, four more years would be required to
deliver the minimum stoichiometric oxygen requirements. Oxygen utilization
efficiency by microbes was difficult to estimate from field measurements.
However, it is likely that less than 100% could be used. Volatilization
rates would also decline over time as ground water quality contiued to
improve. Therefore, more than four years would be required to remove and/or
degrade remainirg fuel residuals from the soil. Switching the entire site
to the spray system would be a practical way of speeding up the clean up.
The estimated non-research cost associated with this effort is $600,000.
1345
-------
Based on current operating costs it is estimated that
costs of $210,000 per year for the
re^ired for
OT thfi
higher.
shculd
,
field and had minimum construction and maintenance costs.
clean up. it siould be noted
of 100% oxjTjen utilisation actual
For those wishing to
SSlite
1. In situ peroxide stability must be greatly improved to provide
o^gen dcwngradient of injection potatsri ^res^loS
per gallon for 35% hydrogen peroxide, the available oxygeTS
- $2.40 per pound. In comparison, the cost of iSSrial
acygenis only $.10 per pouixL therefore, iftt
cc^certaticn achieved with peroxide addition is not sus
than tte 40 ng/l.of oxygen saturation possible with liquid oxygen, then
use of peroxide is not cost effective.
. te given to both vertical and horizontal
of contaminants. Oxygen delivery systems must Titfluence the
a m:j°rity °f the fuel residuals remain. In tbla project
- .was the .only n^fchod which appeared to do this. Other
technologies such as soil venting have a^STgeaber pofesntialfo£
introducing oxygen into unsaturated soUs. poo-nciaj. ror
M methods for predicting in situ biodegradation, peroxide
stablity and geochemical side reactions must .all beiiiproved,.
procedures ^must be refined to account for site specif iTsS? and
water chemistries. On site pilot tests are reconSnded to
srbi peroxide stability, oxygen utilizaion ra
prcoiems.
4. Finally, there is a need for more sharing of meaningful site data by
those ^ experience in the application of this technology. Eata en
S>
v transport, oxygen utilization and toe removaL
residuals from soils are all missing fron the open literature.
1346
-------
1. AEHA, 1985. Standard Methods for the Examination of Water and
Wastewater. 16th ed.
2. Balkwill and Ghiorse, 1985. Characterization of Subsurface
Bacteria Associated With Two Shallow Aquifers in Oklahoma.
Applied Environmental Microbiology. Vol 50. pp 580-588.
3. Britton, L. N. 1985. Feasibility Studies on the Use of
Hydrogen Peroxide to Enhance Microbial Degradation of
Gasoline. American Petroleum Institute Publication 4389.
4. Davis, J.B. et al. 1972. The Migration of Petroleum Products
in Soil and Groundwater. Publication of the API ;
Committee on Environmental Affairs. Dec 1972.
5. Jamison, V.W., Raymond R.L., Hudson J.O., 1976. Biodegradation
of High-octane Gasoline. Proc. 3rd Int. Biodegradation
Symposium. J.M. Sharply and A.M. Kaplan, eds. Applied :
Science Pub. pp 187-196.
6. Lee, M.D. et al. 1988. Biorestoration of Aquifers Contaminated
with Organic Corrpounds. CRC Critical Reviews in Env. Control.
Vol. 18. pp 29-89.
7. Spain J.C., Milligan J.D., Downey D.C., Slaughter J.K. nil Press.
Excessive Bacterial Decomposition of HoC^j During Erihahced
Biodegradation. Accepted for publication in Journal of Ground
Water.
8. Weston, Roy F. Inc. 1984. Response to Fuel in Ground Water at POL
Area Eglin AFB - Table 1 Hydrualic Characteristics.
AIT- Force Engineering and Services Center Tyndall AFB.
9. Wetzel R.S., Durst C.M., Davidson D.H., Sarno D.J. 1985. In Situ
Biological Treatment Test at Kelly AFB, TX, Vol II Field Test
Results and Cost Model. AFESC ESL Tech. Report 85-52.
1347
-------
-------
Appendix 8-D
Microbial Treatment Technology Case Studies
Biological Pre-treatment of Ground Water,
Bunschoten, The Netherlands
1349
-------
Biological treatment of gropndwater polluted with HCH, chlorobenzene
and benzene on a former pesticide production site in Bunschoten, The
Netherlands.
Urlings L.G.C.M., F. Spuij and J.P. van der Hoek
TAW Infra Consult B.V., P.O. Box 479 Deventer, The Netherlands
1. Introduction
From 1945 untill 1949 the factory Dagra (Bunschoten, The Netherlands)
has produced the insecticide Lindane. Lindane (gamma-Hexachlorocy-
clohexane) was made by mixing benzene and chlorine gas under U V
radiation; not only gamma-HCH was produced but also all other HCH-iso-
mers.
As a result of these activities the factory grounds and the surrounding
area were heavily polluted withJiCE^ven the groundwater was contami-
nated with mainly HCH, monochlorobenzene and benzene. An outline of
the production shed and the surrounding is given in figure 1.
Om
2m,
1
CRauNOWATER'H'-V •>'-'• . .'•'• 'SANOY MuihR:--*''^\'-''^s*jr^p*+:'-/:t 7J*,: •.:••*•' CONTAMINATED
.,-T-> = »-.*... : 1^-.-._.. .. ...-.-.• .. • . . • ,'•_... •.- . f .. , .,;-'7.5ROUNDWATER
U SURFACE RUNOFF
Figure 1. Pesticide pollution around the production shed
nnn rlined^1 action of the contaminated soil took place. More
than 11 000 m^ polluted soil was excavated and transported to a Tem-
porary Storage Place waiting; for treatment. Approximately 60 J soil
wMoh ^S re^ded^ P^e HCH (> 90%). The groundwater remediation
±?Ja?o oo nt£U aft6/ ^ excavation wiH °e finished in the second
part of 1988. The groundwater: treatment originally consists of a phy-
sicochemical proces which is made up Pf two sand filters and thrL
activated carbon filters .. During the remedial action of the soil the
inaximum withdrawal was 6.0 n^. In 1986 fche ndwater was 3i ™
h?* b6enp "duced to ?5 ^3Aas a result of the cleanup action in
middle of 1987. In October 1988 the flow was 12 m3/h.
In recent research it appeared that a lot of xenobiotics can be deera-
dated by micro-organisms. Although there is hardly any experience Jith
biological groundwater treatment TAUW Infra Consult B.V. started some
small scale experiments with a trickling filter and a rotating biolo-
gical contactor in 1986 s DIOIO
1350
-------
The results were promising: 50% removal of HCH and 99% removal of
benzene and monochlorobenzene. The required activated carbon would be
reduced in this situation to less than 10% when applying a full scale
biological treatment.
In november 1987 two rotating biological contactors were installed for
groundwater pretreatment (see figure 2). The research was partly
financed by the Ministry of Housing, Physical Planning and
Environment.
R.B.C. = rotating
biological
contractor
Sf. - sandfilter
A.C. = activated carbon
B.F. = biofilfer
:i
I ~ — — — i i
-[TAUWRJ.CJ-
L| Klein R.8.C[-
1 ©
/
(s£Xs£)i
(AX)
"X'
vfx
(A.CJ
i
•-mil.
raw-
water
- infl.
effl.
setting basin
sandfitter
ttt.-B.K.
'.
purifitd
vaftr
1
1
to sewerage
Oostsinget
Ringwetering
Figure 2. Groundwater treatment installation in Bunschoten, the
Netherlands
Although the groundwater treatment installation has been scaled up,
there is still an experimental character in the operation of the
plant.
2. Apparatus and operation
The rotating biological contactor (RBC) installation consists of" two
RBCs. Firstly a TAUW RBC with an effective surface aera of 1000 m2
placed in a trough of 9 m3 and divided into 4 stages (figure 2), and
secondly a KLEIN RBC with an effective surface area of :700 m2 placed
in a trough of 4.4 m3, also divided into 4 stages. Both RBCs rotate
with 1.5 rpm. There are no settling tanks and the effluent of the
RBCs is treated in the sandfilter/activated carbon installation. The
RBCs operate seperately (parallel) but they can be connected in series
if wanted (figure 2).
