vvEPA
           United States
           Environmental Protection
           Agency
           Office of Research and
           Development
           Cincinnati, OH 45268
Office of Solid Waste and
Emergency Response
Washington, PA 20460
           Superfund
           EPA 540-2-90 010
September 1990
Second Forum on
Innovative Hazardous
Waste Treatment
Technologies:   Domestic
and
           Philadelphia, Pennsylvania

           May 15-17, 1990

           Technical Papers

-------
          United States         Office of Research and         Office of Solid Waste an
          Environmental Protection     Development - Risk Reduction      Emergency Response
          Agency            Engineering Laboratory         September 1990
                         EPA/540/2-90/010 September 1990
&EPA   Second Forum on Innovative
          Hazardous Waste Treatment
          Technologies: Domestic
          and  International
          Philadelphia, Pennsylvania
          May 15-17,1990
          Technical  Papers
                             U S. EfWhcmmenta^Prot^on Agency
                             Region 5, Library ^ ;] -•" ^ R .
                             77 West Jackson • ->. -.^ ^^ r.^,
                             Chicago, IL 60604-oo^U

-------
As a result of the high level of interest in innovative hazardous
waste control technologies,  U.S. EPA's Office of Solid Waste and
Emergency  Response   (OSWER)   and  Risk   Reduction  Engineering
Laboratory   (RREL)  jointly  conducted   this  conference.     The
conference consisted  of  presentations  of  technical papers  and
posters by international and domestic vendors of technologies for
the  treatment  of  waste,  sludge,  and  contaminated  soils  at
uncontrolled hazardous waste disposal sites.

The purpose  of the 2  1/2-day conference  was to  help  introduce
promising international technologies  through technical  paper and
poster displays, to showcase the  results of  the U.S. EPA Superfund
Innovative Technology Evaluation (SITE)  program technologies, and
to  present  case  studies  of  applied  technologies  from  EPA's
Superfund contractors.

This compendium  includes  the papers that were  made available by
authors  and  their  institutions.   The  papers  are  published as
received and the quality may vary.
          Although this document  has  been published by
          the U.S. Environmental  Protection Agency, it
          does not necessarily reflect the views of the
          Agency, and no official endorsement should be
          inferred.     Mention   of   trade   names   or
          commercial   products   does   not  constitute
          endorsement or recommendation for use.

-------
                                 CONTENTS


                                                                     Page

Remediation and Treatment of Superfund and RCRA Hazardous
Wastes by Freeze Crystallization
  James A. Heist, Freeze Technologies Corporation	   1-14

ARC/EPRI Process for Clean-up of Contaminated Soil
  T. Ignasiak, Alberta Research Council	   15-25

In-Situ Hot Air/Stream Extraction of Volatile Organic Compounds
  Phillip N. LaMori, Toxic Treatments  (USA) Inc	   26-43

Extraction of PCB from Oil with EXTRAKSOL™
  Diana Mourato, Environcorp Inc	   44-50

DuPont/Oberlin Microfiltration System  for Hazardous Wastewaters
  Ernest Mayer,  E.I. duPont de Nemours,  Inc	   51-68

Tyvek  for Microfiltration Media
  Hyun S. Lim, E.I. DuPont	   69-74

Cleanup of Contaminated  Soil by Ozone  Treatment
  Dr.  E. WeBling, Chemisches Laboratorium	  75-79

BioTrol® Soil Washing System
  Steven B. Valine, BioTrol, Inc	  80-90

Integrated Soil-Vapor/Groundwater Cleaning System at Selected
Sites  in West Germany
  Karlheinz Bohm, Ed. Zu'blin AG	  91-115

Ultraviolet Radiation/Oxidation  of  Organic Contaminants  in Ground,
Waste  and  Drinking  Waters
  Jerome T. Barich, Ultrox  International	  116-129

The Lurgi-Deconterra-Process Wet Mechanical  Site Remediation
  Eckart  F. Hilmer	  130-142

Developments  and Operating  Experience  in Thermal Soilcleaning
  H.J.  van Hasselt, NBM  Bodemsanering	  143-162

Revolving  Fluidized Bed  Technology  for the Treatment of  Hazardous
Materials
  Geoff W. Boraston,  Superburn  Systems Ltd	  163-181

Thermal Gas Phase Reduction of  Organic Hazardous Wastes  in Aqueous
Matrices
  Douglas  J.  Hallett, ELI Eco Technologies,  Inc	  182-191

-------
                            CONTENTS (Continued)

                                                                     Page

X*TRAX™ - Transportable Thermal  Separator for Organic Contaminated
Solids
  Carl Swanstrom, Chemical Waste Management, Inc	   192-204

Anaerobic Pyrolysis for Treatment of Organic Contaminants in Solids
Wastes and Sludges - The Aostra Taciuk Process System Thermal
Treatment
  Robert M. Ritcey, UMATAC Industrial Processes	   205-222

Algasorb®:  A New Technology for Removal and Recovery of Metal  Ions
from Groundwaters
  Dennis W. Darnall, Bio-Recovery Systems, Inc	   223-231

Biosorption - A Potential Mechanism for the Removal of Radio-
nuclides from Nuclear Effluent Streams
  Peter Barratt, Biotreatment Limited	   232-250

Biological Treatment of Wastewaters
  Thomas J. Chresand, BioTrol, Inc	   251-257

Biotechnical Soil  Purification of Soil  Polluted by Oil/Chemicals
  Susanne  Schi$tz  Hansen, A-S Bioteknisk Jordrens	  258-266

In Situ  Physical and Biological  Treatment of  Volatile Organic
Contamination:  A  Case Study Through Closure
  Richard  Brown, Groundwater Technology Canada, Inc	  267-306

Introduction to  the GEODUR  System
  Svend  Mortensen	  307-315

The  SITE Program:   Heavy  Metal  Fixation in  Soil
  Philip N.  Baldwin, Jr., Chemfix Environmental Services	  316-323

Evaluation of  Three Leading Technologies
  Armand A.  Balasco, Arthur D.  Little	  324-342

Soil  Vapor Extraction  and Treatment of VOCs at a Superfund  Site  in
Michigan
  Joseph P.  Danko, CH2M  Hill	'	  343-348

Sludge and Soil  Treatability  Studies at a  Large, Complex  Superfund
Site
  Susan  Roberts  Shultz,  Donohue & Associates, Inc	  349-381

Conceptual Cost  Evaluation  of Volatile Organic Compound Treatment
by  Advanced Oxidation
  Glenn  J. Mayer,  CH2M Hill,  Inc	  382-409

-------
                            CONTENTS (Continued)
Recovery of Metals from Water Using Ion Exchange:   A Case Study
  Thomas A. Hickey, B&V Waste Science and Technology Corp	  410-428

Critical Fluid Solvent Extraction
  Cynthia Kaleri, U.S. Environmental Protection Agency	  429-453

Bench-Scale Solvent Extraction
  Joseph A. Sandrin, CH2M Hill	  454-464

Field Demonstration of a Circulating Bed Combustor (CBC) Operated
by Ogden Environmental Services of San Diego, California
  Nicholas Pangaro, Alliance Technologies Corporation	  465-473

Low Temperature Thermal Treatment (LT3) of Soils Contaminated With
Aviation Fuel and Chlorinated Solvents
  Roger K. Nielson, Weston Services, Inc	  474-485

-------
Remediation
and
Treatment
of
Superfund and RCRA
Hazardous  Wastes

by
Freeze
Crystallization
James A. Heist, President                 	
Ken M. Hunt, Vice-President
Patrick J. Wrobel, Sr. Process Engineer            NOVEMBER 1989
Robert W. Connelly, PhD, Sr. Process Engineer                 •w%*
FREEZE TECHNOLOGIES CORPORATION
Raleigh, North Carolina

-------
INTRODUCTION

Freeze crystallization is a separations process used to remove pure components
from solutions by crystallizing the materials to be removed. The technology has
proven advantages that make it economically attractive as well as uniquely able
to decontaminate many types of aqueous hazardous wastes. This paper will
illustrate how the process can be used in site remediation activities, by-product
recovery activities, and in the handling of mixed industrial wastes.

Freeze Technologies Corporation (FTC) has built a 10 gallons per minute (GPM)
DirCon™ plant to demonstrate the freeze crystallization technology. The plant is
contained in two modules that can be transported and requires less than 1 week
to set up. The company's current demonstration project is scheduled for the
Stringfeltow site in Riverside, California. Stringfellow is  ranked high on the U.S.
EPA's National Priority List (NPL) of Superfund sites targeted for remediation.
The application is for a leachate from interception wells.

Freeze crystallization has several advantages over competing technologies for
remediation and waste recovery applications. First, it is an efficient volume
reduction process, producing a concentrate that has no additional chemicals
added to it. Volume reduction translates into reduced costs for wastes requiring
destruction by incineration or disposal in a landfill. When a large fraction of the
solvent (usually water) is removed from a waste, the remaining impurities often
begin to crystallize as well. These components are often sufficiently pure to have
by-product recovery value. Freeze crystallization has low processing costs
generally ranging from $.03/gal. to $.15/gal. for 40 and 5 GPM plants, respec-
tively.

-------
                                    FREEZE PROCESS DESCRIPTION

                                    The basic operation involved in freeze crystallization is the production of crystals
                                    by removing heat from a solution. Crystals produced in this manner have very
                                    high purities. Once small and uniform crystals have been produced, they must be
                                    washed to remove adhering brine. The brine is recycled to the crystallizer, so that
                                    as much solvent as desired can be recovered. The pure crystals are then melted
                                    in a heat-pump cycle, which further improves the energy efficiency of the proc-
                                    ess.

                                    When one or more of the solutes exceeds its solubility, additional crystal forms
                                    are produced, but they are formed separately from each other and from the
                                    solvent crystals. Consequently is it easy to separate these different crystals by
                                    gravity. This is because, in most waste applications, water is the solvent and with
                                    ice being less dense and the solutes more dense, the separation can be accom-
                                    plished by gravity.

                                    A freeze crystallization process is composed of the following components, as
                                    illustrated in the process flow diagram of Figure 1:
Figure 1:
Freeze Process
Flow Schematic
LfXfl	to  ™
  ^   VlTT—|—U^ REJECTION
  HEAT      f  1    CONDENSER
 REJECTION    I  T
                                                                          *• SLUDGE
                                                                            OUTLET
                                                           ZH1
                                                           *CANTER M STRIPPER
                                                                O—Q^-^
         CONCENTRATE
           OUTLET
                                         CRYSTALLIZER, where heat is removed to tower the temperature of the
                                         material to the freezing temperature of the solution (usually crystallizing the
                                         solvent first);

                                         EUTECTIC SEPARATOR to segregate the crystals of solvent and solute
                                         into different streams, so that each can be recoverd in pure forms;

-------
   • CRYSTAL SEPARATOR/WASHERS to remove the crystals from the
     mother liquor in which they are slurried, and to wash adhering brine to very
     low levels so that the recovered crystals have high purity;

   • HEAT-PUMP REFRIGERATION CYCLE to remove refrigerant vapor from
     the crystallizer, and compress it so that it will condense and give up its
     heat to melt the purified crystals;

   • HEAT EXCHANGERS to recover heat from the cold effluent streams,
     improving the efficiency of the process.

   • DECANTERS and STRIPPERS are required in some processes to remove
     volatile materials and/or refrigerant from the effluent streams before
     discharge;

   • UTILITIES, CONTROLS, ELECTRICAL SWITCH GEAR, PUMPS AND
     PIPING are required to implement the freeze process in a continuous,
     closed system.

The design, operating characteristics, capabilities and limitations are determined
largely by the type of crystallizer used, of which there are three basic types:

    1) INDIRECT CONTACT crystallizers use a scraped surface or similar heat
      exchanger that will crystallize by removing heat through the heat transfer
      surface.

    2) TRIPLE POINT crystallizers use the solvent as the refrigerant at  its triple
      point (where solid, liquid, and vapor phases are all in equilibrium). For
      instance, with aqueous systems, the triple point occurs at less than 3 mm
      Hg. absolute pressure at 30 degrees F, or below.

    3) DIRECT CONTACT SECONDARY REFRIGERANT crystallizers use an
      immiscible refrigerant that is injected directly into the process fluid,  and
      evaporates at several hundred to several thousand times the vapor
      pressure of the solvent.

The direct contact process offers a number of advantages in treating hazardous
wastes. This process is more efficient, requiring less capital to build and less
energy to operate. Also, the direct contact process achieves a greater reduction
of waste volume and is capable of producing higher quality effluents. These are
the reasons FTC  has chosen the direct contact process as the basis for its
technology.

-------
                                     10 GPM DlrCon™ PLANT

                                     The freeze crystallization process that has been accepted into the EPA's Super-
                                     fund Innovative Technologies Evaluation (SITE) Program is a direct contact,
                                     secondary refrigerant process. The mobile DirCon™ plant has been designed and
                                     built for this program, and is illustrated in the photo shown in Figure 2. The
                                     capacity of the plant is 10 GPM of ice produced from an aqueous waste stream
                                     with a freezing point of 20° F.

                                     The plant contains all of the components described earlier, including the crystal-
                                     lizer, eutectic separator, crystal separator/washer, heat exchangers, heat pump
                                     (open cycle screw compressor), decanters and strippers, and ancillary utility
                                     related systems.

                                     The plant is designed for transportability, using modular design concepts devel-
                                     oped by Applied Engineering Co., Orangeburg, SC, the acknowledged leader in
                                    this field. The plant is contained in two modules designed for transport on low-
                                    boy trailers. Each measures approximately 50' long x 13' wide x 11.5' high. Upon
                                     arrival at  a site, a crane is used to stack one on top of the other. The three
                                    electrical and thirty flanged-piping interconnections take about a day to complete.

                                    The plant is self-contained except for electrical supply. Instrument air, cooling
                                    water, electrical distribution, and electrical heating components are all contained
                                    within the plant. The plant is controlled by a distributed digital processor that is
                                    programmed to operate without attendance. On-line quality sensors are used to
                                    evaluate process efficiencies, discern operating problems, and perform corrective
                                    actions such as recycling effluent for re-processing if warranted.
Figure 2:
10 GPM DirCon™ PLANT

-------
DEMONSTRATION PROGRAM - STRINGFELLOW NPL

A demonstration of a commercial scale DirCon™ plant will be conducted at the
Stringfeltow NPL site at Riverside, California in December 1989. This demonstra-
tion is co-sponsored by the U.S. EPA and the state of California. The EPA
participation is through its Superfund Innovative Technology Evaluation (SITE)
program. The California participation is through its Waste Reduction Innovative
Technology Evaluation (WRITE) program sponsored by the Department of
Health Services, Alternative Technologies Office.

The goals of the SITE program are, among other things, to demonstrate applica-
bility of a technology on as general and as difficult a waste stream as can be
found. Procedures are designed to test for economic and technical performance.
Test results are analyzed to determine if the technology is applicable.

The FTC DirCon™ 10 GPM plant will be installed and ready to operate at the
Stringfeltow NPL site in Iate1989. The SITE tests and sampling program will
occur over a two to three week period and a public visitors day will be held while
the plant is at Stringfeltow. Further testing for reliability and longer term perform-
ance will continue for another several weeks before the plant is decontaminated
and removed from the site.

This demonstration program is designed to have minimal impact on the host site,
a condition that is paramount in EPA planning and with local communities.
Wastewater from the NPL site is piped from the leachate collection wells to the
on-site treatment plant. During the demonstration testing period, this flow will be
diverted to FTC's DirCon™ plant. The freeze plant will process this flow, produc-
ing separate streams of clean water and concentrate. These will be recombined
and pumped to the existing on-site treatment  plant for routine processing. The
residence time in the DirCon™ plant, including storage before and after the actual
freeze processing, will be 72 hours.

Since the freeze process operates in closed vessels with recycle of refrigerants
and wastes at pressures less than atmospheric, there are no emmisions from the
process. An option exists at Stringfeltow for treating the concentrated effluent for
metals recovery, for organic by-product extraction, or for solidifying the concen-
trate for subsequent landfilling. The by-product recovery is an attractive option
because it would eliminate a large volume that is now landfilled.

-------
                                   GENERIC APPLICATIONS FOR FREEZE CRYSTALLIZATION

                                   Remediation does offer some opportunities for by-product recovery, but usually
                                   these opportunities occur in on-going RCRA-waste generation facilities. There
                                   are certain generic conditions that favor freeze crystallization. When used under
                                   these generic conditions, the freeze crystallization process contributes signifi-
                                   cantly towards accomplishing a more complete and more economically efficient
                                   remediation through by-product recovery.

                                   Freeze crystallization works by making pure crystals from water, or other compo-
                                   nents, in a waste solution. The waste must first be in liquid form. In the case of
                                   aqueous based wastes, the crystal produced is ice, and all impurities are ex-
                                   cluded and remain in the concentrated liquid portion. The process is effective in
                                   removing water from these wastes and reducing the volume of hazardous
                                   contaminants.

                                   Since crystallization excludes all impurities from the ice, all impurites are equally
                                   removed. The freeze process is capable of treating wastes with heavy metals, all
                                   types of dissolved organics, and radioactive materials. And all of the impurities
                                   are reduced in concentration  in the effluent by a factor of about 10,000.

                                   Figure 3 shows conditions at which alternative treatment technologies are both
                                   technically feasible and cost effective. The chart demonstrates that freezing
                                   becomes more competitive as the waste becomes more concentrated and
                                   complex. For instance, wastes with heavy metals require concentrations of 1,000
                                   to 10,0000 mg/l to be economically recoverable with freezing. Aqueous streams
                                   require organic concentrations of 3 to 7 wt-% before it is economicai to treat them
                                   with freeze crystallization. When the waste contains both organics and heavy
                                   metals at between  .5 and 1.5 wt-% total contaminants, the freeze process
                                   becomes more economical than multi-process treatment trains.
Figure 3:
Technology Comparison
Chart
                           I
                          N
                          C
                          R
                          E
                          A
                          S
                          E
                          D

                          C
                          O
                          N
                          C
                          E
                          N
                          T
                          R
                          A
                          T
                           I
                          O
                          N
                      Type of Contaminant
    Volatile
   Organics
  Heavy
Organics
Salts
Heavy
Metals
   Stripping
                        Ion Exchange
       Carbon Adsorption
Biological & Chemical Oxidation
                        Electrodialysis
                                 Reverse Osmosis
                                             Evaporation
                      Freeze Crystallization

-------
CASE STUDY 1 - Stringfellow Leachates

Freeze Technologies performed treatability studies on Stringfellow leachates with
the feed composition shown in Table 1. These tests were performed on FTC's
batch process laboratory equipment. Test results confirm that freezing will
recover over 90% of the water, concentrating the impurities into a minimal
volume for disposal and/or recovery operations. It has been the experience of
FTC that a continuous process, which  is used in commercial scale plants,
produces even better effluent qualities than those achieved in batch processing.
In the treatability tests, crystal clear melt was produced with a reduction in
conductivity from 39 mS/cm to less than .02, approaching a ratio of 2000:1. The
effluent (melt) qualities expected from  a continuous processing plant at 90%
recovery of water from the leachate are also shown in Table 1.

Currently the leachate from Stringfellow is treated by a system consisting of
hydroxide precipitation of heavy metals and activated carbon adsorption of the
organics. The sludge is disposed of in a hazardous waste landfill, and the carbon
is returned for regeneration. Treatment costs for this  system are $.363 per gallon.
A detailed listing of these costs is outlined in Table 2. Throughput at Stringfellow
has been slowly declining, with current production from the interception wells at
about 225,000 gallons per month. The annual operating  costs to treat this volume
with the current system are $980,000,  excluding equipment amortization.

By incorporating the freeze crystallization process into the existing treatment
system, costs for treatment can be substantially reduced. The costs to operate a
10 GPM DirCon™ plant are $.09 per gallon. (Costs as a function of DirCon™
plant size are summarized in Table 3.) The annual operating costs to treat the
Stringfellow leachate would be only $108,000 for the freeze process and
$300,000 to incinerate the concentrate. Total costs, including equipment amorti-
zation at $.05 per gallon, should not exceed $600,000 per year.
Constituent
PH
Conductivity, mS/cm
Total Organic Carbon
p-CBSA
Volatile Organics
Na
Total heavy metals
Al
Cd
Cr(+6)
Cu
Fe
Mg
S04
Cl
N03
Feed (Mg/l)
3.5
39
1400
2600
11
930
4100
2100
2.98
129
11
417
1180
20000
380
80
Melt (Mg/l)
6.5
0.02
<1.0
<1.0
<0.1
<1.0
<5.0
<1.0
<.01
<.01
<.01
<1.0
<1.0
<5.0
<1.0
<1.0
                                                                            Table 1:
                                                                            Stringfellow Leachate
                                                                            Analyses

-------
Table 2:
Current Stringfellow
Treatment Costs*
'Costs shown do not include
 equipment amortization
Cost Element
I/GAL
                                     Chemicals
                                     Activated Carbon
                                     Supplies
                                     Electricity
                                     Labor
                                     Effluent Post-treatment
                                     Sludge Disposal

                                     Total Cost
                                                 0.015
                                                 0.027
                                                 0.006
                                                 0.005
                                                 0.065
                                                 0.098
                                                 0.147

                                                 0.363

-------
CASE STUDY 2 - Mixed Industrial Wastes

Industrial facilities often have wastes that are collected from a variety of places
and combined for treatment at a common facility. In metal fabrication shops with
electroplating, metal cleaning, pickling, stamping, machining, and other opera-
tions, the combined waste might resemble the analysis shown in Table 4. A
typical treatment train might include an oil-water separator, neutralization,
hydroxide precipitation, and discharge to a POTW for organics removal. How-
ever, municipalities are looking much harder at toxic organic discharges, and
current treatment often fails to meet the effluent standards.

Freeze Technologies has performed laboratory testing on a waste such as this,
collected after an API separator. Over 90% of the water was converted to ice.
Alkalinity and suit ate salts were precipitated with heavy metals, hardness cations,
and probably to some degree with sodium ions. After about 75% water removal,
a second organic phase began to form, and a significant portion of the influent
organics partitioned into this phase that was composed of about 75% organics
and 25% water. The water phase contained less than 10% organics, but had
most of the dissolved salts.

The organic phase that forms this way will have a high heating value so it can be
incinerated directly, or perhaps recovered for its fuel value. The tests in the batch
treatability lab showed that this organic layer had a specific gravity much less
than ice, and it collected in the top of the ice drain column. This will allow its
recovery from a freeze crystallization process.
Cost Component
Amortization, 5 year SLD
Labor
Electricity
Supplies, Chem., etc.
Maintenance
Total Costs
Costs for Varying Plant Sizes ($/gal)
5GPM
0.080
0.040
0.008
0.010
0.007
0.145
10GPM
0.050
0.020
0.008
0.008
0.005
0.090
40GPM
0.015
0.005
0.006
0.005
0.004
0.035
                                                                           Table 3:
                                                                           Freeze Crystallization
                                                                           Cost Summary
    Constituent
    Total Dissolved Solids
    Dissolved Organics
    PH
    Alkalinity
    Sulfates
    Chlorides
    Heavy Metals
Concentration (mg/l)
       12500
       8000
         10
       5000
       4700
       1500
       3000
Table 4:
Typical Integrated
Manufacturing Effluent
                                                      10

-------
TREATABIUTY TESTING

Given the complex waste matrices encountered in most Superfund or RCRA
remedial actions, treatability studies should always be conducted. Wastewater
analysis is not sufficient to define the optimal treatment train for cleaning up liquid
waste streams especially when dealing with new technologies.

From our experience with 8 years of development of the freeze crystallization
process, and another 10 years of development of environmental process technol-
ogy, there are a number of guidelines we would propose for treatability testing in
EPA Superfund and other industrial waste programs.

    1)  Treatability studies should always be done by the technology developer.

    2)  Treatability studies should be incorporated into the RI/FS activities, and
       replace the paper studies that are currently the basis for selecting be-
       tween alternative treatment technologies. The RI/FS process is a way for
       EPA to help the transfer of technology from the developers to EPA
       contractors.

    3)  The EPA and its contractors should not attempt to define the treatment
       technologies for individual  sites. Instead, the EPA should conduct the pro-
       curement activities in a way that elicits  complete strategies from technol-
       ogy  developers. Most remediation sites have the need for a variety of
       treatment technologies. The most efficient method for defining treatment
       trains at individual sites is to first define a clean-up standard and let the
       individual developers provide the  train of technologies to meet those
       standards.

The technology developer should have an institutionalized place in the remedia-
tion activities of Superfund and  private clean-ups. The best process technology is
resident in this community, and this resource should be used for its expertise in a
way that ensures competition in procurement and full assessment of alternatives.
                  11

-------
THE FTC TREATABILITY TESTING LAB

Freeze Technologies performs treatability studies in the apparatus shown in the
photo of Figure 4. A simplified flow diagram for the unit is shown in Figure 5. The
unit operates in a batch mode. Feed is drawn under vacuum into the crystallizer,
and recirculation  between there and the drain column is started with the slurry
pump. Refrigerant is pumped into the crystallizer, where it boils as it is heated by
contact with the waste. The refrigerant vapor passes through a refrigerated
condenser and is pumped back to the crystallizer as a liquid. When the drain
column is packed full of ice, the process is stopped. The concentrate is drained
from the ice and returns via gravity to the crystallizer for reuse. The ice is then
washed and melted. Since only 50% of the water in any batch can be converted
to ice, simulation of 90% recovery requires 4 stages of concentration. Twenty
gallons of sample is required to allow 12 of these staged tests.

These tests accomplish the following:

    •  interactions such as foaming or emulsification between refrigerant and
      waste are observed,

    •  phase equilibria are confirmed, showing the operating temperatures that
      will be required.

    •  the occurance of eutectic crystallization is observed.

    •  physical properties of the waste at the crystallizing conditions are deter-
      mined.

The impact of this information is that more detailed preliminary process designs
are possible, which in turn allows more accurate cost projections. Problems that
would be seen in the field often show up at this stage of testing. Design adapta-
tions made in the batch lab are the first stage  in modifying the full scale equip-
ment for use with new applications.
                                                      12

-------
Table 4:
Batch Treatability Test
Laboratory
Tables:
Batch Treatability Stand
Process Schematic
                                          to vacuum
                                            PRIMARY
                                          COMPRESSOR  CONDENSER
             to
           vacuum
                                                 REFRIGERANT
                                                 CONDENSER
                                                 AND RECEIVER
                                                                   REFHK3ERANT
                                                                     PUMP
drain
                                                                                  SLURRY
                                                                                   PUMP
                                                   13

-------
Freeze Technologies Corporation
2539-C Timbertake Road
Post Office Box 40968
Raleigh, NC 27629-0968

Office 919-850-0600
FAX 919-850-0602
               14

-------
                        ARC/EPRI Process for Clean-up of
                                Contaminated Soil

                      T. Ignasiak, K. Szymocha, B. Ignasiak
                            Alberta Research Council

                            C. Kulik and H. Lebowitz
                        Electric Power Research Institute
1.  Background

    Soils contaminated with oily waste presents a serious environmental problem.
The   management   of  this  depends  upon  many  factors  such  as  contaminant
characteristics,  topography,  climate,  ecology  and soil characteristics.  The
complex  and  diverse  nature  of  oily  wastes also causes severe technical and
economic limitations.

    The Alberta Research Council (ARC) and the Electric Power Research Institute
(EPRI) have responded to this problem by developing a novel process for clean-up
of  oily  waste  soils.    The  potential  of this process has been demonstrated
through  an  extensive  batch  experimental  program followed by verification in
6T/day  pilot  plant  tests.   A wide variety of oil contaminated soils has been
tested  in the program with particular attention directed towards remediation of
soils  from  manufactured  gas  plant  sites.    Except  for  coal  derived tars
(polycyclic  aromatics  hydrocarbons  -  PAH),  such  soils contain considerable
quantities  of  cokes,  chars  and coals as well as slags (molten mineral matter
with  inclusions of organic solvent soluble and/or insoluble hydrocarbons).

    The  work carried out at ARC led to the development of a detailed conceptual
design  for  a  transportable  200T/day demonstration plant for clean-up of oily
waste  soils.    The  unit   is  engineered to handle a wide range of oily waste.
Process economics look very encouraging.
                                         15

-------
2.  Process Description

    The  process  flow  diagram for the ARC/EPRI process for clean-up of tar/oil
contaminated soil is presented in Fig. 1.  A suspension of contaminated soil and
coal  in  water  is  subjected  to  tumbling  in a specially designed drum at an
elevated  temperature  at  optimized  solids  concentration.    The  mixture  is
subsequently  screened  into two fractions: -1mm and +lmm.  The -1mm fraction is
subjected to agloflotation which separates the coal floes (microagglomerates) in
the  form  of  "froth 1".  Depending on the characteristics of the treated soil,
the  "middlings  1"  are  either  reprocessed  or  combined with "froth 1".  The
tailings  from  agloflotation  are  combined  with  a  -1mm  reject derived from
selective  grinding  of  +lmm  fraction.   The combined material is subjected to
reprocessing  in  the  presence  of  small  quantities  of  coal  and a suitable
collector and again is subjected to agloflotation, thus giving rise to "froth 2"
which,  together with "froth 1", forms a  combustible product.  The tailings from
reprocessing  yield clean soil.  Depending on the characteristics of the treated
soil,  the  "middlings  2"  are  combined either with the combustible product or
clean soil.

3.  Development Status

    The  process  scheme  described  above has been developed based on extensive
batch  and continuous 6T/day pilot scale  tests carried out at the ARC.  The flow
diagram  of  the  Continuous  Soil  Clean-up  Pilot Plant as designed, built and
commissioned  at  the  ARC  Devon Facility two years ago  is presented in Fig. 2.
Based  on  results  of those tests, a conceptual design of a 200T/day commercial
demonstration plant was developed in 1989 which has been used for detail process
cost evaluation.

    The  200T/day  demo  plant   is  designed  to  be transportable.  The unit is
engineered  for handling the most difficult soils (for example, manufactured gas
plant sites).

    ARC  and EPRI  intend to form a Consortium of Companies  in the second half of
1990.  The objective will be to demonstrate the ARC/EPRI technology for clean-up
                                       16

-------
        Figure 1:
       Scheme of Processing  Tar/Oil
              Contaminated Soil
             Contaminated soil/coal/water
                      Tumbling
                      Screening
  + 1.0 mm
      Grinding
         I
         -1.0 mm reject
Clean coarse
  reject
                       Collector
                     Reprocessing
         -1.0mm
     Collector
    Agloflotation
Tailings

    Middlings 1   Froth 1

                      Middlings 2    Froth 2
            Clean soil
 ^Combustible product
                          17

-------
       Figure 2:
                                            (I
00
                                  High shear
                                    mixer     0
                                              Flotation cells
                                           oooooo
                 Low shear
                  mixers
                                  Coarse soil
                                   product
Fine soil
product
                       e
                                                                Dewatering screen
                                                                =h
Middlings   Coal product
        Flow Diagram of the 6 T/day Continuous Soil Clean-up Process
                            at ARC Devon Facility

-------
of  contaminated soil during the next 2.5-3 years at a cost of about $7 MM (US).
This  effort will involve detail engineering for the 200T/day unit, construction
of the unit, on-site commissioning and partial clean-up of two to three selected
sites with different soil characteristics.
4.  Media and Pollutants Treated

    The  experimental  work carried out at ARC, emphasized processing soils from
manufactured  gas  plant (MGP) sites.  The experience generated at ARC indicates
that  MGP soils are very difficult to clean.  In total, seventeen samples of MGP
soils  were tested.  These samples originated from various contaminated sites in
the eastern and the northern parts of the United States.

    The tar refuse (MGP) material is composed of tar, solids and water.  Tar has
been  defined  as  any  organic matter extractable with organic solvents such as
benzene,  toluene, xylene or dichloromethane.  The solids represent a mixture of
mineral  matter which contain non-extractable organic carbon (chars, coal, coke)
referred  to  as  combustibles and slag.   The solids are characterized by a wide
range  of  particle  size  distribution  (from  a  few  microns to a very coarse
material) and often contain various types and quantities of clays.

    The  composition  of  tar  refuse sample (concentration of water, solids and
tar)  was determined by Dean-Stark Soxhlet extraction with toluene which yielded
the  water  and  tar concentrations.  After drying and weighing, the solids were
ashed  to  give  the contents of combustibles.  The tar refuse samples that were
subjected  to  clean-up  tests  contained from 1-102 tar, 38-922 mineral matter,
2-222 combustibles and 4-542 water.

    The  tars  extracted  from  tar  refuse  samples  were subjected to detailed
analyses  which indicate that the tars are highly aromatic and are characterized
by  negligible  contents of heteroatoms.   About 60-80% of the tars were volatile
at  temperatures  up  to  500"C  and  the number average molecular weight of the
volatile  fraction was below 300.  The volatile portion of the tar was subjected
                                       19

-------
to  detailed  GC-MS analyses.   The major species identified were various isomers
of  mono-,  di-  and  thri-methylated  naphthalene,  anthracene, phenanthrene and
pyrene and other polycyclic aromatic hydrocarbons.

