k.   600287093
                                PROJECT SUMMARY
               REPORT ON DECONTAMINATION OF PCB-BEARING SEDIMENTS
                                      by

                               Donald L. Wilson
                         Chemical & Biological  Staff
                '  Chemical & Biological Detoxification  Branch
                      Alternative Technologies  Division
                Hazardous Waste Engineering Research Laboratory
                             Cincinnati, OH 45268
               HAZARDOUS WASTE ENGINEERING  RESEARCH LABORATORY
                      OFFICE OF RESEARCH AND  DEVELOPMENT
                     U.S. ENVIRONMENTAL  PROTECTION AGENCY
                           CINCINNATI, OHIO  45268
                                                 ''  H, Environment.?]  Protr-ntlori Agency
                                                   '  -/ J. I ,-:-ar.  (M'L-IC;
                                                     .., :•• ,.__;...^, firec-t, Koom 1670
                                                 L..::..'. o, iL   ?0604

-------
                                  PROJECT  SUMMARY


 Introduction;

      Polychlorinated  biphenyls  (PCBs)  have  in the past been widely used as a
 dielectric  fluid  in transformers  and capacitors as well as in hydraulic and
 heat-transfer  fluid systems  because of their stability and heat resistance.
 During  the  50  years that  PCBs were manufactured and used in this country, an
 estimated 400  million pounds of these  chemicals entered the environment.
 Because of  the broad  use  of  PCBs  (ranging from transformer dielectric oils to
 carbonless  paper  production), they are widespread in the modern environment,
 especially  the aqueous environment.

      Although  PCBs are a  suspected carcinogen, their acute toxicity is consid-
 ered  to be  non-to slightly toxic  by the LD50 Method.  A more significant health
 impact  has  been linked to the incomplete  combustion (burning) of PCBs.  Under
 certain circumstances, oxidation  of PCBs  form dioxins and furans, most toxic of
 all man-made substances.  While PCBs have long been known to be toxic, they have
 only  recently  been acknowledged to be  a general threat to the environment.  The
 characteristics that  make PCBs  such a  problem is their stability and therefore,
 persistence in the environment.  These chemicals have been found in measurable
 concentrations in waterways  and sediments through the world and are widely
 spread  contaminants of fish  and wildlife  resources.  The PCBs contamination
 results from an era when  industrial wastes  were disposed of by flushing them
 directly iflto  waterways,  local ..sewage  plants or landfills.

      The PCB contamination problems in New  Bedford, Massachusetts (EPA Region I),
 the Hudson  River  in New York (EPA Region  II), and in Waukegan, Illinois (EPA
 Region  V) are  reported to be the  worst in the United States in terms of concen-
 tration and total quantity of PCBs.  It is  estimated that 290,000 kg of PCBs are
 contaminating  382,000 M 3(500,000 yd 3) of  sediments of the Hudson River.  Dur-
 ing the 70's approximately 907,000 kg  of  PCBs were used in the New Bedford area
 annually, of which an estimated 45,500 kg were improperly disposed.  Also, there
 are numerous industrial lagoons contaminated with large quantities of PCBs.  For
 example, EPA Region III reports three  sites in Pennsylvania and two in West
 Virginia where over 3200  2447 m3  yd3 of soil/sediments are contaminated with PCBs.
 The PCBs contamination problems pose threats to both drinking water and the
 fishing industry.

      The only  available proven  technology is dredging and expensive incineration.
 Land  disposal  of  the  untreated  sediments  has legal restrictions.  Biodegradation
 is a  possibility, but sufficient  information does not exist to design and operate
 such  a  system. There is  little experience  in the application of encapsulation
 technology  to  PCB-contaminated  sediments.

      Although  a great amount of work has  been done by many research groups in
-the area of PCB-contaminated liquids,  relatively little effort has been directed
 toward  PCB-contaminated sediments and  sludges.  The EPA Regional and Program

-------
                            rnmment on the technical  and economic feasibility
"oT"cKemical/ biological processes for clean-up of tnese_sediment5 and sludges^
 The&tf OfTices do not have adequate data to recommend any of a number of  processes
 proposed or being tested/evaluated for the decontamination of sediments  contain-
 ing PCBs.

   Assessment of Processes

      Because there is little known about the application of chemical  and
 biological processes to the decontamination of PCBs in sediments, Research
 Triangle Institute (RTI) was engaged by EPA to undertake a study to identify
 the most technically feasible processes that have been proposed  by research
 concerns for the removal of PCBs from sediments; to identify their extent of
 development, effectiveness, limitations and probable costs; and  to determine
 needs for their further development.  The study involved three phases: identi-
 fication of the most promising processes and their further development needs;
 evaluation of the unit operations involved in the processes against engineering,
 health, and environmental criteria for further testing and evaluation; and
 definitive assessment of the recommended treatment processes using pilot systems
 and contaminated sediments.

      The first phase involved four steps: data acquisition, screening and selec-
 tion of most technically feasible processes, development of criteria  for process
 assessment, and process assessment.  Under step one, three major sources of data
 were: EPA's file of proposals and correspondence concerning problems  of  PCB
 contamination and possible approaches to alternative solutions;  the open litera-
 ture; and_ direct contacts with proponents of treatment technologies.   In step
'two, the *procesVe'S''*we1"e' categorized according to- their generic technology ; so •-
 that their potential performance could be judged appropriately.   Processes with
 undesirable aspects, for example lack of tolerance for water, were rejected
 from further assessment.  Criteria for assessment*, step three, were chosen which
 relate to a broad range of principles of operation of diverse applied tech-
 nologies, yet can be used effectively in comparing one treatment process with
 another.  Additional factors, specific to a technology, were included to help
 portray the inherent strengths and limitations of a process.  The seven  crite-
 ria used and three additional factors were: estimated residual PCB; available
 capacity; conditions/limitations; concentration range handled; status of develop-
 ment; test and evaluation data needs; estimated cost; unit operations; RCRA
 waste generated; and estimated destruction/detoxification/removal  efficiency.
 In the final step, the processes were assessed by characterization and ranking.
 Characterization provided for objective comparison of the processes.   Ranking
 provided a subjective comparison of the processes based on the seven  criteria.

      In the first phase eleven emerging alternative treatments of the sixty
 four processes considered for PCB contaminated sediments were compared and
 ranked.  Eight candidate treatment processes showed potential  as alternatives
 to chemical waste landfill  and to incineration.  Some had been tested on soils;
 but none were tested specifically on PCB contaminated sediments.  Sediments of
 concern differ from soils in several properties that influence the performance
 of unit operations involved in the treatment processes.

-------
     The second phase built upon the first phase of the assessment so that EPA
may have more complete information before evaluating any technology in the field
This second phase involved three steps: consulting with the treatment process
developers, technical assessment of the process, and the selection of three of
the processes for thorough test and evaluation.  Consultations included visits
to the developers' facilities where further information was developed.  The
most recent performance data were reviewed.  The availability of the system was
determined.  The visits included a checklist of needed information and criteria
for assessment of the processes.  The Technical assessment was based upon the
projection of the performance and coTtjJjra large-scale treatment system of
sHffment_ capacity tQ Jtreat an esfTmated 3&Q-flflQjn3of PCB-contaminated Hudson
R'iver sedimentsin2i5__y_eajis_.  Specific sample data from Hudsgj^River sediments
                                 ~                                 fnr
ment bv_each process assessed.  The processes were compared and rated using the
results of the assessment.  Based on composite ratings, three processes showing
the highest rating were recommended for test and evaluation.

     The objective of the proposed third phase is to provide EPA with a defensi-
ble thorough test and evaluation of the three alternative treatments selected
in Phase 2.  In carrying out this phase. EPA will arrangp tn havp pr.R-c.nntami-
n ajted sediments treated in pilot-scale processing systems selected in Phase 2 .
The~pil"ot tests will evaluatel&erf_Qjmajic£-iindTr flfLLfl concnCTonsT and win  be
suj3£0£ledL- by labopatory tests.  Confirmation of treatment process effectiveness
will make these processes very viable and cost-effective methods of treating
these hazardous wastes.

     The following processes were studied by RTI on Phase 2 of the study:

B.E.S.T.

     The Basic Extraction Sludge Treatment (B.E.S.T.) process has been developed
by the Resources Conservation Co. (RCC), 3101 N.E. Northup Way, Belleview,  WA
98004.  The process uses a solvent having an inverse critica solution point in
water to remove water and oily material from solid matter.  It has been applied
to clean-up PCB-contaminated oily sludges at a CERCLA site (General Refining
Site, Savannah, Georgia).

Ultrasonics/Hydrogen - Ozone/UV Technology

     The process is based upon simultaneous extraction and treatment of a sedi-
ment slurry with ultrasound (above 20KHz) added to increase rates of dissolu-
tion of the PCB contaminants and reduce coalescence of bubbles of gs reactants.
The treatment is achieved by either ozone/ultraviolet irradiation or hydrogen/
ultraviolet irradiation.  The sediments are fed as a slurry containing about
20 wt. percent solids into a mixing tank where predetermined amounts of detergent
and sodium hydroxide are added.  The conditioned mix is pumped to the reactor
where the solids are maintained in suspension while the mixture is exposed to
ultraviolet irradiation and ozone (or hydrogen) is added.  During this treatment,
microscopic turbulence is produced through ultrasonics.  The treated slurry is

-------
 fed into a cyclone where the solids  are removed,  sampled, and tested to certify
.them for discharge.  The separated water is  neutralized and also tested before~
 discharge.

 Bio-Clean Naturally-Adapted  Microbial

      The Bio-Clean Naturally-Adapted Microbial  process has been developed by
 Bio-Clean, Inc.   Bio-Clean is  a  company engaged in developing process systems
 to  clean-up,  remove or  destroy hazardous chemicals in the environment.  The
 process  utilizes  Arthrobacteria  sp and/or other naturally-adapted microbes to
 destroy  PCBs  (and related organics)  under aerobic conditions.

 KPEG with DMSO

      Nucleophilic substitution with  the KPEG process and the Galson Research Corp,
 version  (Terraclean-Cl  process)  was  looked at as a means of decontaming PCB-
 bearing  sediments.   The  potassium polyethylene glycolate (KPEG) process has two
 potential  applications  in the  treatment of PCB-contaminated sediments:  The
 treatment of  the  sediments themselves,  or the treatment of concentrated PCBs
 resulting from extraction processes.  The former is assessed in terms of the
 Galson Modified KPEG process,  that is with dimethyl sulfoxide (DMSO).  The
 latter application  would require treatment of a relatively concentrated solution
 of  PCBs  in a  non-aqueous solvent (triethylamine, kerosene, etc.).

 CFS  Propane Extraction

      The  CFS  Propane Extraction  Process  is a technology available from C. F.
 Systems Corporation,  25 Acorn  Park,  Cambridge, Massachusetts 02140.  The company
 is  a  subsidiary of  Arthur D. Little.  As  applied to PCB-contaminated sediments,
 the  process uses  propane at  ambient  temperature and 1378 kPa (200 lb/in2) to
 extract PCBs  along  with other  oily organics from a water slurry of the sediment.
 The  treated slurry  is discharged after  separation from the liquid propane which
 contains  dissolved  contaminant.  The propane solution is fed to a separator
 where the solvent  is  removed by vaporization and recycled.  The contaminants
 are drawn  off as  a  concentrate for final  treatment. The process has been tested
 for  PCB-containing  refinery  sludge.  The  PCB content of the solids component of
 the  sludge was reduced to-S-ppm-s—    ~

 Modar Supercritical  Water - Oxidation

     The  process  has been developed by  MODAR, Inc., 3200 Wilcrest, Suite 220,
 Houston,  Texas 77042.  The process  utilizes water above critical  conditions
 (374°C and 22.1  MPa) to increase the solubility of organic materials and
oxygen and effect a  rapid oxidation, destroying organic contaminants.   It has
been applied to destroy PCBs  in oils, and to decontaminate dioxin tainted soil.
The company has  a laboratory test unit and a skid-mounted  pilot  test unit having
a nominal organic material flow capacity of 190 Liters/day.   This unit  has bee
operated at MODAR's  research  facility in  Natick, Massachusetts  and at  the CECOS
 International  Niagara Falls  site.

-------
 In Situ Vitrification

? -» ••   In Situ Vitrification was previously discussed-as a process -developed -to '  •
 treat radionuclide-contaminated wastes, and subsequently tested for possible
 adaptation to the treatment of PCB-contaminated soils.  Submerged sediments
 would be dredged before treatment.  The process stabilizes contaminated  soij_s
 by melton into a_durable glass andcrysta'Tline form.  Four electrodes are
 inserted into the sediments in a square array.  A~path for electric current is
 made by placing a mixture of graphite and glass frit between the electrodes.
 Dissipation of power through the starter materials creates temperatures  high
 enough to melt a layer of sediment, which establishes a conductive path. The
 molten zone grown downward through the contaminated soil  At the high tem-
 peratures created (<1700°C) organic materials pyrolyze, diffuse to the sur-
 face, and combust.  Off-gases are collected, monitored, and treated.

 Solvent Extraction Process

      In a separate study, scientists at the New York University's Department of
 Applied Science have been carrying on a three year EPA supported investigational
 program on the development and evaluation of a low energy process technology
 for the solvent extraction and subsequent chemical destruction of PCBs from
 contaminated sediments and sludges.  This is also one of the processes which was
 assessed by RTI in Phase 2.  This program involves experimental studies  on poten-
 tially more cost-effective PCB disposal alternatives to currently accepted
 practices.  The research activities during the first year of this program
 comprised the investigation of suitable solvent extraction and supporting
 equipment, identification of candidate solvents for extraction and concurrently
 developed^ application and verification of analytical methodology for quantitative
 determination of PCBs in contaminated sediments.  Exploratory experiments were
 also carried out on the extraction of PCBs from contaminated sediments obtained
 from Waukegan Harbor, Illinois.

      The concept of using kerosene as an extraction solvent for removing PCBs
 from sludges and sediments was based on two factors: 1) the low cost of  kero-
 sene and 2) some work conducted on extraction of PCBs from soils (M. B.  Saunders
 at Oak Ridge National Laboratory, 1985).  While the cost factor is a driving
 force in any development, it alone cannot overcome technical impracticalities.
 It became quite evident early in this study that the presence of large quanti-
 ties of water in the contaminated samples presented an insurmountable
 barrier to the use of kerosene in this application.  A new approach to
 ±his difficult problem was needed and is herein presented.

      The first step of the process takes advantage of the extremely low solubil-
 ity of PCBs in water.  In a typical PCB contaminated sediment or sludge composed
 primarily of water with generally 20% total solids, virtually all  of the PCBs
 are associated with the sediment.  Thus, the first processing step is a  physical
 separation of water and solids.  In the second step, the PCB contaminated oil
 is disociated from the sediment substrate.  This is accomplished by a multi-
 stage counter-current leaching with a hydrophyllic solvent.  The third step is
 a  stripping operation in which the PCB-containing stream from above is contacted
 in a liquid-liquid extractor with a hydrophobic solvent and additional water.
 The two streams which leave this step are the PCBs concentrated in stripping

-------
 solvent which proceeds to appropriate methods for final  destruction, and the
 hydrophyllic solvent/water mixture containing trace PCBs.   The  solvent/ water
 mixture goes next to,a disti 11-ation-colamn-fn-which the  solvent Is separated '-   "
 and returned to the leaching process.  The water contaminated with trace amounts
 of PCBs is recycled to the front of the stripping process  to close the cycle.

 Conclusions

      Various experimental and supporting activities were carried  out during the
 first year of the EPA-supported research in order to develop an adequate data
 base for achieving an effective low energy process for the extraction of PCBs
 from contaminated sediments and sludges.  Achievements from these investigations
 are summarized as follows:

      1.  Selection of extraction and supporting equipment  for a specific extrac-
          tion.
      2.  Identification of candidate solvent systems for extraction.
      3.  Development of an appropriate analytical methodology for the deter-
          mination of PCBs in contaminated sediments and  sludges.
      4.  Development of an innovative process scheme for the extraction of PCBs
          from sediments and sludges.

      Emerging treatment processes for decontamination of sediments containing
 PCBs that show potential  as alternatives to incinceration  and chemical waste
 landfill have been identified in Phase One of an assessment study. Eleven
 alternative treatments were compared and ranked using technical performance,
 status of development, test and evaluation data needs, and cost as factors.
 The first'eight processes sfrow potential for reduction of  PCB concentrations to
 the desired background levels (l-5ppm) or less, with minimum environmental
 impacts and low to moderate cost.   The sediments must be dredged  for application
 of these treatments.  Based on the weighted ratings of Phase One, eleven
 processes ranked as follows from highest to lowest: KPEG,  LARC, Acurex, Bio-Clean,
 Modar-Supercritical Water, Advanced Electric Reactor, Vitrification, OHM Extraction,
 Soil ex, Composting, and Sybron Bi-Chem 1006.

      At the start of the final assessment study before field tests, some former
 developers indicated their processes should not be considered and some new
 processes became available. Thus,  in the next assessment study-Phase Two, eight
 emerging treatment processes for decontamination of P£B-contaminated sediments
 were evaluated as candidates for thorough test and evaluation using a test
 system of sufficient size to provide performance, co-st,  and seale-u-p data for a
 large commercial plant.  The processes assessed and listed in relative desirability
 of test and evaluation include: (1) Basic Extraction Sludge Treatment (B.E.S.T.)
 developed by Resources Conservation Co.; (2) UV/Ozone or Hydrogen/Ultrasonics
 Technology of Ozonic Technology, Inc.; (3) Bio-Clean Naturally-Adapted Microbial
 Process of Bio-Clean, Inc.; (4) KPEG (Galson) Process modified  by Galson Research
 Corp.; (5) Low Energy-Acetone-Kerosene Extraction Process  developed by the Applied
 Science Dept. of New York University; (6) Modar Supercritical Water Oxidation
.Process of Modar, Inc.; (7) CFS Propane Extraction Process of C.  F. Systems Corp.;
 and (8) In-Situ Vitrification Process by Battelle Pacific  Northwest Research
 Laboratories.

-------
Recommendations

     While all the processes except In Situ Vitrification appear to merit
further development for this application, those three with the highest  compar-
ative ratings are recommended for EPA-supported thorough test and evaluation.
These are the Basic Extraction Sludge Treatment, UV/Ozone or Hydrogen/Ultrasonics
Technology, and Bio-Clean Naturally-Adapted Microbe processes.

     There is, in an evaluation of processes such as this, much uncertainty.
Additional research of the other processes is very important to fill  in the
data gaps.  The other processes, for example, may be more cost effective on a
smaller scale than what the study assumed or with other unit processes. The
results of the EPA funded study at the New York University have led to  a novel
approach to treatment of PCB-contaminated sediments and sludges.  It is
recommended that a study is needed to model, engineer and otherwise fully
characterize this process so that its full potential might be realized.

-------
REPORT ON DECONTAMINATION OF PCB-BEARING SEDIMENTS
                        by

                 Donald L. Wilson
           Chemical  & Biological  Staff
   Chemical  & Biological  Detoxification Branch
        Alternative  Technologies  Division
 Hazardous Waste Engineering Research  Laboratory
               Cincinnati, OH 45268
 HAZARDOUS  WASTE  ENGINEERING  RESEARCH  LABORATORY
        OFFICE  OF  RESEARCH  AND  DEVELOPMENT
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
             CINCINNATI, OHIO  45268

-------
                                    NOTICE
     This report has been reviewed by  the Hazardous Waste Engineering Research



Laboratory, U.S. Environmental Protection Agency, and approved for



publication.  Approval does not signify  that  the contents necessarily reflect



the views and policies of the U.S. Environmental Protection Agency, nor does



mention of trade names or commercial products constitute endorsement or



recommendation for use.
                                     ii

-------
                                      FOREWORD


     Today's rapidly developing and changing technologies and industrial  prod-
ucts and practices frequently carry with them the increased generation of solid
and hazardous wastes.  These materials, if improperly dealt with, can threaten
both public health and the environment.  Abandoned waste sites and accidental
releases of toxic and hazardous substances to the environment also have impor-
tant environmental and public health implications.  The Hazardous Waste Engineer
ing Research Laboratory assists in providing an authoritative and defensible
engineering basis for assessing and solving these problems.  It's products
support the policies, programs and regulations of the Environmental  Protection
Agency, the permitting and other responsibilities of State and local  government
and the needs of both large and small businesses in handling their wastes
responsibly and economically.

     This report describes an in-depth assessment of emerging alternative treat-
ment processes that show potential for decontaminating polychlorinated biphenyl
(PCB)-contaminated sediments and the selection of processes for thorough  test
and evaluation.  The comparison of these processes as to their technical  per-
formance, the availability of a test system, the test and evaluation effort
required, the projected time to commercial availability, and the probable cost
of their application should be useful to EPA Regional Offices, to those con-
cerned with developing hazardous waste treatment regulations, and to those
interested in the development of new and innovative treatments.

     For further information, please contact the Alternative Technologies
Division of the Hazardous Waste Engineering Research Laboratory.
                                    Thomas R. Hauser, Director
                                    Hazardous Waste Engineering
                                      Research Laboratory
                                        iii
                                       v

-------
                                    ABSTRACT


     In recent years, polychlorinated biphenyls (PCBs) have been discovered
as contaminents in sediments throughout the country.  Conventional  and emerg-
ing waste management methods have not been evaluated for decontamination of
PCB-bearing sediments.