To prevent contamination of the.surrounding air by volatilization of
contaminants from the groundwater both RBCs are roofed, and the air
extracted from the RBCs is purified by a compost filter. The air
extraction flow is approximately 150 m3/h.
1351
-------
to raw water basin
RBC's
Figure 3. Flow diagram of the RBCs and compost filters.
3. Groundwater characteristics - . . ..
Table 1 gives an indication of the concentrations of contaminants in
January 1988. ', .-.'...' ,
UP-/!
Benzene
Monochlorobenzene
HCH*
alpha -HCH
beta-HCH
gamma-HCH
delta-HCH
epsilbn-HCH
350
350
200
+ 60
+ 5
+ 80
+ 40
± 15
Other xenobiotics are not substancially present in the groundwater
(GC/MS-analysis).
1352
-------
Some important macroparameters are shown in table 2
Table 2. Macroparameters groundwater
compound
dimension
august 1986
february 1988
COD
BOD
Kjeldalh nitrogen
Ammonia
Nitrite
Nitrate
Total phosphorus
Temperature
PH
mg/1
mg/1
mg N/l
mg N/l
mg N/l
mg N/l
mg P/l
°C.
~ ~
56
4
3,5
4,5
0,02
0,1
0,28
10
7
47
5
3
< 0,01
. 0,35
0,57
9
- 6,8
4. Results and discussion.
4.1 S tart up
After one week of operation (3rd December 1987) a. thin .iiofilm was
observed. In addition sediment from the raw water basin was supplied
to both biocontactors during one day of batch operation to enhance
bacterial atachment. First results, 9th December , showed almost
complete (> 98%) removal of benzene and chlorobenzerie at a low flow
rate of 3 m3/h (hydraulic residence time 3 ,h). After a short freezing
(2 weeks) period and shut down of the RBCs complete removal of benzene
and chlorobenzene was measured after 3 days of the re-start (3 nr/h).
The first HCH biodegradation was found on day, 28.
4.2 Overall removal efficiencies
The applied flow rates through the RBCs are shown in figure 4.
Time In days
+ Klein
Figure 4. Flow rates through the RBCs
1353
-------
- HCH-removal
- 5 ithS ?m°Val effieiencies of alpha and gamma-HCH in the
, is plotted against time, while in figure 6 HCH removal related
to HCH loading is shown.
280
Time la days
Q alpha - HCH + jamma - HCH
Figure 5. Removal efficiencies of alpha and g;
HCH CTAUW Rl
70 . •
30 -
10 -
0 -f
c
Q
n
a
CD n
o & a
30 SO 70
HCU - total lo.dln| In mt/m
80
Figure 6. HCH removal versus HCH loading (TAUW RBC)
Figure 6 shows a linear increase of HCH removal with HCH loadine
However at loadings of 50 - 60 mg/m2d the line deflects and remains
constant. This trend is confirmed by the compartiment sampling (see
figure 12).
1354
-------
- Benzene removal
The performance of the TAUW RBG is given in figure 7.
•B
ci
£
e
-
0
£
6
b
e
N
j.
0
a
c*;u
200 -
180 -
ISO -
140 -
120 -
100 -
80 -
SO -
40 -
20 -
a
a.
o
f
a
j,
0°
«P
DD
•Pa'
a
a
0 40 BO 120 1EO ZOO Z40
Benzene loading In mj/raZ.d
Figure 7. Benzene removal versus benzene loading
Figure 7 indicates that benzene removal is complete and no deflection
is observed.
- Chlorobenzene removal
The removal related to loading of the TAUW RBC is shown in figure 8.
V
g
\
•4
C
C
..
J
-------
- Experiment 1. flow 6.7 m3/h, sampling during 2 h.
Influent
benzene 5,900 mg
chlorobenzene 6,300 mg
Extracted air
benzene 65 mg (1%)
chlorobenzene 101 mg (2%)
BIODEGRADATION
Effluent
benzene 80 mg
chlorobenzene 201 mg
For experiment 2 and 3 the following volatilization percentages are
found.
- Experiment 2:
- Experiment 3:
benzene 2% volatilization
chlorobenzene 1% volatilization
benzene 6% volatilization
chlorobenzene 9% volatilization
The airflow in experiment 3 was extremely high (840 m3/h) which ex-
plains the high volatilization. The HCH volatilization was less than
1%.
4.6. Compartment measurement
To obtain detailed information on removal rates, in individual compart-
ments of the RBCs samples were taken in each compartment of the RBCs
in serie. Both RBCs have 4 compartments equal in volume.
- HCH
The total-HCH concentration is given for the several compartments in
figure 9. In the first two compartments the removal rate is comparable.
camp 1 comp 2 corap 3 Efrt TAUW comp 3 Em KLEIN
20/04
19/06
+ 19/03
X Z9/OS
O OS/06
» 24/08
Figure 9. HCH compartment sampling at different flow rates
In figure 10 and 11 the 5 isomers of HCH are shown at a low and a high
flowrate respectively.
1356
-------
Inrt comp 1 camp 2 comp 3 Em TAUW comp 3 Ern KLEIN
a > + b » c Ad x -1
Figure 10. HCH-isomer compartment sampling on the 24th August
(flow rate - 12 nr3fh) '
to
Figure
lofl camp 1 eomp Z corap 3 Ern TAUW corap a em ELEIN
o « » 6 o c * d « t
11. HCH-isomer compartment sampling on the 29th June
ri
g
\
•
-
0
L
E
U
Figure 12,
(flow rate - 22 rn^/h)
110 -
100 -
80 -
80 -
70 -
60 -
SO -
40 -
30 -
20 -
10 -
0 -
a
q
Q
a a ° a 0
- a a
a o a
0 0
a
a
a°° a
J 100 200 300 400
HCH loidlnt In mg/iuZ.d
HCH loading versus HCH removal: compartment sampling
1357
-------
In figure 10 alpha and gamma HCH breakdown are almost complete after
passage of two compartments while in figure 11 the total disc surface
area is needed to complete the alpha and gamma HGH breakdown. Delta
HCH biodegradation is also observed. The concentrations of the epsilon
and beta HCH remain constant. ,
The results of the compartment measurements (figure 9) is presented in
figure 12 in a loading/removal graph which shows a maximum removal
rate of almost 50 mg HCH/m2d.
- Benzene and chlorobenzene
In figure 13 and 14 the breakdown curves of benzene and chlorobenzene
are given for the RBCs in series. The results for benzene and chloro-
benzene are comparable and the .highest removal rates appeared in
compartment one of the TAUW RBC.
450
D 2O/O4 15.6 mVh
A 15/O6 JlBmVtl
+ 19/OS IWn'/h
O 08/06 22.2 m'/h
V 24/06 13.2 ro'/ti
X 29/O6 12.0 m'/h
Figure 13. Benzene concentration in the compartments for different
flow rates
comp 1 comp Z comp 3 Em TAUW comp 3 ' E1TI KLZIN
20/04 + 19/03
13/06 X 29/06
« 08/08
V 24/08
Figure 14. Chlorobenzene concentration in the compartments for
different flow rates
1358
-------
Figures 15 and 16 give removal rates (mg/m^d) versus losdingrate
(mg/m^d) for benzene and chlorobenzene. And indicate that when
loadingrates exceed 200 mg/rn^d the removal can be estimated as a half
order process. In figures 7 and 8 a zero order removal process was
outlined, which means no limitation.