    In  addition  to  MGP  samples,  a  variety  of   oil  spill  samples  (soils
contaminated  with heavy oils, light oils, gasoline, diesel oil, as well as fuel
oils  and  intermediate  API gravity oils) were tested for cleanability.  All of
the  oil spill samples tested so far could be cleaned much more readily compared
to  the  MGP samples.  Among the oil spill samples,  most of the work was done on
heavy  oil  contaminated  soil  samples.    Compared  to the tar, the heavy oils
contained negligible amounts of low boiling hydrocarbons and only traces of PAH.

    The   ARC/EPRI   clean-up  process  shows  promise  for  cleaning  the  soil
contaminated   with   a   variety  of  organic  compounds  like  creosote  oils,
polychlorinated aromatic hydrocarbons, asphalt, etc.; however, the effectiveness
of this process for these particular compounds has not yet been determined.
Initial and Final Pollutant Concentrations

    Essentially  there  is no upper concentration limit on coal and/or petroleum
derived  pollutants  present  in soil in terms of the clean-up efficiency of the
ARC/EPRI process.  A soil sample containing 50% (by weight) of a contaminant can
be  cleaned as efficiently as a sample which contains only 0.5-5% of contaminant
concentration.    Since,  however,  the  clean-up requires the use of coal as  an
adsorbent   in   quantities   2-4  times  greater  as  compared  to  contaminant
concentration,   it  appears  to  be  economically  advantageous to treat samples
characterized by rather low or intermediate (up to 10-12% by weight) contaminant
concentrations.    Low  concentration of contaminants (0.5-5%) offers particular
opportunities  due to limited application of pyrolytic and combustion techniques
for treatment of such samples.

    Table   1  presents  the  results  of  clean-up  of  soils  contaminated with
tarry/oily  wastes.  The concentration of "contaminant" in the  "as received" soil
and   in  cleaned  soil  is based on the exhaustive extraction of  the feed and clean
product  with  an  organic solvent, as described  above.
                                        20

-------
                                     Table 1
                            CLEAN-UP OF OILY WASTES -
                                  TEST RESULTS
Sample
#
1
2
5
6
9
11
12
Type
Contaminant
MGP Tar #1
MGP Tar #2
MGP Tar #1-1
MGP Tar #2-2
Gasoline Spills
Heavy Oil Spills I
Heavy Oil Spills II
"Contam." cone., wU
Feedstock
8.6
1.2
5.6
10.6
3.1
43.0
0.2
Processed
Soil
0.07a
°-°°a h
0.05?'b
0.17b'c
0.00^
0.08C
0.00C
              a) grinding and reprocessing
              b) froth collector required
              c) results not optimized
    Samples  contaminated  with  coal  derived  tars  (MGP sites) were much more
difficult  to clean.  To bring the concentration of solvent extractable material
in  these  samples  below 0.1% (1000 ppm), optimization of the treatment for the
individual  samples  is  required.   Optimization, grinding and reprocessing was
required  for samples contaminated with petroleum derived products.  One of them
(sample  11) could be readily cleaned to 800 ppm of total organic material.  The
other  two  were  cleaned  below  the  detectibility  level  (100  ppm)  of  the
gravimetric analytical procedure used.

    The  concentration of PAH in cleaned soil samples (Table 1) originating from
MGP  sites  varied  from about a few ppm to about 200 ppm.  The concentration of
PAH's  in clean soil samples from oil spills (Table 1) was below the detectivity
level.

    None  of  the  clean  samples  (MGP  sites  or  oil  spills)  failed the EPA
Teachability  test (TCLP).   The TCLP tests as well as PAH's concentration tests
were carried out by Radian Corporation, Austin, Texas.  Some verification of the
analyses  were  performed  by  the  Laboratories  of  the University of Alberta,
Department of Chemistry, Edmonton, Canada.
                                       21

-------
Process Limitations and Advantages

    The  ARC/EPRI clean-up process is based on utilizing the adsorption capacity
of  coal  with  respect to the oily wastes.  For oils and tars, this capacity is
high  and the selectivity of the process is very good.  This may not be the case
for other organics particularly those characterized by hydrophilic character.

    Since  the  process  requires  that coal (or another solid carbonaceous with
adsorbent   characteristics   material)  be  available  on  a  processing  site,
processing soils contaminated with high concentration of oily contaminants makes
the  processing  economics less favorable.   However, sometimes the contaminated
soil  also  contains  carbonaceous  materials  (MGP sites and some petrochemical
plants)  in addition to tars/oils.  The presence of the carbonaceous material in
contaminated soil offers the opportunity of using this material as the adsorbent
for the ARC/EPRI process.

    Intermediate  scale  combustion  tests  (samples  size  ~5T)  carried out by
combustion  engineers  with the tar enriched coal has not revealed any problems;
most   importantly  no  increase   in  PAH's  presence  in  combustion  gases  was
identified.    Certain  contaminants   (for example polychlorinated hydrocarbons)
adsorbed  on  coal  would  have   to  be  combusted  at  a PCB disposal facility.
However,  it  would  make  more   sense  to  adsorb the polychlorinated aromatics
contaminating    IT  of  soil   (at  1%  concentration)  and  burn  the  50 kg  of
contaminated  coal  material   (40  kg coal  plus 10 kg  contaminants) in a suitable
facility  for  PCB  disposal   rather   than  using  IT  of contaminated soil plus
additional fuel  required for  combustion.

    An   interesting  approach   to  the  utilization   of  the  ARC/EPRI  clean-up
technology   is   for  the  Alberta  and  Saskatchewan  heavy  oil/bitumen industry.
Economics  of  these two provinces rely to  some  extent on their enormously rich,
but  low quality,  oil/bitumen  deposits.  Currently the recovery of oil/bitumen is
being   achieved   by  either   mining  and   hot  water  separation or in-situ steam
flooding,  using  natural   gas  for  steam/power  generation.   It is apparent that
these   methods   present  a   serious  environmental problem  reflected  in the vast
accumulation  of  tailing   ponds   and   oil  spills.   The adaptation  of ARC/EPRI
technology   for   clean-up   of  contaminated ponds and spills  using low cost,  low
                                        22

-------
quality  coal  and  the utilization of this oil laden coal, instead of expensive
natural   gas   for   steam   generation,  offers  a  very  interesting  option.
Furthermore,  the  combustible  product  (oil adsorbed on coal) generated in the
clean-up  process  can  be  thermally treated, releasing light oil, which can be
used  as  diluent  for pipelining heavy oil.  The thermally treated coal product
can then be combusted to generate the steam needed for oil/bitumen recovery.
Process Waste Streams

    The  ARC/EPRI  process  generates three streams:  clean coarse reject, clean
soil  and  combustible  product  (contaminants  adsorbed  on  coal).  Subject to
state/provincial  regulations and laws regarding the cleanliness of the product,
the clean coarse reject and clean soil can be used to generate various products,
for  example, asphalt road, aggregate material or land filled either directly or
after  some  additional  treatment  (ozone  treatment,  bio-treatment).  In most
cases, the oil/tar enriched coal can be combusted.

    When  the ARC/EPRI process  is required to work at elevated temperatures (due
to  high  viscosity  of the contaminant), the process leads to the production of
some  quantity of low boiling hydrocarbons (BTX) which will have to be condensed
and  the  residual  vapors  will have to be quantitatively adsorbed on activated
carbon.    The condensate could be readily combusted or sold to potential users.
The  hydrocarbon-saturated  activated  carbon  would  have  to be reactivated or
combusted.

    The  ARC/EPRI  clean-up  process  is  in most cases a net consumer of water;
about 15T water would be required for 200T/Day facility.  Therefore, no problems
with discharging bulk quantities of process water would be encountered.
Economics of ARC/EPRI Clean-up Process

    Based  on  the  experience  developed  in the course of batch and 6T/D pilot
plant  investigations  carried  out  during the last three years on the ARC/EPRI
clean-up  process,  EPRI  has  undertaken  conceptual  engineering  and economic
                                       23

-------
studies  for  the  demonstration  clean-up  plant.    This work was completed in
November   1989.     Follow-up  work  on  detail  engineering  studies  for  the
demonstration  clean-up  plant  have  commenced   in  January  1990 and should be
completed in July/August 1990.

    Since  the soil clean-up demonstration plant would have to be moved from one
contaminated  site  to  another,  the  plant  with  initial  design  capacity of
200T/day, would be built on trailer mounted modules.  The plant would be capable
of  treating  the  most  intractable  oil/tar contaminated materials (-4in) with
contents  of  organics ranging from 1-20%.  The target for soil clean-up was set
at  0.1%  by  weight,  or less of residual organic material extractable from the
processed  soil  with toluene as described above.  The engineering design of the
plant  is  based  on  the  principles  of batch clean-up presented in Fig. 1 and
verified in a 6T/day pilot plant presented in Fig. 2.

    The  process  of  adsorption  of  organics  on coal, combined with selective
grinding  (limited  to  coke,  char,  slag, etc.  but excluding rocks and mineral
matter),   high   shear  agitation  and  agloflotation  phenomena,  are  process
operations which are utilized in cleaning up the  MGP soils.

    In  the cost analysis, it was assumed that the cost of coal will be equal to
revenue  from  agglomerates   (coal + tar + other  carbonaceous material recovered
from  the  feed)  and  the  cost  of  oil  (heavy  oil)  used in the process was
conservatively estimated at US $20/Bbl.

    The  capital  cost  for   the  200T/day  demonstration  plant  including site
installation  cost  was  estimated  at  $2.7  MM  (US), while the operating cost
(including  electricity,  collector  and oil costs, labor, maintenance, supplies
and  services,  equipment  rental  and  contingency)  arrived  at  was US $35/T,
bringing  the soil clean-up cost both capital and operating to about US $40/T of
soil  processed.    This  cost  includes only the  cost of processing the soil and
does not include, for example,  soil excavation and site restoration.

    The  case   presented  above  is for the most  intractable oil/tar contaminated
material  requiring  significantly  more  extensive  processing compared to, for
instance, heavy oil spills.
                                        24

-------
                                                       sanpisaj ^o [esodsip
                                            pup uoi}djospp
            poog A"j3^          mrt QOT>     -puij6 aAi;o3[3S      Jq/} 01  IHd3/3yV
 ===============================================================================
 +021-06$   POog/jtPj          lurr QOI>     }ua6paa d3H3 UO  ^/^ 02'OT  d3^3 UO
    +091$        P°°9          ^  02>         A"ouanbajj MOT   jq/^ 02-91
      08$   poog A"JSA          uni  Q9>       uonB^°U q}°Jd      jq/^ L2
  OOI~08$        P°°9          i"*  89>           sidujis A"JSA      jq/^ 01
  OIt~0^$        P0°9          u"*  Ł9>       J3}BM aAisua^xg      jq/; 02       ZMH
      jad
                       A"oug6v
                      A"q  passasse  pus  adojn^ ui  pgdo[9A9Q

                      S3I9010NH331  9NIHSVM 1IOS  dO
                                     2
                                            SutqseM [ios  ueadojn^ aq; o; pajeduioo
     6uisiujojd A"jaA 55(001 A"6oiouqoa^  iyd3/3HV 3M^ ^5M^ sjeadde
         iyd3/3JJV 3M} M^^ aiqpjpdujoo  SL   [ios  jo  auuo^  jad
                 "                        passassp s;ueid  aq^ ^o
•oiqo 'i;puuPun 'A"ou96v uoip^^o^d  [P^uguuuojtAU] sn '^uaiudoiaAdQ pup
                 ^sanbaj aq;  uo  s^^asnqopsspw 'pjo^pag 'uon^jodjoo sai6oiouqoai
            8861 ui suoneaiiddv  pun;jadn$  jo;  passassp  pup  adojng
ui padoiaAap sai6o[ouqoa^ 6uiqspM  [tos   >o  A"jpujujns
              6uissaoojd 9q;  6uiJ3MOi  UL  ;[n$9J  'pggpui  'ueo
                 3SOL3 ;e a^qpiiPAp  ajp  ^JPUJ [POO  ps^puiujp^uoo  9L|;  pup
           3J3qM suoi^Pu°3 Japun  LIOS ps^puiuup^uoo  [io  A"Apaq
         ;no paijjpo s9}eun}S9  ;soo  pup  u6is3p  6uiJ3aui6ua

-------
     IN-SITU HOT AIR/STEAM  EXTRACTION
       OF VOLATILE ORGANIC  COMPOUNDS
             Phillip N. La Mori
Senior Vice President and Technical Director
         TOXIC TREATMENTS  (USA)  Inc.
         151 Union  Street,  Suite 150
      San Francisco, California  94111
                     26

-------
                             SUMMARY
This paper presents results on using a new technology for the in-
situ removal of volatile organic compounds (VOCs) from soil.  The
technology  injects hot  air/steam  into  soil  through two  5-ft-
diameter counter rotating blades that are attached to drill stems.
The soil  is cleaned  as the  drill stems  are advanced  into  and
retracted from  the ground.  The  evolved  gases are trapped in a
surface covering  called a  shroud and contaminants are  captured
aboveground via condensation and carbon absorption.  The recovered
contaminants can be reprocessed or destroyed.

The results of the initial  use  of  the commercial prototype show 85
to 99 percent removals of  chlorinated VOCs from clay  soils.  The
technology was able to achieve its goal of less than 100 ppm over
80 percent of the time in three blind tests.  The mass of material
recovered in the  condensate was conservative (89 percent)  of the
mass determined  removed from  the soil  by chemical analysis  and
physical measurements.   The technology also removed  significant
quantities   (50   percent)   of  semi-volatile   hydrocarbons  (as
quantified by EPA 8270).  This was unexpected. Subsequent analysis
has identified potential mechanisms for removal of semi-volatile
components.    Further  testing  is  planned  to  evaluate  these
mechanisms.

The process has not been found to cause undesirable environmental
emissions as a result of  its  operation.   Noise and air emissions
during operation are below the limits set by regional environmental
regulations  in  southern California.  Soil hydrocarbon emissions
during  treatment   are   not   increased  from  background  before
treatment.   Soil  adjacent  to the  treatment has not been found to
have increased VOCs after the process is operated.  Environmental
emissions caused  by operational problems  can  generally  be cured
without shutdown.
                                27

-------
                           INTRODUCTION


This paper describes in detail the  initial results on evaluating
and proving a new technology to remove organic compounds from soil
without excavation.  The process removes volatile organic compounds
(VOC) from soil by hydrocarbons.  The technology is described in
sufficient detail to understand  its  use to  remove chlorinated VOCs
plus some nonchlorinated hydrocarbons from a former tank storage
facility in California.  The unit is call  "The Detoxifier11 and is
operated by Toxic Treatments (USA) Inc.,  (TTUSA).

The  Detoxifier  is a mobile treatment unit used in  the in-situ
remediation of contaminated soils and waste deposits.  The soil is
treated in place  and is not excavated or removed to the surface.
The  Detoxifier  is capable  of  a  wide  range of  site remediation
methods, including;

          Steam-air stripping of volatile  contaminants

          Solidification/stabilization   and   construction   of
          containment  structures by addition  of  chemicals  or
          physical agents  (e.g., pozzolanic materials)

          Neutralization or pH adjustment  by addition of acids or
          bases and oxidizing or reducing  chemicals

          Addition of  nutrients, micro-organisms,  and  oxygen to
          promote in-situ biodegradation.
                    DESCRIPTION OF TECHNOLOGY


The Detoxifier consists of a process tower, a control unit, and a
chemical process treatment train (Figure 1).  These components are
configured to meet site-specific requirements.  The process tower
is essentially a drilling and remediation agent dispensing system,
capable of  penetrating the soil medium  to depths  of  30 feet, as
currently designed.  Remediation agents (in dry,  liquid, vapor, or
slurry form) are added to  and mixed with the soil  at various depths
by the  drill head assembly.  A  box-shaped shroud, under vacuum,
covers environmental  release.   The drill assembly is composed of
two  drill  blades,  each  5  feet  in  diameter,  with  injection
dispensers  (Figure 2).
                                28

-------
ro
vo
  p
  R
  0
  C
  E
  S
  T
  R
  A
  N
Compressor
    Boiler
Condensation
      &
 Separation
 Carbon Beds
  Distillation
      &
   Storage
                V
                A
                P
                0
                R
                    OFFGAS
L
I
Q
U
I
D
                                         CO
                         D
                         R
                          I
                         L
                         L

                         T
                         0
                         W
                         E
                         R
                                              SHROUD

-------
                        OPERATING METHOD


The remediation of a large area  is carried out by a block by block
treatment pattern.  The area to  be remediated is divided into rows
of  blocks,  with  the process  tower,  control  unit,  and  process
treatment train being moved from one  block  to  the next after the
remediation of a block is completed.  To assure complete coverage
of the area to be remediated, the drill assembly is positioned with
a  20  percent  overlap  of  the  previously  treated  block.    The
effective surface  area  of a treatment block is  approximately 30
foot2.   The volume of each block is  determined by  the  depth of
remediation (Figure 3).
                     INSTRUMENTATION AND CONTROL


On-line analytical instruments continuously monitor and record the
treatment conditions.  The monitoring data are used to control the
treatment process  and determine  the achievement of  remediation
objectives.

In  applications   involving  steam/air  stripping  of   volatile
contaminants, the off gas containing the contaminants is captured
in  the  shroud  and  sent  in a  closed  loop  (to  prevent  any
environmental release) to a trailer-mounted chemical process train
for removal  of water and chemical contaminants.   The clean air is
then recycled to the soil treatment zone.

The system  control  consists  of  process  monitoring and  control
instrumentation.   Flame  ionization detectors (FIDs) monitor the
concentration of total hydrocarbons at selected process locations,
including the  process off gas from  the shroud and the purified
return air.    The level of contaminants in the off gas determines
the process  control (Figure 4).  A gas chromatograph (GC) provides
a  periodic   check  of  the  identification and  concentration  of
specific compounds in the off gas stream and can be used to confirm
process control. The output of the FID,  temperature sensors, depth
gauge, and other instrumentation is stored in a computerized data
logging system, displayed  on  a terminal,  and recorded  on a strip
chart  recorder.    This  enables  operator  control  of  process
parameters to  achieve  the most effective treatment in  the least
amount of time.
                               30

-------
TYP NET TREATMENT AREA
> n i • bn>* Mr r»t i
TREATMENT AREA
PER TREATMENT BLOCK
(

1
V
\

y i
7
1
\
\
	 „-• • . */
1
\
\
y ^
7\'
/ \
SHROUD | 1
LIMITS v^^/
"7
1
\
\
'---"' ' • X
/
1
\
\
V
TYP
•\
4
t
\
1
^* S*
/
1
\
\
\
v -
7
1
\
\
,~

/
tm
f
<

/
1 1
/ \
/ \
(ELLY BAR SPACING
X
TYPICAL /
NET TREAT- /
.MENT AREA | '
\ /; \
V"" '"'^"'"
\ CUTTER \
| BIT AREA |

\
! ^
ROW

^A-V^ .-
y^ira-T
iy | IX?) i
\
\
I 'fv>' \ '
\ / ^ KELLY BAR
x^ ^yf ^ ''
-/~\-
\ / v

1 I i
/ \ .
/ \ y
\
X

^

\
\
ROV CENTERLINE
/ 1 '
/ ' TYPICA
\" 1 OF OFf
\ ADJACl
CENTERLINE J *°VS
/
ROV CENTERLINC ,


1
A.
\TION
•SET
•Mt
TYPICAL
SEPARATION
1' OF PARALLEL
, CONSECUTIVE
ROWS

N\ DETDXIFIER TRE/
/' BLOCK PATTERN

-------
      TOTAL LAB VOC vs. FID READING
           AREA A POST TREATMENT DATA
co
ro

c
c
Q.
a
o
o
>
m
<
_i
i
<
i—
o
i
i



zuu
1 Q n
I ou
1 c n
I OU
1 A n
I 4U
1 i n -
1 ZU
1 nn
I UU
on
ou
c n
ou
A n
4U
o n
ZU
n
C

	 -f. . . . . . . .

...4- 	 %^.
V.r»V
	 y« X%A. E?
C^&W "^^"^ 	
^5 ^>^+
	 ' '/*« ' '^*^ 	
1<**S^ +
_/T , .J. i
>^ ^ -h 	
lj^ + T ^ ... 	 	
/ ^1 +
.Xh ... ^. ...... ^_ . + . . +
_i_ j. -+?" y& ^
-4- i*rb_ ~r~ -4-~ ^^ >t: -4-
,• T ^1- _|_ JL^ T- ^ -r
^^^ i ~~pn n
1 500 1000 1500 2000 2500 3000 35(
FID READING (ppm)
+ POST 8 WEEK DATA * PRE 8 WEEK DATA

-------
                   IN-SITU REMOVAL OF VOLATILE
                        ORGANIC COMPOUNDS
The  Detoxifier presently  is remediating  a site  in  San  Pedro,
California, contaminated mostly with chlorinated hydrocarbons.  The
activities at the site were undertaken at the request of a client
and  are  under the  jurisdiction  of the California  Department of
Health Services (CDHS), Toxic Substances Control Division.

The  remedial  design  required that  a  Baseline Calibration  and
Testing Program be conducted to develop operational procedures and
provide the  data  necessary to evaluate the  effectiveness  of the
Detoxifier steam/air stripping process to remove VOCs from the soil
at the San Pedro site.

The Baseline Calibration and Testing Program  involved three phases
of  effort.    The  first   two phases  determined  the  operating
parameters for process control over a wide range of site conditions
and was also used as a run-in for the apparatus.  The third phase
was  a  blind  test  of the  technology  at  three  of  the  most
contaminated portions  of the  site.   Following  the  Baseline Test,
additional testing was performed as part of the  EPA SITE Program in
two  distinct phases.   These  test  efforts and  their  results are
described below.
BASELINE TESTING, 10 BLOCK

     Purposes:

          To  evaluate the  effectiveness  of  the  Detoxifier  in
          removing hydrocarbons from the soil at the San Pedro site

          To  measure  soil  vapor  emissions   before  and  during
          remediation
Treatment Protocol. The Baseline Test was conducted in three areas
of the site (A, B, and D) previously determined to be some of the
most highly contaminated.   The treatment depth was 5  feet,  even
through the contamination was known to extend below that.  To avoid
potential contamination  from  soil below, the  treatment protocol
called for treating to six feet to develop a one foot buffer zone
between the treatment depth and  the  remaining contaminated soil.
This was accomplished by treating the five to six foot zone first
and then  treating the 0' to  5'  soil according to the operating
parameters in Figure 4.
                                33

-------
Pre-Treatroent Sampling and Analysis.  Soil samples were collected
from  24  borings  located in  the 10  treatment blocks  and  in  4
uncontaminated  locations  west  of   the   treatment  areas  (for
background data).  All drilling, sampling,  and chemical analysis
were  conducted  according to legally  defensible  QA/QC procedures
approved by CDHS.   (Sampling  and Chemical analysis protocols are
available upon request.)  A total of  91 soil samples were taken,
including:

          68 samples  collected from  the  20  borings distributed
          among 10 treatment blocks located in Treatment Areas A,
          B, and D

          12 samples collected from the 4 borings located to the
          west of the current treatment areas (as a control)

          11 duplicate samples were selected from the 20 borings

The 68 samples collected from treatment blocks in Areas A, B, and
D, and the 12 samples collected from the 4 borings located to the
west of the current treatment areas were analyzed for VOCs by EPA
Method 8240. Eight of the duplicate samples were also analyzed for
VOCs by EPA Method 8240;  these duplicates  were labeled in a manner
such  that the  laboratory could not distinguish them from the
previously  analyzed  samples.   At  the request  of the  CDHS, the
remaining three duplicate samples were analyzed for semi-volatile
hydrocarbons (SVHs) by EPA Method 8270 and for priority pollutant
metals by EPA Method 6020.

Following  the   above-described  analyses,   additional  chemical
analyses  were  performed on  soil samples  that  had  been  kept in
refrigerated storage.  At the request  of the  CDHS, 65 soil samples
were analyzed for SVHs by EPA Method 8270.  These 65 samples were
the remainder from the 68 samples collected from the 10 treatment
blocks located in Treatment Areas A, B, and D.

The composite average pre-treatment concentrations of VOCs and SVHs
varied among the treatment areas.  Treatment blocks in Area A had
the greatest number of species or groups of VOCs and SVHs detected,
and  treatment  blocks  in  Area D  had the  least.   The  average
concentration  of VOCs  and SVHs  was  also  dissimilar.    Area  D
exhibited the highest concentration of VOCs and Area  A the lowest;
Area B had the highest concentration SVHs and Area D the lowest.

Analyses  for physical  properties   were   also  performed  on soil
samples collected from the 10 treated blocks sampled in Treatment
Areas A,  B,  and D.   A total of  68  soil samples  were analyzed to
evaluate the density and moisture content of the  sample.
                                34

-------
Treatment.  The remediation  protocol  was  designed to achieve the
reduction of  VOCs to less than 100 ppm.   Some of  the  key data
recorded during treatment included:

          Duration of treatment

          Steam injected (total pounds)

          Hot air injected (cubic feet per minute)

          Air temperature

          Concentration of volatile organics  in the off gas from
          the shroud to the process system

          Depth of treatment

TTUSA did not have knowledge  of  the pre-treatment soil analyses of
the 10 blocks until completion of baseline testing.  The decision
to  terminate treatment  of  each  test  block  was  made  when  the
concentration of  VOCs measured by  the FID decreased  to a value
previously determined as equating with less than 100 ppm VOCs in
the soil.   This  concentration  level  had been  determined during
baseline calibration and extended baseline calibration.   TTUSA did
not have the benefit of  any other information  to determine when to
cease treatment.
post-Treatment Sampling  and  Analysis.   The  same  procedures were
used for post-treatment sampling and analysis.  A total of 55 soil
samples tubes collected  from  14  borings  distributed among the 10
treatment  blocks located  in Treatment  Area  A,  B,  and D were
analyzed  for  VOCs  and  SVHs  by   EPA  Methods  8240  and  8270,
respectively.


Results.   Based on  a comparison of  the pre  and  post-treatment
chemical analyses, the major effects of treatment were:

          A very substantial reduction in the concentration of VOCs

          A   significant  and   unexpected   reduction   in  the
          concentration of SVHs
Reduction  of  VOCs.    A   very  substantial  reduction  in  the
concentration of VOCs was calculated from chemical analysis in all
test blocks  (see  Table  1  for summary data).   This  conclusion is
reinforced by  the fact that  the mass  of  VOCs collected  in the
chemical process  train  was roughly  equivalent  to  the calculated
mass removed from he  soil  (see  "Mass Balance"  below).  The major
mechanism that caused the reduction  appears to be volatilization.
                               35

-------
The  treatment  techniques  used  were  designed  to  reduce  the
concentration  of  VOCs  to less  than 100  ppro.    This  level  was
achieved in 8  to 10  treatment  blocks remediated (see Table 1 for
summary data).  As indicated in  Table 1,  all treatment blocks in
Areas A and B  were remediated  to substantially  below this level,
but  only  1  of  3  treatment  blocks  in   Area  D  was  similarly
remediated.   The average  pre  and  post-treatment  concentrations
(ppm) for VOCs are:


     Treatment Area      Pre-Treatment       Post-Treatment

          A                  1,114                 12
          B                  1,353                 30
          D                  3,954                139


The reduced effectiveness of the process  in  achieving a level of
100 ppm  in Area  D is believed to be caused by the  high initial
concentration  of  tetrachloroethylene  and the  presence  of  more
clayey soil than encountered in Areas A and B.


Reduction  of  SVHs.   An  unexpected and  major  reduction  in  the
concentration  of  SVHs  was  indicated  in all  treatment blocks
remediated (see  Table  1).   The  average  pre and  post-treatment
concentrations (ppm)  for SVHs are:


     Treatment Area      Pre-Treatment       Post-Treatment

          A                  3,775                627
          B                 12,116              1,766
          D                  1,014                 85


Mechanisms  that   may   account   for   this   reduction  include
volatilization,  steam  distillation,  hydrolysis,  and oxidation.
These reductions  have been confirmed  by  laboratory  testing.   A
small quantity of SVHs (2.1 Ib)  was collected  in  the process train;
however, the mass of SVHs collected does not account reasonably for
the mass  reduction  (1,018  Ib)  indicated by  post-treatment soil
sample analysis (see "Mass Comparison" below).


Soil  Gas  Emissions  During  and   After  Treatment.    One  concern
expressed by the CDHS was the potential  for increased  emissions of
VOCs  into the  ambient  air  during  and  after  treatment  by  the
Detoxifier.  This  rould occur, for example,  if the  heat used to
volatilize hydrocarbons caused increased soil  emissions adjacent to
the remediation or emissions occurred from just treated  soil after
removal of the shroud.  An evaluation of this was made during
                                36

-------
TABLE 1
SUMMARY OF
CHEMICAL ANALYSES
Treatment 	 Concentration (ppm) 	
Block Pre-Treatment Post-Treatment
Volatile Organic Compounds:
A-8-g 1,149 18
A-9-g 824 7
A-10-g 1,368 11
B-50-n 1 , 1 23 23
B-51-n 1,500 13
B-51-m 1,872 55
B-52-m 917 29
D-92-b 2,305 53
D-93-b 3,720 163
D-94-b 5,383 203
Semi-Volatile Hydrocarbons:
A-8-g 1 ,794 637
A-9-g 2,510 653
A-10-g 7,020 592
B-50-n 22,829 1 ,670
B-51-n 14,924 2,304
B-51-m 10,040 2,495
B-52-m 669 594
D-92-b 707 55
D-93-b 1 ,465 90
D-94-b 869 111



o/
^
Reduction

98
99
99
98
99
97
97
98
96
97

64
74
92
93
85
75
11
92
94
87
37

-------
baseline testing.   The test used was to be qualitative in nature as
a quantitative  test would be  expensive and time  consuming/  per
agreement  with  the  CDHS.    A  background  value  for soil  vapor
emissions was established  for  each area prior  to  treatment,  and
emissions were measured around the perimeter of the shroud during
baseline testing.   A comparison of these data and the data obtained
during baseline testing indicated that the  overall effect of the
hot air/steam stripping process was to increase slightly the rate
of soil gas emissions within and directly adjacent to the treated
area.  Emissions from the just treated uncovered  blocks were always
higher than background for several hours afterward.

A more recent test (under the approval  of the CDHS) using improved
process vacuum  and soil  covering of remediated blocks  showed no
increased  emissions to the  ambient  air.    In  fact, the levels
remained below background during and after treatment.

We believe this  means that soil gas emissions will not be caused by
the treatment process.


Mass Balance. The mass of VOC and SVH liquids collected by the hot
air/steam stripping process  was measured directly.  The  mass of
VOCs  and  SVHs  removed from  the soil  also was calculated using
measured values based on the chemical and physical analyses.  The
mass of VOCs collected is essentially equivalent to the calculated
mass of VOCs removed from the soil.   The following results present
the treatment blocks remediated, the mass of VOCs collected by the
process, the mass  of VOCs  removed  from the soil,  and the percent
VOCs  recovered.    (The  six  identified  blocks  were  analyzed
separately   because  the   CDHS  was   making  also  confirmatory
measurements.)


                                        Calculated VOCs
                       VOCs Collected    Removed from     % VOCs
  Treatment Blocks   by the Process. Ib  the Soil,.  Ib   Recovered

6(A8g, A9g, AlOg,          124.5             142.5         87.4
 B50n, B51n, B51m)
10 (all Blocks)            264.8             296.6         89.3
This comparison  indicates that the  Detoxifier treatment removes
VOCs from the soil and captures them in the process.  The closure
of  the  VOC  mass  balance is  within  what is  possible  for the
constraints of the experimental measurements.   The reasons for the
difference in the mass of VOCs collected and the mass removed from
the soil can be attributed to the following:
                                38

-------
          The size of the test block group compared with the size
          of the process system (i.e., mass possibly lost through
          process variables)

          Uncertainties in sampling

          Other unidentified causes, e.g., hang-up in the process
          train
The  mass  of  SVHs  in  the  collected  hydrocarbon  liquid  was  not
equivalent  to the mass  of SVHs  removed from  the soil.   After
remediation of  six of  SVHs  had  been collected,  whereas  it  was
calculated  that 1,018  Ib  of  SVHs   had  been  removed  from  the
treatment blocks.
                        SVHs in        SVHs Removed       % SVHs
 Treatment Blocks     Process,  Ib   From the Soil, Ib   Recovered

6(A8g, A9g, AlOg,         2.1            1,018            0.21
 B50n, B51n, B51m)


The  reason for  this  difference  in  mass  has  not been  formally
demonstrated.  However, examination of the compounds present, their
potential  chemical  fate in  the presence  of  hot air/steam,  and
chemical analysis suggest that  hydrolysis catalyzed by the clay in
the  soil plays an important role in their  disappearance.   It is
also believed that the hydrolysis reactions  produce compounds that
are  bound  to the  clay and  cannot be  identified by the analytical
techniques employed.
                         SITE DEMONSTRATION


The EPA SITE Program conducted a demonstration of the Detoxif ier at
the San Pedro  site  in September, 1989.  Twelve blocks in Area A
were treated to depths of 5 feet.  Pre and Post-Treatment samples
were taken  from zero  to  5  feet.   Composite samples were analyzed
from the 36  post-treatment cores and discrete samples were analyzed
from four of the  36  cores.   The report from this project is in
preparation.  Preliminary results are given in Tables  5 and 6.  The
results are in general similar to those in the above 10  Block Test.