     The EPA has initiated a research program to Identify chemical/biological
methods as alternatives to incineration and to chemical land disposal  for clean-
up of PCB contaminated sediments.  The overall objective of the program is to
identify, validate, and demonstrate effective and economical chemical/biological
processes for removal/destruction of PCBs in sediments.  This report summarizes
research progress on chemical/biological methods development for the detoxif-
ication/destruction of PCBs in sediments.
                                  iv

-------
                                    CONTENTS
Notice	ii
Foreword	1 i i
Abstract	i v
Tables	vii
   1. INTRODUCTION	1
          PCBs in Sediments	3
              Solvent Extraction Process	4
              Assessment of Processes	5

   2. CONCLUSIONS	9

   3. RECOMMENDATIONS	12

   4. DISCUSSION ON EXTRACTION PROCESS	14
          Selection of Equipment	14
              Survey of Industrial Extraction Equipment	14
              Settler Design	16
              Contractor Selection	17
          Experimental Studies	19
              Extraction of PCBs from Spiked Water Samples	19
              Quantification of PCBs in Environmental Sample	19
              Characterization of Solvents	20
          New Process Devel opment	21
    5. DISCUSSION OF ASSESSMENT STUDY	23
          Screening of Processes	23
          Development of Evaluation Criteria	28
              Problem Definition	28
              Regulatory Factors	28
              Technical Factors	28
              Cost Est imates	29

-------
                       CONTENTS (continued)
   Initial Assessment of Processes	31
       Processes Based on Chemical Technologies	31
       Processes Based on Physical Technologies	38
       Processes Based on Biological Technologies	41
   Initial Characterization and Ranking of Processes	45
       Character^ zati on	45
       Ranki ng	48
   Final Assessment of Processes	53
       Ranki ng Cri ten' a	53
       B .E .S .T	55
       Ultrasonics/Hydrogen-Ozone/UV Technol ogy	56
       Bio-Clean Naturally-Adapted Microbial	57
       KPEG with DMSO	57
       CFS Propane Extraction	59
       Modar Supercritical Water-Oxidation.,	59
       Low Energy Acetone-Kerosene Extraction	60
       In Situ Vitrification	60
   Selection of Three Processes	62

APPLICATION OF PROCESSES TO RCRA WASTES	70
                            VI

-------
                                     TABLES
Number                                                             Page
1      Basic Classification of Industrial Contactors               14
2      Summary of Features and Fields of Application               15
3      PCB Concentration of Waukegan Harbor Sludge                 20
4      Screening of PCB Treatment Processes                        24
5      Unit Cost Estimates                                         30
6      Treatment Processes Characteristics                         46
7      Treatment Processes Ranking                                 51
8      Acceptable and Borderline Values for Processes
       Characteristics                                 .            62
9      Overall  Desirability of Immediate T and E of the Seven
       Candidate Processes          .    .                          64
                                   VII

-------
                                   SECTION I
                                  INTRODUCTION

     Polychlorinated biphenyls (PCBs) are a class of compounds  that  have  209
combinations of chlorine atoms attached to the biphenyl  molecule.  When they
^ere first introduced into the United States by the Monsanto Company in 1929,
PCBs were considered a "wonder" chemical.  PCBs are clear, odorless  liquids
which have thermal, biological and chemical stability, plus high dielectric
constants.  These characteristics made PCBs have widespread use.  Initially,
PCBs were used as dielectric fluids in capacitors and transformers.   A small
amount was used in electromagnets, heat transfer and hydraulic  systems.   PCBs
applications spread to plasticizers in paint, adhesives, dry caulking compounds,
carbonless paper, flame retardants and fillers of all sorts.

     Between 1929 and 1975, over 1.25 billion pounds of PCBs have  been manu-
factured and used in the U.S., primarily in mixtures with chlorobenzenes  (mostly
trichlorobenzene or TCB) known as askarel (typical 40% PCB and  60% TCB).   Askarels
were used as dielectric fluids for electrical equipment and consisted of  about
77% or 965 million pounds of the PCBs produced.  It is estimated that 328 million
pounds remain in service in electrical systems.

     These compounds were not discovered in the environment until  1966 in
Sweden and 1967 in the U.S. despite their use commercially for  about 40 years.
Interest in PCBs arose because they were frequently found in fish, bird,  water,
sediment and other environmental samples when such samples were examined  for
chlorinated pesticide residues.

     Although PCBs have long been known to be toxic, only in recent  years have
they been acknowledged to be a general threat to the environment.  Extensive
research>has shown a link between PCBs and chronic (longterm) effects in  many
species at less than lOppm.  There are well-documented tests on laboratory
animals that show PCBs cause developmental toxicity, reproduction  abnormalities,
and tumors.  Even though PCBs are a suspected carcinogen, their acute toxicity
is considered to be non- to slightly-toxic by the LD50 Method.   Recently, there
have been additional concerns brought on by the finding of other toxic contami-
nants in PCBs and askarel fluids or linked to their incomplete  combustion
(burning).  These toxic compounds include polychlorinated dibenzodioxins  (PCDDs),
polychlorinated dibenzofurans (PCDFs), chlorinated benzenes and other chlorinated
substances.  These compounds are among the most toxic of all man made substances.
Because of these concerns, in 1971, the Monsanto Company, the sole United States
producer, terminated sales of PCBs for all but closed electrical systems  uses.


     Federal regulations of PCBs took several years to develop.  In  1976,
Congress enacted the Toxic Substances Control Act (TSCA), PI 94-469, and  included
special provisions, Section 6e, for the regulation of PCBs.  In the  May  31,  1979
Federal Register 44 (106), the EPA promugated rules which prohibited the  manu-
facture of PCBs and their use in almost all nonenclosed applications.  Since 1979,
many regulations, rules, modifications to existing .regulations, additional rules,
clarification, and re-evaluated EPA positions have been issued.  The position  has
moved toward the direction of tighter regulatory control.  The Chemical  & Bio-
logical Staff conducted a regional survey in early 1987 on the acceptable PCBs

-------
levels in soils and sediments.  There existed some variation between the  EPA
Regional Offices on the clean-up policies for spills of TCBs'on soils and
sediments.  On April 2, a new policy for cleaning spills  of PCBs that sets  the
same standards for every region in the country was released by EPA,  after two
years of negotiations among the Agency, environmental groups and industry (40
CFR, Part 761, April 2, 1987, pp 10688-10710).  The policy calls for decontami-
nation of soil down to lOppm PCBs by weight in residential areas. This soil
would then be covered by a clean-soil cap.  Less restricted areas calls for
clean-up down to 25ppm or BOppm.  Disposal of treated soils/sediments must  meet
the criteria of less than 2ppm PCBs concentration.  The regulations  permit  PCBs
to be landfilled or burned only in special incinerators,  cement kilns or  boilers,
They also allow the EPA Regional Administrator to approve other alternative
technologies to be used if he finds they do not "present  an unreasonable  risk of
injury to health or the environment", and if their "level of performance  is
equivalent to Annex I incinerators or high efficiency boilers".  However, the
March 30, 1983 procedural rule on authority over alternative methods is still in
effect.  Under this ruling, incinerators and alternate methods that  are mobile
or of identical design and intended to be used in more than one region must
obtain approval from the Assistant Administrator for Pesticides and  Toxic
Substances.  Also included are certain research and development methods that
dispose of more than BOOlbs of total PCBs or PCB contaminated material.   This
report discusses several alternative technologies and their abilities to  meet
these regulations.

-------
                               PCBs in Sediments


     Polychlorinated biphenyls (PCBs) are widely used industrial chemicals
that were never  intended for release into the environment, but they have none-
theless been found in rainwater, in human tissue, and in many species of birds
and fish.  PCBs  are released into the environment by several means.  Sewage
outfalls and industrial disposal into waterways are major sources of PCB-contami-
nation.  Rivers  and streams are primary means of transport for PCBs within the
environment, although the atmosphere also is a means of dispersal.  It is esti-
mated that from  1000 to 2000 tons per year escaped into the atmosphere in the
1960s from plasticized materials containing PCBs, and that about 4000 tons per year
entered waterways from dumping and leakage of lubricants, hydraulic fluids, and
heat transfer fluids.  Other sources include leaching from dumps and landfills,
where about half of the PCBs produced eventually end up.  Once in the environment,
PCBs persist for a long time.  Their physical property of low water solubility
caused the PCBs  present in waterways to accumulate in sediments.

     Reported the worst PCBs-sediment contamination sites in the United States,
in terms of concentration and total quantity, are: New Bedford, Massachusetts
(EPA Region I);  Hudson River, New York (EPA Region II); and Waukegan, Illinios
(EPA Region V).  It is estimated that 290,000 kg of PCBs are contaminating
382,000 m3(500,000 yd3) of sediments of the Hudson River.  During the 70s,
approximately 907,000 kg of PCBs were used in the New Bedford area annually,
of which an estimated 45,500 kg were improperly disposed.  Also, there are
numerous industrial lagoons contaminated with large quantities of PCBs.  For
example, Jthe following is a partial list of PCBs removals which have occurred
in EPA Region III:

  Site Name                   Site Location        Approx. cu.yd. Contaminated
Interstate Transformer        Lawrence Co., PA                   200

Paoli Railroad                Paoli, PA                          800

Sandonelle                    Mercer Co., PA                     600

Shaffer Equipment             Minden, WV                        1600

Frazier's Bottom              Putnam Co., WV                    unknown

     These PCB contaminated sites pose threats especially to both drinking
water and the fishing industry.  The only available proven technology for PCB
contaminated sediments is to remove them by dredging and using incineration
which is costly.  Land disposal of the untreated sediments is an option; but,
as previously noted, this option has many legal  restrictions and concentration
limitations.  Biodegradation is a possibility, but sufficient information does
not exist to design and operate such a system.  There is little experience in
the application of encapsulation technology to PCB-contaminated sediments, plus
the large volumes involved may make it cost prohibitive.

-------
     Although a great amount of work  has been done by many research groups
 in the area of PCB-contaminated liquids, relatively little effort has been
 directed toward PCB-contaminated sediments and sludges.  The EPA Regional
 and  Program Offices are being asked to comment on the technical and economic
 feasibility of chemical/biological processes for clean-up of these sediments
 and  sludges.  These Offices do not have adequate data to recommend any of a
 number of processes proposed or being tested/evaluated for the decontamination
 of sediments containing PCBs.  The Hazardous Waste Engineering Research Lab-
 oratory (HWERL) past and current research in the area of decontamination of
 PCB-bearing sediments is addressed to these Agency needs.

 Solvent Extraction Process

     Extraction of PCBs from sediments is viewed as an attractive alternative
 to either incineration or burial.  Solvent extraction processes are defined
 to be separation processes in which two immiscible or partially soluble liquid
 phases are brought in contact for transfer of one or more components.  For
 PCBs removal from sediments, one phase will be the sediments containing various
 amounts of PCBs.  A key element in extraction, the ability of a solvent to
 extract a particular solute, is measured by the distribution coefficient.
 Choice of solvent must be optimized for maximum separation effectiveness.  Low
 energy requirements and low/moderate capital costs are some advantageous
 features of extraction processes.

     PCBs are soluble in most organic solvents and insoluble in water and
 polyhydroxy liquids.  Because of their polar nature, PCBs are more soluble
 in many pdlar solvents than various "hycfrocaYbbns.  The characteristics of a
 good extraction solvent should include high solubility for the PCBs, also a
 low  solubility for oils and a considerable solvent power for water.  Using
 these concepts, experimental studies were initiated about 1980-84 at the
 Union Carbide Corp. Nuclear Division, Oak Ridge, TN, on the extraction of PCBs
 from contaminated sludges in a lagoon using a hydrocarbon as the extraction
 solvent.  It is understood that solvents such as hexane, furfural, triethyl-
 phosphate, etc. have been investigated in the laboratory.  Ultimate disposal of
 the  extracted PCBs was achieved by incineration.  This extraction was used
 essentially only for PCBs contamination in a smaller oil volume.

     During early 1985, exploratory experimental studies being conducted at
 HWERL showed that kerosene is a good solvent for extraction of PCBs from sedi-
ments.  The Department of Applied Science, New York University, was engaged to
 conduct extensive experimental studies which will develop and optimize the
extraction process.  They are to demonstrate both its technical effectiveness
 and economic viability with a pilot plant operation.  The research activities
during the first year of this program comprised the investigation of suitable
 solvent extraction and supporting equipment, identification of candidate solvents
for extraction and concurrently develop application and verification of analytical
methodology for quantitative determination of PCBs in contaminated sediments.
Exploratory experiments were also carried out on the extraction of PCBs from
contaminated sediments obtained from Waukegan Harbor, IL.

                                        \.
                                      .4- -

-------
  In  Situ  Vitrification

- •» -   In  Situ  Vitrifi-cation was  previously discussed as a process -developed -to ' •
  treat  radionuclide-contaminated wastes, and subsequently tested for possible
  adaptation  to the  treatment  of  RGB-contaminated soils.  Submerged sediments
  would  be dredged before treatment.  The process stabilizes contaminated sojjjs
  by  melton into a^durable glass  andcrystanine form.  Four electrodes are
  inserted into tTfe  sediments  in a square array\A~path for electric current is
  made by  placing a  mixture of graphite and glass frit between the electrodes.
  Dissipation of power through the starter materials creates temperatures high
  enough to melt a layer of sediment, which establishes a conductive path.  The
  molten zone grown  downward through the contaminated soil  At the high tem-
  peratures created  (<1700°C) organic materials pyrolyze, diffuse to the sur-
  face,  and combust.  Off-gases are collected, monitored, and treated.

  Solvent Extraction Process

      In a separate study, scientists at the New York University's Department of
  Applied Science have been carrying on a three year EPA supported investigational
  program on the development and evaluation of a low energy process technology
  for the solvent extraction and subsequent chemical destruction of PCBs from
  contaminated  sediments and sludges.  This is also one of the processes which was
 assessed by RTI in Phase 2.  This program involves experimental studies on poten-
 tially more cost-effective PCB disposal  alternatives to currently accepted
  practices.  The research activities during the first year of this program
 comprised the  investigation of suitable solvent extraction and supporting
 .equipment, identification of candidate solvents for extraction and concurrently
 developed^ application and verification of analytical methodology for quantitative
 determination of PCBs in contaminated sediments.   Exploratory experiments were
 also carried  out on the extraction of PCBs from contaminated sediments obtained
 from Waukegan Harbor, Illinois.

      The concept of using kerosene as an extraction solvent for removing PCBs
 from sludges and sediments was based on  two factors: 1) the low cost of kero-
 sene and 2)  some work conducted on extraction of  PCBs  from soils (M. B. Saunders
 at  Oak Ridge National  Laboratory, 1985).  While the cost factor is a driving
 force in any development, it alone cannot overcome technical  impracticalities.
  It  became quite evident early in this study that  the presence of large quanti-
 ties of water in the contaminated samples presented an insurmountable
 barrier to the use of kerosene in this application.  A new approach  to
 this difficult problem was needed and is herein presented.

      The first step of the process takes advantage of  the extremely  low solubil-
 ity of PCBs  in water.   In a typical  PCB  contaminated sediment or sludge composed
 primarily of water with generally 20% total  solids, virtually all  of the PCBs
 are associated with the sediment.  Thus, the first processing step is a physical
 separation  of water and solids.  In  the  second  step, the PCB contaminated oil
 is  disociated from the sediment substrate.  This  is accomplished by  a multi-
 stage counter-current  leaching with  a hydrophyllic solvent.  The third step is
 a stripping  operation in  which the PCB-containing  stream from above  is contacted
 in  a liquid-liquid  extractor with a  hydrophobic solvent and additional  water.
 The two streams which  leave this step are the PCBs concentrated in stripping

-------
ination and possible approaches to alternative solutions proved invaluable in
directing the contacts, .for. further information,,	     ..   .    ...-.. . , _.,	

     A bibliography was prepared, which included approximately 100 references
containing information useful in defining potential treatment technologies and
in analyzing treatment process performance data.  This bibliography was ex-
panded to 171 references by addition of treatment feasibility study reports,
process test and evaluation reports, process development proposals, and
patents.  As processes were identified, direct contacts were made  with  the
investigators for details of their process studies.

     Alternative (non-incineration) destruction/detoxification/removal  (D/D/R)
processes were subjected to screening to identify those to be assessed  fur-
ther.  The processes were categorized according to their generic technology so
that appropriate technical screening criteria could be applied for each.  This
screening yielded 11 emerging processes for further assessment. ,The selected
processes represent six different generic technologies.

     Criteria used to assess the identified emerging technologies  included:
residual PCB concentration in treated sediments, available capacity, condi-
tions/limitations (including range of PCB concentration handled),  the develop-
mental status, test and evaluation (T and E) data needs, and estimated  cost of
treatment.

     The assessment of the different processes required consideration of the
characteristics inherent in the criteria.  Their capability to remove PCBs
adequately*from wet contaminated sediments, the required processing time,  and
process controllability~were all "considered.  Available data were  examined for
limits on the range of PCB concentrations that could be adequately reduced by
the process to yield a treated sediment eligible for delisting.  The status of
development, whether laboratory-scale or pilot-scale, was identified.  The
extent of available data - whether showing only tests of the concept or lists
of the unit processes of a pilot-scale process system - was identified.  A
preliminary estimate of the cost of applying the process was developed  in
conjunction with the Developer.

     Because the processes identified were at various stages of development,
data deficiencies exist, and engineering judgement was used to supplement
available information.

     Using the information obtained and developed, the characteristics  for each
process were given subjective ratings relative to an arbitrarily-defined per-
fect process.  These ratings were then weighted and summed to give a single
number that could represent the overall relative rating of each process.

     The product or report from this study is entitled "PCB Sediment Decon-
tamination-Technical/Economic Assessment of Selected Alternative Treatments",
600/2-86-112, Dec. 1986 (PB87-133112/AS).  Section 3 of this report describes

-------
the  screening of  alternative technologies for technical feasibility to treat
PCB-contaminated  sediments.  Section 4 describes the development of criteria
to assess  selected  processes.   Section 5 describes the assessment of the
processes,  including descriptions,  analyses of performance potential, and cost
estimates.   Section 6  summarizes the characteristics of the selected processes
and  ranks  them  subjectively.   Section 7 presents a bibliography of references
used and other  relevant  sources.  Contacts with developers of the processes
are  noted  throughout the text.

      In November  of 1986,  HWERL funded a second phase study with RTI.  The
purpose of  this study  is to establish suitable factors for further assessment
of the candidate  processes that have been identified, to review these processes
against these factors  and  identify  additional data needs, and to provide a basis
for  the selection of three processes for a defensible, thorough technical assess-
ment (including laboratory experiments and field validation) in a third phase
of assessment.

      This  study builds upon the first phase of the assessment so that EPA may
make better use of  funds by having  more complete information before evaluating
any  technology  in the  field.   Factors considered include:  (a) combinations of
unit processes, for example, a solvent extraction process combined with KPEG
or incineration;  (b) engineering evaluation criteria; and (c) specific data
needs for  full  test and  evaluation  of the applicability of the processes.

      The study  involved  three  phases:  consultations with the treatment process
developers, technical  assessment of the process, and the selection of three of
.the  processes fp.r^t.hp.ro.ush.te.st. and_evaljjattpn. :,, ,  ^_^__   .- „
      Consultations  Included  visits  to  the developers'  facilities where further
 information  was  developed.   The most  recent  performance data were reviewed.
 The  availability of a  system for test  and evaluation was determined.  The
 likely  future commercial  availability  was also  determined.  Preparations for
 the  visits included a  checklist of  needed information  and criteria for assess-
 ment of the  processes. The  checklist  is based  upon the information needs for
 a  permit to  test application.

      Technical assessment of the processes was  based upon the projection of
 the  performance  and cost  of  a large-scale treatment system of sufficient
 capacity to  treat an estimated 380,000 m3 of PCB-contaminated Hudson River
 sediments in 2.5 years.   Sediment characteristics  that may influence the per-
 formance of  an alternative treatment  process were  considered.  Specific sample
 data from Hudson River sediments were  used to guide the projection of operating
 conditions required for treatment by  each process  assessed.

      In this report, each process is  described  as  it would be applied to the
 treatment of sediments.   Material and  energy balances  are applied to the extent
 possible with the limited data available.  Needs for further data are identi-
 fied by type, estimated time, and cost for their acquisition.

-------

     The probable cost of treatment after demonstration and approval of the
process is estimated.  Cost elements given include:  capital, energy, labor,
maintenance, process quality control and testing.  The capital cost is recov-
ered over the 2.5 years of projected operation.  Labor and profit/contigency
are estimated at uniform rates for all processes for purposes of comparative
evaluation.

     Environmental and health characteristics assessed include all process feed
and waste streams, reagents, and operating hazards.

     The lapsed time required to demonstrate process performance, then to design,
construct and check out a full-scale process is projected, based on needs for
additional data and requirements of the developers.

     The processes are compared and rated using the results of the assessment.
Based on composite ratings, three processes showing the highest rating are
recommended for immediate test and evaluation.