400
330
300
230
ZOO
ISO -
100 -
SO -
0
o o
o
a
0.4 O.S
(Taoussnds)
Benzene loidint la m{/mZ.d
Figure 15. Benzene loading versus benzene removal as measured in
compartment sampling
soo -
300 -
200 -
100 -
0 ZOO 400 600 BOO
Cl-benzene londlnj In mj/m2.d
Figure 16. Chlorobenzene loading versus removal as measured in
compartment sampling
4.5 Biofilm characteristics
The growth of biomass was measured in the 1st and 4th compartment of
the TAUW RBC by using exchangeable disc packets. The biomass density
was determined by removing the biofilm from 0.1 m^ disc area. The
biomass-wash out was calculated from the suspended solids
concentration in the effluent.
1359
-------
Table 3 Biomass charactaristics.
1st compartment TAUW
4th compartment TAUW
wash-out TAUW RBC
KLEIN RBC
growth
- 1st compartment TAUW RBC
- 4th compartment TAUW RBC
dry matter
200 g/m2
20 g/m2
10 g/m3
10 g/m3
7 g/m2.wk
2 g/m2.wk
ash
(%
67
62
50
50
content
of dm")
Using an average HCH concentration in the sludge of 15 mg/kg it can be
calculated that 1.6 g HCH adsorbed on the whole TAUW RBC. The HCH
load of the RBC was approximately 100 g/d.. Hence accumulation of HCH
in the biomass does not occur. In the effluent only 0.2 ug HCH/1 is
adsorbed on the suspended solids. .
4.6 Costs
The cost evaluation for the Bunschoten site is based upon the contrac-
tors fee for physicochemical treatment (activated carbon) -and estimated
cost for biological treatment. For the latter the results of the above
discribed experiments are used. The total amount of removed contami-
nants are:
- HCH 100 kg
- Benzene 250 kg
- Chlorobenzene 200 kg
a combined biological/physicochemical treatment three different
For
biological removal efficiencies, as well as for an one-stage physico
chemical treatment the costs are compared in table 4.
Table 4. Treatmentcosts (in Dfl) versus techniques
Removal efficiency Costs
HCH Benzene/Chlorobenzene RBC
Activated Physicochem.
carbon , install.
Total
70%
60%
20%
0
> 95%
95%
60%
0
110.000
90.000
35.000
15.000
20.000
100.000
200.000
80.000
100.000
125.000
150.000
,205.000
210.000
260.000
350.000
Biological pretreatment results in a cost reduction for the groundwater
remedial action of 30 - 40% for the Bunschoten site.
1360
-------
5. Conclusions and recommendations
5.1 Conclusions
Biodegradation:
the removal of HCH, benzene and chlorobenzerie in tile RBC can be
attributed for more than 90% to biodegradation, Volatilization and
adsorption onto the sludge are of minor importance t8 the total
removal;
loadings up t<3 200 mg/m^d for both benzene and chlbrbbSnzene and a
hydraulic residence time of app. 30 min. lead to avertge effluent
concentrations of less than 10 ug/1;
only alpha HCH and gamma HCH show a good biodegradatiStt rate (each
25 mg/m2d). Delta HCH shows little breakdown, while epsilon HCH
and beta HCH concentrations in the RBC remain constantj
the mineralisation of.contaminants is complete, vizi there are no
metabolites found (GC/MS analysis). The applied biotechnology is
environmentally attractive;
The performance of the compost filter for air treatment: shows poor
results.
Costs
for the Bunschoten groundwater remedial action the cbs"t reduction
by RBC pretreatment is at least 30%;
cost minimizing of hazardous waste treatment (less* volume of
contaminated activated carbon).
Application
RBCs need little maintenance;
in wintertime (-10°C) RBCs are applicable when the groundwater
flow rate is not to low;
the adaptatibh time for'benzene and chlorobenzene biodegradation
was only two weeks and for HCH four weeks in the Rfid-aystern.
5.2 Recommendations
More biological techniques like RBCs should be applied iii the ground-
water remedial action, especially for aromatic contaminants.
From literature it is known that a lot of organic iflicrbp*bllutants can
be biodegradated. Probably a combined aerobic-anaerobic: (dechlorina-
tion) system can tackle a wide variety of organic cbritaffiinants in an
environmental friendly way.
Acknowledgement
Thanks to the critical and inspiring contribution of the advising
board. The board consited of dr. ir. A. Klapwijk (chairman,
Agricultural University Wageningen), dr. ir. M. v» Loosdrecht
(Agricultural University Wageningen/Technical University Delft), ing.
P.J.C. Kuiper (Institute for Inland Water Management Stiii Waste Water
Treatment), ing. A.W.J. van Mensvoort, ing. J.Q, Hoogendoorn
(Province of Utrecht), ir. A. v.d. Vlugt (Ministety of Housing,
Physical Planning and Environment).
1361
-------
-------
Appendix 8-E
Microbial Treatment Technology Case Studies
Rotary Composting Reactor of Oily Soils, Soest, The Netherlands
1363
-------
PRODUCTION SCALE TRIALS ON THE DECONTAMINATION OF OIL-POLLUTED SOIL IN A
ROTATING BIOREACTOR AT FIELD CAPACITY
Ger P.M. van den Munckhof and Martin F.X. Veul,
Environmental Technologists, Witteveen+Bos, Consulting Engineers,
P.O. Box 233. 7^00 AE Deventer, The Netherlands
SUMMARY
A biological treatment method has been developed for the decontamina-
tion of oil-polluted'soil at field capacity. Soil is treated in a. rotating
bioreactor in which soil temperature, oxygen, moisture, and nutrient levels
can be adjusted. Following laboratory scale studies to determine the
optimum environmental conditions for oil biodegradation, production scale
trials were carried out with oil-polluted soil (1,000 to 6,000 rag/kg dry
soil) at field capacity. In five batch experiments and two semicontinuous
experiments with 50 tonnes of soil, end concentrations after 1 to 3 weeks'
treatment varied from <50 to 350 mg/kg dry soil. Microbial activity and oil
breakdown was highest in the first 3 to 4 days. This technique warrants
further investigation as it is environmentally friendly, energy saving, and
the end product is a living, fertile soil.
FRAMEWORK
As part of the development of biotechnological methods for cleaning
soil, Witteveen+Bos Consulting Engineers, Deventer, has carried out,
production scale trials on the decontamination of oil-polluted soil in a
rotating bioreactor at field capacity. The aim was to develop a method for
1364
-------
quick cleaning of polluted soil under controlled conditions at field
capacity.
The Microbiology Department of the Wageningen Agricultural University and
Broerius Soil Sanitation Ltd., Voorthuizen, also took part in the study,
which was partially financed by the Netherlands Ministeries of Housing,
Physical Planning and Environment (VROM) and of Economic Affairs (EZ).
FEASIBILITY STUDY
A feasibility study was carried out at the Microbiology Department of
the Wageningen Agricultural University from July through November of 1988.
Optimum environmental conditions for biological decomposition were determi-
ned in incubation experiments with soil in which oil degradation was
monitored by respiration measurements and oil analyses.
Mesophilic decomposition of oil was shown to be most efficient at a
temperature of 30 to 35°C, a moisture content of approximately 10 percent
by weight, and in the presence of 150 mg N nitrogenous fertilizer per kg of
dry soil (oil concentration of 5,000 mg/kg d.s.), Under these conditions,
on a laboratory scale, oil is completely decomposed (to <100 mg/kg d.s.)
within two weeks.
THE ROTATING BIOREACTOR
The reactor used in the production scale trials was a modified DANO-
composting installation for household garbage operated by the Municipality
of Soest and Baarn (Figure 1). Broerius Soil Sanitation Ltd. renovated and
adapted the installation for soil treatment. The reactor is 25 m long and
3.5 m in diameter. As the reactor rotates, the soil is mixed and homogeni-
zed so that biomass and oil substrate come into close contact. Temperature
1365
-------
Figure 1. The bioreactor
Input: Nutrients
contaminated soil
Figure 2. Biological treatment process in the rotating bioreactor
1366
-------
and oxygen levels are maintained with the aid of a warm air blower, and
soil moisture content is adjusted with a sprinkler installed in the
bioreactor. Nutrients are added to the soil on the conveyor before entering
the reactor. The soil treatment process is shown schematically in Figure 2.