Examination  of  the  results showed  several cores with very  high
post-treatment VOC,  e.g., in excess of 300 ppm.  A careful review
of all the chemical analyses suggested that this was a result of
                                39

-------
           Test Parameters
Treatment Depth
Buffer Zone
Cores per Block
Sample type
Deeper Samples
Pre/Post  Sampling
8240/8270
10
Block
6'
Yes
1 & 2
D
4
Both
Both
12
Block
5'
No
3
C/D
No
Both
Both
6
Block
12'
Yes
3
C/D
2
Post
8240
Sampling Depth • 5'

-------
  Post  Treatment 12  Block
          Total VOC's
Block    ppm          Block   ppm
25e     14.4         31e    60.7
26e     12.3         32e    64.0
27e     28.5         33e    104.4
28e     34.1         34e    196.3
29e     81.5         35e    59.9
30e    145.4         36e    55.9

-------
the treatment protocol and not the technology.  The 12 blocks for
the SITE Demonstration were treated  to  5  feet without building a
buffer from  5  feet  to 6 feet between the  clean and contaminated
soil, thus,  the process appeared  to  draw  up materials from below
during treatment.   This interesting result confirmed the need for
a buffer zone.

Shortly after the SITE Demonstration TTUSA  initiated treatment into
the saturated zone to 12 feet of  depth.    The conclusion is that
the process  operates with improved  efficiency in  the saturated
zone.   The  results  for  this  work were favorable and are being
prepared as  part of  another report.   Because of these results in
the saturated zone and a desire to  obtain additional  data with a
buffer  zone, the SITE  Demonstration sampled an additional  six
blocks in February,  1990.  Preliminary results are shown in Figure
7 for VOC.   They are in good  agreement  with the previous work in
that the VOCs are effectively removed.  Both sets of data will be
discussed in detail  in the SITE Demonstration  report scheduled for
summer 1990.
                             SUMMARY
This paper has presented the  first  detailed information on a new
technology for the  in-situ removal  of VOCs from  soil.   The data
indicate that VOCs  are removed substantially from  soil and that
they are captured above ground for disposal without adverse effects
to the ambient environment. No information  is given regarding the
economics  of  this  process,   although  it  is  believed  to  be
competitive with alternate remediation  schemes.   This technology
may be an attractive alternative to landfill  disposal as  landfills
become unavailable  and passive vapor extraction schemes are slow or
uncertain in  their  efficiency.  The  technology  has demonstrated
also that SVHs are removed  from soil by the action of steam and air
in the soil environment.  The  mechanisms for this  are not yet well
understood.   The removal of SVH may  add to the  utility of this
technology in the future.
                                42

-------
Post Treatment 6  Block Data
          Total VOC's
 Block   ppm       Block   ppm
 26n    16        29n    80
 27n    22        30n    119
 28n    36        31n    45

-------
                 EXTRACTION OF PCB FROM SOIL WITH EXTRAKSOL™

                            Diana Mourato and Jean Paquin
                                  ENVIRONCORP INC.
                                  Montreal, Canada
                                       H1K4E4
                                     ABSTRACT

Polychlorinated biphenyls have  been succesfully extracted from clay-bearing soil, sand and
Fuller's earth,  by the Extraksol™ process.  Extraksol™ process is a mobile decontamination
technology which treats soil and sludges by solvent extraction. Treatment with Extraksol™
involves material washing, drying and solvent regeneration. Contaminant removal is achieved
through desorption/dissolution mechanisms. The treated material is dried and acceptable  to be
reinstalled  in its original location.

The process  provides a  fast,  efficient  and  versatile  alternative  for  treatment of
PCB-contaminated soil  and sludge.  The contaminants extracted from the soil matrix are
transferred to the extraction  fluids. These are thereafter concentrated in  the residues of
distillation after solvent regeneration. Removal and concentration of the contaminants ensures
an important waste volume reduction.

Extraksol™ is  a flexible process designed to extract a variety of  organic contaminants from
unconsolidated solids. Polyaromatic hydrocarbons, oils and pentachlorophenols have also been
extracted from sludges, activated carbon, porous  stones and  gravel, with high  removal
efficiencies.
                                           44

-------
INTRODUCTION

Extraksol™  is a mobile decontamination  process which extracts organic contaminants  from
unconsolidated materials such as soil, sludge, gravel, etc. Polar and non-polar contaminants are
extracted  from the soils by  solvent washing through desorption  /  dissolution  mechanisms.
Treatment  with  Extraksol  involves  material  washing  and drying,  which  generatesa
decontaminated,  dry soil, ready  to be returned to its original location.

Extraksol™ has been developed and commercialized by the Sanivan Group to complement its
hazardous waste treatment technologies. Initially developed  to extract polychlorinated biphenyls
(PCBs) from contaminated soil,  Extraksol™ has since demonstrated to  be an efficient solution to
recycle  sand, gravel, mixed soil and  sludge contaminated with oils, greases, polyaromatic
hydrocarbons  (PAHs),  chlorinated organics such  as   pentachlorophenols (PCPs) and other
common organic pollutants.

Extraksol™ has been designed as a closed system (no gaseous or  liquid discharges)  which
considerably reduces  the volume  of contaminants.  The solvents used are non-chlorinated,
non-toxic  and non- persistent.  Solvent regeneration  is an  integral part of Extraksol™  and is
carried in parallel to  the extraction activities. Solvent regeneration allows to associate the
benefits of soil decontamination to the advantages of contaminant volume reduction.

In Eastern Canada, there is no alternative for PCB contaminated soil other than containment and
securisation. Incineration   of PCBs is not yet an accepted solution whereas biodegradation
techniques are not available.
PROCESS DESCRIPTION

The process consists of the 3  phases: soil washing, solvent regeneration and soil drying.

Soil Washing

After  introducing 8  totO  drums  of   soil (or other contaminated  solid material)  within the
extraction vessel, the extractor is purged and the washing cycle is  initiated.

During the  soil washing phase, clean solvent is continuously pumped and withdrawn into and
from the extractor,  creating a dynamic system. To  further  enhance  soil-solvent contact, the
extractor is slowly rotated on its  axis.
                                             45

-------
The  circulating fluid migrates through the soil and dissolves/desorbs the contaminants present
in the matrix  according  to the affinity of the solvent for the contaminant. The  contaminant is
transferred from the soil matrix to the solvent and is carried out of the extractor with the fluid.
The  used solvent is transferred to a storage tank.

The  extraction phase is continued until the soil is thought to be decontaminated.  Visual analysis
of the  solvent  sampled  at  the extractor's  outlet provides an  indication  of the state  of
decontamination. Circulation of  500 gallons of solvent  is equivalent  to one  extraction cycle.
Although the washing time varies with the type and  concentration of contaminant, it ranges from
1 hour to 2 hours for 4 tons of material.

After the soil washing cycles  are terminated, the  solvent is withdrawn from the extractor and
transferred to the contaminated solvent tank.

Solvent regeneration phase

Solvent regeneration is conducted in parallel to the soil washing  phase. Regeneration is carried
by  distillation and  takes  advantage of  the differences in evaporation  temperatures  of the
extraction fluids and the contaminants.  The contaminants are  concentrated and recovered  as
residues  of distillation.

The solvent regeneration phase is an integral part of the Extraksol™ process which considerable
reduces the cost of decontamination. Since the  solvent is regenerated to its original composition,
it can be reused without limitation and without the  need of additives.

Soil drying phase

After treatment with Extraksol™,  the  soil is dry and can  often be  relocated  in its original
location. Soil  drying  consists  of  removing  the  residual solvent  from the soil. This is
accomplished by circulating  hot  inert gas through  the soil within the extractor.  Gas circulation
is induced by a vacuum pump in a closed system operation.

The hot gas evaporates the solvents and carries the vapors out of  the  extractor  and  into a
 refrigerated  condenser  in  which the gaseous solvent becomes  liquid. The  liquid solvent is
 separated from the inert gas and conveyed into a storage tank.
                                              46

-------
The  solvent-free inert gas  is heated by passage through  a hot heat exchanger. The circuit is
completed when the hot gas is reinjected into the extractor. The  drying phase  extends from  1
hour to 1 1/2 hours depending on the type of matrix treated.
TYPICAL RESULTS - PCB EXTRACTION

Typical  PCB extraction results in soils range from 95% to 99.9%, depending on the initial PCB
concentration and the residual PCB allowed to be left in the soil.  The process has the capacity to
treat strongly contaminated material and declassify  these materials as non-hazardous to levels
of PCb  < 5 ppm. The residual PCB concentration left in the soil is a function of the number of
extraction cycles applied. In general, a 30% PCB removal is obtained in each extraction cycle.

Since Extraksol™ is a  batch process, it offers the  flexibilty to treat different types  of materials
to different levels of decontamination.

To date, the  Extraksol process has succesfully treated materials as various as Fuller's  earths,
activated carbons, refinery sludges and wood treatment sludges. Organic contaminants such as
oils and greases, polyaromalic hydrocarbons  (PAH), pentachlorophenols   (PCP) and phenols
have been extracted from a variety of solid matrices.
CONCLUSIONS

Enough work has been done with the  one ton per hour Extraksol™ unit to  confirm  that the
process can effectively extract oils, PCB, PAH, and PCP from soil and generate a "clean", dry
soil, which once mixed with top-soil for  revegetation can be  returned to its original location.

Decontamination of sand, mixed soils clayey soil, sludge, gravel, activated carbon  and stones
have demonstrated that the process is  very flexible and can be adapted  to solve a range of
environmental problems.

The flexibility and mobility of the 1 ton/hour  unit and its 3 day set-up period makes  is a perfect
process to be used on small projects with a  maximum of 300 tons of material to be treated. The
Sanivan Group  is planning to build a larger Extraksol unit which will have the capacity to treat
6 to 8 tons per hour. This mobile unit will be  designed on the same principles of operation as the
smaller unit. This larger unit will be operated by 2 operators as  the smaller unit.
                                            47

-------
TABLE 1, SUMMARY OF RESULTS OBTAINED WITH EXTRAKSOL'S PILOT SCALE UNIT
TYPE OF SOIL


Clay
Clay
Sand
Mixed Soil
Mixed Soil
Mixed Soil
Mixed Soil
TYPE OF FLUID


# 1
# 2
# 1
# 1
# 2
#3
# 4
REMOVAL OF OILS AND GREASES
Initial
Concent
(ppm)
1,970
8,040
470
13,890
14,400
34,300
21,540
Final
Concent
(ppm)
847
590
220
2,270
1,210
1,440
1,880
% Removal

(%)
56.9
92.7
53.2
87.7
92.0
96.0
91.0
REMOVAL OF PCB
Initial
Concent
(ppm)
7,925
2,055
600
3.6
5.3
5.2
4.8
Final
Concent
(ppm)
2,080
48.8
6.3
0.69
0.70
1.0
1.1
% Removal

( % )
73.8
97.6
98.9
89.0
87.0
81.0
77.0
         TABLE  2, SUMMARY  OF RESULTS OBTAINED WITH THE 1 TON/HOUR
                      EXTRAKSOL UNIT - PCB REMOVAL
TYPE OF
SOIL

clay-bearing
clay-bearing
clay-bearing
TYPE OF
FLUID

#2
# 1
# 2
Initial PCB
Concent.
(ppm)
150
163
54
Final PCB
Concent.
(ppm)
1 4
28
4.4
% Removal

(%)
91.0
82.0
92.0
                                       48

-------
TABLE 3, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR
           EXTRAKSOL UNIT - O&G REMOVAL
TYPE OF
SOIL
clay-bearing
clay-bearing
clay-bearing
refinery clayey soil
refinery oily sludge
refinery oily sludge
refinery oily sludge
refinery oily sludge
Fuller's earth
Fuller's earth
Fuller's earth
Fuller's earth
pulp & paper porous gravels
pulp & paper porous gravels
TYPE OF
FLUID
#2
# 1
#2
#2
#2
u r)
IT e.
a. o
TT Ł
" O
Tt Ł.
# 2
# 2
M O
it Ł.
M O
•tt f.
# 2
# 2
Initial O&G
Concent.
(ppm)
1,801
1,789
600
15,000
49,000
72,000
73,000
70,000
366,000
447,000
313,000
332,000
10,000
1 ,040
Final C&G
Concent.
(ppm)
1 82
166
80
800
4,200
2,000
4,800
340
4,200
5,500
3,700
4,000
3,690
207
% Removal
(%)
90
82
92
95
91
97
93
99
99
99
99
99
63
80
 TABLE 4, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR
             EXTRAKSOL UNIT -  PAH  REMOVAL
TYPE OF
SOIL

refinery clayey soil
refinery oily sludge
refinery oily sludge
refinery oily sludge
refinery oily sludge
TYPE OF
FLUID

# 2
2
2
2
2
Initial PAH
Concent.
(ppm)
332
81
240
150
1,739
Final PAH
Concent.
(ppm)
55
1 6
1 0
1 9
"30
% Removal

(%)
83
8 1
96
87
92
                               49

-------
TABLE 5, SUMMARY OF RESULTS OBTAINED WITH THE 1 TON/HOUR
               EXTRAKSOL UNIT - POP REMOVAL

TYPE OF
WASTE

porous gravel
porous gravel
porous stones
activated carbon

TYPE OF
SOLVENT

# 2
# 2
#2
#2

Initial PCP
Concent.
(ppm)
8.2
81.4
38.5
744

Final PCP
Concent.
(ppm)
<0.82
<0.21
19.5
83

% Removal

(%)
>90
>99.7
50
89
                            50

-------
     DU PONT/OBERLIN MICROFILTRATION SYSTEM FOR HAZARDOUS
                             WASTEWATERS
 Paper to  be presented  at U.S. EPA  Second Forum on  Innovative Hazardous
                      Waste Treatment  Technologies

                             May  15-17,  1990
                           Philadelphia,  PA USA

                              Dr. Ernest Mayer
                        E. I. du Pont de Nemours, Inc.
                    Engineering Department Louviers 1359
                               P. O. Box 6090
                             Newark, DE 19714
                              (302) 366-3652
                                  Abstract

      Du Pont has developed and recently commercialized a new filter media based
on Tyvek® flash spinning technology. This new media has an asymmetric pore
structure, a greater number of submicron pores, and a smaller average pore size (1-3).
As a consequence, it has superior filtration properties, longer life, and in many
instances can compete with  microporous membranes, PTFE laminates, and various
melt-blown media.  When coupled with an automatic pressure filter (APF), it provides
an automatic wastewater filtration process that is a dry-cake alternative to conventional
crossflow microfilters and ultrafilters. This Tyvek®/APF process has proved extremely
useful in filtering heavy metal and other hazardous wastewaters to meet strict EPA
NPDES discharge limits. Specific examples and actual case histories will be
highlighted to illustrate its benefits.  In addition, this Tyvek®/APF technology has
recently been selected for EPA's SITE-3 program which is also discussed.
                                      51

-------
TYVEK® T-980 MEDIA

      Du Pont has recently commercialized a new filter media based on Tyvek® flash
spinning technology.  It has an asymmetric pore structure, a greater number of
submicron pores, and a smaller average pore size.(1-3). As a consequence, it has
superior filtration properties and longer life, and in many instances it can compete with
microporous membranes, PTFE laminates, and various melt-blown media (3).  Its key
property is its tight pore structure at a very low cost compared  to competitive products.
Table I outlines the media cost per gallon of waste filtered and shows that Tyvek® T-
980 grade, which is the lowest basis weight manufactured (0.9 oz/yd2) and the only
grade evaluated here, is very cost effective. In most applications the T980 grade was
sufficient so the added cost for a higher basis weight grade was not warranted.  For
example, Tyvek® T-980 produced slightly poorer effluent quality than the 'standard'
0.45p. microporous  membrane at a fraction of the cost; equivalent effluent quality to the
PTFE laminate at a fraction of the cost; and much better effluent quality than typical 1-
and 5- micron melt-blowns at equal or significantly lower cost (depending on
application).  Tests with actual wastes showed the greatest cost benefit (Table I). A
similar cost benefit was obtained in operation with  a low-level radioactive plating waste
at the Savannah River Plant (4,5). This  installation realized almost a $200  M annual
savings when Tyvek® T-980 media was  used with a more efficient filter aid. An added
benefit of the Tyvek® is its superior strength compared to microporous membranes and
the PTFE laminate  media.  This strength permits its use in robust automatic pressure
filters (6). The high strength (7) coupled with the tight, ~1- micron nominal  pore
structure (3) led to Tyvek®'s use in the EPA Superfund SITE program (8).

OBERLIN AUTOMATIC PRESSURE FILTER (APF^

      Tyvek® T-980 media requires  a suitable filter housing.  The Oberlin Filter
Company's automatic pressure filter (APF) was chosen for its  simplicity and fully
automatic operation (9).  The Oberlin APF has other advantages, namely:

      •  Completely automatic, unattended  operation save for Tyvek® roll
         replacement and chemical  treatment makeup.
      •  Enclosed operation tor safety and handling of hazardous wastes.
      •  Fairly high operating pressure (up to 60 psig).
Mayerl                                52

-------
      •   Completely in-line treatment-chemical addition, ie, filter aids and polymer
         flocculants.

      •   Automated shutdown flushing capability.
      •   Cake washing capability to remove hazardous filtrate, if required.
      •   Automatic, positive dry cake discharge.

      •   Direct submicron filtration without the need for further downstream
         processing.
      •   Reliable, low-maintenance performance.
      •   Completely automatic safety interlocks and enclosures and/or purging, if
         required.
      •   Explosion-proof design, if required.
      •   Completely integrated pumping system(s), if required.
      •   Dirty-media takeup and doctoring, brushing, or washing, if required
         (automatically accomplished during takeup).
      •   Standard PLC control.
TYVEK®/OBERLIN APF COMBINATION

      Thus, the Tyvek®/Oberlin APF combination has some unique advantages,
especially its completely automatic submicron, low-cost filtration and its dry cake
discharge (8). This dry cake discharge feature is precisely why the resultant cakes
pass the modified EPA "Paint Filter Test" for land disposal (10) and in some instances
pass the new EPA TCLP (Toxic Characteristic Leaching Procedure) test for hazardous
components (11). Dry cake/submicron filtration in one operation is why the
Tyvek®/Oberlin APF combination was selected at Savannah River over conventional
crossflow microfilters and ultrafilters.  In plating-waste treatment the simple, one-step
Tyvek®/Oberlin APF combination replaced the conventional three-step
clarifier/overflow sand filter/underflow recessed filter press process.  However, the
purpose of this paper is to highlight case histories where the Tyvek®/Oberlin APF
filtration process has been successfully applied. The technology is most suitable for
hazardous wastewater where the solids loading is not too high. Examples include
                                      53
Mayerl

-------
contaminated groundwater, plating wastewaters, low-level radioactive wastes, plant
equipment/floor washings, cyanidic wastes, plant wastewaters that contain heavy
metals, and metal grinding wastes. (7)
SAVANNAH RIVER SIMULATED WASTE

      Table II summarizes a series of Oberlin bench-scale tests with simulated plating
wastewater (to duplicate existing plating line effluents without uranium). The Tyvek®
T-980 media matches all competitive media in rate and filtrate turbidity, but produces a
drier cake because of better air-blowing due to less media blinding. Tyvek® provides
excellent cake release and the lowest  (15 mg) "A Wt Gain" after the cake is removed.
Furthermore, the Tyvek® T-980 is much stronger than the competitive media so it
could be rerolled in the Oberlin APF without a carrying belt.  All the others required
careful handling. This feature simplified the operation and reduced maintenance.

ACTUAL SAVANNAH RIVER OPERATION

      Table III details actual Savannah River plant wastewater treatment specifically
aimed at aluminum and uranium removal (4,5). The data were  obtained from two
Tyvek® T-980/Oberlin APF units which had been operating for about three years.
Aluminum forming and metal finishing  operations generate a high content of solids,
aluminum, and turbidity.  These solids can be reduced below the National Pollution
Discharge Elimination System (NPDES) limits with the T-980/Oberlin combination at
very high 1200 gfd (gallons/sq ft/day) rates.  This rate is about seven  times higher than
the ultrafilter (UF)/reverse osmosis (RO) system (200 gfd) originally considered.  Pilot
testing also showed that the UF/RO system repeatedly fouled with this waste, requiring
aggressive cleaning agents which significantly added to the waste volume.
ELECTRONICS MANUFACTURING PLANT WASTEWATER

      This plant's effluent exceeded the local sewer authority's lead discharge limit.
As a consequence, the plant was  mandated to cease discharge and dispose of their
wastewater in an off-site hazardous waste landfill at a $0.55/gallon cost. The Tyvek®
T-980/Oberlin combination  was installed instead of a crossflow microfilter because its
dry cake feature  significantly reduces waste  volume.  The T-980/Oberlin units repaid
their cost in three months of operation based on disposal cost savings alone.
Mayerl                                54

-------
      Table IV details tests similar to those with the Savannah River plating waste.
Tyvek® T-980 again compares favorably to the competitive media in rate and filtrate
turbidity (except for the 0.45 - micron microporous membrane). However, Tyvek®
T-980 produces the driest cake (75% solids), gives excellent cake release, and has
the lowest media "A Wt Gain" (0 mg). Again, the Tyvek® T-980 media was the only
one that could withstand the rigors of Oberlin APF rerolling without a support belt.

      Table V details actual plant operation and compares the Tyvek® T-980 effluent
with the plant's discharge limits.  As shown, the T-980/Oberlin units reduced the
effluent Total Suspended Solids (TSS) and lead levels to well below the plant's
required discharge limits at very high 1500 gfd flux ratios.  In addition, the Oberlin APF
routinely achieved >60% solids dry cakes that could be disposed of in a RCRA-
approved landfill (at significant cost savings compared to the concentrate from the
crossflow microfilter).
ELECTRONICS PLANT CARTRIDGE REPLACEMENT

      This plant's effluent had to be polished by absolute 0.45-micron cartridge filters
to meet heavy-metal discharge limits.  Cartridge costs exceeded $1200 daily plus
significant labor charges. A Tyvek® T-980/Oberlin APF unit was installed at significant
cost savings compared to these cartridges or a crossflow microfilter which was also
considered.  Payback was on the order of three months; the Oberlin APF produced dry
cakes suitable for landfilling off-site, and significant labor savings resulted.

Table VI demonstrates that the Tyvek®/Oberlin system  easily met the TSS and lead
discharge limits at very high 1500 gfd flux. Cakes were also quite dry at -50% solids.

MUNITIONS PLANT WASTEWATER

      This plant was faced with severe RCRA restrictions on heavy metals discharge.
They hired an environmental engineering  firm to solve their waste problem, namely to
remove all metals (primarily lead, antimony, and barium) to well below discharge limits
and to produce dry cakes that could be hauled off-site to a RCRA-approved landfill.
This firm opted for the  conventional crossflow microfilter/press approach which was
expensive, required manual labor, and was prone to fouling. I proposed the less
costly, single-step, Tyvek® T-980/Oberlin approach which was installed about  a year
ago. Table VII shows that effluent TSS, lead, antimony, and barium levels are well
Mayerl                                55

-------
below the discharge limits. In addition, the system demonstrated very high capacity
(15,000 gallons per day, gpd), very high flux (3700 gfd), and fairly dry, 40%, solids
cakes. The unit operates  at 200° F which would have been troublesome for the
crossflow filter and explains the somewhat higher effluent lead level.
CLARIFIER UNDERFLOW HEAVY METALS REMOVAL

      This plant was faced with a land ban of their main clarifier underflow sludge
(-2% solids) because it did not pass the EPA "Paint Filter Test" (10). This clarifier
treated the entire plant effluent and removed  primarily lead, zinc, and copper.  To
satisfy the RCRA land ban restrictions, the plant hired an expensive ($400/day) mobile
dewaterer who used a manual, recessed-chamber filter press that produced sloppy
cakes. These cakes had to be shovelled into dumpsters for off-site disposal. The
Tyvek®/Oberlin combination was installed about six months ago.  Table VIII shows
excellent effluent quality at very high,  10,000  gpd, capacity and 900 gfd flux. The
cakes were sufficiently dry (50% solids vs. 40% requirement) to pass the "Paint Filter
Test" and were  acceptable to the hauler/landfill operator at a significant cost saving to
the plant. The Tyvek®/ Oberlin system operates automatically at significant labor
savings compared to the manual press.

GROUNDWATER BARIUM REMOVAL

      Contaminated leachate from a landfill required silt and barium removal to pre-
vent  fouling in downstream organics removal  equipment. A Tyvek®/Oberlin system
was chosen for this application because of its automatic operation and dry cake
feature. Table IX shows that TSS and barium levels are well below required limits at
an extraordinarily high 2700 gfd flux.  This was substantially higher than the flux
obtained from an ultrafilter also considered for this application.

BATTERY MANUFACTURING HEAVY METALS REMOVAL

      This plant consistently exceeded their  permitted discharge limits from their con-
ventional clarifier/underflow press system and considered an overflow polishing sand
filter.  Simultaneously, they heard of our Tyvek® T-980/Oberlin APF technology and
decided to evaluate it as a polishing filter. Testing showed that it could replace the
entire clarifier/press installation.  Table X shows excellent metals removal, TSS
reduction, and excellent effluent turbidity (0.5 NTU).  Flux is quite high at 2400 gfd
Mayer 1                                 56

-------
and capacity from the single Tyvek®/Oberlin system exceeds 60,000 gpd. These
benefits are in addition to dry cakes that could be easily disposed of in a RCRA-
approved landfill.

SUMMARY

      These cases and actual operating applications demonstrate the utility of the
Tyvek® T-980/Oberlin APF technology to remove heavy metals at very high flux rates
and to simultaneously produce dry cakes that pass the EPA "Paint Filter Test." This
technology is quite competitive when compared to microfilter cartridges, crossflow
microfilters, and ultrafilters.  In some instances, the Tyvek®/Oberlin system can replace
the conventional three-stage metal treatment process of clarifier, underflow filter press,
and overflow polishing sand filter.  Thus, the waste engineer/consultant now has a
simpler, one-step process for treating hazardous metal-bearing wastewaters.
Mayerl                                 57

-------
                                 REFERENCES

 1.    H. S. Lim and E. Mayer, Paper presented at 1987 Membrane Technology
       Conference, Boston, MA, October 22, 1987.

 2.    H. S. Lim and E. Mayer, Paper presented at the International Technical
       Conference on Filtration and Separation, Ocean City, MD, March 22, 1988.

 3.    H. S. Lim and E. Mayer, Fluid/Particle Separation Journal. 2(1), 17-21, (March,
       1989).

 4.    H. L Martin, Paper presented at 10th Annual AESF/EPA Conference on
       Environmental Control for the Metal Finishing Industry, Orlando, FL,
       January 23-25,  1989.

 5.    H. L. Martin, P. K. Gurney, and L. P. Fernandez, Paper presented at 8th Annual
       AESF/EPA Conference on Pollution Control for the Metal Finishing Industry,
       San Diego, CA, February 11, 1987.

 6.    E. Mayer, Filtration News.  24-27 (May/June, 1988): "New Trends in SLS
       Dewatering Equipment."

 7.    Du Pont Tyvek® Bulletin E-24534, 1988: "Tyvek® Engineered Specifically for
       Filtration."

 8.    E. Mayer, Request for Proposal, SITE-3 Solicitation, Demonstration of
       Alternative and/or Innovative Technologies, "Groundwater Remediation Via
       Low-Cost Microfiltration for Removal of Heavy Metals and Suspended Solids,"
       February, 1988.

 9.    Oberlin Filter Co.'s Bulletin, February, 1988: "Oberlin Pressure Filter."

10.    N. J. Sell; Pollution Eng.. 44-49 (August, 1988).

11.    Federal Register, Vol. 51, No. 114 (June 13, 1986).
 Mayerl
                                       58

-------
                                                  TABLE  I
                                  MEDIA COST PER GALLON WASTE FILTERED
                                                     ActualCosts/Gallon  Filtrat fcents/aair
on
10
Media,
Microporous Membrane
Tyvek® T-980
PTFE Laminate
Melt-blown PP
Melt-blown PP
Nom.
Rating
(um)
0.45
1
0.8
1
~5
Overall
Filtrate
Quality
Excellent
V. Good
V. Good
Good
Poor
60 ppm
ACFTD**
6.7
0.5
8.2
0.7
0.3
600 ppm
ACFTD
29
1.3
41
2.4
1.5
Low-Level
Radioactive
Waste
69
1.7
63
6.5
4.0
Lead-
Bearing
Waste
14
0.3
12
1.3
0.8
    *Based on actual media costs, flux rates, and measured life cycles.


    **AC Fine Test Dust (ACFTD) challenge tests.

-------
                                                  TABLE II
en
o
                                  SAVANNAH RIVER SIMULATED WASTE *

                                    TESTING IN QBERLIN PRESSURE FILTER



Type
Microporous Membrane
Tyvek® T-980
PTFE Laminate
Melt-blown PP
Melt-blown PP
Media**
Nom.
Rating
W
0.45
1
0.8
1
~5

AWt.
Gain
(Mg.)
50
15
60
560
445


Cake
Release
Good
Excellent
Fair
Fair
Good

Cake
o/
/o
Solids
30
34
28
32
29

Filtrate
Turbidity
rNTLM
0.46
0.29
0.51
0.49
5.0

Avg.
Rate
(gfd)
1050
1070
1120
1040
1150
    * Simulated waste with 600 ppm metal hydroxide solids (except uranium) plus 3:1 ratio of Celite 577 filter aid.


    ** A Wt. Gain is media weight gain after cake removed; and cake release judged visually after bending media

       to simulate discharge around roll.

-------
                            TABLE III
              ACTUAL SAVANNAH RIVER PLANT OPERATION
             WITH TYVEK® T-980/OBERLIN PRESSURE FILTER
Propertv
Turbidity (NTU)
TSS (ppm)
Aluminum (ppm)
Lead (ppm)
Zinc (ppm)
Copper (ppm)
Uranium (ppm)
Capacity (gpd)
Flux (gfd)
Raw
Waste
110
687
127
1.6
0.5
2.0
2.3
35,000
—
T-980/
Oberlin
APF
0.32
1.4
0.95
0.2
<0.1
<0.1
0.01
60,000
1,200
NPDES
Limits *
—
31
3.2
0.43
0.32
0.21
0.5
—
._
*Actual State discharge permit values.
                                61

-------
                                                   TABLE IV
CTv
ro
                                     ELECTRONICS WASTEWATER TESTING *

                                        IN OBERLIN  PRESSURE FILTER



Type
Microporous Membrane
Tyvek® T-980
PTFE Laminate
Melt-blown PP
Melt-blown PP
Media**
Nom.
Rating
W
0.45
1
0.8
1
-5

AWt.
Gain
(Ma.)
20
0
20
50
135


Cake
Release
V. Good
Excellent
Excellent
Fair
Good

Cake
%
Solids
63
75
72
64
63

Filtrate
Turbidity
(NTU)
0.47
2.0
1.4
2.6
2.6

Avg.
Rate
(gW)
1730
2160
2050
1900
1900
      *Actual electronics manufacturing wastewater that contains about 6500 ppm heavy metals (Pb, Ba, etc) plus
      1950 ppm Superaid filter aid (0.3 ratio) and 50 ppm high-charge cationic polymer flocculant (Praestol K-122L).


      **A Wt. Gain is media weight gain after cake removed; and cake release judged visually after bending media
        to simulate discharge around roll.