     This report also discusses a phase three study to determine the effective-
ness of the three selected alternative treatment processes to detoxify PCB-
contaminated sediments representative of those present in substantial  quantities
in several U.S. lagoons and waterways.  The purpose of this phase three effort
is to build upon the second phase of the assessment so that EPA has a defensible,
thorough test and evaluation of three alternative treatments selected in phase
two for PCB-contaminated sediments.  Arrangements will be made to have PCB-
contaminated sediments treated in pilot-scale processing systems.  These pilot
tests will evaluate performance under field conditions, and will  be supported -
by laboratory tests.
                                        \
                                      8

-------
                                   SECTION 2
                                  CONCLUSIONS


      Exploratory studies have established that the extraction of PCB-con-
taminated sediments can be done with readily available low cost reagents
such as acetone and kerosene.  These extractions can be carried out at near
ambient temperatures and atmospheric pressures.  Specific conclusions  reached
from the solvent extraction investigations are summarized as follows:

     0  An innovative extraction process scheme was developed to remove PCBs
        from sediments.

     0  Suitable extracting and supporting equipment was selected and  assem-
        blied for the newly designed process.

     0  Appropriate analytical methodology for the determination of PCBs  in
        contaminated sediments was developed.

     0  Leaching experiments conducted with Waukegan Harbor sludge resulted
        in very high effeciencies for the solvents acetone, methanol,  and
        isopropanol.  Kerosene showed much lower effeciencies except  in the
        case of dry sediments.

     0  Settling experiments conducted with dry topsoil containing various
        degrees of water gave the following order of settling rate:  acetone-
        methanol-isopropanol.

     Based on the results from the leaching and settling experiments,  the find-
ings from the first year of research, and taking into consideration various
physical, chemical and toxicological properties, a solvent pair has been  selected.
Acetone has been chosen as the hydrophilic solvent and kerosene the hydrophobic
solvent.

     Stripping experiments conducted with these two solvents, with varying
degrees of sulfate concentration, and with varying water/acetone ratios have
given very promising results.  A partition coefficient between kerosene and
acetone of 30 has been obtained, which allows for the completion of the stripp-
ing operation with a very small number of stages.  An even higher partition
coefficient is expected with further experimentation.

     At the end of a Phase I assessment study-for alternative treatment processes
(Dec. 1986) the following conclusions were reached after evaluating sixty four
processes.

     Emerging treatment processes for decontamination of PCB-contaminated sedi-
ments that show potential as alternatives to incineration and chemical waste
landfill have been identified.  Eight processes -- KPEG, LARC, Acurex, Bio-
Clean, Supercritical Water, Advanced Electric Reactor, Vitrification,  and
O.H.M. Extraction -- ranked highest in terms of potential for cleanup  of contam-
inated sediments.  The sediments must be dredged for application of these treatments,

-------
     Of the 11 processes completely assessed and ranked, all  but the Advanced
Electric Reactor (AER) are in various stages of development from laboratory-scale
through field tests.  The AER is a permitted treatment based  on completed  trial
burns.

     There is no immediately available capacity for any of the treatments.
Further data are needed in most cases to define the final  system design  for
the process.
                            9
     At this stage of development,(1985-86) estimated costs for application of
the emerging treatments are less than or within the range of  the costs of
chemical waste landfill, except the AER process which exceeds that of landfill.
The ranges of estimated costs (1985 dollars) on the eleven processes ranked are
as follows:

     KPEG                              $  211  to    $  378/m3
     O.H.M. Methanol Extraction           400  to       514/m3
     Advanced Electric Reactor            830  to       942/m3
     Acurex solvent wash                  196  to       569/m3
     Bio-Clean                            191  to       370/m3
     Vitrification                        255  to       548/m3
     LARC                                 223  to       336/m3
     Modar supercritical water            250  to       733/m3
     Soilex solvent extraction            856  to       913/m3
     Sybron Bi-Chem 1006 PB            unable  to estimate cost
     Composting                        unable  to estimate cost
     Chemical waste landfill              260  to       490/m3
     Incineration                       1,713  to     1,826/m3

These costs are planning estimates only.  In most cases, further research  is
needed to provide data suitable for more definite cost estimates.

     The emerging treatment processes are based on five different technologies:
one on low-temperature oxidation, two on chlorine removal, one on pyrolysis,
four on removing and concentrating, and three on microorganisms.  The four
types of technologies not yielding competitive emerging processes are:   chlor-
inolysis, stabilization, bottom recovery, and enzymes.  A search of these
technologies yielded no suitable candidate processes.

     The Phase II of the assessment study built upon the information gained
during Phase I, but was not limited to those eight processes  ranked potentially
high since new ones are always emerging.  The closer investigation of the  eight
potential processes listed from the Phase I study and four potential processes
recently developed resulted in the following conclusions.

     Seven emerging treatment processes for decontamination of PCB-contaminated
sediments are candidates for thorough test and evaluation (T  and E) using  a
test system of sufficient size to provide performance, cost,  and scale-up  data
for a large commercial plant.  The processes are:
                                      iov

-------
     0  Basic Extraction Sludge Treatment (B.E.S.T.) Process developed by
        Resources Conservation Co.

     0  UV/Ozone or Hydrogen/Ultrasonics Technology of Ozonic Technology,  Inc.

     0  Bio-Clean Naturally-Adapted Microbial Process of Bio-Clean, Inc.

     0  Potassium Polyethylene Glycolate (KPEG) with dimethyl sulfoxide (DMSO)
        Process Modified by Galson Research Corp.

     0  CFS Propane Extraction Process of C. F. Systems Corp.

     0  Modar Supercritical Water-Oxidation Process of Modar, Inc.

     0  Low Energy Acetone-Kerosene Extraction Process developed by the
        Applied Science Dept. of New York University.

     The processes were evaluated using as criteria:

     0  The probability of cleaning sediments to <2 ppm PCBs;

     0  The probable cost of treatment;

     0  The relative level of Test and Evaluation effort to be supported
        by EPA;
        v\
     0  The availability of a processing system to test; and

     0  The likely future commercial availability of the process.

     While all the processes merit further development, comparative simul-
taneous evaluation of their ratings on a scale of 0 to 1 gave the following
results:
                                            Relative desirability
           Process                          of immediate T and E

     B.E.S.T.                                           0.62
     Ultrasonics/Hydrogen-Ozone/UV Technology           0.62
     Bio-Clean Naturally-Adapted Microbial              0.61
     KPEG with DMS-G              -                      0.59
     CFS Propane Extraction                             0.58
     Modar Supercritical Water-Oxidation                0.57
     Low Energy Acetone-Kerosene Extraction             0.49

     The B.E.S.T., UV/Ozone or Hydrogen/Ultrasonics Technology, and Bio-Clean
Naturally-Adapted Microbial processes have the highest ratings, and are recomm-
ended for immediate, thorough test and evaluation.
                                      11

-------
                                    SECTION  3
                                 RECOMMENDATIONS

      The  result  of  the  previous  EPA funded  study with  NYU entitled  "Develop-
 ment  and  Evaluation of  a  Low  Energy Process Technology  for Extraction and
 Chemical  Destruction of PCBs  from  Soils  and Sludges"  (CR 812123-02-0) has led
 to  a  novel  approach to  treatment of contaminated soils  and sediments with
 potentially a much  broader  applicability.   A two year  study  is proposed to
 model,  engineer  and otherwise  fully characterize this  process so that its full
 potential might  be  realized.   The  study  will,  in the  first year, design and
 assemble  a  bench scale  working model  of  the total process.   The second year
 will  culminate in field tests  at designated sites.  While proof of  concept
 experiments  have established  a clear direction for these studies, this is a
 complex process  which must  be  analytically modeled to  fully  appreciate its
 potential.   Concurrently, engineering design considerations  are required to
 specify particular  equipment.  The  previous study has considered equipment
 which could  be used for the different steps of the process.  The second year
 of  the  proposed  study would address the  issues of process scale-up.  Equipment
 selection and experiment design  will  follow site analyses and discussions with
 the entity  responsible  for  the site clean up.

     We have been seeking a commercial sponsor for this process, and it appears
 that Camp Dresser & McKee (COM), an environmental consulting firm,  is interested
 in  the  EPA/NYU process  for  future development as it relates  to contaminated
 sediments originating in New Bedford  Harbor, Massachusetts.

     The processes  compared in Phase  I of the technical/economical  assessment
 study conducted  by  RTI  for  HWERL were general technologies and not  specific
 unit processes within those technologies, nor were they developed to the point
 of  application to sediments, but rather  PCBs in liquids or soils.   Thus, a Phase
 II  study was started which  would consider:  combinations of  unit processes;
 engineering evaluation  criteria; and specific data needs to  evaluate the
 applicability of  processes.  The assessment study is at a stage where data can
 be  used to commence the longer,  more costly Phase III field type investigation.

     The purpose  of the Phase  III effort is to build upon the second phase of
 the assessment so that  EPA  has a defensible thorough test and evaluation of the
 three alternative treatments selected in Phase II for PCB-contaminated sediments.
 In  carrying out this phase, EPA  will arrange to have PCB-contaminated sediments
 treated in pilot-scale  processing systems selected by EPA in Phase  II.  The
 pilot tests will   evaluate performance under field conditions, and will be
 supported by laboratory tests.

     There is substantial  justification to proceed in the near future with the
test and evaluation of these processes.  The developers of these processes
have the facilities required to begin the tests.  Tests can proceed one at a
time if necessary, with separate funding for each.  Allowing time for permits,
                                      12

-------
set up, and check-out, the results of one or more tests  could  be available
for consideration by those at EPA who may be appropriating  funds for the
Hudson River problem.  Tests and evaluations could be conducted during the
summer and fall  months, when the weather is much more favorable.
                                  13

-------
                                   SECTION 4
                        DISCUSSION OF EXTRACTION PROCESS
                             Selection of Equipment

     The selection of suitable contacting equipment for a specific extraction
involves complex decisions particularly so in the absence of prior data with
the same system on either a pilot plant or industrial  scale operation.  A
study was therefore carried out to ascertain the types of pilot plant and
industrial scale extractors available and to evaluate their applicability for
the extraction of PCBs from sediments and soils.  The results of this study
ares umma ri zed be!ow.

Survey of Industrial Extraction Equipment

     Industrial solvent extraction contactors emphasize equipment with counter-
current flow operational capability.  One convenient classification shown in
Table 1 is that of differential and stagewise contactors.  The former provide
conditions for mass transfer throughout their length, aiming to produce optimum
concentration profiles.  Stagewise contactors provide for a number of discrete
stages in which the two phases are brought to equilibrium, separated and passed
countercurrent to the adjoining stages.  While such contactors avoid certain
problems such as the axial mixing found in some differential contactors, the
necessity of separating the phase between each stage can add substantially to
overall size and solvent inventory.

     Another major division can be made according to whether the countercurrent
flow is produced by gravity action on the density differences between the phases
or by an applied force such as a centrifuge.  A summary of features and fields
of industrially employed extraction equipment is shown in Table 2 and includes
most commercial contactors in use today.
TABLE 1:  BASIC CLASSIFICATION OF INDUSTRIAL CONTACTORS
Forces Producing Phase
  Interdispersion

Gravity only
Pulsation
Mechanical agitation
Centrifugal Force
Differential Contactors


Spray column
Packed column

Pulsed packed column

Pulsating plate column

Rotating disc contactor
Oldshire-Rushton column
Kuhni column*
Graesser contactor

Podbielniak
Quadronic
DeLaval
Stagewise Contactors


Perforated plate column
Pulsed sieve-plate
   column*
Pulsed mixer-settler

Scheibel column
ARD extractor
Mixer-settlers
Westfalia
Robatel
*Could be considered intermediate between differential and stagewise.
                                      14

-------
            TABLE 2:  SUMMARY OF FEATURES AND FIELDS OF APPLICATION
TYPES OF
EXTRACTOR

Unagitated columns
Mixer-settlers
Pulsed columns
Rotary-agitation
columns
Reciprocating-plate
columns
Centrifugal
extractors
           GENERAL FEATURES

Low capital cost, low operating and
maintenance cost, simplicity in
construction, handles corrosive
material.

High stage efficiency, handles wide
solvent ratios, high capacity, good
flexibility, reliable scale-up, handles
liquids with high viscosity.

Low HETS, no internal moving parts,  many
stages possible.
Reasonable capacity, reasonable HETS,
many stages possible, reasonable
construction cost, low operating and
maintenance cost.

High throughput, low HETS, great
versatility and flexibility, simplicity
in construction, handles liquids
containing suspended solids, handles
mixtures with emulsifying tendencies.

Short contacting time for unstable
material, limited space required,
handles easily emulsified material,
handles systems with little liquid
density difference.
FIELDS OF INDUSTRIA
    APPLICATION

  Petrochemical,
  chemical
  Petrochemical,
  nuclear, fertiliz
  metallurgical.
  Nuclear,
  petrochemical,
  metallurgical.

  Petrochemical,
  metallurgical,
  pharmaceutical,
  Pharmaceutical,
  petrochemical,
  metallurgical,
  chemical.
  Pharmaceutical
  nuclear,
  petrochemical
HETS:  Height  equivalent  to  a  theoretical  stage.
                                       15

-------
 Settler Design

      The processes that occur  in  a  settler  are  quite  complex, and interrelated.
•- They include the mechanisms 'of coalescence'and  disengagement.  Coalescence is-
 retarded by an increase in the viscosity of the continuous  phase.   Increases
 in temperature promote coalescence  by lowering  the  viscosity.  Decreases in
 surface tension retard coalescence  because  they produce  a greater area of the
 continuous phase.  Increases  in density differences between the phases promote
 coalescence by increasing the  force acting  to bring about drainage  of the contin-
 uous phase film.  The formation of  larger drops enhances coalescence, while
 smaller drops retard it.  Small droplets can give considerable difficulty in
 practice as they can be easily entrained in product streams where they represent
 both a loss and an impurity.   Small  drops introduced  into the feed  going to the
 settler are particularly troublesome.

      The basic requirements  of efficient settler operation  have been summarized
 as follows:  a) minimize turbulence in a settler as this will retard coalescence
 by keeping small drops in suspension, b) minimize small  drops in the original
 dispersion, c) keep linear velocities along the settler  at  a low value, and d)
 employ coalescence aids such  as removable baskets of  packings at the inlet point
 provided that they do not result  in too large pressure drops.

      High throughput is a most important consideration of settler design.  Exten-
 sive experimentation has shown that there is a  maximum safe throughput for a
 specific extraction in a given mixer-settler, above which quite small changes can
 result in flooding.  Also the  throughput unit volume  of  dispersion  decreases as
 the thickness of the dispersion band at the interface increases.  One approach
 to the problem of increasing  the  throughput of  a settler of given dimensions is
 to divide'the overall dispersion  into a number  of thinner bands.  The Lurgi and
 I.M.I, settler designs attempt to accomplish this by  using  a series of trays or
 plates.

      An alternative approach  to increasing  settler  throughputs is to introduce
 some packing into the settler which should  be preferentially wet by the dispersed
 phase.  A woven-mesh packing  incorporating  materials  that are both  water and
 organic wetted was developed  some time ago  and  has  given quite promising results
 when tested in the Davy Powergas  mixer-settler. The  overall unit has been
 claimed to operate more than  an order of magnitude  better than a simple mixer-
 settler.  Another approach to promote throughput, by  the application of centri-
 fugal contactors, has been well established for many  years.  They offer short
 contact times and efficient phase separation and are  especially desirable for
 applications involving either chemically unstable or  slow sett-Ving  systems.
 The Podbielniak, De Laval and Westfalia extractors  have  an  excellent record
 of performance for such problem liquid-liquid extractions.  The Luweta equip-
 ment is an adaption of the underdriven centrifugal  cream separator  with two
 or more stages mounted vertically in a single assembly.  Each stage comprises
 a spray disk mixer and a centrifugal separator. The  ports  and channels for
 transfer and conveyance are carried in a stationary central tube.   The more
 recently introduced Robatel  contactor features  very short residence times for
 a novel multistage design.


                                       16

-------
Contactor Selection

     As noted above, contactor selection is a complex task as it involves both
technical and economic considerations applied to a specific system.  As a
general guide a chemically unstable system or one with slow settling character-
istics will tend to favor centrifugal equipment.  If however, a system is rather
inexpensive and only a few stages are required, simple gravity columns or mixer-
settlers should be considered.  If a larger number of stages are required, a
mechanically agitated column becomes a candidate.  Mixer-settlers are suitable
for most chemically stable systems, but require more groundspace than do columns.
Also the holdup in mixer-settlers is greater than in columns.  On the other hand,
however, these are better in accepting flow changes from stage to stage due to
the transfer of material between the phases.

Specific technical criteria for selection of extraction equipment include the
following:
Total throughput -
Number of theoretical stages -
Physical properties of system -
Type of reaction -
for typical systems, a minimum total  through-
put of 50 gallons/hour of feed plus extract
phase corresponds to a column of about 10 cm
in diameter; most industrial processes require
capacities of 500 to 50,000 gallons/hour and
more.

the height of most types of vertical  contactors
tends to become excessive with more than,10-)5
stages.  This limitation does not arise with
horizontal equipment especially mixer-settlers,
but floor space requirements become quite large.

contactor performance is affected by  viscosities,
interfacial tensions, densities, boiling and
freezing points, temperature, f1ammabi1ity,
toxicity, etc.

the effect of slow heterogeneous reactions is
always to reduce the rate of mass transfer,
hence either increasing column height or reducing
stage efficiency; interfacial area should be
increased by upping rate of agitation; slow
homogeneous reactions are quite rare  in liquid
extraction.
                                     17

-------
Extreme phase ratios -
Short residence time -
Ability to handle solids -
Tendency to emulsify -
Limited space availability -
the effect of a low value is to reduce the
dispersed phase holdup and hence the inter-
facial area; a high value may cause large back-
mixing; in both cases, performance decreases;
the exceptions are mixer-settlers with provi-
sions for recirculation of the separated phases
back to the mixer.

this may be required with chemically unstable
products to minimize losses and leads to con-
sideration of centrifugal reactors.

some contactors are susceptable to blockage;
the only types which can accept significant
quantities of solids are the rotory and, to a
lesser extent, the raining bucket types (as
well as mixer-settlers).

nonmechanical contactors are generally preferred
to many widely agitated types, with the excep-
tion of the centrifuge units.

mixer-settlers and centrifuges should be con-
sidered.
Special material requirements -  metals applicable to all contactors, non-metals
                             -•--- require-s-pecial study. - -  -

Radioactivity or corrosion present - special studies of safety and material
                                     properties needed.
Ease of cleaning -


Low maintenance -
depends on contactor design and construction
material.

mechanically and pulse agitated contactors
require more careful maintenance due to shaft
timing, seals, pulsing mechanisms, etc.
                                      18

-------
                               Experimental  Studies

      During the first year of the. three year research .effort a significant
 portion of the work was dedicated to the establishment  of  analytical procedures
 for the determination of PCBs in sludges.  These  procedures were then employed
 to analyze a set of PCB spiked water samples covering the  concentration range
 from 2ppm to 123ppm;  details are given below.   Furthermore, an environmental
 sample, a sludge from Waukegan Harbor, Illinois,  was analyzed.  Apart from these
 analytical studies seven different solvents and four additives were evaluated
 for their potential  use in the extraction process.

 Extraction of PCBs from Spiked Water Samples

      Various PCB spiked water samples were  analyzed in  order to determine
 the amounts of the PCB contaminent that could be  recovered by extraction
 methods.   EPA Method  608 was used for the analytical work.  The PCB employed
 was Aroclor 1232.   The experimental  study also  provided information on the
 higher  and lower levels of extractability for known amounts of this particular
 Aroclor.

      The  analytical  data show total  recoveries  in the order of 90% for spiked
 samples containing an initial  Aroclor 1232  concentration of 123 micro grams/ml
 (2  extractions).  Recoveries from the samples spiked with  49.5 micro grams/ml
 Aroclor 1232 were  88% and 90.4% respectively and  about  100% for a 24.5 micro
 grams/ml  sample (2 extractions).  For samples with 12.2 micro grams/ml  initial
 concentrations recoveries were 86.5% and  87.1%  respectively.  The data for the
 two samples with the  initial  very low 2 micro grams/ml  concentration lack
 sufficient precision  to warrant consideration.

 Quantification of  PCBs  in Environmental Sample

      After the analysis of PCB-spiked water  samples had been sucessfully
 completed  the more difficult  problem of analysis of PCB contaminated environmental
 samples remained to be  solved.   A sludge  sample from Waukegan Harbor with an
 approximate PCB concentration  of 5000 ppm was used.  After an initial  separation
 of  the  sludge gave a  liquid  to  solid ratio of about 3:2 it was decided to
 analyze the liquid and  the solid  fraction of the sample by two different methods.
 The liquid fraction was  analyzed  by  the EPA-Method 608.  The wet solids were
 analyzed  following a  soxhlet extraction procedure, which had been used in an
 EPA sponsored  interlaboratory  study.   (Ann L. Alford-Stevens, et al.,  Anna!.
 Chem. 1985,  Vol. 57,  2452-2457).