Witteveen+Bos .has obtained'the patent for this treatment process.,
PRODUCTION SCALETRIALS
Batch Experiments , . :
Five loads (batches) of oil-polluted soil, each weighingj, 50 tonnes
(wet weight), were treated in the reactor as follows: one batch of petrol-
polluted soil, three batches of diesel-polluted soil, and one batch of soil
polluted with both diesel and lubrication oil. The oil concentrations in
the soils varied from 1,000 to 6,000 mg/kg dry soil. Four of the five
batches consisted of fine to loamy fine sand, and one batch consisted of
fairly coarse sand.
Each soil was incubated in the reactor under optimal environmental conditi-
ons as determined in the laboratory feasibility study. This was only partly
successful because the soil temperature in the reactor did not rise above
20 to 22°C. In addition, the structure of the relatively loamy soil batches
partly deteriorated. Some soil particles tended to stick to one another
around small particles of rock, thus forming soil balls sometimes several
centimetres in diameter.
Soil and air samples taken from the reactor were analyzed at regular
intervals to monitor and manage the microbial degradation process. To take
the soil samples, three sampling flaps were made, at a quarter, half, and
1367
-------
three-quarters of the length of the reactor (sampling points !„ 2 and 3,
respectively). In addition to the chemical analyses such as moisture
content, humus content, pH, oil concentration, volatile aromatics, and PAH,
counts were done of oil-degrading microflora, for which the most probable
number (MPN) method was used. The air samples were analyzed for oil and
volatile aromatics to establish a mass balance (degradation versus volati-
lization) .
For comparison with the measurements in the reactor, respiration measure-
ments were also dope on soil samples from each batch in the Microbiology
Laboratory at Wageningen Agricultural University. The results of the
measurements on oil biodegradation in the five batches are summarized in
Table 1, and presented in Figures 3, 4 and 5.
In the rotating bioreactor at a soil temperature of about 20°C, the end
concentration of the oil biodegradation was reached on average within a
period of approximately 2.5 weeks. Biodegradation was most rapid in the
first week. The end concentration of oil varied from <50 to 250 nig/kg dry
soil.
The maximum level of oil-degrading microflora in the reactor was reached
within one to three days. The numbers of microorganisms grew from MPN/106
to MPN/108 per gram of soil,
The laboratory respiration measurements indicated that at a' temperature of
30°C biological degradation was largely completed in a period of 1 to 1.5
weeks. In the humus-poor, fairly coarse sandy soil in batch 1, it took
longer, 2 to 3 weeks. This indicates that, if a soil temperature of 30°C
can be achieved in the bioreactor, then an average degradation period of 1
1368
-------
Table 1. Results of the batch experiments
Batch
1,
2.
3-
4
5-
Soil type
poor coarse
sand
fine sand
loamy fine
sand
fine sand
loamy fine
sand
Oil type
diesel oil
petrol
diesel oil/
lubrication
oil
diesel oil
diesel oil
Oil-
concentration
(mg/kg ds).
start
6,000
920
730
1,000
1,000
1,500
Oil-
concentration
(mg/kg ds)
end
<100
<50
, <100 (GC)
180 egradation
time (weeks)
2.5-3.0
0.5
2.5-3-0
2.5
1.5
Point in
time max.
level
microflora
(days)
, 3
3
2-3
1-2
Results of
respiration
neasurements
(f of total
respiration)
80% after >
2 weeks
90* after
1 week
85X within
It week
90X within ,
1 week
Volatilization
(weight %)
>30
30
0.5
0
1000
Oil cone, (mg/kg dm.)
6
batch 2
1O 15 20
Time (days)
26
batch 3 * batch 4 D batch 6
Figure 3. Decrease in oil concentration in the batch experiments 2, 3,
and 5
1369
-------
600
Number of m.o./g d.m. .CE06)
6 ; 8 10
Time (days)
batch 1
batch 3
12 14
batch 4 --Q- batch 5
16
Figure 4. Development of oil-degrading microflora in batch experiments 1,3,
4 and 5
oxygen consumption (mmol O2/kg soil)
160-
100-
60-
5
• batohl
10 16
Time (days)
batchS
* batch4
batoh6
Figure 5- Oxygen consumption assessed by laboratory respiration measurements
on soil samples from batch experiments 1, 3, 4 and 5.
1370
-------
to 1.5 weeks is feasible.
Based on these results, treatment in the reactor is not guaranteed to
achieve an end concentration of oil below the Dutch Standard for unpolluted
soil (A-level) of 50 mg/kg d.s. However, treatment of petrol- and diesel-
polluted soils in the reactor removed the most critical pollutants. The
remaining oil components are mainly alkanes, which are highly branched
i
and/or of long chain length. Compared to the original oil product, these
components are less volatile and less soluble, do not cause odor nuisance,
and are probably less toxic.
Thus treatment in the bioreactor is environmentally beneficial.
Petrol-polluted soils should not be treated in the rotating bioreactor
without provisions to reduce emission of volatile, aromatics into the air.
There are no significant emissions of aromatics from diesel-polluted soils.
However, sometimes at the beginning of the incubation, hydrocarbon emissi-
ons of up to a few hundred mg per m3 may occur.
Semicontinuous Experiments
In the semicontinuous experiments, 3«5 tonnes of decontaminated soil
were removed daily from the end of the reactor and the same amount of
contaminated soil placed in the front section. The total quantity of soil
in the reactor remained at 50 tonnes and the average retention time was 14
days.
Two semicontinuous experiments were carried out, one for one week (experi-
ment 1) and the other for two weeks (experiment 2). Each day soil samples
were taken at all sampling points, analyzed for oil concentration, and MPN
1371
-------
counts of the oil degrading microflora made. The results are summarized in
table 2.
Table 2. Results of semicontinuous experiments
Parameter
* Mean oil concentration
(IR) (mg/kg ds)
- expt. 1
- expt. 2
* Mean of MPN-counts
(.106/g. soil)
- expt. 1
- expt. 2
Sampling point
input
; 980
995
25
26
1 ..—
sp 1
215
220
183
131
•^^— ^— •— ••
sp 2
370
205
33
46
SP 3
350
190
16
4
output
350
230
16
The soil entering the reactor had an average oil concentration of about
1,000 mg/kg dry soil. At the defined sampling points 1, 2, and 3 and in the
soil leaving the reactor, the oil concentration was the same, namely 200 to
350 mg/kg dry soil. Thus it may be concluded that biodegradation to 200 to
350 mg/kg dry soil occurred quickly (within 3 to 4 days). The highest
microbial activity occurred in the front section of the reactor (up to sp
1). At sp 1 counts of oil degrading microflora were MPN 108 to 1Q9 per gram
of soil. At sp 2 and sp 3, the number of oil degrading microorganisms had
declined to the level in the soil entering the reactor (MPN 107 per gram of
soil).
The results for the semicontinuous experiments are in line with those of
the batch experiments. The highest microbial activity and the greatest
breakdown occurred in the first 3 to 4 days. The end concentration was
slightly above the Dutch A-level for unpolluted soil. On the basis of the
1372
-------
available data, preference, cannot be given to either semicontinuous or
batch incubation.
PROGRESS AND PERSPECTIVE
The production scale trials indicate that, at a soil temperature of
approximately 22°C, oil can be decomposed within one week to an end
concentration varying from <50 to 350 mg/kg dry soil. In this process, the
highest microbial activity occurs in the first 3 to 4 days.
For the present Dutch situation, a soil-cleaning technique is only fully
applicable if the end concentration of a contaminant is below the A-levels
defined in the Interim Act on Soil Sanitation. For mineral oil in a
standard soil (containing 102 organic matter) this A-level is 50 mg/kg d.s.
On the basis of the results of these experiments, soil treatment in the
bioreactor cannot be guaranteed to achieve the required A-level for mineral
oil. But the same is also true for land farming by which oil concentration
is reduced to between 500 and 1,000 mg/kg d.s. after two years.