-------
                            TABLE V
      ACTUAL ELECTRONICS MANUFACTURING PLANT WASTEWATER*
          WITH TYVEK® T-980/OBERLIN PRESSURE FILTER
Property
Turbidity (NTU)
TSS (ppm)
Lead (ppm)
Barium (ppm)
Capacity (gpd)
Flux (gfd)
Cake % Solids
Raw
Waste
>1000
6500
500
-100
2500
—
...
T-980/
Oberlin
x APF
3.4
6.0
<0.25
<1
7500
1500
>60
Discharge
Limits**
N.R.
26
0.7
N.R.
—
—
—
* Includes floor washings also

** N.R. = not regulated.
                               63

-------
                            TABLE VI
         ELECTRONICS PLANT CARTRIDGE FILTER REPLACEMENT*
          WITH TYVEK® T-980/OBERLIN PRESSURE  FILTER

Property
Turbidity (NTU)
TSS (ppm)
Lead (ppm)
Capacity (gpd)
Flux (gfd)
Cake % Solids

Raw
Waste
2400
5500
2300
2000
—
...
T-980/
Oberlin
APF
0.43
1.0
0.18
11,000
1500
50

Discharge
Limits**
N.R.
20
0.7
—
—
...
* Includes floor washings also

** N.R. = not regulated.
                                64

-------
                            TABLE VII
           MUNITIONS MANUFACTURING PLANT WASTEWATER*
             WITH TYVEK® T-980/OBERLIN PRESSURE FILTER
Property
Turbidity (NTU)
TSS (ppm)
Lead (ppm)
Antimony (ppm)
Barium (ppm)
Capacity (gpd)
Flux (gfd)
Cake % Solids
Raw
Waste
>1000
1100
540
4.4
50
3000
—
...
T-980/
Oberlin
APF
0.8
0.5
1.8
0.4
0.6
15,000
3700
40
Discharge
Limits**
—
30
5.0
1.0
50
—
—
•__
* Includes floor washings also

** N.R. = not regulated.
                                65

-------
                              TABLE VIII
             CLARIFIER UNDERFLOW HEAVY METALS REMOVAL*
               WITH TYVEK® T-980/OBERLIN PRESSURE FILTER
Property
Turbidity (NTU)
TSS (ppm)
Lead (ppm)
Zinc (ppm)
Copper (ppm)
Capacity (gpd)
Flux (gfd)
Cake % Solids
Raw
Waste
>1000
17,000
40
410
1050
6800
—
...
T-980/
Oberlin
APF
1.0
2.5
<0.01
0.2
1.2
10,000
900
50
Discharge
Limits**
N.R.
20
5.0
5.0
5.0
—
—
40
* Existing plant clarifier underflow sludge was previously hauled off-site to hazardous
 landfill.
** N.R. = not regulated.
                                   66

-------
                            TABLE IX
           CONTAMINATED GROUNDWATER BARIUM REMOVAL*
            WITH TYVEK® T-980/OBERLIN PRESSURE FILTER
Property
Turbidity (NTU)
TSS (ppm)
Barium (ppm)
Flux (gfd)
Raw
Waste
>1000
1200
500
—
T-980/
Oberlin
APF
9.0
6.6
6
2700
Discharge
Limits**
N.R.
20
50
—
* From hazardous landfill leachate.

** N.R. = not regulated.
                                67

-------
                            TABLE X
            DIRECT FILTRATION FOR HEAVY METALS REMOVAL
               FROM A BATTERY MANUFACTURING PLANT*
             WITH TYVEK® T-98Q/QBERLIN PRESSURE FILTER
Prooertv
Turbidity (NTU)
TSS (ppm)
Nickel (ppm)
Cadmium (ppm)
Zinc (ppm)
Cobalt (ppm)
Capacity (gpd)
Flux (gfd)
Raw
Waste
175
1600
30-50
15-40
0.2-1.0
0.4-1.5
40,000
—
T-980/
Oberlin
APF
0.5
<1
<0.10
<0.05
<0.10
<0.05
60,000
2400
Discharge
Limits**
N.R.
20
2.27
0.4
1.68
0.22
—
...
* Replaced clarifier and underflow press.

** N.R. = not regulated.
                                68

-------
     FLUID/PARTICLE
SEPARATION  JOURNAL
        TYVEK FOR MICROFILTRATION MEDIA

           Hyun S. Lim, E.I.DuPont, Fibers Department
             P.O. Box 27001, Richmond VA 23261

         Ernest Mayer, E.I. DuPont, Engineering Department
             P.O. Box 6090, Newark, DE 19714
     THE AMERICAN FILTRATION SOCIETY
         VOLUME 2, NUMBER 1, MARCH 1989


                    69

-------
 TYVEK  FOR MICROFILTRATION MEDIA
 Hyun S.  Lim, E.I. DuPont, Fibers Department
 P.O. Box 27001, Richmond, VA 23261

 Ernest Mayer, E.I. DuPont, Engineering Department
 P.O. Box 6090, Newark, DE 19714

   Tyvek™ spunbonded olefin is flash spun from high density
 polyethelene and is used in the medical packaging and
 protective apparel markets for its microbial barrier and air
 permeability properties. However, use of Tyvek as a filtration
 media has been limited because of poor permeability and low
 dirt holding capacity. Recently, a new T-980 series Tyvek nas
 been  developed specifically for filtration applications by
 creating greater number of submicron pores and a smaller
 average pore size.
   Permeability and filtration efficiency and capacity of Tyvek
 measured by AC Fine Test Dust challenge tests are discussed
 and its filtration performance parameters are compared with
 melt blown media and microporous membranes. The results of
 actual treatment of waste waters containing metal oxide,
 uranium paniculate, and lead are presented to illustrate the
 benefit of Tyvek for meeting the  strict  Environmental
 Protection Agency discharge limits.
 Fluid/Particle Sep. J,  Vol.2, 17-21  (1989).

 FILTER MEDIA  CHOICES AND
 MANUFACTURING

   A  proper selection of media is the most important factor
 that determines the filler performance. The criteria for selecting
 the optimum media are listed below:

   1.  How much fluid can flow through it (permeability)
   2.  How small a particle can it retain (nitration efficiency)
   3.  How long can it last (filtration capacity or life)
   4.  How much does it cost (nitration cost)

 An extremely  high  standard of separation of submicron
 particles requires  use  of microporous membranes. Most
 membranes can only be used with adequate substrate support
 and must be protected against rough handling. The major
 microporous membranes are manufactured by a wet-cast
 process or biaxial  stretching  which are relatively slow,
 inefficient and generally expensive (costs on the order of $7-
 40/sq yd).  A new  technology  utilizing ultraviolet light  for
 crosslmking of monomer and oligmers at a high  operating
 speed has been developed to lower the media cost However, in
 many  nitration applications requiring  fine separation of
 submicron  particles,  the use of microporous membranes is
 limited due to the high costs.
   An alternative  lower  cost choice is available  with
 nonwoven  media.  Three  basic  technologies  exist in  the
 nonwoven industry to produce filter media: melt spinning.
 melt blowing, and flash spinning.
   Melt  spinning  involves extruding  filaments through
 spinneret orifices followed by quenching the filaments with
cross-flow air and drawing through an aspirator jet by cocurrem
air streams. The descending  filaments are electrically charged to
 separate them and  lay them down in a completely  random
configuration onto  a moving bell, under which is a suction
box. The fibers are relatively coarse as fineness is limited by
the capillary size in the spinneret.
   Fibers of very fine diameter with better uniformity at lower
basis weight than the melt spun fibers can be produced by melt
blowing processes. Melt blowing technology was pioneered by
the Naval Research Laboratory to develop micro-denier fiber to
collect radioactive panicles in the upper atmosphere. This
technology was refined and licensed for commercial use by
Exxon Company.  The process consists of extruding molten
polymer through spinneret orifices and fibrillating the extrudate
by high temperature and high  velocity air streams. Fine
diameter, discontinuous fibers produced on the order of a few
inches in length are  laid onto a moving belt equipped with a
suction box. The melt blown fibers have lower strength than
melt spun fiber and  are often used in combination with other
strong fiber webs. In spite of small diameter fiber, the surface
area  per  unit  weight  is  much lower  than flash spun
plexifilamentary fibrils.
   The high surface area of flash spun fiber results from
internal voids in the fibrils created by rupturing solvent vapor
globules within the polymer solution during the flash spinning
process. The flash spinning process involves decompression of
pure  solvent droplets and highly saturated  polymer/solvent
mixtures through a spin  orifice. The fibrils become highly
oriented which accounts for their high strength.  The fibrils
formed are  of a  micro-denier and  are interconnected in a
continuous network described as a form of plexifilamentary
fiber structure. The  fibril webs are collected on a moving belt
and are convened to a finished form using heat and pressure to
promote self-bonding. The immediate consequence of this fine
plexifilamentary  fibril morphology includes high filtration
efficiency and good sheet tensile and tear strength.
    Tyvek spunbonded olefin is flash spun from high-density
polyethylene and is used in medical packaging and apparel
markets  for its  microbial barrier and air  permeability
properties. However, use of Tyvek  as a filtration media  has
been limited because of poor permeability and low din holding
capacity. Recently a new Tyvek product T-980 has been
developed  specifically for filtration applications. A new
bonding process  allows us to engineer structures to achieve
 higher permeability by creating a greater number of submicron
 pores and a smaller  average pore size. An asymmetric structure
 has been achieved which increases the din holding capacity and
 filler life.
    The selection  of filter media based on cost/performance is
 generally an empirical process that requires thorough testing
 and  evaluation. Rigorous prediction is difficult since many
 variables interact such as fluid properties; contaminant type,
 size and distribution; flow rate; and the  objectives of the
 filtration task. Thus, single-valued characterizations such as
 nominal rating, absolute rating, and contaminant capacity are
 not that useful in media selection (Ostreicher. 1986, refers to
 cartridges). However, certain standards have been more or less
 accepted by industry such that any new media must be
                                                        70

-------
  evaluated by these so called "standard" tests. These can be
  categorized into three areas: permeability, filtration efficiency,
  and capacity (or life).
     The purpose of this paper is to categorize a new filter media
  developed by DuPont (Tyvek™) which we believe is unique
  and exhibits interesting  properties when  evaluated by the
  "standard" industry accepted tests. This new filter media of
  Tyvek will also be compared to a typical  microporous
  membrane,  a   5-micron   "nominal"   melt-blown  PP
  (Polypropylene) media, a 1-micron "nominal" melt-blown PP
  media, and a PTFE (porytetrafluorethylene) membrane laminate
  used extensively in industrial filters.

  PERMEABILITY

    The major deficiency with respect to filtration of the T-10
  Tyvek that is used for envelopes, housewrap and disk sleeves is
  insufficient air and water permeabilities which result in a tight
  structure from micro-denier fibers spun by the flash spinning
  technology. A new T-980 series Tyvek is developed by a new
  bonding  technology which increases the permeability by
  creating greater numbers of submicron size pores and a smaller
  average pore size. A scanning electron microscopy (SEM) of
  T-980 Tyvek shows more  open depth,  finer fibers and an
  asymmetric structure. Consequently, the new T-980 series
  Tyvek exhibits much higher air and water flow rate than the
  typical T-10 products (1042B); and, in fact,  are comparable to
  microporous membranes and PTFE membrane laminates and
 begin to  approach a typical  melt-blown PP media,  which is
 known to have excellent permeability (Fig.l).
      10»c-
   r  10'-
       10*
            Figure 1: Air Flow Rale Comparison

FILTRATION  EFFICIENCY

   Table  I lists typical media properties. The new Tyvek
bubble point (4 psa-which is a general indication of the largest
pore size) is substantially higher than the T-10 Tyvek grades
and  begins to match the "0.8 micron" PTFE  membrane
laminate.  But, it is  still much lower than the 0.45-micron
nominal rated membranes. Table II details some supplementary
latex bead challenge tests at 0.624 and  1.1-micron sizes and
one notes that:
•    0.624-micron and OJ-micron retention efficiencies range
    between 93.0 and 99.6% and 92% and 99.4%, respectively
   depending on basis weight, which are quite good
   considering their 1-micron "nominal" 98% rating (for T-
   980 at 1.1 micron and even higher efficiency for the higher
   basis weight styles).
   Retention efficiencies for the 1.1-micron beads are even
   better (>99.0% for all grades), which indicates that the 1-
   micron  nominal rating is conservative and that single-
   valued ratings are misleading.

                      MEDIA PROPERTIES



Media
Meh-Blown PP
Melt-Blown
Tyvek 1073B
Tyvek 1042B
New Tyvek T-980
New Tyvek T-980
(Hyekophilic)
PTFE Laminate
PTFE Laminate
Microporotu
Membrne
MilliporeHA

Norn.
Micron
Size(fi)
-5
1
1
2
1
1.

0.8
0.5
0.45

0.45
Prizier
Air
Plow
•CTM/ft1)
18
5
0.25
0.35
1.0
1.0

1.0
0.5
0.25



Bubble
PlinHjO
(PSI)
0.4
0.7
2.0
1.7
4
4

5
12
23

33
Water
Permeability
(GPM/h1
@ I PSI)
30
14
0.10
0.12
1.0
2.0

1.5
0.8
0.8

1.0
                                                                   Style No.
                        Table I.

               LATEX AND OOP CHALLENGE TESTS

                          New Tyvek

                  T-980    T-984    T-988   T-989
                                                                                0.9
Baiu Weight
 (otfytP)
%  Retention  93.0
%  Retention  99.0
%  Retention  92.0
                          1.26

                          99.7
                          99. S
                          94.0
                                                                                               1.6
                                                                                              996
                                                                                                      1.9
                                                                                                      99.6  0.624
                                                                                                     >999  1 1»
                                                                                              99.4    994  0.3"
                                                                      I em/i«c  nelyilvrcne Ittex  tpbcra
                                                                    •10 Ofeua OOP Mraol
                        Table n

   Thus, these outside challenge tests  confirm that new
Tyvek™ medias are quite efficient at low micron ratings, are
much tighter than melt-blown PP medias. compare favorably
to PTFE membrane laminates at similar permeabilities, and
begin  to  approach typical microporous  membranes  in
performance. Hence, the next logical step in our program was
to conduct  full scale ACFTD challenge tests (as well  as
comparable 47mm disc tests) to compare the various medias on
an actual commercial filler. The commercial filter chosen was
the Oberlin Pressure Filter (OPF)  which is unique, is fully
automatic,  can use both disposable media and  recleanabte
ckxhs, and can be completely contained to prevent exposure to
toxic materials (Fig. 2).  DuPont  has used many of  these
Oberlin units in a variety of applications, even in radioactive
service. This Oberlin unit is also ideally suited to use Tyvek
disposable  media; and as  a consequence, can achieve tight,
"nominal"  1-micron filtration automatically at very low cost
(more on this later).
                                                        71

-------
    Wasttwiitr
      FMd
              Figure 2: Oberlin Pressure Filter

 OBERLIN  (AND  47MM DISC)  ACFTD  TESTING

    All ACFTD testing was conducted at constant flux (0.75
 but primarily at 1.0 gpm/sq ft) by either peristaltic pump
 control to a Pall filterability kit or by air pressure control to a
 Wilden M-2 double-diaphragm pump which feeds the OPF.
 Filtered (0.2 micron Gelman). deionized water was used  as
 makeup and two ACFTD levels (60 and 600 ppm) were used
 to represent typical wastewater treatment applications. Effluent
 turbidities  were measured with  a  Hach  Model 2100A
 turbidimeter on samples collected throughout the filtration life
 cycle. Composite filtrate samples were collected in particle-free
 bottles for subsequent panicle counting by both a Coulter
 Counter and a Panicle Measuring System (PMS) laser counter.
 Some samples were analyzed for total suspended solids (TSS)
 by  nitration through blanked Nuclepore 0.4  and  0.2-micron
 membranes.
   Fig. 3 shows that the  Tyvek (0.9 oz/sq yd.  Style T-980)
 has effluent turbidities  that are  between the PTFE membrane
 material and a typical 0.45-micron microporous membrane; and
 much better than melt-blown PP medias.
   Fig. 4  plots the Coulter  Beta ratio against the ACFTD
 micron rating (only  for the 47mm  disc tests since  the OPF
 tests were slightly corraminated by stray panicles even though
 unit was thoroughly cleaned prior to use). Nevertheless. Fig. 4
    100 e-
3



I
. 0.41 urn MenporM*
                         TVM (ffltaittl)
indicates that the new Tyvek again has higher Beta ratios (and
hence, higher efficiencies) than both the melt-blown PP and
the PTFE membrane media and approaches the 0.45-micron
membrane.  Thus.  ACFTD  full-scale  tests  essentially
corroborate the outside tests in that Tyvek  is much more
efficient than both melt-blown PP medias. compares favorably
with the PTFE membrane laminate, and begins to approach a
typical 0.45-micron microporous membrane.
                                                               100 c-
                                                                10
                                                                                          Ł	0.45 urn Mkraperaut
                 10
                             too
                        CaUMr B«U RiUo
                                                   10000
       Figure 3: ACFTD Challenge Tests (60 PPM)
        Figure 4: ACFTD Challenge Tests (60 PPM)
FILTRTATION CAPACITY (OR  LIFE)

   The third property of filter media that most end-users are
interested in is din capacity (or life). We conducted such an
evaluation by completing the ACFTD challenge tests (at both
60 and 600 ppm levels) to 30 psi differential  AP (a nominal
value that is sufficient for most applications). These tests were
conducted on both the Pall 47mm disc filterability kit and the
full-scale Oberlin 2.4 sq ft HA filter. Table III summarizes the
ACFTD measured life of each of the medias evaluated here and
basically shows that:

      The new Tyvek  has about  24% the life of the 0.45-
      micron microporous membrane at 60 ppm ACFTD; and
      about 37% at 600 ppm ACFTD. This  shorter life is
      consistent with Tyvek's lower permeability and porosity
      (ie, less depth penetration).

•     HydrophiUc Tyvek has about 70% longer life than the
      untreated Tyvek because of better wetting.

•     Hydrophilic Tyvek compares favorably in life to the
      PTFE membrane laminate, particularly since its cost is
      only a fraction of the PTFE membrane laminate.

•     The two melt-Wowns do not have substantially longer
      life dun the microporous membrane; and their filtrate
      quality it at best marginal.

Thus, Tyvek (both untreated and hydrophilic grades) has a
much shorter life than a 0.45-micron microporous membrane,
but its low cost helps » offset this penalty. Furthermore, cake
filtration is dominant  in many cases such  that media  life
                                                        72

-------
  becomes less important (ie, compare 60 to 600
  and see applications testing below).
ppm results.
      MEDIA FILTRATION CAPACITY BASED ON ACFTD CHALLENGES
60 PPM Tine To
30 PSI (Mini)

Norn.


Riling
Media
Microporoui
membrane
Ty»ek T-910
(Hydrophilic)
PTFE Membrane
Laminate
Melc- Blown PP
Melt-Blown PP
GO
0.45

1

o.s

1
_ r at
47MM
92

23

72

It
140
HA
67

15.5

52.3


-

Avg.
Floor
1.0

0.24

0.7S

0.95
1.50
600 PPM Time To
30 PSI (Mini)


47MM HA
12.0 17.5

6.5

10.0 12.0

13.5 20.0
1.45 21.3

Av|.
Factor
1.0

0.37

0.69

1.14
1 22
                         Table m
 FILTRATION  COSTS

    Any filtration process chosen must meet one's objectives
 as well as be the most economical. Generally speaking, most
 processes are selected based on initial capital investment and
 operating costs are usually ignored. Table IV highlights some
 typical operating costs based on the work here. Basically, new
 Tyvek  is  far more  economical than any of the  other
 competitive medias for four typical applications (based on
 cents per gallon filtered). In fact, its cost is a fraction of the
 microporous membrane cost at only a slight deterioration in
 filtrate quality. And,  it is also cost-competitive to low-cost
 melt-blown PP medias, but produces far  better filtrate quality.
 Tyvek  also exhibits superior strength and cake release, which
 are significant advantages in Oberiin filter operation. Thus,
 Tyvek  provides economical,  low-micron  filtration comparable
 in many instances to microporous membranes  and PTFE
 membrane laminates.

           MEDIA COST PEK GALLON WASTE FILTERED

Media
Microporoui
Tyvek T-9W
Tyvek T-9tO
(Hydropbilic)
PTFEManbnM
Mek-BtowBPP
Mrt-BtowePP
Notn.
Rattaf
00
0.45
i
I
o.t
1.0
-5
General
FUnt
Quality
V Good
V Good
V. Good
Good
Peer

10 PPM
ACPTD
6.7
0.5
0.3
(.2
0.7
0.3

COO PPM
AdTD
29
1.3
O.t
41
2.4
1.5
•»* i.ceowgaa>-
Sua
SPP Vatey
69 14
1.7 0.3
1.0 0.2
63 12
65 13
4.0 01
                COM. (ha raMa, aad meeeured life cyck».


                         Table IV

APPLICATIONS  TESTING

   Table V summarizes a series of Oberiin bench-scale tests
with  simulated  Savannah River M-Area  wastewater (to
duplicate plating line effluents) and the various media evaluated
above. Basically, the  new  Tyvek (hydrophilic) matches all
competitive medias in rate and filtrate turbidity, but produces a
drier cake because of better air-blowing (ie, lower final air-blow
pressure since higher air flow due to less media blinding). As a
result, cake release is excellent and residual media weight gain
is reduced (IS mg vs > SO mg for others).

           SAVANNAH RIVER M-AREA STMIULATED WASTE
          CHALLENGE TESTS IN OBERUN PRESSURE FILTER
                                                                            Media**
                                                                                                                Fowl
TVpe
Mkropoftxat
jjtjLji-J-_rjj|j
miillll BUT
Tyvek T-9W
Tyvek T-9W
(Hydrophilic)
PTFE Membrane
| a»lgng«M-
Melt-Blown PP
Melt-Blown PP
Norn.
Rating
00
0.45
I
1
O.S
1.0
-5
AWL
CUD
(Bf)
SO
10
15
60
560
445
Cake
CUe %
Rfheir Solid*
Good
Excellent
Excellent
Fa»
Fair
Good
30
34
34
28
32
29
FflBite Avg. Cake*"
Tinbtdity Rue An- Blow
(NTU) (jptn/ft2) Press
-------
             CASE HISTORIES SAVANNAH RIVER PLANT

                             TyvckT-MO*
                             OMtoAPF.
                             Fnanol K-144L
                         MH  Pobmo Ptritow 30  NPDES
•Turbidity
•TSSfopn)
•Ahmuwn (ppm)
•Laid (ppm)
•Urmium (ppm)
•Zinc (ppm)
•Copper (ppm)
no
«7
127
2.'3
0.3
02
0.32
1.4
0.95
0.2
0.01
<0, 1
<0. 1

31
3.2
0.43
0.5
0.32
0.21
                               Fhul200GPD

                       TabteVII

ACKNOWLEDGEMENTS

   The authors wish to express their thanks to the following
people who contributed immensely in the data collection on
filtration tests and preparation of this paper: J.G. Wood,
DuPont Engineering Test Center and Dr. T. Oberlin, President
of Oberlin Filter Company.

LITERATURE CITED

Martin, Hi.. Paper presented at 10th Annual AESF/EPA
   Conference on Environment Control for the Metal
   Finishing Industry, Orlando, FL, Jan., 1989.
Ostreicher, Fluid Filtration: Liquid. Vol. II. ASTM STP 975,
   Johnston, P.R. and H.G. Schroeder, eds, Amer. Soc. for
   Test Mat., Philadelphia (1986).
                                                 74

-------
Chemisches Laboratorium Dr. E. WeBling

Oststra(3e2   •   4417 Altenberge  •   Telefon 02505/89-0
     Cleanup  Of Contaminated Soil By Ozone Treatment
1.   Process Description

     Volatile aliphatic  and  aromatic  hydrocarbons  are  easily  removed  with
     high  efficiency  from  contaminated  sites   by   soil  gas  sucking.   This
     method,  however,   fails  when  the  soil  is  polluted  with  less  volatile
     substances  like  mineral  oil  or  polycyclic  aromatic  hydrocarbons.  In
     these  cases  ozone treatment  is  a  suitable  technique  to  reduce  the
     pollutant concentration drastically.

     Ozone,  the most  powerful  technically  applied  oxidizing  agent,  is  able
     to destroy  hydrocarbons  when a  gas  stream enriched  with ozone  is
     passed through the contaminated  soil.  The  pollution may  be reduced by
     up  to  98  % of the  original  concentration but the  efficiency depends on
     several  parameters, e.g. the nature of the pollutants, the  condition  of
     the  soil,   especially  its permeability  to   gas,   and  the  presence  of
     accompanying   substances  (humic  material).   Ozone  treatment  may be
     carried out in-situ  as  well as on-site and off-site.
2.    Experimental Results And Development Status

     Experiments have  been carried  out  so far  in  laboratory and  pilot-plant-
     scale  with  treated  soil masses  of 2  kilograms to 3 tons.  In laboratory
     scale  the contaminated  soil  was placed  in  a  cylindric  glass  reactor.
     Barrels  and skips  were  used  in  larger  scale  treatments. An ozone
     containing  gas  stream was  prepared  from  compressed air or oxygen  in
     an ozone generator by electrical discharge.
                              75

-------
Figures  1  and 2  show typical plots of  pollutant  concentration  versus

treatment time received in  laboratory  experiments.
                            origin: tank farm
                            pollutant: PAH
                            •oil type: sand
                            treated quantity: 2 kilograms
                            specific ozone demand: 2.9 mg/mg PAH
fig  1:  treatment of PAH contaminated  soil

      *
During 20  days the  pollution  of a soil  derived  from a  tank farm  con-

taminated with  polycyclic aromatic  hydrocarbons was  reduced  from about

2 300 mg/kg to 50 mg/kg (fig. 1), which is a reduction by  nearly 98 %.



Similar  results were  obtained  with a mineral  oil containing  soil  from a

refining plant  (figure 2).
                           76

-------
                                 origin: refining plont
                                 pollutant: mineral oil
                                 eoil type: eand
                                 treated quantity: 2 kilograms
                                 •pectfic ozone demand: 3.5 mg/mg oil
                                                  time (h)




fig 2:  treatment of mineral  oil contaminated soil



By ozone treatment  the  concentration  of hydrocarbons  decreased from

12 000  mg/kg to 150 mg/kg within 2,5 days.  Experiments in pilot-plant-

scale gave similar results.



The first  in-situ-treatment  project at  a  former  gasoline  station  was

now started. Figure  3 shows the  corresponding flow diagram.  Pollutants

are mineral oil  and polycyclic aromatic  hydrocarbons in this case.
                           77

-------
gasproof barrier
ft

central control equipment 1

t
k >
\ *
1 emissic

                                                     1 active carbon filter I
                                   control I-
pment I
*
1
1 emission contro
fl
r
1

r^


                                   HL
                            measurement of ozone r"
                            concentration and flow |—
                r injection weU 1
. vacuum well I
fig  3:  in-situ-treatment, flow diagram

After  sucking  off  all volatile  hydrocarbons  ozone  treatment starts.  By
using  multilayer  injection  and  vacuum  wells  it  is  possible  to direct
the  ozone  containing gas stream.  Volatile  pollutants  are  adsorbed  on
active  carbon   filters,   which  also  destroy excess  ozone.  A  central
equipment controls the ozone concentration,  gas  flow  and ozone emission.
                           78

-------
3.   Oxidation  Products And  Influence On  The Soil Microflora

     It  is known that organic compounds  are oxidized by  ozone to alcohols,
     aldehydes,  ketones  and  carboxylic acids depending on  their  molecular
     structur and  reaction time. All these substances,  however, are highly
     biodegradable.

     Oxidation  products  of selected polycyclic  aromatic  hydrocarbons  have
     been  studied   in  more  detail  in  our laboratory.  Primary degradation
     products  are   those,  which  have  been  expected by  mechanistic  con-
     siderations. These  substances are transferred  into  the main products of
     ozone treatment,  oxalic acid,  carbondioxide  and water (figure  4).
                PAH   	>  H02C - C02H  	>  C02 + H20

     fig  H: oxidation  products of polycyclic hydrocarbons

     There  is  no indication  for  the  formation  of any  toxic degradation
     products.

     During  ozone treatment  soil microorganisms  are destroyed  to a  large
     extent.  A  restauration  of  the  soil microflora can, however,  easily be
     achieved for example  by addition of the effluent of a sewage  plant.

     Treatment  Costs

     A considerable amount of the  total treatment costs is  needed  for ozone
     preparation by electrical  discharge.  In  the  Federal  Republic  of Germany
     these costs are  today estimated  at 50 - 300,-  DM (30 - 175  US-Dollar)
     per ton  of treated soil  depending  on the pollutant concentration.  This
     amount  does  not  include  installation costs  and costs for running  the
     equipment  as  well as  analytical control.
                              79

-------
BlOTROC
                                         ... for environmental solutions, naturally!
                    BioTrol® Soil Washing System

                                 by

                          Steven B. Valine
                         Dennis D. Chilcote
                         Thomas J. Chresand

                            BioTrol,  Inc.
                            Presented at
                U.  S.  Environmental Protection Agency
        Second Forum on Innovative Hazardous Waste Treatment
              Technologies:   Domestic and International

                     Philadelphia, Pennsylvania
                          May  15-17,  1990
                              ABSTRACT

      Soil  washing is an innovative, volume  reduction  process for
 treating contaminated soil.  In the process, soil is slurried with
 water and  subjected  to  several  stages  of intensive scrubbing with
 interstage classification.   Separation of the washed, coarser soil
 components from  the  process  water and  contaminated fine particles
  (silt and  clay fraction) results in a significant reduction in the
 volume of  material requiring additional treatment or disposal.

      A  pilot-scale  soil washing  system with a treatment capacity
 of 500 to  1000 pounds per hour was operated at a Superfund site in
 Minnesota  contaminated  with wood preserving  wastes,  including
 pentachlorophenol,   polyaromatic  hydrocarbons,  petroleum  hydro-
 carbons, copper,  chromium,  and  arsenic.  The results  of this on-
 site  pilot  work  are  discussed,  including  preliminary  results
 obtained during  an EPA  Superfund  Innovative Technology Evaluation
  (SITE)  demonstration.    The results  of bench-scale  treatability
 studies covering a variety of contaminants  are presented, as well
 as estimated  costs for  full-scale treatment.
       BioTrol, Inc -11 Peavey Road • Chaska, Minnesota 55318 • 612/448-2515 • FAX 612/448-6050
                                 80

-------
                           INTRODUCTION
Basic Principles

     Soil washing is an innovative volume reduction technology for
the remediation of contaminated soils.  The effectiveness of soil
washing  relies  upon two principal  observations.   First, contam-
inants tend to be concentrated in the fine  size  fraction  (silt and
clay) of  soil.   Second,  contamination associated with the coarse
size fraction (sand and gravel) is primarily surficial.

     In  the  process,  soil  is  mixed with water  and  subjected to
intense attrition scrubbing.  This scrubbing action disintegrates
soil aggregates, freeing contaminated fine particles from coarser
sand and gravel.  Contaminants may also be  solubilized as dictated
by applicable solubility characteristics or partition coefficients.
Concurrently,  the  coarser  soil  particles  are  scoured by  the
abrasive  action  of  the coarse particles themselves,  transferring
surficial contamination  to the fine soil  fraction.   The process
water  containing the  contaminated  fine  soil  material  is  then
separated from  the  washed coarse soil.   These  three mechanisms:
disintegration of soil  aggregates, dissolution of contaminants, and
scouring  of  coarse particle surfaces, each  operate  to varying
degrees, depending upon the characteristics of the soil and contam-
inant (s) .

     To  improve  the efficiency of soil washing,  the process may
include the use  of  surfactants,  detergents,  chelating agents, pH
adjustment,  or  heat.    In many  cases  however,  water  alone is
sufficient to  achieve the  desired level   of  contaminant removal
while minimizing cost.


Process Description

     A simplified flowsheet of the BioTrol® Soil Washing System is
shown in Figure  l.  A multiplicity of unit operations are involved,
most of which are well known in the mineral processing industry.

     Following  excavation,  debris  is  removed  from  the  soil by
screening at 1/2 to 2 inches.  Now free of large debris, the soil
is slurried with water in a mixing trommel or pug mill, depending
upon the nature of the soil and  the amount of energy required to
achieve dispersion of the soil in water.  Next, the soil slurry is
wet  screened at  approximately  1  to  2  mm  to  remove  oversize
material, such as small rocks and pebbles.  This oversize product
can be further washed and/or segregated,  if necessary.

     The undersize product  from  wet screening is fed  to a froth
flotation circuit  where hydrophobic contaminants  (e.g.,  hydro-
carbons) and/or  contaminated fine solids are transferred to a froth

                               81

-------
                       FIGURE 1
              SIMPLIFIED    FLOWSHEET
         BIOTROL SOIL WASHING SYSTEM
  EXCAVATE
  FEED SOIL
   SCREEN
    SLURRY
    SCREEN
    FROTH
   FLOTATION
  MULTI-STAGE
COUNTERCURRENT
   ATTRITION/
 CLASSIFICATION
    CIRCUIT
   DEWATERING
  WASHED SOIL
                                               MAKE-UP
                                                WATER
                        DEBRIS
                              RECYCLE
                    WATER TREATMENT
                        SYSTEM
WASHED
OVERSIZE
  PROCESS
   WATER
 FROTH
FINE SILTS,
CLAYS,AND
ORGANICS
                      THICKENING
CONTAMINATED \
   FINE     |
  SOLIDS    /
                       DEWATERING
                        RESIDUALS
                       MANAGEMENT
                             82

-------
phase and removed.   Flotation underflow then enters an  intensive,
multi-stage,  countercurrent   attrition  scrubbing  circuit  with
interstage  classification.    Each  attrition  scrubbing  cell   is
equipped  with two  or  more propeller-type  impellers  of  opposing
pitch to achieve the desired scrubbing intensity. Hydrocyclones  or
spiral  classifiers are used  to  achieve classification following
each scrubbing  stage.   The soil  exiting the scrubbing  circuit  is
dewatered  in  a  spiral classifier  or  vacuum  belt  filter and
discharged as the  washed product.