      The  sample was centrifuged  for  one hour at 2500 rpm and then separated
 into  a  liquid  and  a wet  solid  fraction.   Two 20 gram aliquots of the wet solid
 fraction were extracted  by soxhlet.

      The moisture  content  of the wet  solid fraction was determined by  air drying
the  sample  4  days  and then  vacuum  drying until constant weight was established.
 PCB concentrations were  calculated for  the liquid  fraction, the wet solid  frac-
tion, the dry  solid fraction and  for  the original  environmental  sample.   The
 results are given  in  Table 3.
                                       19

-------
             TABLE 3:   PCB Concentration of Waukegan Harbor Sludge
                       Liquid     Wet Solid     Dry Solid      Integral
                     Fraction      Fraction      Fraction        Sample

Extracts

1ST                    9.7           13950         32442        5818

2ND                    8.1           14500         33721        6046



     PCB Concentration is given in mg/kg (ppm).

     Moisture content of wet solid fraction was: 57%.

     The mass ratio of liquid fraction to wet solid fraction was:  1.4.

     The mass ratio of liquid fraction to dry solid fraction was:  4.6.


Characterization of Solvents

     A qualitative experimental study was carried out to characterize certain
pertinent properties of candidate solvents for PCB extractions from aqueous
solutions.  Basic criteria for the selection of an extraction solvent have
been described previously.  The objective of this study was to find the  solvent
system which gives a clear phase separation in a very short settling  time.

     Seven hydrophobic solvents with high PCB solubility together  with 4 hydro-
philic solvents as additives at three different solvent to water ratios  (1:9,
1:1, 9:1) were investigated.  Each solvent/water system was studied without
and with additives, whereby the ratio of hydrophobic solvent to additive was
kept at 4:1 for all cases.

     The different solvent systems were subjected to the following experiment:
A total volume of 50 ml of solvent/water mixture was filled in a 65 ml  screw
top sampling bottle and agitated for 2 min, 10 min and 30 min. respectively.
The agitation was performed with a Burrel-Wrist-Action-Shaker, which  was
operated at a constant speed of one-semi-circular-motion per second.  The sep-
aration following the agitation was observed qualitatively after two  minutes
and 10 minutes of settling time.  The quality of the separation was character-
ized using the following criteria.

a)  Quality of phase boundary
b)  Degree of murkiness in either phase
c)  Amount of bubbles in either phase and/or at boundary
d)  Severity of emulsion formed in either phase or at boundary
                                      20

-------
     The hydrophobia solvents and the additives (hydrophilic solvents)
investigated we're:

Hydrophobic Solvents                        Additives
     Hexane                                    Acetone
     Kerosene                                  Carbitol
     Mineral Spirits                           Methanol
     Petroleum Ether (PE)                      Methylene Chloride
     Toluene
     Turpentine
     Xylene

Of the hydrophobia solvents only Hexane, Kerosene and Toluene were combined
with additives.  A total of 288 observations were made.

                            New Process Development


     The concept of using kerosene as an extraction solvent for removing PCBs
from sludges and sediments as originally proposed was based on two factors:
1) the low cost of kerosene and 2) some work underway on extraction of PCBs
from soils.  While the cost factor is a driving force in any development, it
alone cannot overcome technical impracticalities.  It became quite evident
early in the study that the presence of large quantities of water in the
contaminated samples was to present an insurmountable barrier to the use of
kerosene in this application.  A bold new approach to this most difficult
problem was needed and is herein presented.

     This new approach is based on some well known facts about the properties
of PCBs present in sludges and sediments, some common sense about the use of
particular hydrophyllic and hydrophobic solvents and application of a stripping
operation, an existing technology, as a final step for isolation and concentra-
tion of the contaminants.

     The first step of the process takes advantage of the extremely low solu-
bility of PCBs in water.  In a typical PCB contaminated sludge or sediment
composed primarily of water with, for example, 20% total solids, one could •
expect virtually all of the PCBs to be asociated with the sediment in a water
medium.  Thus the first processing step is simply a physical separation of
water and solids.  This can be accomplished with varying degrees of efficiency
using existing equipment.  Practically, the solids fraction from this separation
will  contain on the order of 50% water but will most likely contain 98%+ of the
PCB content.  The solid fraction and the liquid fraction from step 1 require
further treatment, however, it was managed to isolate most of the PCBs and to
reduce the sample size by about 60%.  The water fraction requires subsequent
treatment which will be addressed later.
                                     21

-------
      In the second step of the process, the PCB contaminated oil  must be
 dissociated from the inorganic substrate.  To do this, a water miscible solvent
. (hydrophyllic), acetone, was added in quantities sufficient to break the bond
 between the PCBs and the solid surface.  This is followed by another liquid/solid
 separation.  This description is for one stage of the proposed second step of
 the process.   The amount of PCB transferred to the solvent/water  mixture depends
 on partitioning between the liquid and solid phases and thus we can define a
 partition coefficient and ultimately a stage efficiency. One can  then predict
 the number of  stages required to perform a particular level  of separation given
 the original contamination level and the allowable level in the solid effluent.
 This second step is a multistage countercurrent leaching using a  hydrophyllic
 solvent.

      The products from the second step are: 1) a hydrophyllic solvent/water
 mixture containing nearly all  of the PCB contamination which requires further
 treatment and, 2) a PCB free soil/solvent mixture from which the  solvent must
 be recovered and returned to the process cycle with the soil  now  being decon-
 taminated and  ready to be returned to the environment.  A small fraction of this
 clean soil  is utilized in an absorption column to extract the trace amounts of
 PCBs from the water effluent of step 1.

      The third step of the process is a stripping operation in which the PCB-
 containing  stream from above is contacted in, a liquid-liquid extractor with
 a  hydrophobic solvent and additional  water.  The additional  water drives the
 PCBs out of the solution and into the hydrophobic stripping solvent.  This step
 is required to separate the PCBs from the aqueous phase which is  inconvenient
 for the final  destruction step and also, the PCB can be concentrated in  this
 step further reducing the volume of contaminated sample required  for handling.
 The two streams which leave this step are the PCB concentrated in stripping
 solvent which proceeds to final  destruction either by chemical  means (KPEG
 Reagent) or by other appropriate methods and the hydrophyllic solvent/water
 mixture containing trace PCBs.  The solvent/water mixture goes next to a
 distillation column in which the solvent is released and returned to the leach-
 ing process described in step  2.  The water contaminated with trace amounts of
 PCB is  recycled to the front of the stripping process to close the cycle.

      Proof  of concept experiments have been conducted successfully and experi-
 ments  are now underway to determine process criteria to further advance  this
 concept. Particular attention has been paid to closing the  various liquid and
 solid  stream cycles.
                                       22

-------
                                   SECTION 5
                         DISCUSSION OF ASSESSMENT STUDY
                              Screening of Processes

     The processes were screened by examining each for any undesireable aspects.
Processes with such aspects were rejected from further assessment.   For example,
lack of tolerance for water by a process is undesirable because of  the extensive
sediment drying it necessitates.  There is evidence in the collected data  that
such drying generates fine particulates that must be removed from the exit gases
of the dryers.  The recovered particulates must also be decontaminated. Processes
showing insufficient tolerance for water were therefore rejected from further
consideration as a primary treatment process in favor of more tolerant alter-
natives.

     Table 4 lists the sixty four processes screened, identifies those selected
for further assessment, and gives the reasons for rejection of the  rest.  The
processes are arranged according to the generic types of technologies upon which
they are based.  The use of generic technology types guided the survey and
assessment phases of this study by providing technologies from which treatment
processes could arise.  Some of the technologies (e.g., nucleophilic substitu-
tion) have provided several processes.  Some (e.g., enzymes) have not yet
provided any processes.  These technologies are nevertheless discussed to
describe their status, with respect to PCB-contaminated sediment treatment, so
that their potential may be considered in research planning.

     A process rated "1" in Table 4 for each technology has been selected  for
further assessment.  As noted, eleven processes obtained this rating.  Higher
•rating numbers assigned the rest of the screened processes refer to footnotes
that identify the reason for rejection of"the process for further assessment.
                                     23

-------
     technology
                              References	
                             Centofantl 1971; Chen 1982; Childs 198Tf
                             Craddock 1982; Edwards et al. 1982;
                             Environment Canada 1933; Hornlg 1984;
                             Massey and Walsh 1985; Rogers and Kornel
                             1985; Rogers  1983; Rogers 1985.    '
                             Balllod et al. 1978; Miller and
                             Sev1enton1ewsk1  (n.d.); Miller and Fox
                             1982.
            Process
Evaluation3
leinical
^-temperature oxidation

•t air oxidation
ipercrltlcal water oxidation  Model 1 et al. 1982.
iem1cal oxldants
onation
orlne removal

'hydrochlorlnatlon
                             FMC Corporation (n.d.); March 1968.<
                             AHsman et al. 1981; Lacy and Rice
                             Deschlaeger 1976; Prengle and Mauk 1978.

                             U.S.P. 346. 636

                             Chu and Vick 1985; Laplere et al. 1977.
•dueing agents
                               Chu and Vick 1985; Sworzen and Ackerman
                               1982.
icleophillc substitution
                                                                1
                                                                t
                               Brown et al. 1985a; Brunelle and Single-
                               ton 1985; March 1968; New York University
                               1984; Ruzz et al. 1985; Smith and
                               Gurbacham 1981; Sunohlo (n.d.); Sweeney
                               and Fischer 1970; United States Patent
                               Office 1984b; Weitzman11984; Weltzman
                               1QR4- UpU7man 1QRS              ' '
Uncatalyzed, general                     2
Zlmpro Process, Santa Maria,             4,13
   CA Waste Site
Catalyzed
  Dow Chemical Co. Patent  3,984,311      2
  IT Environmental Science              2

 Modar                                  1

 Potassium permanganate plus  Chromic
    Add and Nitric Add                6
  Chlorolodides                         4,7
  Ruthenium tetroxide                   3,4,8

 GE UV/ozonation process                2
 Molten aluminum/distillation            14

 Catalytic:                              2,3
   Nickel on kleselguhr                 2,3
   Pd on charcoal                        2,3
   Lithium aluminum hydride              2,3
   Butyl lithium                        2,3
   Raney Nickel                         2,3

  Sodium 1n liquid ammonia               7,9
  Nickel-catalyzed zinc reduction        7,9
  Hydrazine                             7,9
  UV light plus hydrogen                 2
  Mildly acidic zinc powder,
    Sweeney and Fisher (1970)            2,14
  Sodium-based processes:
    Goodyear, sodium in naphthalene  (1980)  10
                                                                            Acurex, proprietary solvent
                                                                            PCBX/Sun Ohio
                                                                            PPM
                                                                            Ontario Hydro Power
                                           10
                                           10
                                           10
                                           10
                                                                                                                          CM

-------
                                                    IADLC t
                                                             iuiii
     Generic
     technology
  References
               	Process        	
                                           Potassium poly (ethylene glycolate)
                                           based:
                                             KPEG Terraclean-Cl
                                             GE KOH-PEG
                                             New York University KPEG

Bailin and Hertzler 1977; Bailin and       U/V photolysis
Hertzler 1978; Bailin et al. 1978;         Syntex photolytic
Craft et al. 1975; Dev et al. 1985;        Thermal corona glow
Kalmaz et al. 1981; Meuser and Weimer 1982;Microwave plasma
Plimmer 1978; Rogers and Kernel 1985;      RF in situ heating
Rogers 1985; Trump et al. 1979: West et al.
1983.                                      Gamma radiation (Craft et al. 1975)
                                           LARC
Evaluation3
ucleophilic substitution
 (continued)
adiant energy
lectromechanical  reduction

lorinalysis



rolysis




YSICAL

noving and concentrating
?ated Air Stripping

fraction
 Jsorption.


 'trlfication
Massey and Walsh 1985.

Sworzen and Ackerman 1982.
Boyd 1985; New York State Department of
Environmental Conservation 1985a; New
York State Department of Environmental
Conservation 1985b.
                                           Electromechanical research process

                                           Hoechst process
                                           Goodyear catalytic hydrogenalysis
                                           Exhaustive chlorination

                                           Advanced Electric Reactor
                                           Wright-Malta alkaline catalyst fuel-gas
                                             process
Angiola and Soden 1982; Caron 1985; Gilmer
and Freestone 1978; Githens 1984; Hancher
et al. 1984; Hawthorne 1982; Lee et al.
1979; Saunders 1985; Schwinn et al 1984;
Versar, Inc. 1984.
                                           American Toxics Disposal, Inc.

                                           Critical Fluid Systems, C02
                                           Furfural
                                           Acurex solvent wash
                                           0. H. M. extraction
                                           Soilex process

                                           Carbon adsorption, general
                                           Neoprene rubber adsorption
                                                                                                                    1
                                                                                                                    11
                                                                                                                    12
        3
       3,4
       5
       9,17
        18

        9
        1

       14

        9
        9
        9 '

        1

        12
                                                                                        14

                                                                                        14
                                                                                        15
                                                                                         1
                                                                                         1
                                                                                         1

                                                                                        13
                                                                                        15
tn
csj
Timmerman 1985.
                                                                                             nrt nrnrocc

-------
      Generic
      technology
                               References	   	        	
                               Ghasseml  and  Haro  1985;  Law  Engineering
                               Testing  Company  1982; Stroud et  al.  1978;
                               Subnamanlan and  Mahal 1ngam 1977;  TUtlebaum
                               et al. 1985.

                               Carlch and Tofflemlre 1983;  Hand  and Ford
                               1978; Murakami  and Take1sh1  1978; U.S. Army
                               Corps of Engineers Water Resources Support
                               Center 1983;  Zlmmle and  Tofflemlre 1978.
                               Bedard  et  al.  1985;  Bumpus  et  al.  1985;
                               Clark et al.  1979; Dawes  and Sutherland
                               1976; Furakawa 1982; Isblster  et al.  1984;
                               Kong and Sayler 1983;  McConnlck  1985;

                               New York State Department of Environmental
                               Conservation  1985a;  New York State Depart-
                               ment of Environmental  Conservation 1985b;
                               Rhee et al.  1985b; Rhee et  al. 1985;
                               Unterman et  al. 1985.

                               Catelanl et  al. 1971;  Rochklnd et  al'.
                               Unterman et  al 1985.
  Process
Asphalt with
                                                                                             11me pretreatment
Evaluation3
         16"
         15
         16
         13

         13
.tabllizing




ottom recovery




BIOLOGICAL

Hcroorganisms
nzymes
                                                                               Z-Impremix
                                                                               Sulfur-asphalt blends (K-20)
                                                                               Ground  freezing

                                                                               Dredging
                                                                                Bio-Clean
                                                                                Sybron  B1-Chem 1006 PB
                                                                                Composting
                                                                                B1o-Surf

                                                                                Ecolotrol,  Inc.
                                                                                Wormes  Biochemical's Phenobac
                                                                                Rhee anaerobic degradation
No processes found
                                       1
                                       1
                                       1
                                       4,13

                                       4,13
                                       1,13
                                       14
Explanation  of  process  rating:~
  1.  Identified emerging sediment  treatment  process.

  2.  Destruction efficiency appears  to  be too  low to  meet  environmental  goals.

  3.  Processing time apppears  to be  extremely  long for practical  timely  cleanup.
                                                                 i
  4.  Data available for dioxln, other chlorinated compounds,  or other contaminants,  but  not  PCBs.

  5.  Process has been shown to destroy  PCBs  in gas streams only.   It may be feasible for sediments, but has not been shown
      to be.                                                    .!'
                  to
                  CJ

-------
                                                   TABLE 4 (continued)


6.  PCBs with 5-7 chlorine atoms  per molecule are not destroyed.

7.  Products of partial  degradation may be toxic.

8.  Reagent 1s very costly/toxic  or both.

9.  Process costs appear to be excessively high compared with other emerging  treatment  processes.

10. Water destroys the reagent or Interferes with its action, thus the  process  would  require excessive drying of sediments
    and, probably, extraction in  pretreatments.  The process would therefore  have  application only as a subordinate final
    step to several extraction and concentration operations.

11. This particular process was not evaluated because data were not available for  assessment.

12. This process 1s an alternative to another process using the same generic  technology, but 1t 1s in very early stages of
    development, and data were not available for assessment.

13. This technique 1s basically applicable to preliminary operations prior to treatment or to treatment of wastestreams
    (e.g., wastewaters)  from chemical or physical treatments.                                                         ,

14. This process is in the concept stage and data are Insufficient to assess  It for PCB-contaminated sediments.

15. This process has been found to be ineffective.

16. This technology provides only for encapsulation of the PCB-contaminated sediments.

17. This process supports Incineration of PCBs.
                                                              (
18. The process does not appear to be feasible for submerged sediments.

-------
                       Development of Evaluation Criteria
Problem Definition
     The PCB contamination problem in the Hudson River is representative  of
the type of PCB destruction/detoxification problems focused on in this  study.
The New York State Department of Environmental  Conservation (DEC) recently
concluded a 2-year survey of 40 miles of the upper Hudson River (1976-1977).
The survey, conducted with consultant Malcolm Pirnie, Inc. (White Plains, New
York) identified 40 spots collectively containing more than an estimated
382,000 m3 (500,000 yd3) of sediment contaminated with PCBs at concentrations
greater than 50 ppm.  The depth of contamination seems to be between 0.25 and
0.45 meters.  A more recent survey (1984) was made of a 5-mile reach extending
downstream 5 miles from Ft. Edward, New York.  Samples from the near-shore
area showed from 40 to 60 ppm PCBs.

     A pilot dredging operation, performed with hydraulic dredges between 1978
and 1979, netted 152,900 m3 (200,000 yd3) of material with seemingly little
dispersion of the contaminant downstream.  The contaminated areas are located
in shallow, slow-moving waters near the riverbanks.  Although DEC is investigat-
ing other disposal methods, it is expected to dredge the contaminated sediments
and bury them because of the high costs of alternatives.

     .The wetness of the sediments will preclude the use of many treatment proc-
esses designed to treat PCB-containing oils using reagents that would be  destroyed
or made ineffective 'in the presence of water.

Regulatory, Factors

     The goal for detoxification/destruction is to have no more than 1  to 5 ppm
PCBs in the treated sediment.  At present, PCB disposal  is regulated under TSCA
[40 CFR 761.60 a 5 (1-111)].

     The criterion of 1 to 5 ppm for a promising alternative treatment  should
result in a treated sediment that is disposable without restriction.

Technical Factors

     Technical factors relate to a broad range of principles of operation of
diverse applied technologies.  Factors have been chosen that can be  used
effectively in comparing one treatment process with another.  Additional
factors, specific to a technology, have been chosen to help portray  the inher-
ent strengths and limitations of a process.

     The goal set for process performance is to reduce the PCB concentration in
treated sediments to background levels of 1 to 5 ppm.  Several of the processes
were found to meet this goal.
                                     28

-------
     Available capacity was found not to exist for.any of the processes.  How-
ever, several were developed sufficiently to permit projections of the time
required to build a facility for application of the treatment.

     Conditions/limitations that were considered included tolerance for water,
required processing time, and controllability of process conditions.  Those
treatments that could tolerate water up to about 40 percent would not require
a drying step with its attendant fines control problems.  Those requiring only
1 day for treatment could generally show a faster rate of cleanup than those
requiring 3 days.  Some biological processes required more than 3 weeks.  The
treatments generally provided control of the processing conditions; however, a
few (e.g., composting) would not necessarily do so.

     Concentration range handled in data developed for the processes ranged
from unknown to 3,000 ppm.  Some processes had limits inherent in the tech-
nology; others had no apparent limits.

     Processes were found to range from concept stage to completed field tests.
Most were in the pilot stage.

     Data needs varied with the status of the process.  At worst, data were
available showing tests of the concept.  At best, the process had been field
tested, and only permits and checkout were needed.

     Unit operations employed in each process were identified.  Collectively,
the processes use 18 different unit operations.  The number of operations per
process varied from two to eight, depending upon the complexity of the tech-
nology.             •                            "...

Cost Estimates

     Costs of appplying an alternative technology are very difficult to predict
because of the uncertainties regarding the potentially available new treatment
processes.  Estimates of the quantities of PCB-contaminated sediments are
under review and may change significantly as a result of the reviews.  In most
cases, additional information is needed regarding the yet to be proven effec-
tiveness of emerging treatment processes before the costs of their application
can be assessed accurately.  Therefore, the cost estimates presented here
should be viewed as preliminary with respect to the new treatment processes.
A range of costs is given for supporting operations (dredging and transporta-
tion) to reflect differences in different geographical areas of the country.

     Because the regulations permit the use of incineration or chemical waste
landfill and the application costs of these two methods are available from
firms engaging in their practice, these costs were used as lower and upper
limits with which to compare the costs of applying new alternative technology.
                                    29

-------
     The application of any treatment process  can  involve  the  need for one or
more of the following unit operations:  dredging,  transport, storage, landfill
disposal, land treatment disposal,-Incineration,-and/or alternative treatment.-.
Estimates have been developed for  all of these so  that, in any given process-
ing evaluation, the proper elements  could be added to obtain an estimate of
the cost of application.  The estimates were made  in  terms of  the cost per
cubic meter of sediment treated.  The sediment was assumed to  have a density
of 1.68 Mg/m3.