We must therefore conclude that, on the basis of current Dutch policy,
there are no practical applications for the biotechnological cleaning of
oil-polluted soil. Further research is recommended. In comparison to other
techniques, biotechnological treatment is environmentally friendly and
energy-saving, and the end product is a living, fertile soil. Further
research should be directed not only to microbiological but also to
environmental hygiene aspects.
1373
-------
Microbiological Aspects
Further laboratory and pilot , scale studies would determine 'the
optimal process parameters to increase biological availability and rate of
oil breakdown, and to reduce the end concentration. Attention should be
given to the effects of introducing selected microorganisms and chemical
pretreatment. With a degradation time of up to one week, the bioreactor
treatment is competitively priced with such other techniques as thermal
decontamination.
Environmental Hygiene Aspects
The Dutch A-level for mineral oil was established on the basis of
Physical and chemical properties and the toxicity of oil products. However,
during the biological breakdown process the oil changes in composition.
This means that environmental hygiene characteristics, such as ecotoxicity
and leaching, may have changed to such an extent that it would be appropri-
ate to reconsider the risks to public health and the environment.
Compared with the oil-polluted soil, the components of biologically treated
soil are less volatile, less soluble, and probably less toxic. Thus the
question arises, to what extent is the A-level for mineral oil applicable
to biologically treated soil?
There is draft legislation in the Netherlands (the "Concept Bouwstoffenbe-
sluit" of 1989) which allows for the use of decontaminated soil with an end
concentration above the A-level. The "Concept Bouwstoffenbesluit" states
that decontaminated soil from which no environmental hygiene problem is
expected may be used without restriction. To evaluate more precisely the
1374
-------
environmental hygiene quality of oil-polluted soil that is biologically
decontaminated (and of soil in general), it is important that appropriate
ecotoxicological tests (bioassays) and leaching tests be developed.
ffc-r ••".«
1375
-------
REMARKS
Recently in the Netherlands draft legislation on the use and re-use of
building materials and soil is presented. This latest "Concept Bouwstoffen-
besluit" (of July 9, 1991) allows for the use of decontaminated soil with
oil concentrations up to a maximum of 250 mg/kg dry soil.
The production scale trials in the reactor indicate that end concentrations
below 250 mg/kg ds will be reached. Therefore, based on the latest Dutch
draft legislation, treatment of oil-polluted soil in the bioreactor can be
regarded as a fully applicable soil-cleaning technique.
Qer van den Munckhof
Deventer, November 12, 1991.
1376
-------
Appendix 9
List of NATO/CCMS Pilot Study Participants
1377
-------
List Of NATQ/CCMS Pilot Study Participant.
Harald Kasamas
OEKO Fonds
Reisnerstrasse 4
Vienna, A-1030
AUSTRIA
Deniz Beten
CCMS Programme Director
NATO Scientific Affairs Division
B-1110 Brussels
BELGIUM
J.M. Junger
Commission of the European Communities
Rue de Lou 200
B-1049 Brussels
BELGIUM
Eusebio Murillo Matilla
Commission of the European Communities
DGXI
10, rue Guimard (M23)
B-1040 Brussels
BELGIUM
Hans-Joachim Stietzel
Commission of the European Communities
DGXI
10, Rue Guimard (M22)
1040 Brussels
BELGIUM
Don Bartkiw
Assistant Director
Ontario Ministry of the Environment
Waste Management Branch
40 St. Clair Avenue, W.
Toronto, Ontario M4V 1P5
CANADA
Robert Booth
Mgr. Site Remediation Div.
Wastewater Technology Centre
Operated by Rockcliffe Research
Management Inc.
P.O. Box 5068
867 Lakeshore Road
Burlington, Ontario L7R 4L7
CANADA
Maria Dober ,
Waste Treatment Coordinator
Environment Canada
Contaminants and Assessments Branch
5th Floor, Queens Square
45 Alderney Drive
Dartmouth, Nova Scotia B2Y 2N6i
CANADA <
Ruth Drouin
Ingenieur Charge de Projets ••••••
Ministrie de L'Environnement du Quebec
3900 Rue Marly
Ste-Foy, Quebec G1X 4E4
CANADA
Tony Fernandes ' .
Scientific Officer .;
Alberta Environment • '...••
Industrial Branch, Wastes and
Chemicals Division
A820-106St.
5th Floor
Edmonton, Alberta 15K 2J6
CANADA :
T.W. Foote
Environment Canada
12th Floor
351 St. Joseph Blvd.
Hull, Quebec J8Y 325
CANADA
Bernard Gaboury
Ingenieur Charge de Projets
Ministrie de L'Environnement du Quebec
3900 Rue Marly
Ste-Foy, Quebec G1X 4E4
CANADA
Pierre J. Gelinas
Professor
Department of Geology '
University Laval
University City, Quebec G1K 7P4
CANADA
Marc Halvey
Process Development Engineer
Wastewater Technology Centre
P.O. Box 5050
867 Lakeshore Road
Burlington, Ontario L7R 4A6
CANADA
1378
-------
Brett Ibbotson •••" , ,
Senior Environmental Engineer
SENES Consultants Limited
52 West Beaver Creek Road ,
Unit #4
Richmond Hill,,Ontario L4B 1L9
CANADA
Alain Jolicoeur
Director, Technology Development and
Technical Services Branch
Conservation and Protection
Environment Canada
Ottawa, Ontario K-1A OH3
CANADA
J. Peter Jones
Full Professor
Department of Chemical Engineering
Universite de Sherbrooke
2500 Boulevard Universite
Sherbrooke, Quebec J1K 2R1
CANADA
Suzanne Lesage
Senior Groundwater Chemist
National Water Research Institute
Environment Canada
RRB, CCIW
Burlington, Ontario L7R 4A6
CANADA
Colin Mayfield ... .
Professor
Waterloo Center for Groundwater Research
University of Waterloo
Waterloo, Ontario N2L 3G1
CANADA
Andre Pelletier :
Project Engineer
Environment Canada
Conservation and Protection
1179 de Bleury Street
Montreal, Quebec H3B 3H9
CANADA
Charlie Riggs
Environmental Engineer
Department of Environment
Government of Newfoundland, Canada
Department of Environment and Lands
Industrial Environmental Engineering
Division
P.O. Box 8700
St. John's, Newfoundland A1B 4J6
CANADA
James W. Schmidt
Government Programs
Wastewater Technology Center
Operated by Rockcliffe Research
Management Inc.
867 Lakeshore Road
P.O. Box 5068
Burlington, Ontario
CANADA
Georges Simard
Environment Quebec
3900 rue Marly
B.P. 35
Ste-Foy, Quebec G1X 4E4
CANADA ,
Art Stelzig
A/Chief Chemical Industrial Division
Industrial Programs Branch
Environment Canada
Ottawa, K1A OH3
CANADA
Pierre Sylvestre
Centre St. Laurent
Environment Canada
105 McGill Street .
8th Floor
Montreal, Quebec H2Y 2E7
CANADA
William Taciuk
UMATAC
210-2880 Glenmore Trail, S.E.
Calgary, Alberta
CANADA
Thomas Becker
Ministry of Environment
Strandgade 29
DENMARK
Danna Borg
Ministry of Environment
Strandgade 29
DENMARK
Kim Broholm
Department of Environmental Engineering
Technical University of Denmark
Building 115
2800 Lyngby
DENMARK
1379
-------
Lis Marie Keiding
Miljostyrelsen
Strandgade 29
DK 1401 Copenhagen
DENMARK
Bertel Nilsson
Geological Survey of Denmark
Thoravej 8
2400 Copenhagen NV
DENMARK
Lotte Wammen Sorensen
Waste Office
Miljostyrelsen
Strandgade 29
1401 Copenhagen K
DENMARK
Neel Stroebaek
The County of Sonderoylland
Groundwater Division
Jomfrustien 2
6270 Tender
DENMARK
Suzanne Arup Veltze
Head of Waste Division
National Agency of Environmental
Protection
Strandgade 29
DK 1401 Copenhagen
DENMARK
Troels Wenzel
Hojgaard & Schultz a/s
Jaegersborg 4
DK 2920 Charlottenlund
DENMARK
Florence Blanchard
Agence Nationale Pour la Recuperation et
I'Elimination des Dechets (ANRED)
2 Square la Fayette
BP406
49004 Angers Cedex
FRANCE
Phillippe Boisseau
Chef de la Division Environmental
Industrie!