     Flotation  froth and the  fine soil particles suspended in the
process water from the scrubbing circuit are fed to a  thickening
operation.  During thickening, a polymeric  flocculant is  normally
used to aid in the settling and separation  of the fine solids from
the process water.   Following thickening,  the fine solids can  be
further dewatered using a belt  press or  centrifuge.   Treatment
alternatives  for  the  contaminated  fine   solids  could  include
incineration,   biological   degradation,   stabilization,  on-site
containment, or off-site disposal.

     The clarified process water exiting the dewatering operation
is treated,  if  required,  and recycled back to  the  soil washing
system to minimize water consumption.   Make-up water is  necessary,
since the products exiting the soil washing system usually contain
a higher  level  of  moisture  than the  feed soil.   The  degree  of
process water recycle,  and therefore the make-up water requirement,
is dependent  upon the characteristics of  the  soil  and contamin-
ant (s) .
                 LABORATORY TREATABILITY TESTING

     A  laboratory treatability  protocol has  been  developed to
assess  the  performance of  soil  washing  on  contaminated  soil
samples.  Testing is normally conducted on two samples from a site,
one  that  represents   "average"  site  conditions,   and one  that
typifies "worst" conditions.

     Samples are  screened  to remove  debris,   blended,  and  then
divided  into   representative   aliquots   for  subsequent  tests.
Chemical, particle size, and optionally, mineralogical analyses are
conducted to characterize  the samples prior to  soil washing tests.

     Leaching studies  determine  the degree  to  which contaminants
partition  between the soil  and leaching  solution.   Data  are
generated from a series of low intensity agitation tests covering
a range  of solids concentrations.    Studies can be  designed to
compare  the  effect  of  physical  or  chemical  "additives"  on
contaminant solubility with  baseline  data  from  tests with water
alone.
                               83

-------
     Froth flotation and attrition scrubbing tests are accomplished
with a bench-scale flotation machine.  This type of apparatus has
been in use for several decades  in the mineral processing industry
and can be equipped  with  either a  flotation-type or an attrition
scrubbing impeller.  Design parameters used in scale-up from this
"bench" apparatus to full-scale are well established.

     Results  of treatability  testing with  various  client soil
samples are shown in Table 1.  For organics, levels of contaminant
      TABLE 1.  RESULTS OF BENCH-SCALE TREATABILITY TESTING

                                           mg/kg
Site Description
Wood Preserving
(California)

Wood Preserving
(Florida)
Chemical Plant
(Michigan)
Wire Drawing
(New Jersey)

Parameter
Total PAH (1)
Arsenic
Chromium
Copper
Zinc
Pentachlorophenol
Pentachloropheno 1
Dichlorobenz idine
Benzidine
Azobenzene
TPH (2)
VOC (3)
Copper
Nickel
Silver
Feed
Soil
4,800
89
63
23
345
380
610
770
1,000
2,400
4,700
2
330
110
25
Washed
Soil
230
27
23
13
108
4.0
25
13
6
7
350
0.01
100
60
4
Percent
Reduction
95
70
63
43
69
99
96
98
99
>99
93
>99
70
45
84
Town Gas
(Quebec)

Pesticide
Formulation
(Colorado)


Chemical Plant
(California)
Total PAH (1)        230       11
Chlordane             55      4.7
Aldrin                47      7.5
4,4-DDT               25      5.0
Dieldrin              46      7.0

PCB (4)              290     <0.1
   (Aroclor 1260)
 95
 91
 84
 80
 85

>99
Notes:  (1) Polynuclear aromatic hydrocarbons
        (2) Total petroleum hydrocarbons
        (3) Volatile organic compounds
        (4) Polychlorinated biphenyls
                               84

-------
reduction between the feed and washed soil generally ranged from 80
to greater  than 99 percent.  Reductions  in  the levels of metals
ranged from 45 to 84 percent.  The type of metallic  compound, i.e.,
oxide,  insoluble salt,  etc., will  greatly impact  the degree to
which metals can be removed  from a soil.   Removal efficiency could
be significantly improved by employing common hydrometallurgical
techniques.    The  data  shown  are  insufficient  to  draw  any
conclusions on the effectiveness of soil washing on  metals  removal.


                       PILOT-SCALE TESTING

     A  mobile,   pilot-scale  soil  washing  system with  a  nominal
treatment capacity of 500 to 1000 pounds per hour was constructed
for  use  in  process  development  and  for  conducting  on-site
demonstrations.    The  system  consists   of  pilot-scale  mineral
processing equipment assembled on a  42  foot long drop-down semi-
trailer.  Included in the pilot plant  system are a multi-component
automatic  feed  system  (including   a conveyer belt   scale  and
totalizer), a mixing  trommel, vibrating  screens,  froth flotation
cells, attrition machines, spiral classifiers, hydrocyclones, and
dewatering equipment.

     Clarified  process  water  from  the   soil  washing  system is
treated prior to recycle.   For  biodegradable compounds, a pilot-
scale, fixed-film, biological treatment system is used.  The bio-
reactor, contained in a 20 foot long,  enclosed trailer,  is  a multi-
cell, submerged,  aerated, packed  bed  reactor and uses an amended
bacterial consortium. Each cell is filled with a packing material,
usually structured PVC trickling filter media,  which serves as the
substrate for microbial  attachment.  Porous diffuser pipes mounted
beneath the packing  support grid provide aeration.   The packing
volume is 72 cubic feet, or 540 gallons.   Flow capacity can be up
to 10 gallons per minute.

     From 1987  to 1989,  extensive process development and field
pilot testing  of the  soil  washing system  was conducted at the
MacGillis and Gibbs Company  site  in New Brighton, Minnesota.  This
Superfund  site  is  contaminated  with  wood  preserving  wastes,
including   pentachlorophenol,   creosote   (polynuclear  aromatic
hydrocarbons) , petroleum hydrocarbons, and to a lesser degree, with
arsenic, chromium, and copper.  Only water (no additives) was used
in the test program.

     To  illustrate the  type of  results  achieved during  pilot
testing, a typical mass balance for pentachlorophenol is presented
in Table 2.  Chemical analyses and distribution  are  shown  for both
the dry solid  and aqueous phases.   Note that  47  percent of the
contaminant was solubilized  in  the  aqueous phase of the  fine
particle  slurry.   The  washed soil,  containing 34 mg/kg penta-
chlorophenol,   accounted  for  75 percent of  the soil fed to the
                               85

-------
system on a dry weight basis.  On a mass basis, 94 percent of the
contaminant was removed from the washed soil.
             TABLE  2.   PENTACHLOROPHENOL MASS BALANCE
      PILOT-SCALE TESTING AT MINNESOTA WOOD PRESERVING SITE

                 Dry       Cone., mg/kg     Distribution, percent
                Solids    	  	
                Weight,    Dry              Dry
Process Stream  Percent   Solids   Water   Solids   Water   Total
Feed Soil         100      680      	      100     	     100
Washed Soil        75       34      20       5.0     1.0     6.0

Oversize           12     1000      82        24     3.0      27
 (Mostly Wood)

Fines Slurry       13      820      72        20      47      67
               PRELIMINARY EPA  SITE  PROGRAM RESULTS

     The  U.  S.   Environmental  Protection Agency  conducted  an
evaluation of  BioTrol's  soil washing  technology in the Superfund
Innovative Technology Evaluation (SITE)  program during September
and October of 1989.   The MacGillis  and Gibbs wood preserving site
was chosen as the location for the test.

     The capability of the system was evaluated by testing with two
types of soil from the site.  The first sample was excavated from
the vicinity of a former pole treatment tank and  designated as the
"low" contaminant level soil.  A second sample, designated as the
"high" contaminant level soil,  was excavated from a "hot spot" in
the site disposal area.   Tests  with  the "low" soil lasted two days
while tests  with the "high" soil lasted about  6 days.   The soil
washing  system operated  24  hours  per day for  both tests.   As
before, only water was used (no additives).

     All product  streams  exiting the  system were collected in 55
gallon drums and  weighed  to  determine mass flows.   Feed rate was
determined by  recording  conveyer belt totalizer readings.   Grab
samples  of  all input and output streams were  taken  at roughly 2
hour  intervals;  individual grab samples were  then  composited to
represent about 6 hours of operation.
                               86

-------
     Preliminary results of the demonstration are given in Table 3.
As shown,  pentachlorophenol levels were reduced 90 to 91 percent in
the washed soil  compared to the feed soil.  Reductions  in the level
of total  polynuclear  aromatic hydrocarbons  ranged from 92  to 96
percent while  reductions in  total petroleum  hydrocarbon  levels
ranged from 94 to 95 percent.
           TABLE  3.   COMPARISON OF FEED AND WASHED  SOIL
             PRELIMINARY RESULTS OF EPA SITE PROGRAM
       PILOT-SCALE TESTING AT MINNESOTA WOOD TREATING SITE

                                           mg/kg
5011
Contaminant
Level
Low
(Test 1 of 1)






Parameter
Pentachlorophenol
Total PAH (1)
TPH (2)
Arsenic
Chromium
Copper

Feed
Soil
130
240
3,300
14
17
15

Washed
Soil
12
8.6
210
5.0
9.0
6.2

Percent
Reduction
91
96
94
64
47
59
High
(Test 1 of 2)




Pentachlorophenol
Total PAH (1)
TPH (2)
Arsenic
Chromium
Copper
540
290
8,800
28
49
39
56
23
470
7.2
8.5
5.2
90
92
95
74
83
87
Notes:  (I) Polynuclear aromatic hydrocarbons
        (2) Total petroleum hydrocarbons
     Preliminary Toxicity Characteristic Leaching Procedure  (TCLP)
results for pentachlorophenol are presented in Table 4.  With both
soil types,  the  leachate  solutions contained  less  than  1 mg/L
pentachlorophenol.  Since  the  system is a water-based, intensive
scrubbing process, leachability (TCLP)  of  the washed product will,
in most cases, be very low.  This fact  should allow washed soil to
be placed back on-site.

     Test results demonstrating the effectiveness of the biological
water treatment system on the process water from the soil washing
system are  illustrated in Figure 2.   For days 1 through  3,  the
system treated water  generated from the  "low"  soil test.   Water
from the "high" test was fed to the unit on days 4 through 10.
                               87

-------
              TABLE 4.  RESULTS OF TCLP TESTS
COMPARISON OP PENTACHLOROPHENOL IN WASHED SOIL AND LEACHATE
          PRELIMINARY RESULTS OF EPA SITE PROGRAM
   PILOT-SCALE TESTING AT MINNESOTA WOOD PRESERVING SITE
Soil
Contaminant
Level
                                  Pentachlorophenol
Washed Soil,
   mg/kg
Washed Soil TCLP
 Leachate, ntg/L
Low (Test 1 of 1)
High (Test 1 of 2)
    10
    19

    59
    70
     0.23
     0.32

     0.74
     0.92
    50
    40
    30
    20
    10
            FIGURE 2. BIOLOGICAL TREATMENT OF
            PROCESS WATER FROM SOIL WASHING
           (PRELIMINARY DATA - EPA SITE PROGRAM)

      Pentachlorophenol Concentration (ppm)
             Influent
      012345678
                            Day
                      10   11

-------
                               COST

     Estimated treatment costs for a mobile, commercial-scale, 20
ton per hour soil washing system are  shown graphically in Figure 3
as a function of tons  processed.   Costs include capital recovery
(charged as an equipment leasing rate) and water treatment; not
included are costs for  excavation, debris removal,  and treatment or
disposal of residuals generated during treatment.

     Total cost per ton is the sum of the mobilization and treat-
ment cost components.   The estimated  treatment cost of roughly $60
per ton varies  only slightly with tons  processed.   Mobilization
cost per ton has the most impact at relatively small soil volumes
(less than 3,000 to 4,000 tons).

     For illustration, from  Figure 3,  the  estimated soil washing
cost for 20,000 tons of soil  would be approximately $71 per ton or
$1.42 million.   Costs  for  treatment  or disposal of the residuals
generated during soil washing must also be calculated to estimate
total remediation costs.   For example,  soil washing of 20,000 tons
of soil with the characteristics shown in  Table 2 would generate
2,400 tons  of  oversize and  2,600  tons of  fines,  both  requiring
additional treatment.  Using incineration at $200 per ton for the
oversize and slurry-phase biodegradation for the fines at $100 per
ton,  treatment of  residuals  would  cost an  additional  $740,000 or
$37 per ton based on  20,000 tons.   Total  soil treatment costs,
including treatment of residuals,  is therefore  estimated at $108
per ton or $2.16 million.

     Soil  washing unit  costs will  be  significantly  lower  for
relatively large soil volumes (greater than  100,000 to 250,00 tons)
or for a fixed  facility with  an expected operating life of greater
than 5  years.   Larger  throughput  capacities  and  use  of  longer
periods for capital recovery would lower total treatment costs to
the range of $25 to $40 per ton.
                               89

-------
        120
             FIGURE 3.  ESTIMATED TREATMENT COST FOR
               20 TON PER HOUR SOIL WASHING SYSTEM
                              (TYPICAL)

           COST PER TON, $
        100 -


         80


         60


         40


         20


          0
. Total Cost (Treatment + Mobilization)
      Treatment Cost

           „ Mobilization Cost
                  10      20     30      40      50
                      TONS PROCESSED (Thousands)
                         (Cost excludes excavation, debris
                         removal, or disposal of residuals.)
                                  60
                            CONCLUSION

     Laboratory  treatability   studies   and  on-site  pilot-scale
testing have  demonstrated the effectiveness  of  the BioTrol® Soil
Washing System.  Removal  of  organic contaminants typically ranged
from 90 to greater than 99 percent.  Removal of metals was somewhat
lower, but could be improved through the use of hydrometallurgical
techniques.

     Since  the  system  is  a  water-based,  intensive  scrubbing
process,  leachability  (TCLP)  of the washed product will,  in most
cases,  be very low.   This  fact should allow  washed soil  to be
placed  back  on-site  without  endangering human  health  or  the
environment.   As  a  cost effective,  volume  reduction  step,  soil
washing  will  play  a   significant  role  in  the   remediation  of
contaminated  soil.
                                90

-------
INTEGRATED SOIL-VAPOR/GROUNDWATER CLEANING SYSTEM

                       AT

         SELECTED SITES IN WEST GERMANY
                     Authors

              Karlheinz Bbhm,  Ph.D.
                       and
                   Gerrit Rost

     Department of Environmental  Engineering
                  Ed.  Ztiblin AG
                  Albstadtweg 3
                7000 Stuttgart 80
           Federal Republic of Germany

                       and

                 Martin R. Griek
            ZDI Construction Services
                607 Herndon Pkwy.
                    Suite 100
               Herndon, VA  22070
                  Presented at
               The Second Forum on
    Innovative Hazardous Waste Technologies:
           Domestic and International
           Philadelphia, Pennsylvania
                May 15 - 17,  1990
                         91

-------
                   INTEGRATED SOIL-VAPOR/GROUNDWATER CLEANING SYSTEM
 I.  INTRODUCTION

     Ed.  Ziiblin  AG is  an international   civil  engineering   firm headquartered  in
     Stuttgart, West  Germany.    It is   represented in   the  U.S.   by  its  operating
     subsidiary, ZDI Construction Services,  Herndon,  Virginia.

     Ed.   Zu'blin   has  provided  the   European   and   International   markets   with
     construction related services for  over  92 years.    For  the   past decade  their
     Department of  Environmental Engineering has been   involved in  the  treatment
     and  containment of hazardous material.   These areas include:

          Asbestos - removal  and disposal
          Hazardous waste site  containment - HOPE lined  slurry walls
          Thermal incineration  of contaminated soil with rotary  kilns
          Integrated soil-vapor/groundwater  cleaning  systems

     Vacuum extraction of soil-vapor is not  a unique  process.  It is, in  fact, the
     standard   method  in  Europe  today  for cleaning  up sites  contaminated with
     VOC's.     [Chlorinated  Hydrocarbons   (Solvents   TCE-PCE-DCE) and  Aromatics
     (Petroleum Products - BTEX)]

     Combining   a  soil-vapor  extraction   and treatment system  with  a groundwater
     treatment  system for the  remediation  of  a  VOC's  contaminated  site   is a
     sensible decision.   The  vadose zone  can  be remediated by  vacuum extraction
     with relative ease and economy.   This   process   removes the source  of the
     groundwater contamination.    Thus  the groundwater only  needs to  be cleaned  of
     residual   contamination.    Removal   of  the  source pollution   shortens the
     duration of  the treatment  process since  percolation  of volatile substances
     through the soil horizons  into the groundwater can   take decades.   The  costs
     and  effectiveness of remediation can  be vastly improved by  using the combined
     phase treatment process.

     This paper presents two case studies  utilizing Zliblin's  integrated  treatment
     process incorporating  this  procedure.    Zublin  has  installed and operated
     dozens of   these  systems   during   the   past  decade.    With typical  German
     efficiency, the  company designed  a highly  flexible, efficient and functional
     system to   recover  solvent  contaminants  and  minimize discharge   into the
     biosphere.

II.  SCHERWIESEN SITE

     This study  deals with a landfill, for  industrial  sludges,  animal processing,
     and  household waste which  had  been  operating  since  1967.    Following its
     closure in  1980, the landfill was excavated to  a depth of  three (3) feet and
     capped with a one meter thick compacted clay seal.

     Engineering studies performed in   early 1984,  showed that   hydroxide sludges
     and  chlorinated hydrocarbons were  present throughout the site and had started
     to contaminate the groundwater.   The  contamination plume   had  extended over
     650  meters  (2,100 feet  from the  site).  At this point it  had contaminated a
     natural spring which was  the  primary  drinking  water  source   of  a nearby
     village.   Remediation of the site  was ordered by the appropriate authorities.
                                            92

-------
     In  1987  Zublin's  Department   of  Environmental  Engineering was  awarded the
     project to extract and treat both  the soil-vapor   and  the   contaminated water
     in the  perched water  table  (source  contaminants)  as well  as the contaminated
     spring water.   Starting with commencement   of  the  operation,  a  study was
     conducted to record the effects  of the treatment  process.   The data  collected
     during this study is presented herein.   The   plant   is   still  in  operation
     today.  Figure  1 is a  schematic  diagram of  the  treatment system.   A  site plan
     of the project  is shown in Figure  9.
III. TREATMENT PROCESS

     A.   Soil-Vapor

     Initially,  ten (10)  soil-vapor  extraction wells were  constructed with  varying
     depths  of  13  to   17 feet  at  the  points  of  highest  contaminant  concentrations
     at  the  site.   A set  of vacuum pumps  extracted  250 cubic meters  per   hour  from
     a five   (5) well  manifold  with  a  negative pressure  of approximately  200 mbar.
     The manifold was  shifted to different  wellheads when   the  vapor concentration
     decreased.   The  soil  at the  site is a heavy clay.   Its zone  of influence for
     each soil-vapor well  was only 10  to  16 feet.   With  a  permeability of the  soil
     of  10"6  cm/s, the need to drill  additional wells was apparent.  As  a  result,
     in  1988,   a year   later, twenty  (20)  additional  soil-vapor  extraction wells
     were installed.

     The extracted   soil-vapor  passes   through a water separator which removes the
     water particles picked-up  by  the  vacuum, along with the soil-vapor,  from the
     soil.   The water from the   separator,  which  may contain  trace contaminants,
     mixes with  and is cleaned  in  the  water  treatment phase  of the system.   The
     vapor,  which   has been heated  by the energy  produced by  the vacuum pump,  is
     cooled  prior to entering the  activated  carbon filter for maximum adsorption
     by  the  carbon.

     Two carbon  filters  are used:    As  one  filters the influent  air the other is
     either  on stand-by  or being  regenerated.    Prior   to   the  on-line filter
     reaching  break-through, the order is reversed.  In  this fashion the  treatment
     continues uninterrupted without  the  need  to  exchange and   dispose of spent
     carbon  filter  material.

     During  the  regeneration  process,  steam  desorbes   the  solvent   from   the
     activated carbon.     The   contaminant  laden  steam is   then  condensed.
     Subsequently,  the solvent and  water  are   separated  resulting  in recovery of
     the solvent.   This solvent is 99% pure and  can then  be  recycled or properly
     disposed.   The contact water is  fed back into the  liquid  phase of the system
     for cleaning.   During  the  first year over 3,000 kg  of solvent  was  recovered
     at  this site.
                                        93

-------
Contaminant concentrations  of the  soil-vapor were  reduced from  8,000 ppm to
between 30 - 50 ppm.  Figure 3 shows the  influent concentrations  of the soil-
vapor.   Air volume  was measured  with an integrated flywheel  anemometer.   The
total volume extracted and  treated  over  this  period  of  time  exceeded 2.5
million cubic  meters.   This is graphically demonstrated in Figure 4.   The air
discharge concentrations were well below  the  German  standards  of  100 ug/1.
Figures  5  and  6  show  the  air discharge concentrations and quantity of air
treated.  Figures 7  and 8  illustrate quantities  of solvent  recovered by the
plant.

As can  be seen  in looking  at these  curves there  are instances of several
upward shifts.  The primary cause for these shifts are two.  The first is the
result of  climatic conditions.   As  the clay dries, cracking  occurs:   These
cracks provide  paths  for  the  extraction  of  ambient  air  and  lower the
concentration measured  in the influent air stream.  Further evidence of this
occurrence is the  "decrease" in the negative pressure measured at the vacuum
pump.

The second  reason for  a shift  is caused  by interruption of  the extraction
process at one or more of the wells.  The phenomena of  increased contaminant
concentration  in  the  extracted  soil-vapor  after  a shut-off well resumes
extraction is common.  During  this  shut-off  period  the  volatiles  have a
chance to  re-establish equilibrium  in the  "vacuumed" section  of the soil.
That is, the volatiles  inside the  soil have  a chance  to diffuse  into the
soil-air space.   Consequently,  concentrations following start-up are higher
than those recorded prior to shut down.

Other minor factors affecting the peaks are due to:  shut downs  of the plant
which occurred  in December  1987, and  June 1989  depicted on  Figure 6; and,
precipitation - since a partial saturation of the vadose zone impedes removal
of the  VOC's.   The  important  point of  the graphs, however,  is to note the
trend of decreasing concentration in the contaminants.

B.  Water

In order to expedite  remediation of  the site,  a groundwater  extraction and
treatment phase was implemented in conjunction with the soil-vapor treatment.
Five  (5) of the  original ten  (10) wells  were used  to extract  the perched
groundwater.    A  shallow  ditch  was  constructed  encompassing the site to
capture and treat the surface run-off.

The perched water is  part of the  aquifer  that  feeds  the  spring.     It is
greatly influenced  by precipitation as shown in Figure 2.  The perched water
with a flow rate of 200 liters per  hour it  is pumped  into a  storage tank.
From there  it is fed through a sand filter into two activated carbon filters
operating in series.  The initial concentration of the water  measured at the
plant was  23,500 ug/1  with point measurements at the monitoring well  (P-12)
as high as 166,000  ug/1 of  chlorinated hydrocarbons.   Following treatment,
effluent concentrations  of the  discharged water were well within the German
drinking water standards of total CMC's less than 20 ug/1•

It should be noted  that after  constant  and  continuous   remediation   of 2-1/2
years,  the  contaminant concentration  of the  perched water was reduced to  1.9
mg/1.
                                       94

-------
IV.  ECONOMIC ANALYSIS

     A  cost  analysis  of  the   plant   follows:    Costs are  1n 1987 dollars at a
     conversion  ratio of 1.6 DM  to  $1.00 US.

          Site Operations                                  DM            US

          1.   Soil-Vapor System
              a.    Installation                           17.025 DM     $10,640
              b.    Operation and Maintenance  (12 mos.)      5.235          3,275

          2.   Solvent Recovery  System
              a.    Installation                           52.200         32,625
              b.    Operation and Maintenance  (12 mos.)     18.070         11,295

          3.   GAC  System
              a.    Installation                           13.500          8,440
              b.    Operation and Maintenance  (12 mos.)      2.160          1,350

                                      Total               108.190 DM     $67,625

          Installation Costs

          1.   Site
              a.    Soil-Vapor System                       17.025 DM     $10,640
              b.    Solvent  Recovery System                 52.200         32,625
              c.    GAC System                              13.500          8,440

                                      Total                82.725 DM     $51,705

          Operation  and Maintenance

          1.   Site
              a.    Soil-Vapor                               5.230 DM     $ 3,270
              b.    Solvent  Recovery                        18.070         11,295
              c.    GAC                                     2.160          1.350

                                      Total                25.460 DM     $15,915

          These  Costs  Include:

          1.   Site  preparation  for the  plant
          2.   Mechanical  and electrical  installation charges
          3.   Labor,  material and equipment
          4.   Operating  supplies, utility costs, and maintenance

          These  Costs  Exclude:

          1.   Well  installation
          2.   Permitting  costs
          3.   Replacement charges (depreciation)
          4.   Carbon  replacement - liquid phase
          5.   Engineering charges for the sampling/testing study
                                           95

-------
conlominaled
water
                            condensolor
 ii Ii tin n
conk
air

II
no
0

r|
SI





ted
r


DCK










^L_




***•
**~*
^






1
i
!
i
:
•n —
---
\
-r-
\
•
•





1
1
IL
V
^
-~-~
; |


I


IT

--







—I/
j||
*
nJn
1- •' '•; activated
• . ! I • carbon filler
fZ^J 'or a'r
:--jj y
! •
* ".r.r<"VJL..""fi*ff"**r""j*"".T.i".'."i"
Sl(


r water -seporalor
— 1
                                                                sand filler
                                          sleam generator
            pump
                                                                                sloragelank
                                                                            Irealed waler
                                                             waler softener
tundensale    solvents
                                                      disposal
                                                                         	  water

                                                                         i":  oir

                                                                         v/.v.v.:  sleam
  SOIL-AIR  AND WATER TREATMENT -  SYSTEM  DIAGRAM
                                                                            ZUBLIN
 Figure 1.  Schematic  of  the soil  - air and water  treatment system

-------
Ł.
X
1.-:


l.ii


I.-I


i.::
       0,6 -
       0,4 H
        0.2 -
         0
                                                  O±s=»i>osts.."JL  Si


                                           Wntcr quantity
                      '4
                               fi
                              Wf
G/87    9/87    12/87    3/88     6/88    9/88


                                    Date
12/88   3/89
                                                                           6/89
                                                                            9/89   12/89
             Figure 2.  Quantity oT perched  wal.or extracted  and  treated by the  plant.

-------
                                            86

                                         CHC  concentration   (ppm)
                                               (thousand)
n
3
O
0
~s

f

9
n
      o  c
   O  *--  N
                                         o   o
                                     01  --
            00
            ~J
            so

            CO
            CO

            CO
            CO
            CO
            CD


            CO
        3
        r*

        !3
            CO
            CO
            CO
35
            co  •
            CSJ
            30
            13
                                                         '   |    '
Li   fJ:   --I  Ct1

J_._l	L_
        J

                                                                                      O
                                                                                      3
                                                                          o
                                                                          0
                                                                          3
                                                                          3


                                                                          C8
                                                                                      O
                                                                                      3
                    w
                    0
                                                                               H-
                                                                               (5
                     H-
                     ffl
                                                                               P
                                                                               H

                                                                               W
                                                                               H-
                                                                          -    (D

-------
                                        66


                                     Soil vapor (cbm)
                                       (million}
 13
 O

 i
 a
CO
o
o
**


'")
                                                                            o
                                                                            c
                                                                            3
                                                                            3
                                                                            9
                                                                            r-
                                                                            w
                                                                            o
                                                                            0
                                                                            1

                                                                            fO
                                                                            X
                                                                            o
                                                                            f
                                                                            **•
                                                                            o

-------
                                                                     S:L tea
                                        EffluonL air  concontrntions
          100
o
o
   o
   a:
   o


    p.
    o.
              6/87     9/87    12/87   3/88    0/88     9/88   12/88   3/89     G/89     9/89   12/89
                 Figure  5.   Efriuont air conccnl.ral.iotis of  Lhc soiJ  vapor exl.rnc:t i on  system.

-------
                                                                   . .1.  S i. 1;. e»
                                        Soil  vapor quantity  treated
  JC
  X


K-. .0
o  u

400 -
3Oi.i -
.•".i1 -
... _

100 -
r..o -


U.iO -


00 -


}
1













'|
i
\
\
\
\ }

'-/ '|
i
















1
1
r.
"\ n A '-I ',-

V ' V ',
\r-\j\fa;
^V









W,A
i V'
















. .' ^
1 \ /;
t 1
1 1
V






,
'(•'
If
ff



rt

V !
\
i



, l;

s

i.
Mil
' i















v J\
V-/ f.f '
'1 'l Irf







6/87    9/87    12/87    3/88    G/88     9/88   12/88



                                    Da I o
                                                                              0/89    9/89    12/89
                 Figure (5.   Soil  vapor quantity  l.roal.rd  over t i mo.

-------
                                                   O.i.Ł5|posei.J.  Si. to
                                    Recovered chlorinated hydrocarbons
O
ro
    
-------
                                 Total  recovered chlorinated hydrocarbons
o
CO
O
a:
u
          1  -
          0
           6/87    9/87    12/87    :«/88     G/88    0/88   12/88    3/89    G/89    9/89    12/89



                                               I)a l.o
             Figure 8.   Cummul alive solvent rocovorc»d by  soil  vnpor trr-ntmont.

-------
Ubersichtsplan Scherwiesen
                                   Aktiv-x,
                                   kohle-
                                   filter
                                  Biotop
ZOBL!
                     FIGURE 0

                       104

-------
V.   NUREMBERG SITE

     The second  study involves  a company located in Nuremberg,  West Germany.   It
     was a  used  oil  recycling  plant  for  a  number  of  years  which declared
     bankruptcy in  the early  1970's.   Several years later in 1979, the buildings
     were razed and the number of hazardous chemicals removed from this site along
     with several  feet of  surface soil.   Eight (8) groundwater monitoring wells
     were installed to monitor potential contamination of the underground aquifer.
     During the  drilling operations,  it was  discovered that the groundwater  was
     highly contaminated to a depth of 11 meters (33 feet) however, no  action  was
     taken at this time to remediate the site.

     Four  (4)  years  later,  following  a period of substantial precipitation,  a
     large neighboring sand pit filled with oil due to a  rise in  the groundwater
     level.  Once visual contamination became readily apparent more investigations
     were  ordered  which  showed  that  the  groundwater  was  contaminated with
     chlorinated hydrocarbons  and residual  petroleum products.    The groundwater
     concentration was calculated as high as 18,000 M9/1•

     In  addition,  located  under  a  former  barrel  storage   area,  soil-vapor
     concentration as  high as 8,000 ug/1 consisting mostly of PCE, TCE and CIS 1,
     2-DCE were noted.  The area underneath the site was contaminated in the range
     of 100 - 700 ug/1.

     In  September  1988,  almost  10  years  after  the  buildings  were razed,  a
     treatment  contract  to  remediate  the   site   was   awarded   to  ZUblin's
     Environmental  Engineering  Department.    Figure  10  is  a plan view of  the
     Durbanol site.   Figure  11  is  a  larger  scale  map  showing  the relative
     location of the extraction wells and the plant.


VI.  TREATMENT PROCESS

     The treatment  process at  this site  consisted of soil-vapor and groundwater
     extraction and treatment of the VOC's and oil.   The  process was complicated
     by a  substantial amount  of iron  and manganese  present in the groundwater.
     Figure 12 is a  schematic diagram  of the  various treatment  process at this
     particular site.  Figure 13 is a plan view of the plant itself.

     Soil-vapor extracted  from well  B9 using  a vacuum blower with a capacity of
     300 cubic meters per hour at a pressure of 50 mbar.   The soil  is very sandy
     with a  permeability of  10~3 cm/s.  Range of influence for each well exceeds
     50 feet.  Wells Bl, 2, 5, 6  and 8  are used  to monitor  the groundwater  and
     offer additional vacuum extraction points, if required.

     The soil-vapor  is passed  through a water separator to precipitate any water
     picked-up by the suction.  In this instance, the soil-vapor does not  need to
     be cooled  because the  blower, using  less energy, does not heat this air up
     significantly.

     Wells SI and S2 are the two collection wells constructed in 1988 northeast of
     the original  plant site  in the  sand pit.  These wells were each drilled 15
     meters (45 feet) deep.  Groundwater, at a flow rate of 1.5  liters per second
     for each  pump (measured  by in-line  gauges), is  continually extracted from
     both wells and fed into a two-stage air stripping plant.
                                             105

-------
       As the water trickles down through the towers,  the counter  flowing  air  strips
       out the VOC's.   The process  air from  the stripper  is  mixed   with the soil-
       vapor.  The combined flow is then fed through the activated carbon  filter
       system which contains dual filters operating alternately in a  fully automatic
       environment.  The system includes an  on-site steam  regeneration process  for
       the  carbon.    It  is  designed  so  that prior  to reaching break-through it
       switches over to the clean  filter.    At  that  point,   the  steam  which is
       generated by  an oil  fired burner, desorbes the  VOC's from the carbon  in  the
       off-line filter.  Once the carbon is desorbed,  which takes approximately an
       hour,  the  filter  remains  off-line  while  the  carbon  cools down. At a
       calculated predetermined time the automatic cycle  switches'the  clean  filter
       into the on-line mode.