     Table 5 shows the unit cost estimates used to develop cost ranges for the
emerging treatments.

                          TABLE 5: UNIT COST ESTIMATES


	Operation  	Cost, $/m3	

   Dredging                                               20
   Transport                                           13  to 126
   Storage                                                10
   Landfill disposal    .                              260  to 490
   Land treatment                                      33  to 111
   Incineration                                         1,680


     Baseline costs are taken as the range of costs for sediment  removal, trans-
port, and disposal in a chemical waste landfill.   This range is calculated using
the unit costs from Table 5.

     Baseline costs, $/m3 = dredging cost and transport cost + disposal cost

                          = $20   +    $13 to $126   +   $260  to  $490

                          = $293 to $636.

     Incineration costs include those for dredging, transport, and incineration.
The cost of residue disposal is included in the incineration cost:

     Incineration cost, $/m3 = dredging cost + transport cost  + incineration
                               cost

                             = $20  +  $13 to $126    +   $1,680

                             = $1,713 to $1,826.

     The costs of alternative treatments are calculated in the same way, by
summing the costs for the separate operations involved.
                                    30

-------
 • .                      Initial Assessment of Processes

     The processes assessed represent six different technologies;  each  of which
is briefly described prior to discussing the processes to which it applies.
For the processes themselves, the depth of the summaries varies because of the
varing extent of available,, information.^ .,_,,...  ,...._,._ . _.,   ...... ••-_-.•_*._

Processes Based on Chemical Technologies

     The MODAR supercritical water (SCW) oxidation process is  a relatively low-
temperature oxidation process that utilizes temperatures and pressures  of super-
critical water (above 374 °C and over 22.09 MPa) to break down hazardous organics
to carbon dioxide, water, and other simple, less harmful molecules.  The treat-
ment is not selective but rather applicable to a broad range of compounds.
Normally water-insoluble organics become highly soluble in supercritical water.
The water also acts to reform complex organics, and heteroatoms, including
halogens, phosphorous, sulfur, and metals, are readily precipitated as  salts
when present with appropriate counter-ions.

     The MODAR process as applied to soil-slurry decontamination consists of
the following steps:

     1.  PCB-contaminated sediments are fed as a slurry with 20 to 40 percent
         solids, as determined by some continuous flow bench-scale testing and
         preliminary economic studies.  The slurry is pressurized to super-
         critical pressure then heated by heat exchange with oxidizer effluent
         generated in a subsequent step.

     2.  At the oxidizer, pressurized oxygen, organic fuel, preheated slurry
         and oxidizer recycle are broght together.  Because the water is
         supercritical, the oxidant is completely miscible with the solution  ' -
         (i.e., the mixture is a single, homogeneous phase).  Organics  are
         oxidized in a controlled but rapid reaction.  Because the oxidizer
         operates adiabatically, the heat released by combustion of readily
         oxidized components is sufficient to raise the fluid phase to  tempera-
         tures at which all organics are oxidized rapidly.  For a feed  of 5
         percent organics by weight, the heat of combustion is sufficient to
         raise the oxidizer effluent to at least 550 °C.

     3.  The effluent from the oxidizer is fed to a salt and sediment separa-
         tor, where inorganics and sediments originally present in the  feed  are
         removed as a solid slurry.  At 500 °C and above, the solubility of
         inorganics in SCW is extremely low.
                                      -31-

-------
     4.  A portion of the superheated SCW is recycled to the SCW oxidizer by a
         high temperature, high pressure pump.  This operation provides for
         sufficient heating of the feed to bring the oxidizer influent to
         supercritical conditions.

     5.  The remainder of the superheated SCW (with some C02) is used to
         preheat the incoming soil slurry.

     As a waste destruction process, the Modar concept has several advantages.
The chemical reactions that occur are carried out in a closed system, making
it possible to maintain total physical control of waste materials from stor-
age, through the oxidation process, to the eventual discharge of the products
of combustion and any associated wastes.

     The Modar System has been tested for soils, and major problems in handling
this type of process feed have been identified.  The components  would be con-
structed from corrosion-resistant, high-nickel alloys.  Projected costs of
treatment alone are $184 to $554/m3 of PCB-contaminated sediment.  Dredging
would be required, as well as transport for treatment and redeposition of
treated sediments.  The overall cost estimate is thus:

     Cost, $/m3 = ($20 dredging) + ($13 - $126 transportation)
                  + ($184 - $554 treatment) + ($33 deposition)

                = $250 - $733/m.3.

     Nucleophilic (electron-donating) substitution removes chlorine from aro-
matic compounds by two mechanisms, the intermediate complex mechanism and the
benzyne mechanism.' The intermediate complex mechanism consists  of thef follow-
ing two steps, with the first usually rate-determining:
                                    Fast
                                    32

-------
      The  resonance  hybrid  formed  in  the  first  reaction  is represented by the
 structures  within the  brackets.   These are shown in an  alternative representa-
 tion  for  reaction 2.   The  overall  reaction rate  is increased by the presence
 of  activating  groups  (e.g.,  NO, N02)  on  the molecule.

      The  benzyne mechanisms  occur on  aryl halides such  as PCBs, which require
 stronger  bases than normally used because they lack activity groups on the
 molecule:
                                             NH3 + cr
                                               or
 If the aryl halide contains two ortho  substituents, this mechanism cannot
 occur.

     Nucleophilic substitution using alkali metal hydroxides in poly (ethylene
 glycol) [PEG], or poly  (ethylene glycol methyl ether) [PEGM] has been shown to
 remove chlorine from PCBs.  Chlorine removal probably involves alkoxide reac-
 tion, as illustrated by"the following mechanism:

          ^       ArCln + ROK -> ArCln_i OR + KC1
                  Chlorinated
                   aromatic
Potassium
glycollate
-> PEG ether
The partially dechlorinated, water-soluble reaction product may continue to
undergo dechlorination, depending on the reaction conditions.

     The KPEG treatment process is identified by Gal son Research Corporation  as
Terraclean-Cl.  Evaluation of this process is based upon data for the Terraclean-
Cl version.

     The Terraclean-Cl process involves the chemical dehalogenation of PCBs
under mild conditions.  This is achieved by mixing the soil with an equal
volume of hot (150 °C) reagent in a rotating mixer such as a converted cement
mixer.  Water coming in with the soil is vollatilized and recovered for later
use in the process.  The reagent consists of a mixture of polyglycols and
                                      33

-------
capped polyglycols (PEG and PEGM), potassium hydroxide (KOH), and dimethyl
sulfoxide  (DMSO).  The exact formulation of the reagent is varied according
to the specific soil and contaminant combination.  The DMSO does not take part
in the reaction but acts as a catalyst and phase transfer agent to extract  the
PCBs from the sediments.

     At the end of the reaction, usually 30 to 120 min, the bulk of the reagent
(>80 percent in field tests) is decanted from the soil.  A small part of the
reagent remains on the soil, along with some of the dechlorinated reaction
products.  The residual reagent and dechlorinated by-products are removed from
the soil by mixing the soil with an equal volume of water and decanting the
water.  This washing is done two or three times and provides >99 percent over-
all recovery of reagent.  The washwater from the last wash is passed through a
bed of activated carbon, which preferentially removes the dehalogenated prod-
ucts.  The contaminated carbon is burned in a PCB incinerator.  The partially
dechlorinated PCB continues to dechlorinate, with >98 percent of the chlorine
associated with the PCB being recovered as KC1 under some conditions; the OMSO
acts as a catalyst and phase transfer agent by increasing both the rate of
reaction and the rate of transport of PCB from the soil into the reagent.

     The total cost of application of the treatment process is estimated as
fol1ows:

       Cost item                                      Cost, $/m3
                                             4-h cycle         8-h cycle

     Capital                                   12.56            25.11
     Reagent                                   25.66            25.66
     Energ^                                    33.20            33.20
     Maintenance                                1.26             2.51
     Labor                                     22.00            44.00
     Water disposal                             2.20             2.20

     Subtotal                                  96.88            66.34

     Profit/contingency (50 percent)           48.44            66.34

     Total                                     145.32           199.02

     These costs do not include dredging or permitting costs.  With dredging
costs of $20/m3, transportation costs of $13 to $126/m3, and final placement
costs of $33/m3, total  costs in the range of $211 to $378/m3 are anticipated.
Permitting costs are a separate item and would depend on the degree of cooper-
ation given by the various regulatory bodies involved.  However, the total
costs are not highly sensitive to this item because permitting costs of
$2,000,000 would only add $5/m3 to the costs of processing.
                                     34

-------
     Ultraviolet (UV) radiation may drive chemical reactions in PCBs.  The
structure of the products of UV radiation alone are not well defined.  The
-primary reaction at wavelengths greater than 290 nm is stepwise dechlorina-
tion.  Photolysis experiments in methanol at 298 to 313 K for 10 to 15 h
showed that HC1 was evolved.  The products contained in the menthanol solution
consisted of dechlorinated PCB and certain methanol substitution products.
The amount of methoxylated products did not exceed 5 percent of the total
amount of products formed.

     UV light energy, combined with a reducing environment, has been used  to
dechlorinate PCBs essentially completely in 1.5 to 2 h.  Because the photo-
chemical reaction is initiated by the absorption of light energy, the irradia-
tion wavelength must match the absorption band associated with the chemical
bond of interest in the molecule and the solvent must not absorb significantly
at the irradiation wavelength.  For PCBs, low-pressure mercury lamps that  emit
approximately 95 percent of their energy at 253.7 nm provide adequate irradia-
tion.

     Suitable solvents include water, alcohols, and hydrocarbons.  However,
these have very different effects on the process efficiency and mechanism.  In
heavy hydrocarbons, the dominant PCB reaction is the polymerization of the
biphenyl radicals to yield polyphenylenes.  In basic alcohol solutions, a
stepwise dechlorination of the molecule occurs.  This reaction occurs via  the
triplet state, and the reaction rate is dependent on the lifetime of that
state.  The hydrogen for the hydrodehalogenation initially comes from the
solvent.  In alcohol solutions, sparged hydrogen gas increases the reaction
rate by removing triplet quenchers from the solution by combination with the
sblvent from radicals formed.  In-addition to "the hydrogen, small amounts  of
sodium hydroxide are added to the alcoholic solution as a chlorine scavenger.
         \
     These photochemical reaction conditions lead to a rapid, highly control-
lable destruction of the PCBs, yielding only biphenyl and sodium chloride  as
the final products.  Oxygenated derivatives, chlorinated dibenzofurans, or
chlorinated dioxins have never been observed in the gas chromatographs or  the
mass spectra of the intermediates or products in the LARC degradation of PCBs.

     The LARC process (Light Activated Reduction of Chemicals) uses UV light
and an optimized reducing environment to dehalogenate various chlorinated
compounds extracted from soils.

     Studies were performed to determine the extraction efficiency from soil of
Aroclor 1260 using isopropanol as the extraction solvent.  Isopropanol was
chosen because it is a good LARC solvent, dissolves PCBs readily, and is
relatively inexpensive.
                                      35

-------
      The cost of treatment using  this  process  has been estimated  (1984)
 using an average concentration  in-the-soil of  1,500 ppm.  -Cost items include  r-;-
 daily operation, labor,  analytical,  travel,  per  diem, and profit.  These total
 approximately $157/m3 of soil.  Adding costs of  dredging, transportation, and
 placement gives a cost range  of $223 to $336/m3.

      Pyrolysis  is a  thermal  rupture  of the chemical bonds of a molecule that
 destroys 1t  without  oxidation.  The  energy requirements vary with the material
 being processed.  The use of  pyrolysis requires  the adoption of process designs
 that  provide for transport of material  that  can  melt, char, or become sticky
 during passage  through a  reactor. The design  must provide for efficient heat
 transfer to  the material  so that  it  can be heated to 2,000 to 2,300 °C.
 Pollution controls are required to remove particulates and toxic components
 from  exit gases.  Feeds  to pyrolysis units must  usually be predried.

      The advanced electric reactor (AER) is  a  thermal treatment process.  Pre-
 treatment of the feed material  is required.  It must be dried to a moisture
 content  of less than 3 percent  and ground to 35 mesh particle size.

      The treatment process is based  on  a high-temperature fluid-wall reactor.
 The reactor  heats organic compounds  rapidly  to temperatures in the range of
 2,200 °C using  intense thermal  radiation in  the  near infrared.  The reactants
 (in this case,  nonliquid  PCBs)  are isolated  from the reactor core walls by a
 gaseous  blanket of nitrogen flowing  radially Inward through the porous core
 walls.   Carbon  electrodes are heated and, in turn, heat the reactor core to
 incandescence so that heat transfer  is  accomplished by radiative coupling from
.the core to  the feed  materials.   The only feed streams to the reactor are the  -
 solid waste  containing PCBs and the  blanket  gas  (nitrogen).  PCBs are destroyed
 by pyrolysxis  rather  than  oxidation.  Therefore, typical products produced by
 incineration such as  carbon monoxide, carbon dioxide, and oxides of nitrogen
 are not  formed  in significant concentrations.  The principal  products of soil-
 borne PCB destruction are H2, Cl2, HC1, elemental carbon, and a granular free-
 flowing  solid-derived waste.

      The solid  feed  stream is introduced at  the top of the reactor by a metered
 screw feeder  connecting the air-tight feed hopper to the reactor.  Nitrogen is
 introduced primarily  at two points in the reactor annulus formed by the external •
 containing vessel  and  the  porous  graphite core to create the  fluid wall.

      Small nitrogen  streams are used to sweep  sight glasses,  prevent oxygen
 intrusion at  the  electrode  ports, and similar nonprocess uses.  Nitrogen flow
 is monitored  by  calibrated  orifices  in  each  feed line.  The solid feed passes
through  the  reactor where  pyrolysis occurs.  After leaving the reactor, the
product  gas  and  waste  solids  pass through two  post-reactor treatment zones (PRTZ)
The first  PRTZ  is  an  insulated vessel that provides additional  high-temperature
 ( 1,095  °C)  residence  time  ( 5 s), and the second PRTZ is water-cooled.  It
also  provides additional   residence time ( 10 s), but it primarily cools the
gas to less than  538  °C prior to downstream  particulate cleanup.
                                   36

-------
     The product gas then enters a baghouse for fine particle removal followed
 by an aqueous caustic scrubber  for chlorine removal.  Any residual organics
 and  chlorine are~removed by" passing 'the 'product gas through two banks" in*
 series of five parallel activated carbon beds just upstream of the emissions
 stack.  The organic-, particulate-, and chlorine-free product gas, composed
 almost entirely of nitrogen, is then emitted to the atmosphere through the
 process stack.

     Solids exiting the PRTZ are collected in a solids bin that is sealed to
 the  atomsphere.  Additional  solids in the product gas are removed by a cyclone
 and  routed back to the solids bin.  The solids collected in the solids bin are
 removed by plant personnel after each test or pair of tests.

     For very large volume cleanups, a cost estimate is $412/ton ($763/m3)
 strictly as a budgetary cost figure and not as an absolute cost or quotation.
 This amount does not include the costs of dredging, transporting soil to the
 AER  site, or facilities for  storage of dredged soil.  It also does not provide
 for  cost of landfilling or other disposal of the treated sediment.

     The capital cost for construction and initial testing of a single 25,000-
 ton/year transportable AER is approximately $4 million, not including permits
 and trial burns.  With six units, it would require approximately 3.5 years to
 treat 382,000 m3 of sediment.

     Dredging would be required, also transport for treatment and redeposition
 of treated sediments.  The overall cost estimate is thus:
     Cost, $/m3 =  ($20 dredging) + ($13 - $126 transportation)
                   + ($763 treatment) + ($33 deposition)

                =  $829 - $942/m3.

     Thionation introduces sulfur into an organic molecule to displace chlo-
rine.  The possible use of sulfur reactions to degrade pollutants has not been
systematically exploited.  It has been reported that, at 150-170 °C, sulfur
and sodium carbonate react with p-dichlorobenzene, removing the chlorine to
form an insoluble  polymer, sodium chloride, and carbon dioxide.  Limited tests
were conducted in  1986 by W. C. Von Meyer (EPA Contract No. 68-02-3992, Task 53,
Project Officer John Glaser) on a mixture of dried loam soil sparged with
mixed dichlorobenzenes (DCBs), sodium carbonate, and sulfur.  The dichloro-
benzenes were added as a €-.131 g/ml solution in acetone.  Reaction was carried
out in a tilted 500-ml flask fitted with a condenser and sodium carbonate
trap.  The trap was required to remove hydrogen sulfide gas evolved during the
reaction.  The reactants were mixed periodically by rotating the flask a
quarter turn every two hours over the 9 hour reaction period.  The temperature
reanged to 88 °C, and rose in the eighth hour to 260 °C when an exothermic
excursion followed evolution of volatiles formed from the sulfur.  The re-
sults, based on analyses by thin layer chromatography and gas chromatography,
were as follows:
                                      37

-------
     1,2 dichlorobenzene:                      8.4 percent dechlorinated
                                               3.4 percent volatilized   .   ..-..-.

     1,3 dichlorobenzene:                      90.8 percent dechlorinated
                                               3.4 percent volatilized

     1,4 dichlorobenzene:                      91 percent dechlorinated
                                               1.4 percent volatilized.

     This is the extent of Von Meyer's investigation of the application of
thionation to contaminated soils.  As part of his study, DCB's and three chlo-
rinated pesticides (Dieldrin, Lindane, and Heptachlor) were treated directly
with two reagents; sodium polysulfide in an ethylene glycol-water media, and
sulfur in ethylene glycol.  With water present, the organics were all  con-
sistently volatilized at the 108-114 °C reaction temperature.  The sulfur-
ethylene glycol reagent degraded Dieldrin 99 percent and Lindane 90 percent.
Heptachlor was not tested with this reagent.

     Thionation is in the early stages of investigation, and a process for
treatment of PCBs has not been sufficiently identified to permit further
evaluation.

Processes Based on Physical Technologies

     The term "extraction" is employed to cover not only liquid-liquid extrac-
tion but also leaching of solids with organic solvents.  Extraction of PCBs
from a contaminated soil or sediment is emerging as a prerequisite for most
chemical processes.  PCBs are first extracted using a suitable solvent, and
the obtained PCB solution is then subjected to one of the processes for
destruction of the PCBs.  Proven solvents include kerosene, methanol,  ethanol,
isopropanol, furfural, dimenthyl formamide (DMF), dimethyl sulfoxide,  ethylene
diamine, and Freon mixtures.  Supercritical fluids such as carbon dioxide have
also been considered.  Destruction of the dissolved PCBs is accomplished by a
chemical reagent such as APE6 detoxification, or the PCBs are concentrated  by
distillation or adsorption on carbon to give a RCRA waste that can be  incin-
erated.

     "Soilex" is a process developed at the Oak Ridge National Laboratory and
reported on in 1985.  Kerosene and water were determined to be the solvent  of
choice.  Kerosene can be recovered for recycle by steam-stripping.  Soil-to-
water ratios of 3 to 5 and soil-to-kerosene ratios of 3 to 5 were found to  be
best.  No difference in extraction performance could be determined between  15-
and 30-min mixing times.  The kerosene retention in the soil was about 25 vol
percent.  A three-stage batch pilot unit operating with a 6-to-l volume ratio
experienced soil feed PCB concentrations of 180 to 350 ppm and discharge soil
levels of 6 to 9 ppm.  Lower soil values could be obtained using additional
stages.  The process generates a RCRA waste (concentrated PCBs) requiring treat-
ment and disposal.
                                      38

-------
     The "Soilex" process reported pilot plant data showing an estimated  85
percent removal of PCBs from soils using a kerosene-water mixture.   The  kerosene
solubilized the PCBs and attendant other oils, and the water helped break up  the
soil particles.  The mixture ratio found to give the most complete  extraction of
PCB was one part of soil to three parts of kerosene and three parts of water.

     A rough cost estimate based on increasing the number of extraction  stages
to achieve desired background levels of PCBs in the treated soil, is:  $856 to
$913/m3.  This includes dredging, transportation, treatment, and redeposition
of the treated sediments.

     In 1985, The Galson Research Corporation showed that the rate  of  liquid-
liquid extraction of PCBs from oils solutions by APEG reagents could be  substan-
tially increased by adding dimethyl sulfoxide (DMSO) or ethylene diamine  to the
hydroxide/PEG phase.  The reagent with DMSO reached <2 ppm PCB after 30  min
versus about 400 min for the reagent without DMSO.  Details of the  APEG  process
have been described under Processes Based on Chemical Technologies.

     The Acurex Corporation designed and built a solid/liquid extraction  process
in late 1984.  A proprietary solvent is used, having been selected  by  comparisons
of adsorption isotherms and PCB diffusion rates into several fluids: pure hexane,
pure FC-113 (1,1,2-trichlorotrifluoroethane), a proprietary solvent blend, and
an FC-113/hexane blend.  With the pure Freon and topsoil, the equilibrium was
favorable; however, it took up to 18 h for the system to reach equilibrium.
Other solvents and solvent combinations were found that reduced the time  to
reach equilibrium to 30 to 40 min, a practical value.