Direction Regionale de ('Industrie et
de la Recherche (Region Midi-Pyrenees)
84 Rue du Feretra
31078 Toulouse Cedex
FRANCE
Thierry Dumesnil
Attache de Direction
E.i.f. Ecology
97 Rue Pierre de Montreuil
93100Montreuil
FRANCE
Rene Goubier
Head of Hazardous Sites Team
Agence Nationale Pour la Recuperation et
1 'Elimination des Dechets (ANRED)
2 Square la Fayette
BP406
49004 Angers Cedex
FRANCE
Pierre Lieben
Organization for Economic Co-Operation
and Development
15 Boulevard Amiral Bruix
Paris 16E
FRANCE
Marline Louvrier
Agence Nationale Pour la Recuperation et
I'Elimination des Dechets
2 Square la Fayette
BP406
49004 Angers Cedex
FRANCE
Claude Mouton
Agence Nationale Pour la Recuperation et
I'Elimination des Dechets (ANRED)
2 Square la Fayette
BP406
49004 Angers Cedex
FRANCE
Dominique Poiroux
Direction Regionale de ('Industrie
et de la Recherche Midi-Pyrenees
84 Rue du Feretradex
31078 Toulouse Cedex
FRANCE
Sylvie Pommelec
Agence Nationale Pour la Recuperation et
1 'Elimination des Dechets (ANRED)
2 Square la Fayette
BP 406
49004 Angers Cedex
FRANCE
1380
-------
Heinz-Jurgen Brauch
Engler-Bunte-lnstitute of the
University of Karlsruhe
D-7500 Karlsruhe 1
Richard-Willstatter-Allee 5
GERMANY
Winfried Brull
Klockner Oecotec GmbH
4100 Duisburg
Nendorfer Str, 3-5
GERMANY
H. Czech
Bergdirektor
Oberbergamt des Landes
Nordrhein-Westfalen
Goebenstr. 25
D-4600 Dortmund 1
GERMANY
Jurgen Fortmann
Ruhrkohl Umwelttechnik
4300 Essen 15
GERMANY
Volker Franzius
Umweltbundesamt
Bismarckplatz 1
D-1000 Berlin 33
GERMANY
Peter Fuhrmann
Landesanstalt fur Umweltschutz
Griesbachstr, 3
GERMANY
Prof Gellinek
Projektgruppe
Dorstfeld-Sud
Sudwall 2-4
D-4600 Dortmund 1
GERMANY
Hans-Jurgen Heimhard
Klockner Oecotec GmbH
Neudorfer Str. 3-5
D-4100 Duisburg
GERMANY
Fritz Holzwarth
Budesministerium fur Umwelt
Naturschutz und Reaktursicherheit
AHRSTRASSE 20
Postfach 120629
5300 BONN1
GERMANY
Dietrich Hueber
Landesanstalt fur Umweltschutz
Griesbachstrasse 3
D-7500 Karlsruhe
GERMANY
Herr Kaufmann
Der Senator fur Bau- und
Wohnungswesen - HG
Wurttembergischestr. 6-10
D-1000 Berlin 31
GERMANY
Ulrich Krauss
Ruhrkohle Aktiengesellschaft
Abteilung P 7
Postfach 103262
D 4300 Essen 1
GERMANY
Gerd Kuhnel
Federal Minister for the Environment
Nature Conservation and Nuclear Safety Div.
- WA II 4
Postfach 12 06 29
D-5300 Bonn 1
GERMANY
Gerhard Lehmann
Chief of the Dept. "Decontamination,
Reclamation & Env. Protection"
Ruhrkohle Westfalen AG
(Ruhr-Coal Co. Westphalia)
Deutsche Strasse - D4600 Dortmund-16
GERMANY
Dr. Melsheimer
Der Senator fur Stadtentwicklung
und Umweltschutz
Lindenstr. 20-25
D-1000 Berlin 61
GERMANY
Margaret Nels
Friedrichsthaler Weg 28A
D-1000 Berlin 28
GERMANY
Manfred Nussbaumer
Ed. Zublin AG
Albstadt Weg 3
7000 Stuttgard 80
GERMANY
1381
-------
Joachim Ronge
Ruhrkohle Umwelttechnik GmbH
Rellinghauser Str. 1
D-4300 Essen 1
GERMANY
Rolf Roth
Dekonta GmbH
Lotharstrape 26
D-6500 Mainz/Rhein
GERMANY
Hansjorg Seng
Landesanstalt fur Umweltschutz
Baden-Wurttemberg
Institut fur Wasser-und Abfallwirtschaft
P.O. Box 210752
Griesbach Str. 3
D-7500 Karlsruhe 1
GERMANY
Hans Sonnen
Harbauer GmbH & Co.
Ingenieurburo fur Umwelttechnik
Bismarckstr. 10-12
D-1000 Berlin 33
GERMANY
Wilko Werner
Harbauer GmbH & Cokg
Bismarckstrasse 10-12
D-1000 Berlin 12
GERMANY
Peter Walter Werner
DVGW-Forachungsstelle AM Engler-Bunte-
Institute de Universitat
Karsruhe
D 7500 Karlsruhe 1
GERMANY
Klaus Wolf
Umweltbehorde Hamburg
Hermannstrasse 40, IV
D-2000 Hamburg 1
GERMANY
Martin Zarth
Umweltbehoerde Hamburg
Hermannstrasse 40, IV
D-2000 Hamburg 1
GERMANY
Peter I. Richter
Technical University of Budapest
Budafoki ut 8
H-1111
HUNGARY
Alessandro di Domenico
Laboratory of Comparative
Toxicology and Ecotoxicology
Institute Superiore di Sanita
Viale Regina Elena, 299
00161 Rome
ITALY
Giuseppe Giuliano
Institute di Ricerca Sulle Acqua
Consiglio Nazionale delle Richerche
Via Reno, N. 1
1-00198 Rome
ITALY
Sukehiro Gotoh
Senior Research Officer
National Inst. for Environmental Studies
Systems Analysis and Planning Division
162, Onogawa, Yatabe-machi, Tsukuba
Ibaraki 305
JAPAN
Takashi Ikaguchi
Senior Research Scientist
The Institute of Public Health
Department of Sanitary Engineering
6-1 Shirokanedai 4 Chrome, Minato-Ku
108 Tokyo
JAPAN
Per Antonsen
Statens Forurensningstilsyn
P.O. Box 8100 Dep
N-0032 OSLO 1
NORWAY
James Berg
Senior Scientist
Aquateam Norwegian Water Technology Center
Water Technology Center A/S
P.O. Box 6326, Etterstad
0604 Oslo 6
NORWAY
Gijs Breedveld
Norwegian Geotechnical Institute
Sognsveien 72
P.O. Box 40
0801 Oslo
NORWAY
Harald Brunstad
Geological Survey of Norway
P.O. Box 3006 Lade
7002 Trondeim
NORWAY
1382
-------
Beate Folkestad
Executive Officer
Statens Forurensningstilsyn
P.O. Box 8100 DEP
N-0032 Oslo 1
NORWAY
Fetter D. Jenssen
Institute for Georesources
and Pollution Research
P.O. Box 9
N-1432AS-NLH
NORWAY
Jan Johansen
Statens Forurensningstilsyn
P.O. Box 8100Dep.