       Water for the steam, both to feed the boiler and  condense the  steam,  consists
       of contaminated water which has been  cleaned  in  the  liquid  phase  of  the
       operation.  On  this particular site, due to hardness problems, it is softened
       before heating  to eliminate scale build-up.

       The steam, after it cleans the carbon, is condensed and  flows  into   a cyclone
       which separates  the solvents  from the condensate.  The condensate is  pumped
       back into the system, mixed with the groundwater  prior to the  stripping tower
       where its trace VOC's are removed.

       After the   stripping  columns, the  water flows   into a collecting basin  and
       then through a  sand filter which removes the flocculating iron and  manganese.
       The  sand   filter  is  piped  so  it  can  be  back-flushed  to remove  the
       flocculation when a drop in pressure is noticed.

       In order to remove the trace VOC's and oil remaining in  the water,  a  two  stage
       liquid phase  carbon filter  of approximately 5 cubic meters is utilized.   Both
       this discharge  water and the cooling water used in the solvent recovery process
       are subsequently discharged into a local sewage system.

       The  oil  that   is  found  floating  on the surface of the  groundwater  in  the
       extraction wells is skimmed off by  a membrane  pump.   The oil-water mixture
       is fed  into a  separation tank which removes the  oil from the  water.  The  oil
       is collected in containers and  recycled.    The   water   is  pumped  into  the
       treatment system prior to the stripping columns.


VII.   ANALYTICAL MONITORING

       Routine sampling  and analysis  of the system was performed during  the study.
       Four (4) sample ports measured the following information:

          BL1 -  Soil-vapor from  well  B9  after  the  vacuum   blower but
                 before the activated charcoal filter.
          BL2 -  Consisted  of   a  mixture  of  soil-vapor  from   B9  and
                 stripping  air  from  the  towers  before  the  activated
                 charcoal filter.
          BL3 -  The discharge after the activated charcoal filter.
          BL4 -  Stripping air only before the activated charcoal  filter.
                                            106

-------
       From  start-up  through  mid-December  1988,   only  three  sampling  ports  were
       available;  BL1,  2 and  3.    Sampling  port  BL4 was  installed  in  mid-December
       1988.   Once it was installed no further sampling  was taken from port  BL2.
VIII.  OPERATION
       Installation of  the plant commenced  on November 8,  1988.   Ten  days  later  the
       first test run was conducted.   On  November  24  (3 weeks  later) the  groundwater
       and soil-vapor treatment system commenced full  operation  and has been  running
       continuously since.   During the period  of November  24,  1988 through December
       21, 1989,  approximately 13 months,  the  following results  have been achieved.
       Soil  Vapor

       Hours of operation
       Volume of soil-vapor
       Total vapor extracted (approx.)
       Recovered solvent
June 1989

4,854 hours
250 cu meters
1,213,500 cu meters
2,300 kg
December 1989

9,065 hours
250 cu meters
2,266,250 cu meters
2,700 kg
       With   the   site   concentration   averaging   500   ug/1.    The  total  quantity  of
       solvent received  during  the   13  months   of operation   was  approximately  2,700
       kilograms;  about  2.5  kilograms per  day.
       Treated Groundwater

            SI
            S2
                           TOTAL:
            Stripped  VOC's
June 1989

18,315 cu meters
17,206 cu meters

35,521 cu meters

145 kg
December 1989

40,470 cu meters
34.210 cu meters

74,680 cu meters

175 kg
       The  results   of  the   treatment  program   are  pictured   on  several  graphs.   The
       liquid  extraction  wells  show  a substantial   contaminant  reduction   on  Figure
       14:   SI  initially  increased in  concentration  from  5,050  ug/1  to  over  6,980
       ug/1  in December 1988 and  then  dropped  to  approximately 700  yg/1  in December
       1989.   S2  concentration  was reduced  from 11,500  ug/1  to under 700  ug/1  in  the
       same  time  frame.

       It should  be  noted that  in February  and August 1989,  clogging of   the packing
       material,  due   to  the  flocculating   iron   and manganese,   resulted   in an
       inefficient stripping process which  is  readily   apparent  on  Figure 14.    The
       packing material  was  exchanged  - cleaned and  the   stripping  efficiency
       resumed.   At  all   times  the  discharge  concentration   was below   1  ug/1  after
       treatment.

       On Figure  15, the  obvious and expected decrease in  soil-vapor  concentration
       from  6,900 ug/1  to values  of approximately 500 ug/1 after one year  is readily
       apparent.
                                        107

-------
IX.  ECONOMIC ANALYSIS

     The cost  analysis for this project is  divided  up  into  six  distinct  phases as
     follows:
                                                         DM            US

     A.    Installation of Plant                   35.110  DM      $21,945

     B.    Soil-Vapor System
          1.    Capital Expense                    33.845          21,155
          2.    Operation and Maintenance          17.200         -10,750

     C.    Stripping Tower
          1.    Capital Expense                    45,920          28,700
          2.    Operation and Maintenance          11.830           7,395

     D.    Oil Stripping / Separator
          1.    Capital Expense                    66.790          41,745
          2.    Operation and Maintenance          8.485           5,305

     E.    Iron / Manganese Removal
          1.    Capital Expense                    40.565          25,355
          2.    Operation and Maintenance          8.855           5,535

     F.    Groundwater - ACS
          1.    Capital Expense                    22.085          13,805
          2.    Operation and Maintenance          5.375           3,360

     Total Capital  Expense                       244.315  DM    $152,705

     Total Operation and Maintenance              51.825  DM    $ 32,345


     These Costs Include:

     1.    Site preparation for the  plant
     2.    Mechanical and electrical installation charges
     3.    Labor, material and equipment
     4.    Operating supplies, utility costs,  and maintenance

     These Costs Exclude:

     1.    Well  installation
     2.    Permitting costs
     3.    Replacement charges (depreciation)
     4.    Carbon replacement - liquid phase
     5.    Engineering charges for the sampling/testing  study
        Costs are in 1989 dollars with a conversion rate of 1.6 DM to $1.00 US
                                             108

-------
  o
  c:

 m
t

'5
\
\
\
\
\
\
\
\
\
to \
*• \
^ \
00 \
o \
s \

^
.'
/
{'•.
\ \
\ '•
j
i

-------

-------
  FORCED AIR
  INJECTION
                                          DISCHARGED
                                          TREATED AIR
                                                                                  FEEDING WATER
                                                                                           ACTIVATED CHARCOAL
                                                                                                  FILTERS
OIL SEPARATOR
                                              STRIPPING
                                               COLUMNS
                                                                                                            PRESSURE
                                                                                                             KETTLE
                                                                                          DISCHARGE-TREATED WATER
                               DURBANOL-PROCESS  DIAGRAM

-------
                                                                         10.3 METERS
    AIR TREATMENT
     COLLECTION  BASIN I
      COLLECTION BASIN II
t-> "•  WATER STORAGE
ro .•
                                                             ACTIVIATED
                                                             CARBON
                                                             FILTERS
      STEAM GENERATOR
       WATER-SOFTENING
               SCALE  I > 29
               DATE   10-28-1988
PLAN  VIEW

-------
  10000


   5OOO
   1000

    500
   100
  m
co
    10

     9
              N
            1986
                         M
M
J        J
  1989
                                                                                                                 N
J
1990
                    LEGEND
                                                                     DURBANOL
                                                          VOC  CONCENTRATIONS  IN  WATER
            o	
            •	
VOC CONCENTRATION SI
VOC CONCENTRATION 52
VOC CONCENTRATION AFTER AIR STRIPPING
VOC CONCENTRATION AFTER FIRST A.C. FILTER
VOC CONCENTRATION AFTER DISCHARGE WATER
                                                                                      DATE:   1/31/1990

-------
 ug/l



 5000
 4OOO
 3000
 2000
O
c
3>
tn
^
Ol
 1000
     1989
         LEGEND
     BL I      SOIL VAPOR
     BL 2     OFF GAS AFTER TREATMENT
       DURBANOL
    ANALYTICAL RESULTS

TOTAL VOC CONCENTRATION IN AIR
                                     1990

-------
X.   CONCLUSIONS

     The following conclusions can be drawn from these  two  case  studies:

     A.    The combined soil-vapor/groundwater extraction  and  treatment  system
          proved very  effective  in  remediating  a site  contaminated  with
          chlorinated hydrocarbons.   One major objective of  this  process was
          to  minimize  impact   on   the   surrounding  environment  during
          remediation.  This goal  was admirably achieved.

     B.    The in-situ combination (water/air treatment) permits  a  wider range
          of remediation.   It also provides for recycling of  the contaminated
          condensate produced  during on-site regeneration  so that a complete
          clean-up is achieved.

     C.    The on-site  regeneration of  activated carbon  eliminates the  cost
          and problem of exchanging or disposing of the spent filter material
          thus eliminating the "transfer of pollution"  so  common   in today's
          treatment strategy.

     D.    Concentrations of both the discharge water and  air  were  well  within
          the regulatory requirements.

     E.    Residual soil  and water concentrations at  the  site  have continued
          to  decrease.     It  is  certain  that complete remediation will  be
          obtained with  this process in the very near future.

     F.    Precipitation  decreases the efficiency  of a  soil-vapor extraction
          system.

     G.    An integrated  extraction system is effective  in both sandy and  clay
          soil.   However,  clay soil will be more expensive  to treat.
     SUWARY

     The  successful   remediation  of   sites,   containing   both  soil-vapor  and
     groundwater contaminant  problems,  can be  best accomplished by the integrated
     extraction and treatment system described  in this paper.   Dozens of projects,
     throughout  West  Germany  and  Western  Europe,   have  proven  it  to be the
     economical method to remediate sites contaminated with volatile organics.
                                        115

-------
ULTROX
INTERNATIONAL

S435 South Anne Street
Senta Ana. California 927O4
TEL: (714] 545-5557
FAX: (714) 557-5396
            ULTRAVIOLET  RADIATION/OXIDATION

               OF ORGANIC CONTAMINANTS   IN

          GROUND, WASTE AND DRINKING WATERS



                      By: Jerome T. Barich
             Presented at the EPA'a SECOND FORUM ON


        INNOVATIVE HAZARDOUS WASTE TREATMENT TECHNOLOGIES:


                   DOMESTIC AND INTERNATIONAL



                        May 15-17,  1990
                  An ON-SITE Toxic Control, Inc. company

                                116

-------
                     TABLE OF CONTENTS
                                                            Page
INTRODUCTION	  1

DESCRIPTION OF THE UV/OXIDATION PROCESS  	  2

ULTROX® UV/OXIDATION EQUIPMENT	  2

APPLICATION OF UV/OXIDATION 	  3

CASE STUDY:    EPA SITE PROGRAM - LORENTZ  BARREL AND
               DRUM SITE,  SAN JOSE,  CALIFORNIA	  4

TYPICAL O&M COST GUIDELINES	  5

PROCESS INFLUENT DESIGN CONSIDERATIONS	  5

SUMMARY   	  6



Table I.     Compounds Treated with  UV/Oxidation  	  7

Table II.     Applications  of ULTROX® Technology 	  7

Table III.   Typical O&M Costs by Compound Type 	 10

Table IV.     Process Influent Design Consideration	 10


Figure 1.     Isometric View of System	 11

Figure 2.     System Flow Diagram 	 12
                                 117

-------
INTRODUCTION

Regulatory  standards  for environmental  clean-up  are  becoming
increasingly stringent.  The U.S. EPA requires greater reliance on
permanent remedies.  Local air quality  districts  are  refusing to
allow water-borne contaminants to be transferred to the air.  The
more conventional technologies (such as  air stripping or activated
carbon) which  transfer contaminants from  one physical  state to
another are becoming less well suited for solving today's problems.
In  order  to keep  step  with these  tougher  standards,  Ultrox
International  has  developed  its ULTROX®  UV/oxidation technology
for the treatment of organic contaminants in water.

The ULTROX®  UV/oxidation  process is currently  being applied to a
wide range of groundwater, wastewater,  drinking water  and process
water problems.  Chlorinated solvents,  BTEX compound,  pesticides,
PCBs, phenols,  and many other organic compounds, as well as BOD and
TOC, can be  economically reduced to mandated levels.

Conventional chemical oxidation has been used in the treatment of
various waters polluted by organic chemicals for a  number of years.
Hydrogen peroxide with  a catalyst such as ferrous sulfate (Fenton's
Reagent)  has been  used  for oxidizing  phenol  and  other  benzene
derivatives.   Processes  utilizing  iron catalyzed  peroxides and
chlorine compounds are attractive in that they utilize relatively
low-cost treatment equipment.  The disadvantages of these processes
are  that they  can  attack  only  a  limited  number  of refractory
organics, and they  produce  iron  sludges or chlorinated organics.

Commercial  application of the  patented  ULTROX® process utilizing
ultraviolet   light   catalyzed   ozone   plus   hydrogen   peroxide
(UV/oxidation) as a  water treatment technique is rapidly expanding.
It  offers  a means of destroying on site many  of  the  toxic water
soluble organic chemicals found today in groundwater,  wastewater,
leachate and drinking water supplies.

This  paper  describes the experience  of Ultrox  International in
applying the patented ULTROX® process to the full-scale treatment
of organic chemicals in wastewaters, drinking waters, leachates and
groundwaters.  Ultrox International was  issued  a process patent in
1988  covering  the  application  of  UV  light, ozone and  hydrogen
peroxide to  the treatment of a broad range of organic  compounds in
water.
                                118

-------
DESCRIPTION  OF THE  UV/OXIDATIQN  PROCESS

Ultraviolet  light,  when combined with  O3  and/or H202 produces  a
highly oxidative environment significantly more destructive  than
that created with O, or  H,02 by themselves  or in combination.
UV light significantly enhances ozone or H2O2 reactivity by:

    I.       Transformation of 03  or  H2O2 to highly reactive
             (OH) radicals

   II.       Excitation of the target organic solute to a higher
             energy level

  III.       Initial attack of the target  organic by UV light

The importance of the conversion of the ozone or H202 to  (OH) "can be
more easily understood after studying the relative oxidation power
of oxidizing species. Hydroxyl radicals have significantly higher
oxidation power than either hydrogen peroxide or ozone.

                                 Oxidation       Relative
                                 Potential       Oxidation
       Species                     Volts          Power*

       Fluorine                    3.06             2.25
       Hydroxyl Radical (OH)"      2.80             2.05
       Atomic Oxygen               2.42             1.78
       Ozone                       2.07             1.52
       Chlorine Dioxide            1.96             1.44
       Hydrogen Peroxide           1.77             1.30
       Perhydroxyl Radicals        1.70             1.25
       Hypochlorous Acid           1.49             1.10
       Chlorine                    1.36             1.00

          * based on chlorine as reference (= 1.00)
ULTROX® UV/OXIDATION  EQUIPMENT

The equipment components  in  ULTROX® systems:   (1) have very  few
moving parts, (2) operate at low pressure,  (3) require  a  minimum
of maintenance,  (4)  operate full-time or intermittently  in either
a continuous or batch treatment mode, (5) utilize energy efficient,
low temperature,  long life UV lamps, and (6) can employ the use of
a micro-processor to control and automate  the  treatment  process.


A typical process schematic and a reactor side view are illustrated
in Figures l and 2.  The process incorporates highly reliable, time
                                119

-------
tested  components to  ensure  high equipment  utilization  rates
(minimal down time) and continuous operation at performance levels
satisfying or exceeding performance specifications.


APPLICATION  OF UV/QXIDATION

UV/oxidation equipment developed by Ultrox in recent years has been
used to treat a wide variety of waste streams.  Table I lists toxic
compounds found in groundwaters, wastewaters and  drinking waters
that have been successfully treated with the ULTROX® process.  Case
histories from commercial systems and pilot and design studies for
applications in private industry and municipalities  are presented
in Table II.

Modular  equipment  designs  are  used   in  ULTROX's  commercial
installations.   Reactor size  varies  from  325  gallons to  5,200
gallons.  Ozone generators range from 21  Ib. to 250  Ib.  per day.
In  most  applications,  hydrogen  peroxide  is  used  with  ozone.
Commercial ULTROX® systems now treat flow rates  from 1.0 gpm to 950
gpm.

Generally,  laboratory treatability  studies  and/or on-site  pilot
plant  studies are  performed  to  determine the  feasibility  of
treating the water with UV/05,  UV/H2O2,  or UV/0,/H202.   The ULTROX®
data base on  the  treatment  of a wide range of compounds  has now
expanded to where a  minimum amount of  laboratory or  pilot  plant
testing is required to design a commercial system for some priority
pollutants.

Full-scale  systems,  in  most  cases, are  automated  using  micro-
processor control.  Generally, the  system is monitored (once per
shift or once per day).   We have built commercial systems that are
monitored only once or twice a week.  The systems  are designed to
operate  in  a batch or  continuous  mode depending upon treatment
requirements.

In  a  number of cases,  UV-oxidation is  used as a component  in a
larger treatment train.   For example, at wood treating sites prior
to  the  UV-oxidation  treatment,  the  wastewater  or  groundwater
requires  upstream processing  to break  oil/water emulsions  and
remove  suspended  matter.  Process  influent  design considerations
are listed in Table IV.
CASE STUDY:    EPA SITE PROGRAM - LORENTZ BARREL AND DRUM SITE,
                 SAN JOSE, CA

The  EPA  has established  a  formal  program to  accelerate  the
development,  demonstration,   and   use  of  new   or  innovative
technologies to be used  in site cleanups. This program, called the
                                120

-------
Superfund Innovative Technology Evaluation (SITE) program, has four
goals:

•    TO identify  and,  where possible, remove  impediments  to the
     development and commercial use of alternative technologies.

•    To conduct  a  demonstration  program of  the more  promising
     innovative  technologies  for the  purpose  of  establishing
     reliable  performance  and   cost   information  for   site
     characterization and cleanup decision-making.

•    To develop procedures  and  policies  that  encourage selection
     of available  alternative  treatment remedies   at  Superfund
     sites.

•    To structure a development  program that nurtures  emerging
     technologies.

Ultrox International applied and was selected for inclusion in the
SITE Program and  in 1989 demonstrated  its UV/oxidation technology
at  the Lorentz  Barrel  and Drum Superfund  site  in San  Jose,
California.

The Lorentz site was used for drum recycling for nearly 40 years.
Over this  period  of time,  the site received drums  from  over 800
private companies, military  bases,   research  laboratories,  and
county agencies  in California  and Nevada.  Drums arrived  at the
site containing residual aqueous wastes,  organic solvents,  acids,
metal  oxides,  and  oils.   The  groundwater  beneath   the  site was
contaminated with a number of chlorinated solvents,  chlordane,
toxaphene and PCBs.

An ULTROX®  P-150  pilot plant was moved  in  on  February 21, 1989.
Thirteen  (13) tests were conducted between  February 24 and March
9, 1989,  on extracted groundwater  from the site.  The SITE Program
evaluation criteria were:

•    Compliance   of  the  treated  water  with   regulatory
     requirements

•    Efficiency of the off gas  (DECOMPOZON™)
     Treatment Unit in Removing Ozone and VOCs

The organic compounds analyzed during the evaluation were:

•  Aqueous Phase:
          TCE            1,1-DCA        1,1,1-TCA

•  Vapor Phase:
          Vinyl Chloride      1,1-DCE        1,1-DCA
          1,2-DCE             1,1,1-TCA      1,1,2,2-TCA
          TCE                 Acetone        Benzene
                                121

-------
Upon completion of the pilot plant demonstration and evaluation of
the analytical  data from  samples  of the  influent and effluent
(treated) water and the off gas  into and out of the DECOMPOZON™
unit,  it was concluded in the EPA's  Technology Evaluation Report
that:

•    The treated ground water  met the applicable NPDES
     standards for  discharge into Coyote  Creek

•    There were no  volatile organics  detected in the exhaust
     from the DECOMPOZON™ unit

•    The DECOMPOZON™ unit reduced ozone in the ULTROX® reactor
     off gas to levels  less than 0.1 p.p.m.  (OSHA) standards.

The U.S.  EPA's Technology Evaluation Report fully explains the
details of this study.

The remainder of this paper is focussed  on  full-scale  applications
of the ULTROX® process as shown in Table  II,  Cases  A through  L.


TYPICAL  O&M  COST GUIDELINES

Operating and maintenance costs can vary  widely depending upon:

   Electrical Power    .  Type and Concentration of Contaminants
   H202 Cost            .  Concentration of  Competing Species
   Oxygen Cost         .  Treatment Goals

O&M costs  can vary from as low  as $.10/1000 gallons  in municipal
drinking water applications to as high as $.10/gallon  for leachate-
type or  highly  concentrated organic-laden waste waters.  See Table
III.
PROCESS INFLUENT  DESIGN CONSIDERATIONS

UV/oxidation  is  used often  as  a stand-alone treatment  process.
Many  times,  however,  other  characteristics  of  a waste water  or
ground water dictate that UV/oxidation be combined into a treatment
train with other technologies to obtain the desired result.

At one wood treating site,  oil removal and filtration precedes the
ULTROX® process.  Metals precipitation and free hydrocarbon removal
equipment  is  upstream of  an ULTROX® system  in a marine  terminal
evaluation.  pH adjustment of the influent to the ULTROX® process
can,  at times, result in  faster reaction times  and lower oxidant
requirements.  Table IV lists several parameters that could impact
UV/oxidation  efficiency and  necessitate the  use  of  additional
processes.
                                122

-------
SUMMARY

The  ULTROX®  process  is  used  commercially  in  a  variety  of
applications.  We  expect this range of applications  to increase
greatly based on  the results of many recently completed laboratory
and pilot  plant studies.   The process  destroys or  detoxifies
contaminants ON SITE without the generation of sludges or the use
of adsorbents  that require additional  treatment.   UV/oxidation
process economics are competitive on a life cycle cost basis in a
direct comparison with these older processes that have the residual
problems cited above.  Yet, these economic comparisons usually do
not assign any credit for the fact that residual  liability is not
associated with  UV/oxidation.   Any economic analysis  that would
assign credit for the elimination of residual liability (certainly
this must be considered  in the case of real estate transfer) would
further improve the competitiveness of the technology.
                              123

-------
   TABLE I:  COMPOUNDS TREATED  WITH  UV/OXIDATION
ETHERS

BTEX

PHENOL

TCE

PCE

DCE

POLYNITROPHENOLS

KETONES

VINYL CHLORIDE
PESTICIDES

CITRIC ACID

TCA

DCA
MeCl2

CRESOLS

PCBS

PCP

TNT
AROMATIC AMINES

COMPLEXED CYANIDES

POLYNUCLEAR AROMATICS

DIOXINS

HYDRAZINE

RDX

1,4 DIOXANE

EDTA

HYDRAZINE
     TABLE II:  APPLICATIONS  OF ULTROX® TECHNOLOGY
          CASE A:

          SOURCE:
          CONTAMINANT:
          GOAL:
          RESULT:
          DISPOSITION:
 GROUNDWATER

 METAL PLATER
 450 p.p.b.  COMPLEXED CYANIDES
 <20 p.p.b.
 < 2 p.p.b.
 P.O.T.W.
          CASE B:

          SOURCE:
          CONTAMINANT:
          GOAL:
          RESULT:
          FLOW:
          DISPOSITION:
 WASTEWATER
 CHEMICAL MANUFACTURER
 BOD5   70 p.p.m.
 <30 p.p.m.
 BOD5  <20 p.p.m.
 650 g.p.m.
 SURFACE WATER
          CASE C:

          SOURCE:
          CONTAMINANT:
          GOAL:
          RESULT:
          FLOW:
          DISPOSITION:
 WASTEWATER

 PETROCHEMICAL MANUFACTURER
 PHENOLS >135 p.p.m.
 PHENOLS <5.0 p.p.m.
 PHENOLS <1.0 p.p.m.
 70 GPM
 BIOTREATMENT
                              124

-------
TABLE II: APPLICATIONS  OF ULTROX®  TECHNOLOGY
CASE D:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
DISPOSITION:
WASTEWATER

PHENOLIC RESIN PLANT
BIOTREATMENT EFFLUENT PHENOLS >3.0 p.p.m.
14 p.p.b.
 8 p.p.b.
25 GPM
P.O.T.W.
CASE E:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
DISPOSITION:
GROUNDWATER

AUTOMOTIVE FOUNDRY
CHLORINATED VOCs >6000 p.p.b.
<20 p.p.b.
< 5 p.p.b.
200 GPM
SURFACE WATER
CASE F:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
DISPOSITION:
GROUNDWATER

SEMICONDUCTOR MANUFACTURER
12.5 p.p.m. BTEX
<20 p.p.b.
< 5 p.p.b.
5.0 GPM
P.O.T.W.
CASE G:

SOURCE:
CONTAMINANT;
GOAL:
RESULT:
FLOW:
DISPOSITION:
GROUNDWATER

PHOTOCOPIER MANUFACTURER
TOLUENE + TCE >6000 p.p.b.
99% DESTRUCTION
99+%
150 GPM
STORM SEWER
CASE H:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
DISPOSITION
WASTEWATER

PHARMACEUTICAL FIRM
COD >700  p.p.m.
COD <250  p.p.m.
COD <200  p.p.m.
7.0 GPM
P.O.T.W.
                   125

-------
TABLE II: APPLICATIONS OF ULTROX* TECHNOLOGY
CASE I:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
DISPOSITION:
PROCESS WATER

STEAM CONDENSATE
TOC >1000 p.p.b.
TOC <  50 p.p.b.
TOC <  30 p.p.b.
25 GPM
BOILER FEED WATER
CASE J:

SOURCE:
CONTAMINANT:

GOAL:
RESULT:
FLOW:
WASTEWATER

SPECIALTY CHEMICAL MANUFACTURER
150+ p.p.a.  1,4  DIOXANE  IN BIOTREATER
EFFLUENT
90% - 95% DESTRUCTION
98+%
10 GPM
CASE K:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
MUNICIPAL  DRINKING WATER

MANUFACTURING FIRMS
PCE  25 p.p.b.
< 5 p.p.b.
0.5 p.p.b.
950 GPM
CASE L:

SOURCE:
CONTAMINANT:
GOAL:
RESULT:
FLOW:
GROUNDWATER

WOOD TREATING FACILITY
6000 p.p.b. PHENOL
<200 p.p.b.
<100 p.p.b.
40 GPM
                    126

-------
   TABLE III:  TYPICAL  O&M COSTS BY COMPOUND TYPE
                     S/1000 GALLONS*

TCE         BTEX         BASE NEUTRALS        KETONES
PCE         _r	         	=	          -
$.45         $.80           $1.10             $1.50

* Costs based on influent contaminant concentration * 5.0 mg/1
  TABLE IV:  PROCESS INFLUENT DESIGN CONSIDERATIONS
         OPACITY                       IRON
         SUSPENDED SOLIDS               MANGANESE
         FREE HYDROCARBON               ALKALINITY
         HARDNESS                      pH
                            127

-------
ro
CO
     ISOMETRIC VIEW OF
       ULTROX SYSTEM
TREATED OFF GAS
                        CATALYTIC OZONE
                        DECOMPOSER

                       REACTOR OFF GAS
          OZONE GENERATOR
                        DRYER
                                                                      TREATED EFFLUENT
                                                         ULTROX
                                                     UV/OXIDATION REACTOR
                                                     HYDROGEN PEROXIDE
                                                     FROM FEED TANK
                                 FIGURE 1

-------
                             ftotemtter
 Ozone
from OZOM
generator
                                     1
                                                                              from SMtam
                                                                              Ground-Water
                                                                              Uonttoring Weta
                                        f    Effluent
                                         1	 Sample Tap
             Hydrogen Peroxide
               Feed Tank
Contaminated Water
Feed Tank (Bladder)
                                         FIGURE 2

                                   ULTROX*  SYSTEM
                                     FLOW DIAGRAM
CREATED: 11/4/88 ) REVBED: 02/02/89 |  PIOT.CRF

-------
The   Liaxrgi — Deoonteir ra— Piroc<
Wet   Mechanical   Site  RemecL i si t i on
                   Eckart   F.   Hilner
                              Philadelphia
                                May 1990
                            130

-------
1.   Introduction







2.   Soil washing strategy







3.   Process description







4.   Comparison to other soil decontamination systems







5.   Cleaning results







6.   Costs
                          131

-------
Introduction

Contaminated soils are destined to become a major problem for
all densely populated countries. For instance this problem is
one of the most urgent political and technological challenges
facing Germany today.

The protection of the environment from pollution and the reme-
dial treatment of contaminated soil is an important objective
for the years to come.

In future natural land will hardly be available for new hou-
sing areas or industrial complexes. Old buildings give way to
new buildings. For new industries only old industrial sites
will be available.

These old sites often are highly contaminated being partly
dangerous for the inhabitants.

Before building up new industries in many caces the site has
to be decontaminated including soil and groundwater.

Because landfill capacities in Europe are more and more
limited, land disposal prohibitions will soon require that
these wastes be treated using best demonstrated available
technology.
Soil washing strategy

An important strategy in operating a soil washing system de-
pends on the fact/ that contaminants are primarily adsorbed
by clay and silt-minerals, hydrous oxides, and organic mat-
ter and are therefore concentrated in the fines fraction of
the soil. The cation exchange capacity of clay minerals
                         132

-------
enables large amounts of certain elements, inorganic and or-
ganic matter, to be adsorbed. On the other hand contaminants
may stick on the coarser material, which may be scrubbed by
either water alone or mixtures of water with some detergents.

In general separating organic and inorganic contaminants from
the soil is a purely pysical procedure.

The main problem is the input of sufficient energy into the
soil to liberate the contaminants and to transfer these into
a fine material suspension.

Lurgi's Deconterra-Process was developed using extensive ex-
perience in the field of wet mechanical processing of mine-
rals, especially of ores and salts, as well as of sludge from
loaden waters.

The separation into a clean soil fraction and a concentrate
of contaminants takes place in an attritioner drum. Conta-
minated soils undergo a severe scrubbing process, which in-
termingles coarse and fine material, using the coarser pebbles
like grinding bodies with a net energy input of 4 to 16 kW
hr/ton of throughput, during which the rotation speed of the
drum is between 50 and 90 % of the critical speed of

                "  42'4   (min'1)
The suspension of the drum discharge will be screened in
coarser and finer fractions. These fractions are further
treated.
                     133

-------
Process description

The Deconterra-process for soil washing is illustrated in the
following flow sheet (picture 1).

The contaminated soil is reclaimed/ classified at 300 mm and
screened into a 0 - 100 mm fraction and a 100 - 300 mm frac-
tion.

The coarse fraction (100 - 300 mm) is crushed to under 100 mm
in a crusher and, together with the fine fraction, is trans-
ferred to a wet attritioner drum.

In the attritioner drum, the bulk of the adhering contaminants
is rubbed off with a controlled input of energy and first of
all suspended in the liquid and then bound adsorptively to
the fine particles of the soil.

The energy required for the attrition can be matched to the
type of soil and the character of the contamination, up to 16
kW hr/ton being transferred to the material.

The material discharged from the drum is screened in several
stages.

The fraction over 20 mm is returned via the braker cycle once
again to the attritioner drum.

The 1 - 20 mm fraction is either discharged from  the process
as purified end product or subjected to a gravimetric sizing,
the heavy material being discharged clean, whereas the light
material, such as contaminated coarser coal, tar  and wooden
particles, being obtained as a contaminants concentrate.
                         134

-------
Of the fraction under 1 nun, the finest fraction is cleared
from mud and transferred to a hydrocyclone, the overflow of
which is concentrated.

The cyclone underflow, together with the de-slimed coarse
fraction of the classifier, is introduced into the 2nd attri-
tion step and subjected to a further purification.

The material discharged from the 2nd attrition drum is pumped
to a froth-flotation system, the special reagents of which
are matched to particular requirements of the contaminants
involved.

The suspension discharged from the flotation system contains
the purified material, which is subsequently dewatered to a
residual moisture content of about 15 - 20 % and can be dis-
posed or piled as cleaned product.

The contaminants are concentrated in the foam discharged du-
ring the floatation. This foam reaches the thickener for pre-
dewatering. The thickener underflow is brought to a moisture
content of about 30 % in a dewatering step and represents the
bulk of the so-called contaminants concentrate, which con-
tains contaminants removed from the bottom in a concentrated
form.

The contaminants concentrate is composed of the following
materials, which are separated in the process:

1.   Light material from the gravimetric sizing (such as
     wood, tar fragments, coal, coke, etc.)

2.   Overflow from the hydrocyclone, esentially the fines
     portion of the soil with the adsorbed contaminants.

3.   Froth from the flotation.
                     135

-------
Fractions 2 and 3 are obtained together as filter cake.