     The Acurex Corporation reported that repeated washings of PCB-contaminated
soils with Freon-type solvents reduced the residual loading to <2 ppm.  The soil
types tested were sand-clay mixtures and a dark loam.  PCB loadings to 1,983  ppm
were leached in an agitated extractor.  The pilot soil-washing system  consists
of a soil contactor, dirty and clean solvent storage tanks, a solvent  reclamation
system, steam generator, and ancillary piping equipment.  It also includes vent
condensers and an activated carbon adsorption system for air pollution control.

     Basic estimated costs for the treatment alone range from $130  to  390/m3.
Adding costs for dredging, transport, and placement of the treated  sediments
gives a total cost estimate of $196 to $569/m3.

     In 1985, 0. H. Materials Company used methanol to extract PCBs from pre-
dried contaminated soils in a Superfund cleanup project at Minden,  West  Virginia.
The soil was reduced to <25 ppm PCB and land-farmed.  It is assumed that  further
reduction in PCB concentration could be achieved using more stages  of  extrac-
tion.  The process involves slurrying the ground and predried soil  with  methanol,
                                      39

-------
   separating and redrying the treated soil, and solvent cleanup for resale.   The
   cleanup employs activated carbon, which is subjected to incineration (other
 -- treatments could-be used) as a RCRA wastes- -The-cleaned -soil  is subjected"'to--:
   light land farming for degradation of any residual methanol.   Wastewaters  are
   treated in a holding pond.

        Based on the conceptual process and the experience thus  far, the process
   is estimated to cost $400 to $514/m3, including dredging and  transport costs.
   In the field application (1985), methanol was purchased at $0.69/gal ($0.18/L)
   and resold at $0.23/gal ($0.06/L).

        Vitrification stabilizes radioactive-contaminated soils  by melting.
   Large-scale tests (1985) have demonstrated the vitrification  of 300 Mg per
   setting.  At the high temperatures created (>1700 °C), the process pyrolyzes
   organic materials.  The products of pyrolysis diffuse to the  surface and combust,
   Any off-gases must be collected, monitored, and treated.  Remaining ash, along
   with other noncombustible materials, dissolves or becomes encapsulated in the
   melt.  Natural convective currents within the molten mass help distribute the
   stabilized materials more uniformly.  The molten soil or sediment is allowed  to
   cool to a durable glass and crystalline form.

        In-situ vitrification (ISV) is a patented process developed at Battelle
   Pacific Northwest Laboratory for the U.S. Department of Energy as an in-place
   stabilization technique.  Submerged sediments would be dredged before being
   subjected to this treatment.  In evaluation of its potential  application to
   soils contaminated with PCBs, an engineering-scale in-situ vitrification test
..,-fWJth PCB-contaminated soil  has been performed for the Electric Power Research
   Institute to determine the fate of PCBs and their byproducts  when the process
   is applied.
           \
        The cost of using the vitrification as a technique for in-situ treatment
   has been estimated by the developers.  The estimate includes  expenses from
   four categories:   site preparation and closure activities, annual  equipment
   charges, operational  costs, and consumable supplies including electrical power
   and molybdenum electrodes.   These costs are estimated to be between $150 and
   $330/m3 depending on electrical  power rates and soil  moisture content.  Soil
   moisture content  increases  the cost by requiring more energy  and time to
   vitrify a given volume of contaminated soil  because the water in the soil must
   be evaporated.  The total  cost of using this  approach requires,  for submerged
   sediments, the addition of dredging, transportation,  and  redeposition costs.

        The solidified blocks  of treated sediment could  not  be disposed of by  land
   treatment.  If treated at  a hazardous waste facility, they could be landfilled
   there.   The cost  of such disposal  would probably be the regular  charge for
   hazardous waste disposal,  $260 - $490/m3.  The treated material  is estimated  to
   be eligible for delisting,  however, and redeposition  at a lower  cost should be
   possible.   Given  that the  physical  form of the treated sediments will  limit
   somewhat the available locations for its redeposition, a  cost of $72/m3 has
 •  been  estimated.  This cost  is less  than that  of controlled, managed land
                                         40

-------
treatment  ($111)  since the material would require no control or management.
It is greater than the $33/m3 cost of short-term land treatment.  The overall
cost estimate is  thus: •-•••-••--:   .-•-.-  ------  = =--.— -=•-.•-.      ;-  ..--       ._ —

      Cost,  $/m3  = ($20 dredging) + ($13 - $126 transportation)
                   +  ($150 - $330 treatment) + ($72 redeposition)

                  = $255 to $548.

Processes  Based on Biological Technologies

     Biological technologies for PCB treatment include the use of free microbes
or their enzymes.  The microbes may be bacteria or fungi and may act aerobically
(or anaerobically) or facultatively.  They may be indigenous microorganisms,
conventional chemical mutants, or recombinant microorganisms.  Their action may
result in  partial or total degradation of the PCBs.  Total degradation, termed
mineralization, occurs when the metabolic products are C02, H20, and HC1.   Their
action may take place within the cell (bacteria) or external thereto (fungi).

     Enzymes may  be applied as free-flowing solutions or bonded to a solid
substrate.   The use of enzymes alone has great conceptual  potential but has
not progressed very far at this time.

     Thirty-four  bacterial strains and five fungal strains have been shown to
have varying degrees of competence in degrading PCBs.  These microorganisms
metabolize PCBs using enzymes specific to the strain.  Since most enzymes  are
notably specific  in .their actions,-:differ.ent,,enzymes .catalyzing the reaction-of-
different  PCB isomers  (termed cogeners in the biotechnology literature).  In
some cases,  the cogener may be metabolized only partially by a particular
species, ^and a product (e.g., a chlorobenzoate) may accumulate.  In a parallel
situation, another species may be able to metabolize that  product further,
although the second species may lack enzymes needed to metabolize the parent
PCB.  By themselves, neither species could mineralize the cogener.  However,
a mixed culture might act in concert mineralizing the product resulting from
metabolism of the substrate by the other species.  A consortia of more than
two species may be required to mineralize a substrate and the effective species
may be bacteria,  fungi, or a mixture of the two.

     Some  cells cometabolize PCBs; that is, they metabolize a PCB cogener  while
obtaining  their carbon and energy from other sources.  Such metabolism may be
partial or complete and depends upon enzymes already active in the cell.

     Biological technologies include a very broad spectrum of potential
capabilities, many aspects of which are being explored.  True measures of  their
ultimate capabilities for PCB-destruction cannot yet be made.  Recent develop-
ments in mutant and enzyme technology are described, in the following nine
paragraphs.
                                      41

-------
     Indigenous and conventional chemically mutated microorganisms (those found
growing naturally in diverse PCB-containing soils and sediments) have been
isolated, identified, and tested for PCB-degradation capability using an assay
mixture containing five types of cogeners selected to provide different resist-
ance to enzyme attack.

     The results obtained using 25 strains, including 15 isolates and 11 unidenti-
fied field isolates (e.g., H337 is an unidentified gram-negative organism isolated
from Hudson River sediments), identify Alcall genes eutrophus H850 and Pseudomonas
putida LB400 as the best microorganisms, both being able to degrade 13 test
cogeners 80 to 100 percent.  Ten additional strains of bacteria were studied by
different investigations.  Of these, the Arthrobacteria degraded the most PCB
cogeners.

     The degradative pathways and cogener preferences differ among the micro-
organisms.  Many bacteria oxidize PCBs via chlorobenzoic acid intermediates
and often accumulate the chlorobenzoate products.  The major pathway proposed
for A. eutrophus H850 involves a dioxygenase that preferentially attacks at
carboH positions 3,4.  In contrast, Corynebacterium sp MB! probably employs a
more common 2,3 dioxygenase mechanism.

     Genetics studies of PCB-degrading bacteria have demonstrated that some of
the effective strains contain one or more plasmids.  Plasmids are stable extra-
chromosomal circular double-stranded DNA replicons that are inheritable but are
also dispensable.  Mutants of H850 that have lost their plasmid have also lost
the ability to metabolize PCBs. .__-._  ...-,^  	.  ......   ,.. .  ..„_ . ^ _.

     Mutations that living cells may undergo alter the genetic message.  Some-
times the alterations in the'DNA code may lead to alteration of the cell's
metabolism.  Some mutations allow the cell to survive in the presence of
potentially toxic compounds.  In other cases, the cell acquires the ability to
use previously unsuitable substrates.  Some agents, including UV light, some
kinds of radiation, and some chemical agents, cause increased mutagenesis.
These mutations are characterized by being randomly distributed across the
DNA.  Mutations can be selected by applying selective pressures to a popula-
tion, creating an unfavorable environment in which only those cells that have
adaptation may take months before the altered population is large enough to be
observed.

     Limited studies on the degradation of specific chlorinated biphenyl com-
pounds by pure strains of bacteria have established some general features of
PCB metabolism.  A few strains have been shown to mineralize some chlorinated
biphenyls.  In most cases, bacteria can degrade one ring of a chlorinated
biphenyl but are unable to degrade the resulting chlorinated benzoates.  These
compounds accumulate in pure cultures.  Other strains have been shown to be
capable of completely mineralizing chlorinated benzoates (e.g., the hybrid
strain of Pseudomonas sp B13 carrying the relatively nonspecific oxygenase of
P. putida mt-2 carried on the TOL plasmid).  Mixed cultures of bacteria have
been shown to mineralize PCBs with four or fewer chlorines per molecule.  More
heavily substituted PCBs appear to resist degradation to a greater extent.
                                 42

-------
     The mechanism of hydroxylation of PCBs by bacteria has not been eluci-
dated, nor have the enzymes mediating the steps in the proposed pathways been
isolated.  Two pathways have been proposed.  The first involves initial
hydroxylation in~the"2,3-pdsitibh of the less substituted"ring, followed by
meta cleavage and subsequent degradation of the aliphatic portion of the
molecule to form substituted benzoic acids.  Chlorines on the aliphatic  car-
bons are lost during this process.

     However, this may not be the mechanism for degradation of PCBs substituted
in all the ortho positions.  A second pathway, based on the action of a  mono-
oxygenase in bacteria has been proposed after discovery of 4-hydroxy-4'-
chlorophenyl in extracts of bacteria cultures incubated with 4-chlorobiphenyl.
More evidence corroborating this mechanism is needed.  It is proposed that  the
major pathway of PCB metabolism in A. eutrophus H850 utilizes a dioxygenase
that preferentially attacks at carbon positions 3,4; while Corynebacterium  sp.
MB1 probably employs a more common 2,3 dioxygenase.  Limited evidence on
fungal metabolism of PCBs indicates activity of a monooxygenase.

     Little has been done to prepare or work with enzymes alone, separated  from
the organisms, despite the conceptual attractiveness of the approach. The
separation processes for PCB-degrading enzymes are yet to be developed.   Some
enzymes may be cell wall bound, requiring special  handling in their prepara-
tion.  At present, little is known concerning the use of enzymes alone for  PCB
decontamination of sediments.

     The Bio-Clean process is a biological  treatment process developed by
Bio-Clean, Inc., Bloomington, Minnesota.  It has been well demonstrated  for
PCPs.  It requires further testing with, perhaps,  the use of some additional
strains of microorganisms to bring it through the pilot stage for PCBs.   It is
listed here as a representative bio-process that has potential  application  and
can be made available commercially.  It does not require pretreatment to dewater
sediments, and it does not appear to generate any RCRA wastes or emissions.
The process involves two steps:  (1) extraction, sterilization, and solubilizing
the contaminants using high pH and temperature, and (2) bacterial  destruction
of the contaminant.  These steps generally require 3 days.  Standard digesters
that can handle 22 to 28 m3 of sediment (dry bases) are used in sets of  three
to provide a daily capacity in multiples of 22 to 28 m3/day.
     The processing cycle requires four steps:
     1.  Contaminated sediment is charged to the digester,  in slurry  form  if
         so received.  The final  digester charge will  be approximately  2/3
         water and 1/3 solids.  The charge is made alkaline with  sodium hydroxide
         (or other alkaline hydroxide), then heated to approximately  82 °C, at
         which temperature it is held for one hour.  The mix is agitated to
         promote solubilization of the PCBs.
                                      43

-------
     2.  After the extraction, the slurry is cooled to 30 °C,  neutralized and
         inoculated with a selected microorganism(s) to initiate  decontamina-
         tion.  The slurry is kept under these conditions for  48  to 72  h as
         required, until the desired degradation has been attained.  The exact
         time requirements must be determined by tests for each particular
         decent aminant.  The degradation is an aerobic process and  uses sterile
         filtered air for oxygen.

     3.  A treated batch of sediment is discharged to a dewatering  pit  where
         the sediment is separated for redeposition.  The material  is sampled
         and tested to determine the concentration of residual PCB  prior to its
         relocation.  The water is also analyzed for PCBs and  thus  disposed of
         in a manner suitable for the particular site (e.g., through sanitary
         sewer systems, storm sewer systems, or percolation into  the ground or
         recycle).

     4.  At the end of the decontamination period, the cycle begins again at 1.

     The estimated cost of treatment using the Bio-Clean process  on floating
barges is $191 to $370/m3.  This includes dredging at $20/m3,  and treatment
costs at $171 to $350/m3.

     The Sybron process is a conceptual process still in the laboratory scale.
Information obtained from the company was very limited, and its evaluation is
correspondingly limited.  The process may be able to operate under  both aerobic
conditions or anaerobic -conditions with facultative aerobes and supplements of
nitrates.  Further studies, to define an anaerobic process concept and principles
have been nroposed.  This process has not been developed sufficiently to permit
a description in terms of unit operations.  Information provided  by Sybron
includes tests of their Bi-Chem 1006 PB for removal of PCBs from  a  municipal
sewage sludge.  Cost estimates could not be made, due to the limited available
information.

     Aerobic and anaerobic composting of PCB-contaminated soil has  been studied
on a laboratory scale.  Experiments were carried out with soil that was intention-
ally spiked with PCBs (Aroclor 1242).  Aerobic composting resulted  in the great-
est decrease in PCB concentration, with an average reduction of 62  percent in
four weeks.  Residual concentrations ranged from 504 to 688 ppm.  Anaerobic
composting resulted in 27 to 47 percent reduction in four weeks,  with residual
concentrations ranging from 825 to 1,120 ppm.

     Composting would appear to require considerable work-site space to handle
the sediment, with considerable monitoring to determine the progress of the
process.  Lack of control of weather and other conditions would make the proc-
ess uncertain in terms of time and effort required.
                                    44

-------
— :  Data for composting are insufficient to" provide" a basis  for  cost  estimates.
 Considerable extrapolation of the performance data  would  be required to project
 process conditions for desired residual  PCB concentrations.

                Initial  Characterization  and Ranking of Processes


      This subsection provides two comparisons of the alternative  treatment
 processes previously discussed:   characterization and ranking.  The character-
 ization provides for objective comparison of processes.  The  ranking provides
 a subjective comparison of the processes based on the seven criteria previously
 described.

 Characterization

      Table 6 summarizes five characteristics of the processes:  unit operations,
 available capacity, conditions/limitations, concentration handled, and any
 generated RCRA wastes.   The unit  operations employed are  given, and each is
 identified by a number.  Generally,  a  greater number of unit  operations will
 mean a greater effect on treatment costs.  None of  the processes  has currently
 available capacity approaching that  required for major cleanups.  Therefore,
 the time required to build capacity  is listed.  Construction  time ranges from
 12 to 24 months.  Certain conditions that typify the process  or limit  its
 versatility are given in column 4 of Table 6.

      The data from studies of the processes-'were-examined for ranges of'PCB
 concentrations handled  to date.  .Generally, the values are not limitations on
 the process, but only on the data acquired.  The value <300 ppm for the
 Bio-Clean process may,  however, be a limitation requiring process adjustment
 to control.
                                   45

-------
lABLt  b. IKEATMENT  PROCESSES  CHARACTERISTICS
;i
Process t
Chemical
Supercritical water oxidation
KPEG, Terraclean-Cl
KPEG, NYU
t
LARC I
•i
Advanced electric
reactor (J.M. Huber)
Physical
0. H. Materials
methanol extraction
"Sollex" kerosene/water
Acurex solvent wash
Vitrification
Available
capacity
Unit (or time
Operations to provide)
1.4,10 	
1,3,4,7 (24 mo)
1.2,3,4,5,
6,7,9
1,2,5,15 (24 mo)
7,8,12,13 (16 mo)
14
•
2, 7,8, ,14
15
1,2,5,15 	
2,4,5,6,
10,11
8,12,14
; Conditions
and limits
20-40% solids; 374 °C,
23.3 MPa organic
content >5% or supple-
mental fuel.
150 °C, 0.5-2 h
t
i
tolerates 25% water.
2204 °C, 2,400 kWh/m3
needs predryer
,
predry to <1% moisture
26% of kerosene sol-
vent retained in soil;
3 d per batch
t
3*-12 washes, tolerates
<40% water.
Electrical power usage
Concentration
handled
>3000 ppm
500 ppm or
greater


480 ppm
>3000 ppm

>400 ppm
to 350 ppm
tested
up to 1 ,983
ppm
500 ppm
RCRA
waste
generated
None
w.w.tr.
act.
carbon


None
None
tO
PCB-loaded
carbon from
solvent
cleanup
Concentrat-
ed PCB from
still to
Incinera-
tion
Concentrat-
ed PCB's to
KPEG
None
                              Increases with soil
                              moisture: submerged
                              sediments dredged
                              and treated

-------
                                                   TABLE  6 (continued)
1 ,
Process
ological
mposting
Unit
Operations
15,16
Available
capacity
(or time
to provide)
(16 mo)
\
Conditions
and limits
i
Seasonal effects,
reaction time must
>4 weeks
Concentration
handled
1590 ppm
be
RCRA
waste
generated

Treated
material
is still a
RCRA waste
o-Clean
bron Bi-Chem 1006
1,2,17
15,17
27 m3/d avail-
able, 12 mo for
full-size
Proved for PCP, labor-   <300 ppm
atory confirmed for
PCB's

Unknown                   Unknown
None
                                                               Unknown
Dte-Unit operations key:

   1.  Liquid/solids separation
   2.  Extraction/solubilization  (liquid-solids)
   3.  Liquid/liquid extraction
   4.  Chemical  reactor
   5.  Stripping still
   6.  Solvent  recovery still
   7.  Adsorption
   8.  Dryer (solids
   9.  Dryer (liquids)
                                  10.  Filtration
                                  11.  Steam cleaning
                                  12.  Thermal  reactor
                                  13.  Grinding
                                  14.  Air pollution controls
                                  15.  Landfarm
                                  16.  Inoculation/digestion
                                  17.  UV light reactor

-------
Ranking

     In contrast tT> process characterization, ranking is subjective.   An
attempt was made to define and determine a single number that could represent
the overall position of each process relative to an arbitrarily defined per-
fect process.  To do so, factors considered were given ratings as described
below.

     The goal set for process performance is to reduce the PCS concentration  in
treated sediments to background levels of 1 to 5 ppm.  Several of the proc-
esses were found to meet this goal.  Those that showed reduction to less  than
2 ppm were assigned a rating of "6".  Those that attained a level between 2
and 10 ppm were assigned a "4".  Those with residual concentrations greater
than 10 ppm were rated "2".

     Available capacity was found not to exist for any of the processes.   How-
ever, several were developed sufficiently to permit projections of the time
required to build a facility for application of the treatment.  Those for
which such projections could not be made were rated "2".  Those requiring 24
or more months were rated "4".  Those requiring 12 to 16 months were  rated
"6".

     Conditions/limitations that were rated were tolerance for water, required
processing time, and controllability of process conditions.  Those treatments
that could tolerate water up to about 40 percent would not require a  drying
step with its attendant fines control problems.  Those requiring only 1 day
for treatment could generally show a faster rate of cleanup.than those requir-
ing 3 days.  Some biological processes required more than 3 weeks. The treat-
ments generally provided control of the processing conditions; however, a few
(e.g., composting) would not necessarily do so.  The three conditions/limita-
tions were ranked as follows:

         Conditions/limitations                                 Rank

     Tolerates to 40 percent water and treats in 1 day           6
     Sediment needs to be dried                                  5
     Tolerates to 40 percent water and treats in 3 days          4
     Tolerates water and treats in >3 weeks                      3
     Sediment needs to be dried, treats in >3 weeks              2
     Processing conditions uncontrollable                        1

     Concentration range handled in data developed for the processes  ranged
from unknown to 3,000 ppm.  Ratings were assigned based on the upper  limit of
feed concentration.  The ratings were as follows:
                                   48

-------
     PCB concentration treated, ppm                               Rank

             >3,000                                                6
             2,000 to 3,000                                        5
             1,500 to 2,000                                        4
             500                                                   3
             250 to 350                                            2
             Unknown                                               1

     Status of development ratings were "1" for no data, "2" for laboratory-
scale tests completed, "3" for bench-scale tests completed, "4" for pilot-
scale tests completed, "5" for field tests completed; and "6" for commercial
system designed and ready for construction.