N-0032 Oslo 1
NORWAY
Bernt Malme
Statens Forurensningstilsyn
P.O. Box 8100Dep
N-0032 Oslo 1
NORWAY
Arve Misund
Geological Survey of Norway
P.O. Box 3006 Lade
7002 Trondheim
NORWAY
Ola Nordal
Statens Forurensgingstilsyn
P.O. Box 8100
N-0032 Oslo 1
NORWAY
Gunnar Randers
First Chairman for NATO/CCMS and
Former Assistant Secretary General
for Scientific Matters in NATO
Trosterudstien 4
0386 Oslo 3
NORWAY
Harald Solberg
Statens Forurensningstilsyn
P.O. Box 8100Dep
N-0032 Oslo 1
NORWAY
Guus Annokkee
Department of Environmental Technology
TNO - Division of Technology for Society
P.O. Box 342
7300 AH Apeldoorn
THE NETHERLANDS
Jan W. Assink
TNO
P.O. Box 342
7300 AH Apeldoorn
THE NETHERLANDS
A.L. Batstra
Heidemij Uitvoering BV
afd. Milieutechniek
THE NETHERLANDS
Herman Bavinck
Ministry of the Environment
P.O. Box 450
2260 MB Leidschendam
THE NETHERLANDS
H.C.M. Breek
Hollandse Wegenbouw Zanen BV
afd. Bodemsanering
Vanadiumweg 5
3812 PX Amersfoort
THE NETHERLANDS
A. Costerus
Ecotechniek
B.V. Aannemingsbedrgt, NBM
P.O. Box 16032
2500 BA, The Hague
THE NETHERLANDS
Bram de Borst
TAUW Infra Consult b.v.
P.O. Box 479
7400 AL Deventer
THE NETHERLANDS
J.F. de Kreuk
TNO/MT
P.O. Box 217
2600 AE Delft
THE NETHERLANDS
E.W.B. de Leer
TU Delft
Lab. Voor Analytische Scheidkunde
2625 RZ Delft
THE NETHERLANDS
1383
-------
D.V. Egmond
Rijksinstituut voor
volksgezondheid en milieuhygiene (RIVM)
Antonie Van Leeuwenhoeklaan 9
Postbus 1, 3720 BA Bilthoven
THE NETHERLANDS
H.J.P. Eijsackers
Program Director
Netherlands Integrated Soil Research
Programme
P.O. Box 37
6700 AA Wagemmegen
THE NETHERLANDS
D.H. Eikelboom
TNO/MT
P.O. Box 217
2600 AE Delft
THE NETHERLANDS
J.J. Gaastra
Mourik Groot-Ammers BV
THE NETHERLANDS
Josephine Hagenaars
Rijksinstitut voor volksgezonheid
en milieuhygiene (RIVM)
Laboratory for Waste Material and
Emissions (LAE)
Antonie Van Leeuwenhoeklaan 9
Postbus 1
3720 BA Bilthoven
THE NETHERLANDS
R. Kabos
Delft Geotechnics
P.O. Box 69
2600 AB Delft
THE NETHERLANDS
K. Keuzenkamp
Ministerie VROM
P.O. Box 450
2260 MB Leidschendam
THE NETHERLANDS
Rene Kleijntjens
Department of Biochemical Engineering
Delft University of Technology
Julianalaan 67
2628 BC Delft
THE NETHERLANDS
Reinout Lageman
Hydrogeophysicist/Hydrogeologist
Geokinetics
Poortweg 4
2612 PA Delft
THE NETHERLANDS
Prof. Karel Ch.A.M. Luyben
Department of Biochemical Engineering
Delft University of Technology
Julianalaan 67
2628 BC Delft
THE NETHERLANDS
P. Massink
Provincie Waterstaat Utrecht
Galileilaan 15
3584 BC Utrecht
THE NETHERLANDS
J.E.T. Moen
Ministerie VROM
P.O. Box 450
2260 MB Leidschendam
THE NETHERLANDS- '
J. Roels
Ministerie VROM
P.O. Box 450
2260 MB Leidschendam
THE NETHERLANDS
C. Schuler
Ecotechniek BV
P.O. Box 8447
3503 RK Utrecht
THE NETHERLANDS
Esther Soczo
NATO/CCMS Pilot Study Co-Director
Rijksinstituut voor volksgezondheid
en milieubeheer (RIVM/LAE)
Antonie Van Leeuwenhoeklaam 9
Postbus 1, 3720 BA Bilthoven
THE NETHERLANDS
Frank Spuij
TAUW Infra Consult
Handelskade 11
P.O. Box 479
7400 Deventer
THE NETHERLANDS
1384
-------
Sjef J.J.M. Staps
Grontmij nv
Afdeling Bodem en Water
P.O. Box 203
3730 AE De Bilt
THE NETHERLANDS
B. Tuin
TU Eindhoven
Afd. FT-HAL
P.O. Box 513
5600 MB Eindhoven
THE NETHERLANDS
Leon Urlings
TAUW Infra Consult
Handelskade 11
P.O. Box 479
7400 AL Deventer
THE NETHERLANDS
Reinier van de Berg •
Rijksinstitut voor volksgezonheid
en milieuhygiene (RIVM)
Laboratory for Soil and Groundwater (LAE)
P.O. Box 1
3720 BA Bilthoven
THE NETHERLANDS .
W.J. van den Brink '
Netherlands Organization for Applied
Scientific Research TNO
P.O. Box 297
2501 BD The Hague
THE NETHERLANDS
Ger van den Munckhof
Witteveen & Bos Consulting Engineers
Van Twickelostraat
P.O. Box 233
7400 Deventer
THE NETHERLANDS
B.L. van der Ven
Ministerie VROM
Directie Bestuurszaken
P.O. Box 450
2260 Leidschendam
THE NETHERLANDS
Jaap van Eyk
Delft Geotechnics
P.O. Box 69
2600 AB Delft
THE NETHERLANDS
A.B. van Luin
DBW/RIZA
P.O. Box 17
8200 AA Lelystad
THE NETHERLANDS
J.H.A.M. Verheul
RIVM/LBG
P.O. Box 1
3720 BA Bilthoven
THE NETHERLANDS
C.W. Versluijs
RIVM/LAE
P.O. Box 1
3720 BA Bilthoven
THE NETHERLANDS
M.F.X. Veul
Witteveen & Bos Consulting Engineers
Van Twickelostraat 2
P.O. Box 233
7400 AE Deventer
THE NETHERLANDS
John Vijgen
TAUW Infra Consult B.V.
P.O. Box 479
7400 AL Deventer
THE NETHERLANDS
J.G. Wessels Boer
Ministry VROM
Plv. dir. DWB
P.O. Box 450
2260 MB Leidschendam
THE NETHERLANDS
Martien W.F. Yland
DHV Consulting Engineers
P.O. Box 85
3800 AB Amersfoort
THE NETHERLANDS
M. Resat Apak
Associate Professor of Analytical Chemistry
Istanbul University
Faculty of Engineering
Avcilar Campus, Avicilar 34840, Istanbul
TURKEY
Aysen Turkman
Associate Professor
Dokuz Eylul Universitesi
Cevre Muhendisligi Bolumu
Faculty of Engienering and Architecture
Dept. of Environmental Engineering
Bornova Izmir
TURKEY
1385
-------
R. Paul Bardos
Warren Spring Laboratory
Department of Trade and Industry
Gunnels Wood Road
Stevenage SG1 2BX
UNITED KINGDOM
Robert Bell
Managing Director, Environmental
Advisory Unit. Ltd.
Yorkshire House
Chapel Street
Liverpool L3 9AG
UNITED KINGDOM
C.H. Bowden
Department of the Environment Rm. A 315
Romney House
43 Marsham St.
London SW1 SPY
UNITED KINGDOM
John D. Mather
Hydrology Research Group
British Geological Survey
Maclean Building
Crowmarsh Gifford
Wallingford, Oxon
OX11 8 BB
UNITED KINGDOM
Michael A. Smith
Clayton Environmental Consultants, Ltd.
68 Bridgewater Road
Berkhamsted
Hertfordshire HP4 IJB
UNITED KINGDOM
Yalcin B. Acar
Associate Professor,
Civil Engineering Dept.
Louisiana State University
and President of Electrokinetics, Inc.