It has turned out that the process can also effectively clean
the fine grain fraction. The process therefore is also able
to decontaminate soils with a high proportion of fine grain
(less than 63 microns in diameter). Depending on the composi-
tion of the soil and the nature of the contamination, the
yield of decontaminated soil is between 65 % and 85 % of the
untreated soil.

The scrubbing water, resulting from the Deconterra-process,
is slightly contaminated, so that the bulk of the process
water can be pumped back into the plant. To avoid accumula-
tion of the contaminants in the water, a portion of it is
diverted to a water processing stage, in which the very fine
solid materials, as well as the dissolved contaminants are
reduced to the required minimum amount.
Comparison to other soil decontamination systems

Besides soil washing systems thermal methods suit very well
for organic contaminants to be destroyed by incineration. On
the other hand inorganic contaminants can be removed only
partly by incineration. Heavy metals can be solidified or
leached by extraction methods. The costs of treatment inclu-
ding exhaust gas cleaning and residue disposal will be very
significant.

Another alternative for decontamination of soil is the bio-
logical treatment. But this methods is limited to some simple
build up organic compounds and to a very small amount of in-
organic compounds.
                     136

-------
In general all various soil washing systems operate on the
principle that contaminants are concentrated in the fine com-
ponents of the soil. Materials coarser than for instance
100 y m are discharged essentially free of contamination. The
systems differ by the method of putting energy into the con-
taminated soil and by using different washing agents.

Two companies from the Netherlands use hydroclones and scrub-
bing with detergents and oxidants.

A dutch company and a german company use a high pressure wa-
ter beam to remove the contaminants from other soil partic-
les. Problems exist in high contents of fines in the soil.

One german company is using a vibrating screw for energy
transfer while another german company uses a centrifugal se-
parator.

The Lurgi-Deconterra-Process is using only water as extrac-
ting agent without addition of detergents, solvents, acids,
bases or similar materials. Included into the process is the
further treatment of the + 1 mm soil fraction.

After classification and the second attritioner step the
fraction 20 um to 1 nun is purified by pneumatic flotation.
During this step an optimum of different surface characte-
ristics of the soil components and the actual contaminants
can be achieved by dosing agents into the three flotation
cells.

This is a major difference and an essential advantage of the
Lurgi-Deconterra-Process compared to other soil washing pro-
cesses, which are mainly based on classification as a sepa-
rating process and not so much on a sorting process.
                      137

-------
 Cleaning results

 Lurgi has investigated and purified  various samples from many
 different contaminated soils,  for  instance from sites of old
 coking plants, old ammunition  plants,  various chemicr1  indu-
 strial sites and contaminated  waste  locations.

 The following tables give some results out of various test-
 work.

      material from different sites of  old coking plants in
      the Ruhr Area and the Saar Area,  Germany
Component    contaminated soil   decontaminated
            (mg/kg)           soil  (mg/kg)
Dutch reference
 (mg/kg)
  A      B
Naphthalene
Phenanthrene
Anthracene
Fluor anthene
Pyrene
Benzo-a-pyrene
I Polyaromatic
hydrocarbons
(PAH)
Ł Hydrocarbons
Cyanide (total)
18
237
40
220
170
49


1357
4000
94
- 0,2
1,5
- 0,2
0,3
- 0,2
- 0,2


7,1
30
17
0,1
0,1
0,1
0,1
0,1
0,05


1,0
100
5
5
10
10
10
10
10


20
1000
50
                          138

-------
Component     contaminated soil   decontaminated
             (mg/kg)            soil  (mg/kg)
                              Dutch reference
                               (mg/kg)
                                A       B
Arsenic
Lead
Copper
Nickel
Zinc
Vanadium
Chromium
Mercury
435
80 - 250
60 - 200
30
330
50 - 70
1400
700
22
< 50
< 50
< 20
<100
< 40
130
22
20
50
50
50
200

100
0,5
30
150
100
100
500

250
2
 During  the second world war  several large  ammunition works  in
 germany produced explosives/  mainly Trinitrotoluene TNT.  To-
 day these  old sites are highly contaminated  with Nitrotoluene,
 Soil from  some of these works was treated  in a Deconterra-
 pilot-plant with good results.
component
contaminated
soil  (mg/kg)
decontaminated
soil  (mg/kg)
Nitrotoluene
 106,000
   0.34
                       139

-------
Several possibilities are available how to treat the conta-
minants fines concentrate. The non satisfying method is ha-
zardous wast landfilling.

The thermal treatment of the concentrates in incineration
plants may be a good solution specially for organic contami-
nants. For inorganic contaminants solidification is a solu-
tion. In Germany thermal methods bear a lack of acceptance by
the population and therefore are very difficult to prevail.

At present Lurgi develops a process to destroy organic conta-
minants from the Deconterra washing concentrate by pressure
oxidation in an autoclave.
Costs
The economics for a Deconterra-system operating in the United
States of America are very difficult to estimate.

The costs are not only caused by the process itself but also
to a large extend by the local conditions, the contaminants,
the analytical expense, the safety conditions and environmen-
tal restrictions.
                      140

-------
Based on a Deconterra plant, which is a modular construction
and semimobile with a capacity of 25 t/hr wet material or
80.000 tons/year the treatment fee per ton is in the range of
$80-90. Excluded from this amount are the costs for re-
sidue disposal.
                     141

-------
tss
            CONTAMINATED SOL (FEED)
                                                                      ION
                                                                    EXCHANGE ACTIVATED
                                                                            CARBON FILTER
             LIGHT MATERIAL
             (CONTAMMATEO)
HEAVY MATERIAL
   (CLEAN)
CONTAMINATION
     DECONTERRA Process

-------
NBM BODEMSANERING
 Developments and operating experience in thermal soilcleaning

                  NBM Bodemsanering B.V.
             P.O.  Box 16032, 2500 BA The Hague
                       Netherlands
           Phone 31 70 3814331 / Fax 31 70 3834013
            Ir. H.J.  van Hasselt/Ir.  A. Costerus
    1.     Abstract

    2.     Introduction

    3.     System description

    3.1    Process

    3.2    Equipement

    4.     Operational Costs and Production factors

    5.     Legislation

    6.     Development
                      143

-------
 NBM BODEMSANERING
Developments and operating experience in thermal  soilcleaninq


1.   Abstract
This paper discusses the development and operating experience of
an indirect heated  soil cleaning plant.

In The Netherlands, the soil pollution in the village of Lekker-
kerk,  was one of  the news highlights in 1979.
As a direct result  of  this,  the Dutch  government decided to start
a soil sanitation project for The Netherlands.

NBM Bodemsanering  B.V.,  subsidiary  of NBM Amstelland  N.V.,  the
second building  and  construction  group in The  Netherlands,  has
developed an  indirect  heated  thermal  treament  system and became
one of the  leading firms in The  Netherlands in  the cleaning of
soil and  the execution of projects for  soil  pollution.

The equipment, operational  costs,  production factors and legisla-
tion are  presented  in this paper, and have,  in 3 years of operati-
on, produces over 200.000 tons of cleaned  soil.

The gas  conditions in the  rotary tube furnace,   the  absence of
oxygen and the presence of hydrogen,  create  optimal conditions for
the cleaning of soil containing halogenated  hydrocarbons.

At the Technical  University  of  Delft, laboratory work was done to
determine the process conditions required to  bring the pollution
level   below  the  accepted  ("A" value,  <  0,1  mg organochlorine
compound/kg)  level  and  to test  the process conditions for incine-
ration. Special  attention was  paid  to the formation of  CDD and
CDF.

Based   on  the  results  of this  laboratory  work,   NBM  decided to
carry  out practical experiments.

First,  in  1989,  the incineration conditions  were tested  in the
full scale  plant in Schiedam.  While  running  the plant  with an
input   of  soil contaminated with  fuel  oil a  known amount  of  a
mixture of  trichloroethene  and  tetrachloromethane  was injected
into  the  incinerator,  and  stack  emissions  were measured.  The
destruction efficiency was higher than  99,999 %  .
                         144

-------
 NBM BODEMSANERING
Further experiments  were  carried out  in April  1990.  About  700
tons of soil  containing  a.o.  aldrin,  dieldrin,  and lindane were
processed  during a three  day test run.  In these  tests, the remai-
ning rest  concentrations  of contaminants  were below the detection
limit of 10 micrograms/kg dry solids.


2.  Introduction


In The Netherlands, the soil pollution in the village of Lekker-
kerk, was  one of the news highlights in 1979. New houses had been
built  on  an old  dumping  site,   and the  soil was  polluted with
benzene,   among  other substances. The  public was  alerted by  a
stream of  news about  other sites that were contaminated.

As a direct result of this,  the  Dutch Government decided to start
a soil sanitation project for The Netherlands.

A first general investigation was made, and the Dutch Government
published  the reference values  for soil  contamination,  known as
the ABC  values.

The government made Hfl.  2,000,000,000  available for the investi-
gation of suspected  sites,  to make plans for the sanitation, to
develop cleaning technologies,  and to  execute the first priority
jobs.

The  road  construction company  NBM,   a  subsidiary  of the NBM-
Amstelland group,  was one of the firms that started the develop-
ment of a  thermal  cleaning technology.

The system had to  meet the following requirements:
- It must  be possible to  clean any type of soil  (sand,  clay,
  peat).
- It must  be possible to remove all organic compounds and pollu-
  tants that can be volatilized  at temperatures up to 600 - 650 C
  or can be removed by pyrolysis reactions at temperatures of 300
  - 650 C.

  Thermal  process  are  not  expected  to remove  heavy  metals from
  soil. Their very high boiling  ooints  require the development of
  other cleaning process.
  NBM decided to choose the  indirect heated thermal process.
                         145

-------
NBM BODEMSANERING
 See figure  1:  schematic presentation  of  direct and  indirect
 heated thermal process.




Soil input
























_

r--^^
r i



Rotary tub* lurnac*


II 1 1 II I! || ||









_










chlmrwv












nclncntoi






\







.^-^
r^
S
n
•oil input
burntr

Schematic indirect heating
, 	 	 Rotary drum


~~*
chimney _
Inclntrttor

-H
\
          Schematic direct heating

 The principal arguments  for  the  choice  between  direct  and
 indirect heated process are given in the following table.
Direct heated
Advantage
Unlimited
heat tansfer
capacity
relative short
heating and
cooling time
of equipment

Drawback
High volume of
gas stream in
rotary kiln
High volume of
gas steam in
incinerator
Dust difficult
to remove from
gas stream
Indirect heated
Advantage
Low volume of
gas stream
Low volume of
gas stream in
incinerator
Filtration of
dust at high
temperature
possible
Drawback
Limited heat
transfer
Long heating
and cooling
time of the
equipment

                        146

-------
NBM BODEMSANERING
3.  System description
3.1  Process
When  a  soil  contaminated  with organic  compounds is  heated  the
evaporation of volatile compounds will occur. The  initial evapora-
tion  rate  will depend on  the  vapor pressure of  the  compound at
the surface of the matrix (soil),  which is  a  direct  function of
the temperature, the  physical  chacteristics  of  the compound,  and
the adsorbtive interactions  (physical  and chemical)  between  the
compounds and the soil.

The aim  of the equipment  is  to achieve a soil  temperature high
enough to obtain a vapour pressure of the compound above. 1 atmosp-
here.  In that  case the transport  in the vapour phase  will  be by
convection transport even in the pores of the soil particles.
The temperature  of  the  tube   furnace  and  the  soil  is given in
figure 2.

    Figure 2
    Temperature distribution of soil in tube
    furnace for pilot - and commercial plant
1000
    Temperature
 800
 600
 400i
 200
   0      15      30      45

            Residence time (min)

    1               3
   	 Pilot pltnt       	 Comm. plant norm*)

   	 Comm. plant max.   	 Wall ttmptratur*
60
                          147

-------
 NBM BODEMSANERING
For  compounds  which  are  not  thermostable,  thermolytic  fission
reactions occur before the vapour pressure  of  1 atmosphere will be
achieved. It  is known that  thermolylic fission  reactions  occur
for the following compounds:

- Iron cyanide complex ( Prussian blue)
- hexachloro cyclohexane
- longer aliphatic hydrocarbons
- natural humic materials
The thermolytic fission reaction of  the  natural  humic materials
plays an important role in the process in  the rotary tube furnace.

The following reaction is suspected  to take place :

Formula
                     +H2OT
Mol.
quant.  1    —> 31    +16
+COT    +H2T  +CH4T +C2H6T
                                 +5
       +2    +2    +4     +1
Weight
kg.    1   --> 0,30  +0,22  +0,17 +0,12  +0   +0,03 +0,09  +0,08

Gasprod.out of
1 kg gas volume
in liters   -->  -     +311  +100  +100   +40  +40   +80    +40


The rotary tube  furnace is a highly gastight furnace, therefore no
explosion occurs with the inflammable gases because of the absence
of oxygen. The oxygen which however could enter the system reacts
instantaneously with the hydrogen present.
The  absence  of oxygen  in  the  polluted  gas  stream  is also an
important factor  in treating chlorinated hydrocarbons.  In the gas
phase dioxins  and  dibenzofurans  could easily  be formed in  the
temperature range of 400 - 700 C ,  but  the absence of oxygen makes
these formations impossible.

The  polluted  gas   stream which  has a  small  volume  (300  -  500
nm3/hr)  is treated  in  an incinerator  at  a temperature  of  1100-
1200 C.  All  organic compounds  will be oxidized to CO-  and  H-0,
with small quantities of NO
                           148

-------
 NBM BODEMSANERING
At a  temperature of  1100  -  1200  C,  with a residual  time of at
least 1  second  and  good  turbulence  (a  small  relative  residence
time  distribution)   are  required  to  destroy  chlorinated hydro-
carbons   too,  without  forming  undesired  products  of  incomplete
combustion.
Laboratory tests were conducted  to establish the undesired forma-
tion of dioxins and  dibenzofurans in the oxidation  of  hexachloro
cyclohexane  and of drins as a function of  residence time,  tempera-
ture  and  oxygen content.  The  formation  of  1  ng  of  undesired
product   of  incomplete  combustion  from  1 gram  of chlorinated
hydrocarbon  was accepted as maximum value.

         Formation of PCDD  + PCDF by combustion of dieldrin
time /temp.
0,5 s
0,7 s
1,2 s
1,5 s
1,7 s
750° C
• • • •
0+17
• • • •
21+2700
0+0
950° C
• • • •
93+21000
7+357
0+0

1150° C
50+1100
0+20

0+0

         Formation of DCDD + PCDF by combustion of HCH
time/temp.
0,5 s
0,9 a
0,9 s
1,2 s
1,5 s
2,1 s
2,3 s
680° C
• • • *
• • * •
• • • •
• • • •
* • • •
• * • *
4500+274000
750° C
• • • •
• * « •
• • * •
* • • •
0+0
0+95

900° C
* * • •
0+95
0+19
10+465



1100° C
0+0
0+0





The values are  the  quantities  expressed  in  ng of PCDD  +  PCDF
formulated out of 1 gram Dieldrin or  1 gram  HCH.
                        149

-------
 NBM BODEMSANERING
In the commercial  plant,  the average temperature in the incinera-
tor is 1100  C, and the average residence time 2 seconds. The  stack
emissions of the incinerator meet the limits of the licenses.
Before the oxidation of the polluted gas stream out of the rotary
tube furnace the gas stream has  to be filtrated from dust. Other-
wise dust will be melted and settled down in the incinerator.
High temperature gas filtration  is necessary because cooling down
the polluted gas  stream  will  condense  the high  boiling organic
pollutants as polyaromatic hydrocarbons.
Ceramic pipes filtrate the gas stream at temperature levels  above
550  C. The  filtered dust is clean of  the  organic  pollutants if
the temperature has been high enough.
3.2  Equipment
3.2.0  Pilot  plant

Before realising a full scale plant  a pilot plant was built in  the
first half of 1984. It had a capacity of 0.5 tons of soil/hour.
In the middle of 1984 a great number of test were carried out with
soil originating from former gasworks sites.
Figure 3 shows some characteristic  results of  testing  at various
temperatures.
    Sou Mfflo«r«tur*
  1000
  800
  800 h - -
CN content of polluted soil before
and after cleaning as a function of
the furnace temperature.
  400]	
  200i	,
                        1000
                            150

-------
NBM BODEMSANERING
The  pilot  plant  tests  demonstrated,  that  the  selected  system
yields to the values  of  the Dutch Government, that  is  to  say near
the detection limit. Accordingly, this system formed the basis for
the concept developed for a production plant.
In general  it must be stated that no plant can be so designed that
it  is  an optimal design for  any  type  of  soil  and  any  sort of
pollution.   The  basic principles  -  evaporating in  an  indirect
heated rotary tube and oxydation of the pollutants in an incinera-
tor  will  always apply.   However,  the  optimum  design will show
differences,  depending  on the specific conditions  for any given
proj ect.

Plant of  Schiedam
The design  was made for:

maximum 13  tons/hour
moisture  content 25%
energy content 1200 kJ/kg.d.s
any type  of soil
continuous  operating, 7000 hours a year
3.2.1  Commercial plant


The commercial  plant consists of the  following items:


1.  Soil input

The polluted soil,  from which  very  coarse  pieces,  exceeding  100
nun, have been removed, is loaded into a  feed unit.


2.  Feed unit

The soil  coming  from the bin  drops into a picking belt,  where
coarse pieces and non-ferrous metals  may be removed. Then the soil
is moved upwards by a conveyor belt to the dryer. A magnet mounted
above the belt  removes ferrous metals as much as possible from the
soil.
                         151

-------
 NBM BODEMSANERING
3.   Dryer

The dryer comprises mainly a rotary drum having a diameter of 3 m
and a length of  21,5  m.  The drum is heated externally, that is
indirectly,  by  means of hot gases. The heat sources  for  the dryer
are:
- burner gases  exiting from the indirect heated rotary tube furna-
  ce,
- the  flue gases  from the incinerator, and
- auxiliary  burners.


4.   Screen

When leaving the  dryer, practically all of the free water has been
evaporated from the soil. At this stage the structure of the soil
is such that fragments of stone, rubble and pieces of wood may be
removed  by screening.  This is necessary because the transportation
and sealing  means of the rotary tube furnace cannot cope  with very
coarse pieces.


5.   Transfer to the tube furnace

The fine  fraction  passing  the  screen,  together with  the  dust
filtered from  the exhaust  gas  from the  dryer is  passed  through a
screw  conveyor  to an elevator. The elevator carries  the  soil into
an intermediate bin, where it is introduced into the  tube furnace.


6.   Rotary tube furnace

The dried and  screened soil is  heated  in  the  tube  furnace  to a
temperature  between 450 and 600  C. The achievable maximum tempe-
rature of the  soil  and the residence time of the soil in the drum
are, obviously, a function of the quantity of soil introduced per
hour.  The tube furnace is provided with gas-sealing  gaskets and
seals.
                              152

-------
 NBM BODEMSANERING
7.  Soil cooler

The cooler consist of  a  rotating  drum.  A  plurality  of pipes  are
located inside the drum, through  wich cooling  air is  blown.  The
cleaned soil  introduced into the cooler is  cooled to approximately
150  C. The air leaving the cooler has a  temperature of approxima-
tely  250   C  and  is used  as  preheated combustion  air for  the
burners of the tube furnace and the dryer.


8.  Mixer humidifier

The soil is continuously humidified in a mixer. The  soil leaving
the mixer has a temperature of  approximately 50  C and a moisture
content that  can be adjusted  in advance. A conveyor  belt carries
the soil to a bin.


9.  Treatment of the gas leaving the dryer

The gases  exhausted  from the dryer  include mainly  steam,  light
pollutants evaporated from the soil and air sucked in via the feed
mechanism  and  some  of  the  seals. This gas  contains  a  certain
quantity of dust which  is removed from the  gas stream by a filter.
The dust collected in the filter is fed back to  the main stream of
soil.

Depending on  the type of  the  soil  and  the  type  of the pollutants,
the gas stream from the  dryer  may be processed by either  one of
two routes.

Steam condensation
By lowering the temperature of  the gases exiting in the dryer the
major part of the steam is condensed in a condenser. The condensed
water is passed to a water treatment unit.

Direct exhaust to the incinerator
If it is not desirable  to condense the steam  from the dryer,  the
dryer effluent passes  directly  from the  filters to the incinera-
tor.  The gas  stream to  the incinerator is now much more substanti-
al and so is  the energy consumption.
                         153

-------
 NBM BODEMSANERING
10.  Treatment of  the gas leaving the tube  furnace

The  gas  stream from  the tube  furnace comprises  the  following
groups of substances:
- pollutants evaporated from the soil.
- moisture evaporated from the soil.
- products  of pyrolysis  of  organic substances  present  in  the
  soil.
- inert  gas  entering  into the  gas  stream through  the  flushing
  device of  the ceramic  filter  located on the  line  between  the
  tube furnace and the incinerator.

The temperature of the gases leaving  the tube furnace is a functi-
on of  the selected  soil temperature.  It  has  the same order  of
magnitude as  the  temperature  of the existing soil.  The  gaseous
mixture is then passed to the ceramic filter unit.


11.  Ceramic filter unit

The ceramic filter unit comprises the following parts:
a. A settling chamber where the coarser dust particles  contained
   in the gas stream and entering the chamber are pre-separated by
   settling.

b. Two parallel filter units,  each  provided  with valves and  an
   exhaust fan. The  gases pass from  the settling chamber  through
   one or both of  the filters and go  on  to  the incinererator.


12.  Incinerator

All the  gases  having been in  contact with the  polluted soil  are
burnt in the incinerator. If they are burnt in the incinerator for
a sufficient period of time at an adequately high temperature and
with an adequate oxygen percentage, all  pollutants will be conver-
ted into non-toxic compounds such as  H-0, C02 and N2«
The composition of the gas stream fed  to  tne  incinerator depend
greatly on the composition  of the soil, its  moisture content  and
the type and quantity of the pollutants.
                            154

-------
 NBM BODEMSANERING
A number of variables  are known:
a. The plant is operated with or without the condensers.
   The amount  of  steam,  and thus the  fuel consumption,  is very
   much dependent  on this variable.
b. Type and quantity of the pollutants.
   All pollutants  in the  soil  will,  through the gas phase, end up
   in the incinerator. The  energy  contents  of the substances will
   be more or less positive.
c. The quantity of organic matter present in the soil  (besides the
   pollutants).
   Under  the  process  conditions  in the  tube  furnace,  organic
   matter  will be  broken  down  to  carbon,  gases  and  tar-like
   materials,  which will  be  exhausted  to  the  incinerator.  The
   energy contents of  these materials may be considerable.
d. The final temperature of the soil in  the furnace.
   The pyrolysis process  described above will also be influenced
   by the process  temperature.

Analysis of  the above parameters shows  that the gas stream ente-
ring  the  incinerator  is  subject  to  considerable  variations.  If
requirements are established for the residence time and/or the end
temperature of the gases  in  the  incinerator, the quantity of soil
introduced into the installation must be adjusted as a function of
the process conditions.


13.  Exhaust

When the installation  is  in normal operation,  practically all of
the gases  produced in the incinerator,  together with  the gases
from  the  tube  furnace,  will  be  exhausted through  the  heating
jacket of the dryer.

An exhaust  fan removes the gases  from the system  and  ensure  a
sub-atmospheric pressure in the entire system.

If the gas stream  from the  incinerator  is too great to be handled
through the dryer, this stream (or a part of it)  will be removed
from the installation  by another exhaust fan.
                         155

-------
 NBM BODEMSANERING
4.   Operational costs and production factors

The industrial plant  in Schiedam  starts cleaning  in April 1987.
The production hours and  tonnages  of  polluted soils  are given
below.
      Production indirect thermal cleaning plant Schiedam

1987
1988
1989
Production
hours
4.400
6.080
6.800
Tons of polluted soils
cleaned
50.200
76.000
82.000
The cleaning results proves to be independent of the input concen-
tration. In  the following  table the average value for the most
analysed pollutants in the cleaned soil are given.
Cleaning results
Input Output
CN total
Aliphatic hydrocarbons
Aromates
Naptalene
PAH's (EPA)
0
0
0
0
0
- 10,000
- 20,000
- 5,000
- 10,000
- 20,000
< 5
< 50
0.05 -
1
1


0.1


                           156

-------
 NBM BODEMSANERING
The operational costs  are principally energy,  labor, maintenance
and investment.  The relative costs of the selling price are shown
below.
Breakdown of cleaning price
Item
Risk/benefit
Energy
Labor
Maintenance
Capital cost
Analyses
%
15
20
17
15
30
3
The operational costs greatly depend on the production capacity.
Only the energy cost is related to the  production. The production
capacity depends on:
.  type of soil
.  moisture content
.  energy content
.  type of pollutants

The total selling price under Dutch conditions and at an exchange
rate of 1 $ - fl. 2.00   is $ 100 to $ 125 per ton.
                       157

-------
NBM BODEMSANERING
5.   Legislation and licenses

In   the  case  of  the indirect  thermal  soil  cleaning  plant  the
following licences were  necessary:

. water outlet licences  ("WVO vergunning"),
. general environmental  protection  ("hinderwetvergunning")  valid
 to August 1988,
. waste handling licences ("afvalstoffenwet")  from August 1988.

The principal  limits were  put  on  wateroutlet  emissions,  noise
emission  and  stack  emission.  The  actual  licences  permits  the
following values concerning stack emissions.
Indirect thermal soil cleaning
Stack emission values of the actuel licences
Pollutants
Fluoride
Sulphur-
dioxide
Chloride
acid
Cyanides
total
Total
Aroma tes
Total
PAH's
Cadmium
Mercury
Dust
Limit value of waste
licence
5 mg/mo3
460 mg/mo3
75 mg/mo3
1 mg/mo3
5 mg/mo3
2 ug/mo3
0.1 mg/mo3
0.1 mg/mo3
75 mg/mo3
Measured value
( mean )
0.12
200
9.15
0.05
0.28
3.5
0.29
0.001
12
                           158

-------
 NBM BODEMSANERING
6.  Development

The results of the laboratory studies proved that the cleaning of
chlorinated hydrocarbons  is  possible at the  NBM indirect heated
thermal cleaning plant.
The central and  local  authorities  recognized  this conclusion but
asked for trials on the full  scale  plant.
After three years of discussion  the administration agreed with the
trials proposed,  total  cost a million Dutch guilders.
The trials are carried  out in two phases.
6.1  Trial on incinerator
The original  maximum value of products  of  incomplete combustion
(PIC's) was 1 ng  out of 1 gram chlorinated hydrocarbons.  It does
not appear to be  feasible to determine this rest formation during
trials while the time fluctuations were too important.
It was  decided  to run  tests  on the  incinerator  using principal
organic hazardous constituers  (POHC's) during normal production.
The following limits  were set  up:

1) stable temperatures  in  the incinerator in time  and in space,
   and stable gas velocities
2) combustion efficiency of more then  99,9%
3) destruction efficiency of POHC's of more then 99,99%

A  schematic  presentation of  the  incinerator  is shown in figure
(4).  During the  test  the temperature and gas velocity proves to be
stable indicating that  the gas flow conditions  were good  and the
residence time distribution small enough.
The gas velocities are shown in figure (5).
Figure 4
   umpllng
Figure 5
Gas velocity distribution in incinerator
Schematic Incinerator NBM Indirect thermal cleaning plant
                                          40  ta   76  »»
                                            Distance ( cm )
                         159

-------
NBM BODEMSANERING
The residence time varies between 1,8 to 2,1 seconds  depending on
temperature and gas-flow.
The combustion efficiency was so good that the results mostly meet
the limits of 99,9%

The destruction efficiency was  determined  under  the following
conditions:
POCH's           trichloro ethylene
                 tetrachlore methane
Temperatures :    1100  C and 1200  C
Residence time:   1,8 —> 2,1 sec.
Oxygen content:   +_ 10%
C1CH-CC1.
 CCL,
The calculation of the destruction efficiency takes account of:
1) Values  below  detection limit,  are counted at  a minimum  - 0,
   maximum - detection limit
2) thrichloro ethylene  concentration in cooling  sampling nitrogen
   gas  of 115 ng/m3
3) tetrachloro methane  concentration in cooling  sampling nitrogen
   gas  of 57 ng/m3
4) sampling in triplo

The results are shown in the table below:

Destruction efficiency of trichloro ethlynene
sampling
1
2
3
Temperature 1100° C DE%
min . max .
99.9999 99.9998
99.9999 99.9999
99.9999 99.9999
Temperature 1200° C DE%
min. max.
99.9999 99.9999
99.9999 99.9999
99.9999 99.9999
                            160

-------
 NBM BODEMSANERING
Destruction efficiency of tetra chloro methane
sampling
1
2
3
Temperature 1100° C DE%
min . max .
99.9999 99.9999
99.9970 99.9970
99.9900 99.9900
Temperature 1200° C DE%
min. max.
99.9999 99.9999
99.9999 99.9999
99.9999 99.9999
All the  results meet the limits mentioned above.  The administrati-
on, according to these results, allows phase 2.
6.2  Cleaning a batch of 200 tons HCH polluted soils and 500 tons
     drins polluted soils

The following limits were set up:

1) cleaning results soil
                  "A"-value: HCH < 0.01 mg/kg d.s.
                            Drins < 0.01 mg/kg d.s.
2)  Stack emission
                   Equivalent dioxines < 0.1 ng/nm3
The soils have been cleaned from 3 to 5 April 1990.
The cleaning results were as follows:
Pollutant
HCH's som
Drins som
Input value
mg/kg d.s.
5 - 50
500 - 1500
Output value
mg/kg d.s.
< detect lim 0.01 mg/kg
< detect lim 0.05 mg/kg
The stack emission values are not known  at  the moment of writing
this paper.
                       161

-------




_






LI
lf
-------
  REVOLVING FLUIDIZED BED TECHNOLOGY

for the TREATMENT OF HAZARDOUS MATERIALS
             BY: Geoff W. Boraston
            Superburn Systems Ltd.
   a Div. of Consolidated Environmental Technologies
               Vancouver, B.C.
           FOR PRESENTATION AT

 SECOND FORUM ON INNOVATIVE HAZARDOUS

     WASTE TREATMENT TECHNOLOGIES

        DOMESTIC OR INTERNATIONAL

               May 15 -17,1990
                      163

-------
Introduction
The 1990's is the decade of the environment and it is not surprising that technology
has advanced to meet the demands of our current crisis of waste disposal.
Incineration has been used for over a hundred years for waste disposal and because
of stricter environmental requirements it too has advanced to become recognized as a
safe and effective method of waste disposal.

Fluidized beds are also a combustion technique which has been used for decades
and have always been recognized as an efficient system. They are now being used
for waste disposal with excellent results.  The Revolving Fluidized bed is a unique
type of fluidized bed which was developed particularly for waste material.

Background to Fluidized Bed Combustion

Fluidized beds have been around for a long time and date back to the early 1920's.
Because of certain unique features such as very high mass and heat transfer rates
between solids and gases/liquids, fluidized beds  have been useful in a variety of
process  applications including catalytic processes, drying and combustion. A fluid
bed is simply a bed of granular particles through which a flow of gas or liquid passes
upwards. The particles are suspended in the upwardly flowing stream and have the
appearance of a boiling liquid.
                        Feed Stock
                       Red Hot Bed
                       OfSand
                   Air Distributor

                 Plenum
           AIR
 Figure 1. Basic Principal of Fluidized
             Bed Combustion.
With fluidized bed combustion fuel or waste is
burned in a turbulent bed of red hot inert
particles, normally sand, which is fluidized by
the combustion air. The basic fluid bed
combustor is shown in Figure 1. There are four
major components (1) plenum, or windbox
which first receives the combustion air from
the forced draft fan (2) the air distribution
plate which transmits the air from the windbox
to the fluidized bed (3) the fluidized itself
(which is usually sand, ash, limestone or a
mixture of these) into which the feed is injected
and (4) the freeboard area above the bed.

Fluid beds are designed with different bed
depth which can be anywhere between
                                        164

-------
6 in. and 4 ft. deep. The bed of particles is suspended by the upcoming air and
expands to 1.1 to 1.8 times its static height, depending on air velocities, bed depths
and particle characteristics.

Prior to introducing the feed stock the bed is heated from the start-up temperature
to operating temperatures with the light up burners.  Once the bed is hot, say
between 1400 F to 1800 F, the feed is introduced into the bed. The feedstock then
burns in the turbulent mixture of red hot solids and combustion air. Because of
the high heat transfer area the granular particles present and the mechanical action
in the bed the combustion is very rapid and very thorough and occurs in a controlled,
near ideal environment. At any given time, only a small portion of fuel is in the bed
and the bed material acts as a huge thermal flywheel which smooths out process
variations.

Most of the combustion occurs in the bed; however, the freeboard supplies additional
gas residence time for complete burn out of the flue gases. Secondary air or
recirculated flue gases are often added in the freeboard. This bed/freeboard
combination is really an integral combustor/after burner combination. The gas
residence time in the freeboard varies between 3 to 8 sec.

Ash is usually reduced to a sufficiently small size that it exits the combustor with
the flue gases. Larger objects which are introduced to the bed sink to the distributor
plate and are eventually discharged through a bed cleaning port.