     Test and evaluation data needs could be rated differently, depending upon
the purpose.  For indicating the extent to which a treatment process is readied
for use, the more data that are available the better.  For indicating the need
to support a very promising technology that lacks sufficient progress, the poten-
tial and the data needs should be rated in combination.  The ratings used here
are for the former purpose and are as follows:

     Test and evaluation data needs                                Rank

     None except permits and checkout                              6
     Field tests                                                   5
     Pilot tests and costs       _     .._  „   .  	- ,  . u      -   f  .
     Laboratory and bench tests*  "   	   ~   "'~'  ~^~         3
     Conceptual treatment process design                           2
     O'.D.R data, residual PCB data, RCRA waste data                1

     The estimated costs were rated by comparing the range of the cost estimates
obtained with the cost of placing them into a chemical waste landfill.  Treatment
processes showing the lowest estimated cost range were rated "6";  those showing
a probable cost lower than landfill were rated "4"; those showing  an estimated
cost equal to landfill were rated "2"; and those showing an estimated cost range
greater than landfill  were rated "1".
     Overall ranking was accomplished through the use of weighting factors
assigned to each rated factor.  The weighted average rank was then obtained  by
summing the products of the weighting factors and the ratings and dividing  by
the sum of the weighting factors.  The weighting factors were:
                                   49

-------
             Factor     	_.                             Weight

     Residual PCB concentration                                    5
     Capacity                                                      2
     Conditions/limitations                                        3
     Concentration range handled                                   2
     Status of development                                         2
     Test and evaluation data needs                                1
     Estimated costs                                               4

     Table 7 lists five additional characteristics of the processes and  the
rating developed in the ranking process.  The characteristics shown here
relate to the needs for further process development, listing, and  evaluation.
The listing begins with an indication of the process status in terms of  stages
of development completed.  The processes range in stages completed from  con-
cept to pilot plant.

     Based on the weighted ratings, the eleven processes rank as follows from
highest to lowest:  KPEG, LARC, Acurex, Bio-Clean, Modar-Supercritical Water,
Advanced Electric Reactor, Vitrification, OHM Extraction, Soilex,  Composting,
and Sybron Bi-Chem 1006 PB/Hudson River Isolates.
                                    50

-------
                                           TABLE 7. TREATMENT PROCESSES RANKING
Process
Chemical /physical
Supercritical water
oxidation, Modar
KPEG Terraclean-Cl
LARC
Advanced electric
reactor
Physical
0. H. Materials,
Methanol extraction
Soilex
Acurex solvent wash
Status3
Field test with
PCB liquids
Pilot tests
Lab tests
Pilot tests


Field tests under
way
Pilot tests
Pilot-scale
Estimated
D/D/R
efflclency.% D
>99.9995

>98
>90
>99.9999


97

95
(3 stages)
e
Estimated
residual
PCB, ppm
<0.1 ppb

<1 ppm
38-50
<1 ppb


<25 ppm

6-9 ppm
<2 ppm
Test & evaluation
data needs
1,2,3,4,5,6,7,

1,6
2,3,4,5,6,7
Noned


2,3,6,7

5.6,7
Identity of
Estimated
costs, $/m3
250-733

208-375
223-336
830-943


401-514

856-913
196-569
Rating0
4.58

5.42
5.26
4.58


4.16

3.26
5.21 C
                      (field  tests
                      planned)

In-s1tu  vitrification  Pilot test  of  soil
 Battelle Pacific
 NW for  EPRI
                              mixed solvent,
                              6.7
99.9
None 1n vltrl-    6
field block. 0.7
ppm 1n adjacent
soil
255-548
4.53
See footnotes at end of  table.

-------
                                                     i nuui-  / .  y mil L i nucu /
Process
Biological
Composting, aerobic
anaerobic
Bio-Clean, aerobic
Sybron Bl-Chem 1006
Status3
Lab-scale
Lab-scale
Bench-scale
Lab-scale and
Estimated
D/O/R
efficiency ,%b
62
18-47
99.99
concept 50
Estimated
residual
PCB, ppm
504-908
825-1268
25 ppb

Test & evaluation Estimated
data needs costs, $/m3
4,5,6
4,5,6
3,5,6,7 191-370
3,4,5.6.7

Rating0
1.79
1.79
4.84
1.47
NOTE-Data needs key:

    1. D/O/R data
    2. Residual PCB data
    3. Unit operations data
    4. Bench-scale data
    5. Pilot-scale data
    6. Field test data
    7. Cost data
    8. RCRA waste
 aStatus Is defined 1n terms  of the types of studies  completed.

 bD/D/R=destruct1on/detoxif1cat1on/removal.

 cThe rating was obtained as  shown  by the example,  under Characterization.

 dAER 1s fully permitted under TSCA 1n EPA Region IV  for destruction of PCB,

 treatment is continued until a residual of <2 ppm PCB's is  obtained.
                                                                                                                          C\J
                                                                                                                          ur>

-------
                          Final  Assessment of  Processes

      This  part  of  the^final  assessment, building upon the  initial phase, describes
 the development of criteria  for the  final ranking of the processes, elaborates on
 description  of  processes  previously  mentioned, and discusses each new process
 being assessed.

      Under the  "Initial Assessment of  Processes" we discussed the identification
 of  eleven  candidate  processes,  out of  sixty four evaluated, with potential for
 decontamination of PCB-contaminated  sediments.  Listed in  rank from highest to
 lowest, the  top eight  processes were:  KPEG,  LARC, EPRI (Acurex), Bio-Clean,
 Modar-Supercritical  Water, Advanced  Electric  Reactor, Vitrification and OHM
 Extraction.   The three  processes  ranked but considered non-candidates were:
 Soilex, Composting,  and Sybron  Bi-Chem 1006.  At the start of the final assess-
 ment,  four developers  indicated their  processes should be eliminated from further
 consideration.   These  four processes were: LARC, EPRI (Acurex), Advanced Electric
 Reactor, and  OHM Extraction.  During the course of the final assessment, four
 new technologies became available:   The Basic Extraction Sludge Treatment
 (B.E.S.T.) by Resources Conservation Co.; the UV/Ozone or Hydrogen-Ultrasonics
 Technology by Ozonic Technologies, Inc.; the  CFS Propane Extraction Process
 by  C.  F. Systems Corp.; and  the Low-Energy-Acetone-Kerosene Extraction Process
 by  Applied Science Dept.  of  New York University.  The UV/Ozone or Hydrogen
 Ultrasonics  Technology  provides continuity for the radient-energy approach
 previously represented  by the LARC process.   The other three processes provide
 improved approaches  to  extraction technology.  These eight processes, four
 initial and  four new, were assessed  as final  candidates for thorough test and
 evaluation.

 Ranking Criteria

     The .following  criteria were  used  for the final  selection of treatment
 processes  for test  and evaluation (pilot-field tests):

     0  The  likelihood that the process will  acceptably clean-up the PCB-
        contaminated sediments;

     0  The  probable cost of the  application;

     0  The  relative level of T and E  effort;

     0  The  availability  of a processing system; and

     0  The likely future commercial  availability of the process.

     Acceptable  clean-up  requires a standard.  The standard selected is a PCB
concentration in treated  sediments of  2 ppm or less.  The regulation require-
ments related to this standard  were previously discussed under the initial
assessment.
                                      53

-------
     The probable cost of treatment 1s presented as the cost per cubic meter of
sediment treated, based on a system sufficiently large to process 380,000 m3
of Hudson River sediments in 2.5 years.  By focusing on a specific site and
-stze of clean-up-task,-each process could be assessed using .available data from
sampling and analyses to characterize the feed materials to the processes, and
comparative cost estimates for a specific application could be obtained.  The
sediments from the Hudson River also meet the requirement for use of a variety
of soil/sediment types in testing PCB-treatment processes.

     The probable cost of treatment was obtained from the developers for those
processes sufficiently supported by commercial firms, or was estimated using
as major cost elements capital costs, and operation and maintenance costs.
Treatment process requirements determined capital, energy, and maintenance
costs.  Labor rates, overhead, contingency, profit, and health and safety were
costed using standard unit values for all the processes.

     Maintenance was estimated at 10 to 15 percent of the capital cost, depend-
ing on the number of unit operations involved and engineering practices for
the operation.  The allowance for safety equipment was generally 0.30/m3 of
sediment treated ($114,000 for the clean-up of 380,000 m3 of sediment).

     The treatment system cost estimate was capitalized (recovered) over the
2.5 years of operation taken as the base period.  Some developers provided
treatment costs that were correspondingly lower for subsequent applications.

     Engineering costs were not included in the estimates.  An additional 7 to
15 percent of the subtotal cost (before profit and contingency) should be
included to provide for engineering in the design and construction stage of
the process application.

     The test and evaluation effort required has been estimated based on a
comparison of available process data with the requirements for thorough test
and evaluation.  A checklist of information requirements was developed to
identify the data categories to be supplied to qualify the processes for a
permit to test.  The checklist identifies the following information as basic
to assessment of each process:

     1.  Waste characteristics;

     2.  Process engineering description;

     3.  Sampling and monitoring plan;

     4.  Accident and spill  prevention and countermeasure; and

     5.  Demonstration test plan.
                                    54

-------
'/'    For these assessments, Hudson River sediments  were selected  as  the char-
 acterized wastes.  Hudson River sediment material has  been  classified according
 to its content of clay, silt, muck, muck and wood chips, sand,  sand  and wood
 chips, coarse sand, and coarse sand and wood chips.  Sediments  have  been shown
 to range from clay to cobbles, with the largest mass fraction being  in the sand
 sizes.  The coarse-fractien-^0-,42 mm) of the's'ediments typfdaily contained" -"
 wood chips, sawdust, shale chips, cinders, and coal fragments.  The  fine size
 fractions contained some fragments of the above, plus  sand  (containing quartz
 and feldspar), silt, clay, and organic material. The  highest PCS concentration
 was the muck with wood chips class, which typically had over 30 percent silt and
 clay, high volatile solids and some small but visible  wood  chips. The size low-
 est in PCS was medium sized sand or gravel without  wood chips.

      For most of the processes assessed, test systems  are available  from the
 developer.  Most developers would need financial support of the testing time
 and effort.  A few of the processes (e.g., Low Energy  Extraction)  lack a suit-
 able test system, but their estimated costs of application  justify their
 consideration herein.  The availability of a suitable  test  system and any condi-
 tions/restrictions on its use are discussed for each process.

      The likely future commercial  availability of the  processes vary with the
 strength and extent of their sponsorship.  Some developers  are  commercial firms
 in  the waste treatment business with resources committed to further  commercial-
 ization of their process.  Some developers have a need  for  financial support to
 achieve commercialization.  In all  cases, the short-term (2.5 years) effort
 projected in this evaluation, and the uncertainties of  further  markets make the
 construction of a full-scale treatment system contingent upon completion of the
 T and E (with attendant EPA approval  of the process) followed by  a contract for
 the clean-up work itself.  Under these necessary conditions, all  the processes
 assessed would be commercially available.  The estimated time required to make
 them available varies from process  to process; however,  this has  been taken
 into consideration in this evaluation.
          •>
 B.E.S.T.

      The Basic Extraction Sludge Treatment (B.E.S.T.) process has been developed
 by  the Resources Conservation Co.  (RCC), 3101 N.E. Northup  Way, Belleview, WA
 98004.  The company  was founded in  1968 and is a subsidiary of  Reading and
 Bates.  The process  uses  a solvent  having an  inverse critica solution point in
 water to remove water and oily material  from  solid matter.  It  has been applied
 to  clean-up PCB-contaminated oily  sludges at  a CERCLA site  (General Refining
 Site,  Savannah, Georgia).

      The company has bench units to perform glassware simulations of the process,
 required to establish parameters for  its application to  a particular sediment.
 The company also has a process components test unit sufficiently  large to process
 91  kg  of sediment  feed in seven days,  and a large-scale  skid mounted unit designed
 to  process  91  metric tons per day  (24  hours)  of  feed.
                                     55

-------
      The process Is described generally as it  is  applied  to  PCB-contaminated
'sediment, sludge, or other feed material  containing solid matter, oily con-
 taminants, and bound and unbound water.  The feed is first pretreated with an
 alkaline composition, then admixed with triethylamine (TEA)  while cooling
 below the critical  solution temperature (CST). A single  liquid  phase is formed
-from which the solid matter i-s separated.  The liquid is  then  heated to a temp-
 erature above the CST, to form an amine phase  and a water phase, after which
 the water phase is decanted from the amine phase.  The amine phase  contains
 substantially all of the oily material  including  organic  contaminants.  It is
 processed to recover the oil and contaminants, and the TEA is  recycled for the
 processing of additional feed material.  The pretreatment of the feed with an
 alkali reduces substantially the amount of residual  amine carried over into the
 solid and water products.

      This process is essentially ready  for T and  E.   The  program should include
 three "glassware" evaluations at a cost of $3,500 each.  If  the  glassware test-
 ing is successful, a pilot test at a cost of $50,000 would then  be  conducted.
 Total cost:  $50,500, with the equipment  provided by RCC. EPA would supply
 the sampling and analyses, and evaluation of the  results  as  a  separate program.
 The estimated cost of T and E is $148,530 (analyses - $13,030; support including
 permits, sampling/testing, and report - $75,000;  test operation  - $60,500).

 Ultrasonics/Hydrogen - Ozone/UV Technology

      The LARC process, discussed under  initial  assessment, is  not now under
 further consideration by the developers (George Anspach,  Atlantic Research Corp.,
 5390 Cherokee Ave., Alexandria, VA  22312). Another firm (Edward A. Pedzy,
 Ozonic Technology Inc., 90 Herbert Ave.,  P. 0.  Box 85, Closter,  NO  07624) has
 capabilities and interest in conducting tests  of  PCB-contaminated sediments.
 Thus, their Ultrasonics/Hydrogen-Ozpne/UV Technology Became  a  replacement^in -
 approach for the similar LARC-UV process.  This Ozonic process has  been shown
 to be effective in cleaning PCBs from solid surfaces.
          \
      The process is based upon simultaneous extraction and treatment of a sedi-
 ment slurry with ultrasound (above 20KHz) added to increase  rates of dissolu-
 tion of the PCB contaminants and reduce coalescence of bubbles of gas reactants.
 The treatment is achieved by either ozone/ultraviolet irradiation or hydrogen/
 ultraviolet irradiation.  The sediments are fed as a slurry  containing about
 20 wt. percent solids into a mixing tank  where predetermined amounts of detergent
 and sodium hydroxide are added.  The conditioned  mix is pumped to the reactor
 where the solids are maintained in suspension  while the mixture  is  exposed to
 ultraviolet irradiation and ozone (or hydrogen) is added. During this treatment,
 microscopic turbulence is produced through ultrasonics.  The treated slurry is
 fed into a cyclone where the solids are removed,  sampled, and  tested to certify
 them for discharge.  The separated water  is neutralized and  also tested before
 discharge.

      The process needs to be tested using PCB-contaminated sediments to estab-
 lish the process feasibility and to determine  its potential  performance.  The
 tests would involve assessing both UV/hydrogen/ultrasonics and UV/ozone/ultra-
 sonics.  The total  estimated T and E cost is $151,000 (analyses  - $21,000;
 T and E support, permits, and report -  $75,000; system operation -  $55,000).
                                       56

-------
 'Bio-Clean Naturally-Adapted Microbial

      The Bio-Clean Naturally-Adapted Microbial process has been developed by
 Bio-Clean,  Inc., 201 W. Burnsville Pkwy., 11306, Burnsville, MN 55337.  A
-patent is pending-on ttie-process*  Bio-Clean 1s-a-company-engaged in develop-------
 1ng process systems to clean-up, remove or destroy hazardous chemicals 1n the
 environment.  The company has a suitable bench-system to support pilot and
 field tests (32-45 metric tons/day capacity), and a laboratory test system
 for conducting preliminary tests to establish specific conditions for a field
 test.  Support would be needed for the pilot and field work.  The process, once
 proven, would be available from the firm.

      The process utilizes Arthrobacteria sp and/or other naturally-adapted
 microbes to destroy PCBs (and related organics) under aerobic conditions.
 Specifics of the process have been discussed under Initial  Assessment of Proc-
 esses-Processes Based on Biological Technologies.  Preliminary tests would be
 required to adapt the biological treatment to the specific waste.  The total
 estimated T and E cost is $163,600 (analyses - $33,600; T and E support, permits,
 and report - $75,000; system operation - $55,000).

 KPEG with DMSO

      Nucleophilic substitution with the KPEG process and the Galson Research  Corp,
 version (Terraclean-Cl process) was previously discussed as a means of decon-
 taming PCB-bearing sediments.  Presented here is an elaboration of that discus-
 sion pertaining to recent developments in the KPEG process.  Although the
 Corporation identified their version of the KPEG treatment  process as the
 "Terraclean-Cl" process, a more descriptive name is "KPEG with DMSO", since the
 DMSO is the major ingrediant in their modification to the KPEG process.

      The potassium polyethylene glycolate (KPEG)  process has two potential  '
 applications in the treatment of PCB-contaminated sediments:  The treatment of
 the sediments  themselves, or the treatment of concentrated  PCBs resulting from
 extraction processes.  The former is assessed herein in terms of the Galson
 Modified KPEG  process, that  is with dimethyl  sulfoxide (DMSO).  The latter
 application would require treatment of a relatively concentrated solution of
 PCBs in a non-aqueous solvent (triethylamine, kerosene, etc.).

      The process developed by the Galson Research Corporation, 6601  Kirkville
 Road,  E.  Syracuse,  New York  13057,  is  scheduled to  demonstrate treatment  of
 contaminated sandy  soil  on Guam, and is assessed  here based on data from the
 Phase  1 study  and new data developed to qualify for this demonstration.

      Gal sort has laboratory test facilities necessary to-set process parameters
 for a  demonstration,  and  a 150 L (40 gal)  pilot system.  For treatment  of Guam
 soils,  a  1.5 m3(40  gal)  system will  soon be constructed and available for
 various demonstrations.   It  is this  potential  system for Guam soils that  was
 considered  in  this  assessment  study.
                                      57

-------
      As  previously  discussed  under  initial  assessment of processes, the Gal son
 or "KPEG  with  DMSO processrdestroys "PCBs  by'hucleVphillc'Vubstitution/ '"Potassium
 hydroxide is  reacted  with  polyethylene glycol  (PEG) of about molecular weight
 400 to form an  alkoxide.   The alkoxide reacts  with one of the chlorine atoms on
 a chlorinated molecule  to  produce an  ether  and potassium chloride.  Addition of
 an RO-group enhances  the  solubility of the  molecule and makes 1t less toxic.
 The reaction  may  continue  until  several  chlorine atoms are removed from the PCB
 molecule.  The  reagent  consists  of  a  mixture of PEG, potassium hydroxide, and
 dimethyl  sulfoxide  (DMSO).  The  DMSO  acts as a solvent or phase transfer agent
 to extract the  PCBs from the  sediments.

      Equipment  to be  used  to  implement the  above chemical process consists of a
 carbon steel  reactor  (1.5  m3  capacity),  storage tanks for pre- and post-
 treatment  bulk  storage, and a carbon  adsorption unit for collection and control
 of gaseous byproducts of the  process.

      Contaminated sediment is fed to  the reactor from 55 gallon drums.  An
 equal volume  of reagent is added to the  soil in the reactor.  The reagent is
 blended  with  the  soil using a carbon  steel  bladed mixer.

      During operation of the  system,  reagent is mixed with make-up reagent
 from  the reagent  storage tank and recirculated into the reaction vessel
 containing contaminated soil.  The  reaction vessel  is heated and the soil and
 reagent  are kept  mixed until  the reaction is complete.  Volatilized material
 from  the bulk storage tank and the  reaction vessel  are vented through a  char-
 coal  adsorption unit.  Water  vapor  is condensed and used as wash water.   At
 the end  of the  reaction, the  slurry is pumped into a similarly vented bulk  •
 storage  tank".   The  reagent is  decanted,  weighed, and stored for reuse.  The
 soil  is  washed  twice with water to  remove excess reagent, and the wash water
 is held  for analysis and possible treatment with activated carbon.

      For this assessment study, temperature in the reaction vessel is controlled
 at 150 °C as measured on the  system's temperature gauges.  Mixing blade  speeds
 are controlled  between 500 to  1,500 rpm.  Energy to operate the blades and heat
the reaction  vessel is supplied by a diesel generator.

     The treated  sediment is  held for analysis.  If PCB concentration is greater
than  2 ppm, the sediment is retreated.   Byproducts  1n the sediment will  be less
than  2 ppm KPEG with DMSO plus reaction  products.   All  of which have been proven
to be non-toxic.  The total T and E cost is estimated at $196,000 (analyses -
 $21,000;  T and  E  support, reports, and permits - $75,000; system operation -
$100,000).
                                    58

-------
 CFS Propane Extraction

      The CFS Propane Extraction Process is a  technology  available  from C. F.
 Systems Corporation, 25 Acorn Park, Cambridge,  Massachusetts 02140.  The company
 is a subsidiary of Arthur D. Little.   As  applied  to  PCB-contaminated sediments,
 the process uses propane at ambient temperature and 1378 kPa (200  lb/in2) to
 extract PCBs along with other oily organics from a water slurry of the sediment.
 The treated slurry is discharged after separation  from the  liquid  propane which
 contains dissolved contaminant.  The propane  solution is fed to a  separator
 where the solvent is removed by vaporization  and recycled.  The contaminants
 are drawn off as a concentrate for final treatment.   The process has been tested
 for PCB-containing refinery sludge.  The PCB  content  of  the solids component of
 the sludge was reduced to 5 ppm.