The Louisiana Business & Technology Center
So. Stadium Drive, LSU
Baton Rouge, LA 70803
UNITED STATES
• Douglas Ammon
Project Manager
Clean Sites Inc.
1199N. Fairfax Street
Alexandria, VA 22314
UNITED STATES
Naomi Barkley
Environmental Scientist
U.S. Environmental Protection Agency
RREL Superfund Technology Demonstration
Div.
26 West Martin Luther King Drive
Cincinnati, OH 45268
UNITED STATES
Olav Berstad
First Secretary
Embassy of Norway
2720 34th Street, N.W.
Washington, DC 20008-2714
UNITED STATES
Edward Burk, Jr.
On Scene Coordinator
U.S. Environmental Protection Agency
Response Section I
9311 Groh Road
Grosse lie, Ml 48138
UNITED STATES
Thomas O. Dahl
U.S. Environmental Protection Agency
National Enforcement Investigations Center
Denver Federal Center, Building 53
Denver, Colorado 80225
UNITED STATES
Paul de Percin
Chemical Engineer
U.S. Environmental Protection Agency
RREL Superfund Technology Demonstration
Division
26 W. Martin Luther Kind Drive
Cincinnati, OH 45268
UNITED STATES
Jonas Dikinis
U.S. Environmental Protection Agency
Waste Management Division
Remedial & Enforcement Response Branch
230 S. Dearborn
Chicago, IL 60604
UNITED STATES
Jacolyn Dziuban
U.S. Environmental Protection Agency
402 M Street, SW
WH-548B
Washington, DC 20460
UNITED STATES
1386
-------
Timothy Fields, Jr.
Deputy Director
Office of Emergency and Remedial
Response
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460 ,
UNITED STATES
Cynthia French
Installation Services and
Environmental Protection
Defense Logistics Agency
Attn: DLW-W, Room 4-D-446 , . ..
Cameron Station -
Alexandria, VA 22304-6100
UNITED STATES
Linda Galer
U.S. Environmental Protection Agency
WH562A
401 M Street, SW
Washington, DC 20460
UNITED STATES
John M. Gilbert
U.S. Environmental Protection Agency/ERT
26 W. Martin Luther King Drive
Cincinnati, OH 45268
UNITED STATES
Wendy Grieder
Office of International Activities {A-106}
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
UNITED STATES
Margaret Guerriero .
Remedial Project Manager
U.S. Environmental Protection Agency
230 S. Dearborn
5HS-11
Chicago, IL 60604
UNITED STATES
Virginia Hathaway
JACA Corp.
550 Pinetown Road
Fort Washington, PA 19034
UNITED STATES
Roy C. Herndon
Director of the Center for
Biomedical and Toxicology Research
and Waste Management
Florida State University
2035 E. Paul Dirac Drive
Morgan Bid. Suite 226
Tallahasse, FL 32310
UNITED STATES
Merten Hinsenveld
University of Cincinnati
Department of Civil and Environmental
Engineering
741 Baldwin Hall {ML #71)
Cincinnati, OH 45221-0071
UNITED STATES
Lawrence D. Hokanson
Director, Engineering and Services Lab.-
AFESC/RD '•'.,-
Tyndall Air Force Base
Panama City, FL 32404
UNITED STATES
Stephen C. James
Chief, Site Demonstration
and Evaluation Branch
Superfund Technology Demonstration Div.
U.S. Environmental Protection Agency ...
RREL
26 W. Martin Luther King Drive
Cincinnati, OH 45268
UNITED STATES
Lynn Johnson
U.S. Environmental Protection Agency
Office of Radiation Programs
401 M Street, S.W.
Washington, DC 20460
UNITED STATES .
Herbert King
Environmental Engineer .
U.S. Environmental Protection Agency
26 Federal Plaza
Room 747
New York, NY 10278
UNITED STATES
Walter W. Kovalick, Jr.
Director, Technology Innovation Office
Office of Solid Waste and
Emergency Response
U.S. EPA (01100W)
401 M Street, S.W.
Washington, DC 20460
UNITED STATES
1387
-------
Norma Lewis
Environmental Scientist
Superfund Technology Demonstration Div.
RREL
U.S. Environmental Protection Agency
26 W. Martin Luther King Drive
Cincinnati, OH 45268
UNITED STATES
Steve A. Lingle
Deputy Director, Office of Environmental
Engineering
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
UNITED STATES
Helmut Luders
Embassy of the Federal Republic of Germany
4645 Reservoir Road, N.W.
Washington, DC 20008
UNITED STATES
Carolyn McGill
JACA Corporation
550 Pinetown Road
Fort Washington, PA 19034
UNITED STATES
Sort Metz
Science Advisor
Royal Netherlands Embassy
4200 Linnean Avenue, N.W.
Washington, DC 20008
UNITED STATES
John Moerlins
Associate Director
Biomedical and Toxicological Research
and Waste Management
Florida State University
2035 E. Paul Dirac Drive
Morgan Bldg. Suite 226
Tallahasse, FL 32310
UNITED STATES
Norm Neidergang
U.S. Environmental Protection Agency
Region V
230 S. Dearborn Street
Chicago, IL 60604
UNITED STATES
Robert F. Olfenbuttel
Director Waste Minimization and
Treatment
Battelle, Columbus Division
505 King Avenue
Columbus, OH 43201
UNITED STATES
Gregory G. Ondich
U.S. Environmental Protection Agency
401 M Street, S.W. (RD-681)
Washington, DC 20460
UNITED STATES
Wayne A. Pettyjohn
Regents Professor, Sun Chair and
Head, School of Geology
Oklahoma State University
105 Noble Research Center
Stillwater, OK 74078
UNITED STATES
Eydie Pines
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
UNITED STATES
Ronald Probstein
Massachusettes Institute of Technology
Dept. of Mechanical Engineering
Room 3-246
Cambridge, MA 02139
UNITED STATES
John Quander
U.S. Environmental Protection Agency
Technology Innovation Office
401 M Street, S.W., OS-110W
Washington, DC 20460
UNITED STATES
Donald Sanning
NATO/CCMS Pilot Study Director
Chief, Emerging Technology Section
U.S. EPA, RREL
Superfund Technology Demonstration Div.
26 W. Martin Luther King Drive
Cincinnati, OH 45268
UNITED STATES
Lynn Schoolfield
CCMS Projects Officer
International Activities A-106
U.S. Environmental Protection Agency
Washington, DC 20460
UNITED STATES
1388
-------
Robert L. Siegrist
Oak Ridge National Laboratory
U.S. Department of Energy
P.O. Box 02008
Oak Ridge, TN 37830
UNITED STATES
Alan Sielen
NATO/CCMS Coordinator/
International Affairs Office
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
UNITED STATES
John Skinner
Director, Office of Environmental Engineering
and Technology Demonstration
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
UNITED STATES
Fred Stroud
U.S. Environmental Protection Agency
Region IV
345 Courtland Street, N.E.
Atlanta, GA 30365
UNITED STATES
Bruce Thomson
Assoc. Prof.
University of New Mexico
Dept. of Civil Engr.
Albuquerque, NM 87131
UNITED STATES
Mehmet T. Tumay
Director, Geomechanics Program
Directorate of Engineering
National Science Foundation
1800 G Street, N.W. Room 1108
Washington, DC 20050
UNITED STATES
Catherine M. Vogel
Environmental Engineer
Bioremediation Treatment Technology
HQ AFCESA/RAVW
Tyndall Air Force Base, FL 32403
UNITED STATES
Thomas J. Walker
Chief, Environics Division
U.S. Air Force Engineering and
Services Center
HQ AFESC/RDV
Tyndall AFB
Panama City, FL 32403
UNITED STATES
Andre P. Zownir
U.S. Environmental Protection Agency
Woodbridge Avenue
Edison, NJ 08837
UNITED STATES
*U.S.COVERNMEVTPRlNTINGOmC£:1993 -750-002/ 60956
1389
-------
-------
-------
-------
-------
------- |