      The company has a small portable 1-liter test unit  for preliminary evalu-
 ations  of the potential  of the process using  0.56  kg  of  feed, and will have a
 mobile  propane pilot system soon.   Preliminary  tests  of  Hudson River sediments
 using the portable test  unit would be conducted  by C. F.  Systems at their
 expense.  Large-scale tests would  require  financial support.

      The technology is available for preliminary testing  to establish the proc-
 essing  parameters and system needs for PCB-contaminated  sediments.  A full-
 scale system would be designed based upon  the results of  testing using the
 bench-scale unit.  C. F.  Systems would not be able to conduct the tests at
 Cambridge,  but could make the unit available  for testing  elsewhere.  The total
 estimated cost for T and  E is $122,660 (T  and E  including permits and report
 $75,000;  analyses - $17,660; system operation -  $30,000).

 Modar Supercritical  Water - Oxidation

      The foodar Supercritical  Water - Oxidation  process is another process that
 has been  previously discussed under the initial  assessment of chemical  tech-
 nologies.   The process is described here based upon discussions with the
 developers  and data  supplied by them during the  final  assessment.

      The  process  has been developed by MODAR, Inc., 3200 Wilcrest, Suite 220,
 Houston,  Texas 77042.  The process  utilizes water  above critical  conditions
 (374  °C  and  22.1  MPa)  to  increase  the solubility of organic materials and
 oxygen and  effect  a  rapid  oxidation,  destroying  organic contaminants.  It has
 been  applied  to destroy PCBs  in oils,  and  to  decontaminate dioxin tainted soil.
 The company  has a  laboratory test  unit and a  skid-mounted pilot test unit having
 a nominal organic  material  flow capacity of 190  Liters/day.  This unit has been
 operated at  MODAR's  research  facility in Natick, Massachusetts  and at the CECOS
 International  Niagara  Falls  site.

     The  following  describes  the preocess  as  it would be applied  to sediments.
 If not done as part  of the dredging  operation, the system would need to include
a screen to  remove  rocks  and  large  pebbles.   A feed pump would  handle the range
of sediment particle  sizes  up  to about 2 mm diameter,  at 20-40  percent  solids.
                                      59

-------
 Feed to the process is controlled  to an upper limit  of  heating  value of 4187
 kj/kg (1800 Btu/lb).  The Hudson River sediments  lack sufficient  heating
 value, therefore-fuel  addition "will  be necessary.  A combination  of preheat
 by exchange with process effluent  and fuel  addition  is  a more cost-effective
 option.  A portion of  the supercritical process effluent will be  recycled to
 the reactor by a high-temperature, high-pressure  pump to raise  the combined
 fluids to a high enough temperature  to maintain rapid oxidation reactions in
 the countinuously fed  reactor.  Oxygen, stored as  a  liquid,  is  pumped to
 system pressure, preheated,  and  metered into  the  reaction  vessel.  Alterna-
 tively, air can be compressed  and  used as the oxidant.  It would  be best to
 demonstrate viability  on their bench-scale  unit.   This  would avoid the costs
 of modification of the pilot system  for slurry handling and  onsite demonstra-
 tion until  the process has been  proven. To accommodate the  particle-size
 restrictions,  the sediments  may  need to be  pulverized and  hydrosieved to
 provide a feed of size less  than 38  urn. The  total estimated cost of T and E
 is $483,000 (analyses-$8,000;  T  and  E support including sample  preparation,
 permits,  and final  report-$50,000; operating  cost-$425,000).  The support
 cost is high because of the  anticipated effort to  be expended in  resolving
 design and  safety considerations.

 Low Energy  Acetone-Kerosene  Extraction

      The  Low Energy-Acetone-Kerosene Extraction Process is being developed by
 the Applied Science Department of  New York  University, 26-36 Stuyvesant St.,
 NY,  NY 10003.   The extracted contaminants are concentrated for  subsequent
 disposal.   The basic principles  of the process have  been confirmed experimentally
 using  PCB-contaminated sediments from Waukegan Harbor,  Illinois.

      The  process is based  on an  EPA-funded  study to  develop and evaluate a low-
 energy technology for  extraction of  PCBs and  other organic contaminants from
 sediments and  concentrating  the  extract  in  a  manner  suitable for the application
 of  a chemical  destruction  process.   Details of the process and the study have
 been  previously discussed  under  Section  4 (Discussion on Extraction Process).
 The  development of  this process  through  field test and evaluation is estimated
 by  New York  University to  require  $827,000.   This  includes the construction and
 operation of a field-scale unit.   A  possible  commercial  sponsorship is being
 actively  sought  which,  if  confirmed,  would  reduce this cost to an estimated
 $170,000.

 In Situ vitrification

      In Situ Vitrification was previously discussed  as a process developed to
treat  radionuclide-contaminated  wastes,  and subsequently tested for possible
adaptation  to  the treatment  of PCB-contaminated soils.  Submerged sediments
would  be dredged  before  treatment.   The  process stabilizes contaminated soils
by melting  into  a durable  glass  and  crystalline form.  Four electrodes are
inserted into  the sediments  in a square  array.  A path for electric current is
made by placing  a mixture  of graphite  and glass frit between the electrodes.
                                      60

-------
Dissipation of power through the starter materials creates temperatures  high
enough to melt a layer of sediment, which establishes a conductive path.   The
molten zone grows downward through the contaminated soil. ~At the high tem-
peratures created (<1700 °C) organic materials pyrolyze, diffuse to the  sur-
face, and combust.  Off-gases are collected, monitored, and treated.

     The process has been developed by Battelle Pacific Northwest Laboratories,
P. 0. Box 999, Richland, Washington.  Battelle has indicated that they are
performing an Eastern soils demonstration in July, 1987 at the Oak Ridge
National Laboratory site in Tennessee, and have scheduled a pilot-scale  PCB
test for September, 1987.

     The estimated total T and E cost is $400,000 (Analyses, $14,200;  T  and E
support, $75,000; test system operation, $310,800).  The test system includes
exit gas, capture canopies, and pollution controls.

     The estimated treatment cost for this process is higher than that for
other processes evaluated, primarily due to the cost of consumable electrodes
and electric power.
                                      61

-------
                          Selection of Three Processes

     All the processes assessed have merit.  In selecting among them, a  ranking
system has been employed for comparative simultaneous evaluation of all  five
criteria characteristics.  For each process, the desirability of immediate
thorough test and evaluation was expressed by a desirability value, Dj:
              Dj = (dd
 •
!og Dj.modified s 1/n (*i log d^ + X2 log d2j + x3 log d3j  + ... + xn log dnj
          "x
                           0 < d < 1


     The value found for each characteristic, y, was transformed to a value of
d according to the following judgements:

     Value of d

     1.0-0.99         Represents the ultimate level of the characteristic y.
                      Improvement beyond this point would have no appreciable
                      value.

     0.99-0.80        Acceptable and excellent.  Unusually good performance.

     0.80-0.63        Acceptable and good.

     0.63-0.40        Acceptable.  Some improvement is desirable.

     0.40-0.30        Borderline acceptability.

     0.30-0.01        Unacceptable.  This one characteristic could lead to
                      rejection of the process.

-------
     The scale of d so developed is a dimensionless scale to  which  any charac-
teristic may be transformed so that it may be interpreted in  terms  of its
desirability for the intended application.  In this evaluation, the most cost-
effective final process was sought that could be available in the shortest
reasonable time.

     Where a characteristic was assessed on some scale of measurement scale,
mathematical transformation from the measurement scale of the characteristic
to the scale of "d" was accomplished by the basic equation:

                   d. ..-

In this equation:  y, is a value of a treatment process characteristic i;
                   yih is the acceptable valuable of yi ; and
                       is the borderline value of y .
     Table 8 shows the acceptable and borderline values of y^  for  each
characteristic rated.

             TABLE 8.  ACCEPTABLE AND BORDERLINE VALUES FOR PROCESS
                                CHARACTERISTICS
Characteristic
Probability of cleaning to < 2 ppm
Probable cost of treatment, $/m3
T and E effort, $/1000
Test system availability, rating
Time to pxrovide commercial system, months
Acceptable
Value3
0.9
100
300
0.9
18
Borderline
Valueb
0.3
300
900
0.3
36
ad = 0.63 for these values.
bd = 0.37 for these values.

     The probability of cleaning to £ 2 ppm was rated 0.9 if such  performance
had been demonstrated with soils of any type, 0.8 if such performance was
projected from test data, and 0.3 if no data were available.

     The probably cost of treatment was considered acceptable if $100/m^,  and
borderline if $300/m3.

     T and E effort was considered acceptable up to $300,000.  Values above
$900,000 were considered borderline, but could be justified  if the process had
potential for lowering the cost of treatment, or rated extremely well on other
desired characteristics.
                                      63

-------
     Test  system availability was  rated 0.9 for an available company-provided
system with experienced operating  staff and resources to commercialize the
process; 0.7 for an available government-provided system; and 0.3 if a suit-  "
able test  system were not available.

     Time  to provide a commercial  system sized to effect cleanup of 152,000 m3
of sediment per year was considered acceptable at 18 months, but borderline if
36 months  were required.

     Using the values of the characteristics shown in Table 8, Equation 4 was
applied to calculate the individual ratings shown for each process in Table
9.  This table also shows the overall desirability rating of the process,
calculated using Equation 1.

     All the processes showed acceptable "D" values.  The B.E.S.T., UV/Ozone or
Hydrogen/Ultrasonics Technology, and Bio-Clean Naturally-Adapted Microbial processes
showed the highest values, and are recommended as the three processes for immediate
test and evaluation.

     While this project was under way, an alternative process selection method-
ology became available at EPA's Hazardous Waste Engineering Research Laboratory.
The methodology is available as a  computer program entitled "d-SSYS, A Computer
Model for the Evaluation of Competing Uncertainties," (Klee, 1987).  This method
was applied by RTI in addition to the Desirability Function approach.

     The D-SSYS calculates weights for each evaluation factor using values of
weight ratios assigned, by the user.  Weight ratios were assigned to emphasi-ze
the importance of the range of treatment costs and test-and-evaluation costs.
The range of ratings for probability of cleaning to <2 ppm PCBs is only 0.1
(Table 2)> indicating that all the processes might reasonably be expected to
meet the requirement.  The availability of a test system was not considered as
important as the total  test and evaluation cost.  The time required to make a
commercial  process available showed a range of only six months, and was judged
of lesser importance than the two major costs assessed.  All ratios among the
five factors that resulted from these assignments are shown below as a matrix.
For example, the ratio (test system availability)/(T and E cost) is shown as
the intersection of Row 4 and Column 3 as 0.2.
                                      64

-------
                      TABLE 9.  OVERALL DESIRABILITY OF IMMEDIATE T AND E OF THE EIGHT CANDIDATE PROCESSES

probability of clean-
ing to < 2 ppm <
\ rating
3robable cost of
treatment, $/m^
J rating
r and E effort '
$1000
i rating

\vailab1lity of a
•ystem for a test
future purchase by
govern, required
future purchase by
govern, not required
d rating ::
.ikely future avall-
ibility of the process
months ';
d rating j
't
)verall desirability, D
'
L
earliest future avail.
latest future avail.
average
KPEG,Galson

0.9
0.63

160-191
0.54a

216
0.66






0.9
0.63


19.5
0.62




0.615
0.615
0.615
Modar UV/Ozone-Hydrogen
Supercritical Ultrasonics CFS
Water Bio-Clean Technology Extraction

0.8
0.59

86-136
0.62

483
0.56






0.9
0.63


21.5
0.59




0.60
0.60
0.60
l,
r
0.9 0.8
0.63 0.59

156 - 90-120
0.57 0.63
-
166 '/ 151
0.68 ; 0.69
i
,.
t'
V
>'•

0.9 ?• 0.9
0.63 i 0.63
r
•'•
19 • 21-24
0.62 0.59-0.55
*
»•
-5
' <
0.617 0.625
0.617 0.616
0.617 : 0.621
S

0.8
0.59

153-264
0.50

123
0.69






0.9
0.63


25
0.54




0.59
0.59
0.59
Low-Energy In Situ
B.E.S.T. Extraction Vitrification

0.8
0.59

133
0.59

149
0.69






0.9
0.63


19
0.62




0.623
0.623
0.623

0.9
0.63

76-83
0.65

170-827b
0.67




0.8


0.59


25
0.54




0.614
0.614
0.614

0.9
0.63

293-332
.035

400
0.59




0.8
L
t.

0.59


19-24
0.62-0.55




0.54
0.52
0.53
'Average  cost  used  for rating.
'Cost;^)f'$170,000 1f developed  by sponsoring firm

-------
 Clean to  2 ppm
 Cost
 T  and E Cost
 Test system availability
 Early com. availability
Clean
to
2 ppm .,
1
5
5
1
1.25


.XostL
0.2
1
1
0.2
0.25

T & E
Cost
0.2
1
1
0.2
0.25
                      Test System
                      Availability

                            1
                            5
                            5
                            1
                            1.25
                                        Early
                                      Commercial
                                     Availability

                                          0.8
                                           4
                                           4
                                          0.8
                                           1
     From these  ratios and the  following tabular algorithm, the factor weights
 (W) were generated.
    Factors

Clean to 2 ppm
T and E Cost
Future commercial proc.
Test system availability
Cost
Ratios
0.2000
                   W
4,
1
  000
  25
0.2000
5,000
0.20
1.00
0.25
0.20
1
Weights. W

  0.0755
  0.3774
  0.0943
  0.0755
  0.3774
                                               2.65

The procedure for weight generation is as follows:.

     0  Construct an  intermediate weighting scale (the w-column) by the
        following procedure.  Opposite that last factor enter a "1".  The
        remaining numbers  in this column are formed by the product of its
        predecessor and Ratio value opposite it in a sort of zigzag route up
        the column.   For example, the first w-value, 0.20 is the product of the
        second w-value (1.00) and the first Ratio-value (0.2000).

     0  Total the w-values.  This total is 2.65.  Construct a column of
        standardized  weights by dividing each element of the w-column by this
        total to obtain the W-column.  The elements in the W-column will,
        perforce, total one.

     The program then scales the factor scores to obtain a linear utility func-
tion:

                          y1 = bg + bi (factor score)

                                  0 < y' _< 1.

     In applying the scaling procedure to the two factors "Probability of
Cleaning to < 2 ppm" and "Availability of a Test System," it is noted that
                                              (5)
                                     66

-------
 these are positive factors (the higher the factor  score, the better the
 process) and the 'y'  values are obtained  by:
                      .score-jj  - minimum
              ...............................................
              maximum score \j  - minimum score^j                        (6)

 The other three factors are noted  to  be negative  (the  higher the  factor score,
 the worse the process)  and the y1  values are obtained  by:
                        maximum  score-jj  -
                    — : ............... :•: ......... - ......
                    maximum  score-jj  -  minimum score^ j                   (7)


      The yij  values for the five  factors by which each of the eight processes
were  assessed are  given in  Table  4.


               TABLE 4.   SCALED  RATINGS  OF EIGHT TREATMENT PROCESSES
Clean
to
2
KPEG (Gal son)
MODAR
Bio-Clean
UV/OZ or H2/Ultrasonics
CFS Propane Extraction
B.E.S.T. s
Low Energy Extraction
In Situ Vitrification
ppm
1
0
0
0
0
0
1
1
Probable
Tr. Cost
0.59
0.87
0.67
0.89
0.45
0.77
1
0
T * E Test System
Cost Availability
0.75
0.043
0.89
0.93
1
0.93
0
0.26
1
1
1
1
1
1
0
0
Early
Commercial
Availability
0.92
0.58
1
0.42
0
1
0
0.58
     Depending on the users degree of risk that he is willing to accept, d-SSYS
fits a utility function to the y1 values via the following function:


                      utility = y'f .                                    (8)


The exponent f is evaluated by presenting the user with a structured lottery.
     d-SSYS requires s comparison between two simple lotteries for each factor
rated.
                                      67

-------
         Lottery 1 =
         Lottery 2 =
50% chance of most undesirable rating,

50% chance of-most desirable rating.



X value of the rating for certain.
     Using probable treatment cost as an example, RTI selected for Lottery  1

               50% chance of a treatment cost of $313/m3
               50% change of a treatment cost of $80/m3

and an X value equal to the mathematical expectation of Lottery 1  for
Lottery 2.

               (0.5 x $313) + (0.5 x 80) = $196.50/m3.


     The value of $196.50/m3 on the y' scale is
                          $313 - $196.5

                              $313 - $80
                        =  0.5
     The utility of Lottery 2 is easily determined, since it  is  equal to the
utility of Lottery 1:
          \


               (0.5)(utility of $313/m3) + (0.5)(utility of $80/m3)=
                     (0.5 x 0.0) + (0.5 x 1) = 0.5.
From Equation 8:
                     f = (In utility)/ln y'

                     f = (In 0.5)/ln 0.5 =
                                                 (9)
     Note that if Lottery 2 had been set at a lower cost for certain, f would
have been greater than 1 and function would have been a risk-taking  one, in
that one would be willing to pay more for Lottery 1 in the hope  of gaining a
treatment cost of $80/m3.
                                    68

-------
.< ,A   The remaining utilities for each factor are then calculated using Equa-
tion 8.
  ^

     Using the scores scaled by Equation 8, the program computes an overall
deterministic score for each treatment process as the sum of the scaled factor
 scores times the scaled factor weights.  Using the factor scores of Table 4
 (which equal the utility when f - 1) and the weights cited above, the follow-
 ing deterministic scores were obtained for the treatment processes (Table 5).


             TABLE 5.  DETERMINISTIC SCORES FOR TREATMENT PROCESSES
	Process	Score

Basic Extraction Sludge Treatment                                     0.8127
UV/Ozone or Hydrogen/Ultrasonics                                      0.8010
Bio-Clean Naturally-Adapted Microbe                                   0.7583
Potassium Polyethylene Glycolate, Galson                              0.7434
Critical Fluid Systems (CFS) Propane Extraction                       0.6214
MODAR Supercritical Water Oxidation                                   0.4738
Low Energy Extraction, New York University                            0.4529
In Situ Vitrification, Battelle                                       0.2299
     The highest scores were attained by the Basic Extraction Sludge Treatment,
UV/Ozone or Hydrogen/Ultrasonics Technology, and Bio-Clean Naturally-Adapted
Microbe processes, the same processes that ranked highest using the Desir-
ability Function ranking methodology.  These are recommended for immediate
test and evaluation.

     In the application of this ranking, probable treatment cost and test and
evaluation cost were assigned weights 4 to 5 times those of the other three
factors.  This increased emphasis on the costs involved did not change the top
three processes.  With different weights assigned, it would be possible to
obtain a different ranking of the processes.
                                     69

-------
                                    SECTION  6
                     APPLICATION OF  PROCESSES TO RCRA.WASTES    .


      The seven  processes  evaluated  in  the final assessment have potential applica-
 tion  to  treatment  of some RCRA  listed  hazardous wastes, as well as for removal  of
 PCBs  from sediments.

      The B.E.S.T.  process applies generally to the separation of oils, solids,
 and aqueous  components  of hazardous  wastes.. It has been applied to the clean-up
 of  a  CERCLA  site at  the General  Refinery near Savannah, Georgia.

      The Ultrasonics/Hydrogen-Ozone/UV Technology has been applied to the treat-
 ment  of  waste slurries  from the  pulp and paper industry, and to the cleaning of
 many  types of materials.   The UV energy must penetrate the treatment media
 sufficiently to reach the molecules  of the hazardous compound.  Filtration or
 cyclone-separation may  be required  following extraction in order to provide a
 suitable medium for  treatment.

      The Bio-Clean Naturally-Adapted Microbial  process can be applied to any
 contaminant  that is  extractable  and biodegradable.  It has been applied to the
 decontamination of pentachlorophenol-bearing soils.  The process requires selec-
 tion  and adaptation  of  a  suitable microorganism for the compounds being treated.
 Complete mineralization of the compound may be achieved, or if this is not
 possible, then non-hazardous residues  may be produced.  The process needs to be
 evaluated, using bench-scale tests,  for the different hazardous materials to be
 treated.  --     - —.«.„.-_-•'      -....--•

      The  Galson KPEG with  DMSO process, based on nucleophilic substitution, has
 been  applied to "dioxins".  It has  potential application to a number of listed
 halogenated  organic  hazardous wastes from both non-specific and specific sources,
 for example  waste  numbers  F020,  F021,  F024, F027, and K042.  The treatment leaves
 dehalogenated compounds as a residue,  which may or may not be toxic.  Further
 study is  required to determine the performance of the process for these additional
 applications.

     The CFS Propane Extraction  Process has been applied to the removal and concen-
tration  of oily contaminants.  Hazardous solvents and oxygenated compounds are
 extracted using pressurized carbon dioxide.

     Ttre Modar Supercritical Water-Oxidation process can apply generally to any
organic waste that can  be  oxidized at  supercritical conditions.  It has been
applied to decontaminate dioxin  tainted soil.  A test unit was operated at the
CECOS International  Niagara Falls site.

     The Low Energy  Acetone-Kerosene Extraction process is applicable to any
organic waste that can be  dissolved into a hydrophilic solvent.  It produces a
concentrated solution suitable for incineration or other treatment.
                                      70

-------