Types of Fluidized Beds

For combustion  applications there is generally two types of fluid beds, atmospheric
fluid beds and pressurized fluid beds. Atmospheric fluid beds operate at near
atmospheric pressures and usually at a slight vacuum produced by using both an
induced draft and forced draft fan. Pressurized fluid bed operate at high pressures
and are used in  combined cycle power generation. Although pressurized fluid bed
seem promising as a clean coal burning technology with high efficiencies they have
not yet received commercial popularity. Atmospheric fluid beds are generally used
for power generation and waste disposal application.

There are different types and variations of atmospheric fluid bed, but historically
they have fallen into two categories, bubble beds and circulating fluidized beds.
Superburn Systems Ltd. uses a third type called the revolving fluidized bed which
is an enhancement of the bubble bed technique.
                                     165

-------
A conventional bubble bed is shown in Figure 2.  Bubble beds may operate with
fluidizing velocities between 4 and 8 ftVsec. These velocities are sufficient to keep
the bed fluidized and bubbling but are not high enough to carry over the bed
                                                     material into the flue
                                                     gases.
                                                     A circulating fluidized bed
                                                     is shown in Figure 2.
                                                     These beds operate at
                                                     much higher gas
                                                     velocities, than the bubble
                                                     beds, up to 30 ft/sec. As a
                                                     result a large portion of
                                                     the bed is carried over
                                                     with the flue gases.  A
                                                     cyclone is used to remove
                                                     the bed material from the
                                                     flue gases so that it can be
                                                     returned to the bed.
Bubble bed
Revolving Fluidized
      Bed
Circulating Fluidizes
      Bed
Figure 2. Types of Fluidized Beds.

Revolving Fluidized Bed

The revolving fluidized bed (RFB) shown in Figure 3 is a major development of the
bubbling bed technique. Its enhanced mixing both laterally and vertically and inbed
circulation of material lends the revolving fluidized bed to applications which have
been more difficult for conventional fluidized beds. It is closer in characteristics to a
bubble bed than a circulating fluidized bed, but it is absolutely distinct in the way the
bed is caused to revolve in an elliptical pattern.

The characteristic features of the RFB are an inclined air distributor plate,
differential air velocities across the bed and an inwardly inclined deflector plate
above the bed of sand. The gas velocities on the low side of the distributor plate are
much higher than on the high side. As a result the sand rises from the high velocity
side and is deflected across the bed towards the low velocity side. This combination of
a rising and sinking bed along with the sloped distributor plate causes the bed to
revolve (circulate) in an elliptical pattern.  The bed action is actually a churning
action where waves of sand move at regular intervals across the bed.
                                        166

-------
The flow pattern of the bed material is such that the waste material introduced
onto the sinking bed region is engulfed by the red hot sand and mixed throughout
the bed volume. This ensures the most rapid heat transfer between the sand and
the waste material and an accelerated rate of combustion.  Consistent agitation of
the bed material and burning feed stock ensures that no boundary conditions exists
and there is a high degree of combustion.  The improved lateral mixing distributes
the feed stock evenly throughout the reactor volume and the inbed circulation allows
non-combustible objects to be easily discharged from the bed.

Light debris does not float on the surface of the bed but is embedded in the sand and
covered by the circulated sand. This ensures that a high portion of material is burned
within the bed where the most ideal combustion conditions exist.

As combustibles descend with the sinking portion of the bed, they undergo drying,
gasification and combustion. Volatiles are released fairly rapidly; however, the
majority of combustion still occurs within the bed due to the engulfing of the waste
material and the settling of sand on the sinking region.
          AIR
                Non-Combustibales
Figure 3. Single Revolving Fluidized Bed
Non-combustibles which could be anything from
tin cans to rocks travel with the moving bed of
sand as shown in Figure 3. They sink through
the low velocity zone and are carried
horizontally above the air distributor plate.
Once the non-combustibles reach the high
velocity zone they are not fluidized because of
their weight but are discharged through the ash
outlet. The residence time of the non-
combustibles in the bed is sufficient to ensure
complete burnout of any combustible portion
and can be controlled by the rate of ash
discharge and the circulation rate of the sand.

The inherent ability of the revolving fluidized
bed to easily discharge non-combustibles from
the bed distinctly separates it from the other
fluidized bed. The accumulation of rocks and
metal objects in fluidized beds can cause serious
problems and result in defluidization and
agglomeration of material.
                                    167

-------
Foreign material can pile up along the air distributor resulting in deuuidization and
impairment of the combustion process. The revolving fluidized bed was originally
developed to overcome these weaknesses, and as a result, the system became popular
for municipal waste which contain a high percentage of non-combustibles.

The enhanced lateral mixing of the revolving fluidized bed improves the overall
quality of combustion. Bed temperatures are extremely even and the fuel is
distributed throughout the reactor volume. The enhanced mixing eliminates hot
and cold spots and greatly reduce the possibility of bed agglomeration.  Any
                                     agglomerated material which  may form is
                                     quickly broken down by the increased
                                     turbulence or discharged from the furnace.
                                     With the high degree of turbulence in the
                                     bed, the revolving fluidized bed can handle
                                     solids, liquid and sludges, while retaining
                                     the majority of the combustion in the sand
                                     bed.  This leaves the freeboard volume for
                                     the final stages of combustion in order to
                                     ensure adequate destruction of organic
                                     pollutants.

                                     The single revolving fluidized bed shown in
                                     Figure 3 can be placed face to face in order
                                     to achieve the twin revolving fluidized bed
                                     as  shown in Figure 4.  The performance of
                                     these two bed configurations is identical.
                                     The twin revolving fluidized bed is built
                                     twice as wide as a single revolving fluidized
                                     bed.  For small applications, a single bed is
                                     used and for larger applications, a twin bed
                                     is used.
       t     t    1
      AIR   AIR   AIR

      Non-combustables
Figure 4. Twin Revolving Fluidized Bed.
Advantages of the Revolving Fluidized Bed Combustion

1.    Efficient Combustion

The intimate mixing in the fluid bed results in high oxygen availability and hence
combustion efficiency. The revolving action of the bed ensures that the feedstock is
                                        168

-------
carried away from feed location and mixed throughout the combustion chamber. The
high heat transfer area on the inert bed material and large thermal mass results in
rapid combustion at even temperatures. The constant collision of particles in the bed
reduces the ash layer which may be formed on organic particles exposing additional
combustible surface.

2.    Feed Stock Variation

The revolving fluidized bed can handle a wide variety of feed stocks including
hazardous waste, industrial waste, municipal waste, sewage sludge and
contaminated soil.  In addition the system is very tolerate to fluctuations in the
quality of the feedstocks. It can be designed to burn subautogenous or
superautogenous fuel or shift from one to  the other.

Because the revolving fluidized bed contains such a large thermal flywheel changes
in the calorific valve or moisture content do not immediately upset process
temperatures.  Changes in fuel consistency can be monitored by the (^reading and
drifts in operating temperatures.  The process control system can detect slight
changes and make adjustments without upsetting the process.

3.    Easily controlled

With the excellent mixing uniform temperatures and combustion conditions exist in
the fluid bed. This allows for simple and accurate process monitoring and simplified
autommation. Because the combustion is so rapid process response to changes in feed
rate and air supply is immediate.

4.    Shift Operation and Shut-down

The combustion is rapid and very little fuel is held in the bed at any time so that shut-
down can be achieved almost instantly. This has advantages for both shift operation
and emergency shut-down.

The bed of sand loses its heat very slowly  during shut down conditions,
approximately 40 F per hour, so start-up  time can be very short.

5.    Emissions

As with any combustion system air pollution control can be accomplished with a
                                        169

-------
variety of air pollution control devices and the design of a complete system will
depend on the emission levels required by regulation.  By nature of the fluid bed
many pollutants can be minimized at the combustion source.

In-bed control of sulphur gas emissions by direct lime addition to the bed is a well
known technique in fluid bed combustion. The calcium in the limestone reacts with
the SQs. to form gypsum, similarly, halide capture as calcium halide may also be
effected, reducing acid gas production. Sulphur removal depends on the sulphur in
the feedstock and the limestone characteristics, however 80% sulphur removal can
normally be achieved quite easily.

Fluid beds operate at relatively low temperatures so the formation of thermal oxides
of nitrogen from nitrogen in air is minimized.  Fuel bound nitrogen primarily
contributes to the formation of NOx.

The fluid bed is unique in that the majority of ash is carried over as flyash and there
is little or no bottom ash. The amount of bottom ash will  depend on the larger non-
combustible objects in the fuel. The flyash will be removed using either a baghouse or
electrostatic precipitator.

6.    Maintenance

There are no moving parts in the combustion chamber itself since all the mixing and
material movement is achieved by air. This reduces the maintenance costs.

The highest maintenance item is the repair and replacement of refractory but with
the elimination of thermal shock resulting from the large thermal flywheel and
reasonable care during start-up and shut down, damage to the refractory is
minimized.  Many fluid beds operate for over 10 years with only minor refractory
repairs.

7.    Capital Cost

The fluid bed is a fairly simple system with relatively few easy to construct
components, so provided an elaborate waste treatment system is not required the
capital cost will be 60% to 85% that of a rotary kiln of equal capacity.
                                       170

-------
8.    Specific Advantages of the Revolving Fluidized Bed over other Fluidized Bed.

The revolving fluidized bed was developed in order to overcome specific weakness of
conventional fluidized bed, namely to improve lateral mixing and also to improve the
movement of large non- combustible objects out of the bed.  This ability to easily
discharge inert lumps allows the revolving fluidized bed to burn waste material
which contains rocks, metal etc. Also because of the revolving action the feedstock is
drawn into the bed where the best combustion conditions occur. It is because of these
strengths that the revolving fluidized bed has found acceptance for feedstocks such a
sludge or soil containing rocks and municipal waste containing a variety of non-
combustible objects.

Development Status of the Revolving Fluidized Bed

Fluid bed combustion is a well known technique and has been used for decades. The
revolving fluidized bed was developed in the late  1960's particularly for waste
material and has been used commercially since the early 1970's. There are over 60
incineration and steam generation plants existing using the revolving bed technique.

Existing Installations:                                Number Installations:

Great Britain     Coal and Industrial Waste                       8
                 Coal and Hospital Waste                        2

West Germany    Soil Remediation                        1 (under const.)
                 Wood                                   1 (under const.)
                 Municipal Waste                         1 (under const.)

Japan      Industrial Waste and Sewage Sludge             9
                 Municipal Waste                         31
                                                         9 (under const.)

Tunisia          Industrial Waste                               1

Canada          Various Waste                                 1
                 Hazardous Waste                        1 (under const.)
                 Oil Refinery Waste                       1 (under const.)
                 Lumber Industry Waste                   1 (under const.)
                                       171

-------
                              CASE STUDY

               SYDNEY TAR PONDS CLEAN-UP PROJECT

Background

The Sydney Tar Ponds in Sydney, Nova Scotia, are one of the most hazardous
chemical waste sites in eastern Canada, and the largest chemical waste site in the
country. More than 700,000 Tons of coke oven residue was discharged over 80
years by the local steel mill into a creek bed. The results are oily black mud flats.

The coke oven residue contains a high concentration of hazardous organic
compounds, principally polyaromatic hydrocarbons and heterocyclic nitrogen
compounds. It has the appearance of wet tarry, muddy sand and has a high ash
content and a moisture content between 30 and 70%.

The Provincial and Federal Governments have sponsored a program to remediate
the site by excavating the waste product over 7 year period and treating it in an
incineration plant.  The principal objective is to destroy the hazardous organics
compounds with an efficiency of 99.99%. As a bonus, the incineration process will
generate up to 10 MW of electrical energy which will be used by the steel mills new
electric arc furnaces.

The incineration plant currently under construction, is a revolving fluidized bed
supplied by Superburn Contractors and designed by Superburn Systems Ltd.
Superburn Contractors was selected because a series of test burns confirmed the
effectiveness of the  system and it had the lowest construction costs, lowest
operating costs and best delivery schedule.

Test Burns

During the summer of 1988, Superburn Systems Ltd.  conducted a series of trial
burns on the tar pond sludge in a 6  MBtu per hour, demonstration plant located in
Vancouver, B.C. The purpose of the tests were to demonstrate the contaminated
sludge could sustain combustion under steady state conditions and that 99.99%
destruction of the principal hazardous organic compounds could be achieved.

Approximately 29 tons of tar pond sludge were shipped to Vancouver from Sydney
under the hazardous waste transportation regulations. The sludge delivered
                                      172

-------
represented a characteristic range of material consistency.

The test program concluded that the sludge could be burned under stable conditions
and that between 99.99% and 99.9999% destruction efficiency could be achieved on
the PAH's with an operating temperature of 1750 F and 1.7 sec gas residence time.
See Table 1, Appendix 1.
Sydney Tar Pond Project
Revolving Fluidized Bed Incineration Plant
      Location:
      Rated Throughput:
      Steam Production:

      Power Production:
      Major Equipment:
      Capital Cost:
Sydney, Nova Scotia
31,000 Ib/hr Tar Pond Sludge
100,000 Ib/hr
at 450 psi and 750 F
Up to 10 MW
Three day sludge storage lagoon
Two piston pumps
Two furnaces with F.D & Sec. air fans
One Boiler
One Baghouse
One ID. Fan
Stack

$16.25 million Canadian
      Emission Criteria:
      1. Stack emissions:
                 DRE
                 Sulphur Removal
                 Particulate
                 Carbon Monoxide
           99.99%
           80%
           0.02 gr/scf
           80   ppm
                                     173

-------
      2.  Ground Level Concentration
            Maximum Ground Level concentration at point of impingement as
            calculated by the Ontario Ministry of Environment Dispersion Model.
Heavy Metals
Lead
Arsenic
Cadnium
Chromium
Copper
Zinc
(ug/m3)
10
1
5
5
100
100
PCDD's andPCDFs
                       JL  x
                       450     (450X50)
Where X is the ground level concentration of PCDD's in pg/m3 and Y is the ground
level concentrations of PCDF's in pg/m3.

Plant Description (see Figure 5)

The primary purpose of this plant is to effectively dispose of the tar pond sludge in
an environmentally sound manner. Power will be produced, but the plant will
operate in a turbine follow mode which will ensure that the plant will always
operate under the most ideal conditions for sludge destruction.

The sludge is delivered to a storage lagoon divided into three cells, each cell
having one day storage capacity. There are two piston pumps, one for each
furnace which are fed from the storage cells by a gantry crane. The sludge is fed
directly to the furnaces and spread over the fluid bed by a steam injection nozzle.

A large freeboard section provides a 3.5 sec gas residence time in order to ensure
complete burn-out of the organic pollutants.  Secondary air and flue gas is
introduced into the freeboard which both controls the temperature and provides
turbulence to aid complete burn-out. The flue gas from the two furnaces are then
ducted into a boiler.

The boiler used is an existing piece of equipment which is being refurbished. The
                                      174

-------
incineration plant is built adjacent to the power house which contains the boiler
and the hot gas duct penetrates the building wall to enter the boiler.  The steam
from the boiler is fed to an existing turbine.

The cool gas from the boiler is ducted out of the power house to a baghouse. The
treated gas then passes through a final heat recovery unit before being emitted to
atmosphere by the stack.
                                    Steam Turbine
                                                                    Stack
                                                                    Induced
                                                                     Draft
                                                                     Fan
                                                         *
                                                         Ash
                                                       Collection
 Figure 5. Flow Chart of Sydney Tar Pond Incineration Plant.
Lime is fed to the fluid bed in order to capture the sulphur and convert it to gypsum.
A fine lime powder will be carried over to the baghouse where it will form a fine
coating on the bags. This process will ensure maximum effectiveness of the lime.

Coal is used for plant start-up and also as a sweet fuel in the event the calorific
value of the tar pond sludge drops too low for good operation.

There are three ash stream, bottom ash for the furnace, economizer ash and baghouse
ash. The bottom ash from the fluid beds is extracted with water cooled screws. This
ash consists of the rocks contained in the sludge and bed material. The bed material
is separated and returned to the furnace. The economizer and baghouse ash is the
fine ash and lime which will make up approximately 95% of the total ash.

The control system is a Bailey Network 90 DCS and is design for fully automatic
operation with all the necessary safe guards to ensure ideal  combustion condition
and safe plant shutdown. The overall plant is designed to operate 24 hours per day
310 days per year and dispose of the tarpond sludge over a 7 year period.
                                        175

-------
                           APPENDIX I
              TABLE 1 SYDNEY TAR PONDS TEST BURN
                    ORGANIC EMISSION RESULTS

Parameter
Date
Test Time
Duration
PAH DUE %
Moisture (Vol%)
C02 (Vol% Dry)
02 (Vol% Dry)
PCDD (ng/Nm3)
PCDF (ng/Nm3)
Gas Residence Time (sec)
Operating Temp. (oF)
Run 1
Furnace Outlet
July 23/88
1828-1924
56.0
99.996
10.6
14.0
9.7
-
-
1.78
1886 °F
Run 2
Furnace Outlet
July 24/88
1358-1454
56.0
99.996
11.7
7.5
14.0
.
-
1.57
1870 °F
Run 3
Furnace Outlet
July 25/88
123-1622
200
>99.99
9.1
10.0
10.7
-
-
1.68
1560 °F
Run 3
Stack
July 25/88
1223-1622
200
99.99
31.6
4.3
15.7
N.D.
N.D.
1.68
1560 °F
Note:All tests were taken at the furnace outlet before the air pollution control
equipment, except for Run 3 where a stack test was taken as well.

N.D. - Not Detected
                                  176

-------
                              APPENDKII

 TAR POND SEDIMENT CHEMICAL AND THERMAL CHARACTERISTICS

A.    PAH CONTENT

The primary POHC's (Principal Organic Hazardous Constituents) to be destructed in
the tar ponds sediments are PAH's (Polynuclear Aromatic Hydrocarbons).  A review
of the analytical data for the tar pond sediment PAH content indicated that average
PAH content of the sediment decreases with distance downstream of Coke Oven
Brook. This parallels the overall decrease in grain size in the downstream direction.
However, considerable fluctuation occurs in any given area. The average total PAH's
at the 0.6m depth by location, are summarized below:
                     Area            Average total PAH (mg/kg)

          Wash Brook Arm                     3,700
          Remainder of South Pond              12,000
          North Pond to Narrows                8,000
          North Pond beyond Narrows            5,500

Reviewing trends with depth, the data (as summarized below) show little variation
over the top two depth intervals and with a noticeable reduction below 1.2m. Even
taking into account the fact that some of the samples have a small portion of the
underlying clean materials incorporated in with the tar pond sediments, total PAH
generally declines with depth.

                       Depth       Average total PAH (mg/kg)

                      0.00-0.61               9,000
                      0.61-1.22               8,200
                      1.22-1.83               2,300
                      1.83-2.44              6,100*
                      2.44-3.05               1,400

* = one sample enormously high
                                     177

-------
The general order of PAHs in terms of their relative abundance is as follows:

      1. Naphthalene               2. Phenanthrene
      3. Benzo (a) anthracene        4. Pyrene
      5. Fluoranthene              6. Anthracene

Naphthalene and phenanthrene account for about 60 to 80% of the total PAH content.
These two compounds are also the two most variable.

The minimum, maximum and average values for the PAH's analyzed are tabulated in
Table 1 included in the Appendix to this section.

B.    HNC CONTENT

The concentration of total HNC's is less variable than total PAH's with no clearly
discernable trend except to note that HNC's generally correlate with PAH's.

The general order of HNC's in terms of their relative abundance is as follows:

      1. Benzacridines              2. Benzocarbazole
      3. Acridine                  4. 7,8 Benzoquinoline
      S.Methylated Benzoacridines   6. Methylated Benzocarbazoles

The first three HNC compounds comprise 60-70% of the total HNC content. It is
therefore to be expected that these compounds will be named as POHC'c for which a
99.99% DRE must be demonstrated during trial burns.

The minimum, maximum and average concentration of selected HNC's obtained from
13 analyses is as follows:

                                    HNC's (mg/kg)
                                          14
                      Maximum           1,510
                      Average             169
                                      178

-------
The selected UNO's were composed of:
      Methylated Benzocarbazoles
      2, 4 Dimethylquinoline
      Acridine
      Benzacridines
      Dimethylquinoline
Methylated Benzoacridines
Carbazole
Quinoline
Benzocarbazol
A more detailed breakdown of tar pond sediment HNC content is given in Table 2 of
this section.

C.    HEAVY METALS CONTENT

The most distinctive trend for heavy metal concentrations in the tar ponds is for zinc,
lead and to a lesser extent, copper to increase in concentration near the mouth of
Muggah Creek. This suggests an association with the finer-grained sediments (silts
and clays), which in turn may reflect adsorption due to charge and increased surface
area with smaller sized sediment particles. The concentrations of arsenic and
chromium are lower and more variable over the Tar Ponds.

Zinc is generally the most abundant heavy metal followed by lead, copper or arsenic,
andthen chromium.

The minimum, maximum, and average concentration for the following heavy metals
is as tabulated below:
Metal
Arsenic
Chromium
Copper
Lead
Zinc
Average
Concentration
(mg/kg)
11
6
48.4
131
218
Average
Concentration
(mg/kg)
197
58
321.8
741
2,112
Average
Concentration
(mg/kg)
102.4
15.7
118.7
309
560.5
                                      179

-------
D.    HEATING VALUE AND CHEMICAL ANALYSIS

The following tar properties represent the results of some sample analyses conducted
on tar pond sediment samples obtained from the tar pond sampling program.

Moisture content:  The air dried moisture content of the samples
                 range from 10.64 to 69.2% on a weight basis.
                 Average content was 35.3%. this moisture
                 content represents the weight percentage of
                 the original sample that was evaporated
                 when the sample was air dried.
Proximate Analysis:     (% w/w, air-dried)

                                Range       Average

             Moisture        1.06 -  9.57%      2.81
             Ash             19.33 - 70.68     40.16
             Volatiles        13.42 - 33.20     25.84
             Fixed Carbon    8.61  -  48.27      31.19
                                               100


Ultimate Analysis (% w/w, air-dried)

                                Range       Average

             Moisture        1.06  -  9.57*
             Ash             19.33 - 70.68
             Sulphur          /1.00 - 5.66
             Carbon          /21.1Q • 70.78
             Hydrogen        1.49-4.47
             Nitrogen         0.25 - 3.02
             Oxygen
H.H.V.kj/kg air-dried    7,073 - 2,9510.0        22,036
      (Btu/lb)          (3,140 - 12,687)        ( 9,474)
                                      180

-------
"This moisture represents that quantity of moisture which remained after air drying,
but which was subsequently removed when the samples were dried in accordance
withASTMD3173.

Note that the proximate analysis is based on the first 13 samples analyzed while the
ultimate analysis is based on some 63 tar pond sediment samples that were analyzed.

Additional Chemical Analysis:

Several chemical analyses were performed on the six samples noted above under
Laboratory Determined Values, Tar Ponds Sediments and Physical Characteristics
The results of these additional chemical analyses are as follows:

            Chlorine content:       0.25 to 0.66% by weight
            Sodium content:         0.22 to 0.90% by weight
            Vanadium content:      29 to 30 ppm

Tar Pond Sediment Ash Characteristics:

a) Physical

   Ash Fusion Temperatures

                                           Range     Average

         Initial Deformation temperatures   1160 -1216    1187
         Softening Temperature            1232 -1338    1276
         Hemispherical Temperature        1266 -1360    1320
         Fluid Temperature                1349 -1449    1390
                                    181

-------
                                                                     TOXIC CHEMICAL
                                                                         SPECIALISTS
ECO  LOGIC	000
                          Thermal Gas Phase  Reduction of

                             Organic Hazardous Wastes

                                in Aqueous Matrices


                                        by
                            Douglas J. Hallett,  Ph.D.,
                                     President
                            Kelvin R. Campbell,  P.Eng.
                                  General Manager
                             Wayland R. Swain, Ph.D.,
                                U.S.  Vice-President
                             ELI Eco Technologies Inc.


                                   prepared for
                      EPA Second Forum on Innovative Hazardous
                           Waste Treatment Technologies:
                            Domestic and  International
                                 Philadelphia,  PA

                                  May 15-17, 1990
                                         182

-------
      Eco Logic International, through its subsidiary ELI Eco Technologies, will
be demonstrating a new type of hazardous waste processor for Canada's Department
of National Defence as part of the Defence Industrial Research Program.  The test
program will be  taking  place during the summer  and fall  of 1990.   Of  direct
importance to the military worldwide is the ability to clean up and process PCBs,
chlorinated solvents, chlorinated dioxins, etc. which may be contained in harbour
sediments, soil, hydraulic  fluids,  cleaning  fluids, defoliants,  mustard,  and
residual chemical warfare agents.

      The thermo-chemical reaction  that makes the Eco Logic process possible is
the  ability  of  hydrogen  to  dechlorinate  organic  molecules  at  elevated
temperatures.   Bench-scale  tests have shown  that a well-mixed  combination of
hydrogen and chlorinated organic waste subjected to 850°C or higher for a period
of 1 second will result in 99.9999% destruction or better.  Consistent results
have been  obtained  for  over 100  tests  during  the last  16  months.   The  end
products of the  reaction are HC1 and  dechlorinated organics themselves  broken
down by hydrogenation to methane and ethylene.

      The Eco Logic  process is not  an incineration technology.   Destruction of
chlorinated  organic  waste  using   incineration  and  pyrolytic  processes  is
accomplished by breaking contaminant molecules apart with high temperatures and
combining  them  with  oxygen, usually   from  air.    PCBs  first  fragment  into
chlorobenzenes. which sometimes  combine with oxygen  to form dioxins and furans,
which are more  toxic than the original  PCBs.  The Eco Logic  process uses hydrogen
at elevated temperatures to reduce, rather than oxidize, chlorinated organics.
Since there is  no free  oxygen in the  reducing atmosphere,  no dioxin or furan
formation is possible.   As well, since  combustion air is not  required, there is
no nitrogen to  use up reactor volume and heating,  resulting in the reactor being
much smaller than an incinerator handling the same throughput.

      The chemical reaction is shown on Figure 1.  Hydrogen and a PCB molecule
with 4 chlorine atoms attached are  reacted above 850°C to form HC1 and benzene.
In the second reaction, which occurs at  the same time, benzene and hydrogen react
to produce methane (and some ethylene).   In the third reaction, a non-chlorinated
alkane hydrocarbon reacts with hydrogen to form methane.  The presence of water
enhances all of these reactions, as does an excess of hydrogen, and an increased
residence time in the reactor.

      Figure 2 shows a schematic of the reactor module designed to accommodate
the thenr.o-cherr.ical  reaction.   A mixture of preheated waste and  hydrogen is

                                         183

-------
injected by  nozzles  mounted tangentially  near  the top of  the reactor.   The
mixture swirls around a central ceramic tube past glo-bar heaters and is heated
to 850°C by the time  it passes through the ports at  the bottom of  the ceramic
tube.  Particulate matter up to 3 mm diameter not  entrained  in the gas stream
will impact  the hot  ceramic  walls  of the  reactor,  thereby volatilizing  any
organic material associated with the particulate.  That particulate will exit
out the reactor bottom  to a quench  tank,  while  finer particulate  entrained in
the gas stream will flow up the ceramic tube into the  exit elbow and through the
retention zone. The reduction reaction takes place from the  bottom of the ceramic
tube onwards, and the organic waste is completely dechlorinated and reduced to
methane, ethylene,  and hydrogen chloride.  Depending on the water content of the
waste, carbon monoxide and hydrogen may also be formed from  the reaction of water
with methane.

      The reactor module is six feet in diameter and  nine feet tall without the
exit elbow.   Photos  of the  reactor steel shell under construction are shown
following Figure 2.   Figure  3  shows the  process equipment layout  for mounting
on  the  two  drop-deck transport trailers.  Photographs of  the  trailers follow
Figure 3.  The layout shows the reactor (R) with the attached  exit elbow  (E) and
extension (EXT) leading to an HC1  and particulate scrubbing system  (SCRUB).  In
the scrubber, the gases are first cooled by direct  injection of  water spray from
a later condensation step.  Removal  of HC1 and particulate  is  accomplished using
a wet contact  scrubber  and lime scrubber water.

      An  amount  of water  approximately equal  to the  water in the waste stream
is  carried through the scrubber as humidity in the  hydrogen and hydrocarbon gas
stream bv controlling the scrubber temperature.   This prevents rapid accumulation
of  scrubber  water  waste.   Most of the water carried through is then  condensed
out  using a  plate  heat  exchanger, leaving  relatively  dry excess  hydrogen,
methane,  ethylene  and carbon monoxide.   Approximately 90%  of  this dry gas is
recycled  into  the reactor in order to use up excess hydrogen, and the  remaining
10% is  sent  to the propane boiler as supplementary fuel.   The boiler produces
steam  for the  heat exchanger  that preheats the  waste stream  prior  to  injection
into the  reactor.

       The only gas emissions are from the  boiler, which  will require  a small
stack,  but  since the fuel  going into  the boiler is very clean,  and contains no
chlorine,  the emissions will  be insignificant.  Effluent  from  the  process will
be  stored and  analvsed in batches  before  release.   It will consist of grits  from
the bottom exit of the reactor,  calcium chloride and particulate sludge from the

                                         184

-------
HC1 scrubber,  and  condensed water.  Some of that water will be recycled into the
process for primary cooling  in  the scrubber.  Any effluent not  conforming to
emission standards can be processed through the system again.

      The process  control and emission monitoring systems for the waste processor
are interlocked  to provide  an  extra measure  of  safety,  and to continuously
document proper destruction efficiencies.   The V&F CIMS-500 chemical ionization
mass spectrometer is capable of measuring PCBs and their breakdown products on
a continuous basis, so that  any  potential  process upsets result in an immediate
automatic halt  to waste processing.   As  well, when the process is operating
normally, a continuous  readout  of  destruction  efficiency  is  obtained.   During
the demonstration program, regulatory test methods using Modified  Method 5 stack
testing trains will be used to confirm destruction efficiencies.
Net Benefits of the Eco Logic Process
      Mobility/size -  Incineration systems are  carried  on a number  of large
transport vehicles, require weeks  of  set-up  time,  occupy large  areas when set
up (football fields) and as a consequence of such overheads, must be used with
high volume, lengthy burns.  Eco Logic's waste processor requires two standard
tractor trailers,  is completely mobile  (compared to transportable),  requires only
davs to set up, and occupies little more area than the vehicles.  Minimum runs
may be less than a single unit's daily capacity.

      Scale of  job - The  continuous throughput  process is well-suited to high
volume long run jobs,  compared to sodium  and potassium  stripping methods, which
are  typically  small batch  processes.    Throughput  capacity of the  Eco Logic
process can be increased by ganging reactor units on a  single ancillary support
and control system,  allowing flexibility  of operation,  and redundancy of design.

      Aqueous Content - Some of the largest and most serious contaminant clean-
up  requirements  are  soils  and  sediments  having   a  high  water  content.
Incineration technologies consume  very large  amounts  of  energy  to heat up the
water component to the incineration temperature.  As well,  because they use air
(79% nitrogen)  for combustion, and must  combust  all the organic material, they
require  approximately  10  times  the  volume  for the  same residence  time  of
reaction.   Sodium and  potassium processes are  precluded from treating water-
bearing wastes because of explosive reaction  potential.  Eco Logic's destructor

                                         185

-------
must also heat water,  but  to temperatures which are 30% lower,  and  the water
component actually enhances  the chemical  reduction process.   As  well,  the use
of hvdrogen (no excess nitrogen) in the reaction reduces the gas  volume of the
products and therefore the size of the reactor required.

      No Dioxin/Furan  Emissions - Since  oxygen  is  not used  in  the  reaction,
formation of chlorinated dioxin or furan molecules is precluded.   Furthermore,
if a process upset  does occur,  the CIMS-500 continuous organic emission monitor
will automatically divert the  sidestream  gas  flow to  the boiler  back into the
reactor, so that no  air emissions occur.   The waste stream would  be shut off
automatically and the gas stream recirculated until  the problem was rectified.
A charcoal filtration unit prior to a divert stack is available if the reactor
had to be purged.

      Cost - The combination of  hardware requirements  and process characteristics
suggest  lower  capital  cost  for the Eco Logic  system,  by  a factor of  5 to 10
tir.es, compared to incineration processes.   Operating economies  are predicted
to  be  frorc 3  to  5 times lower than  incineration technologies  of comparable
capacities.
                                         186

-------
             FIGURE 1
THERMD-CHEMICAL REDUCTION  REACTIDNS
Cl
 /
Cl
     5 H
Cl    Cl
+  4 HCl
           9 H
                  6 CH4
       -l>  H
                            CH
                   187

-------
             FIGURE  2
THERMQ-CHEMICAL REDUCTIDN  REACTOR

              /
                       REFRACTORY
                       GLDBAR HEATER
                       CERAMIC TUBE
                    188

-------

-------
IO
o
     FIGURE  3  -  PROCESS EQUIPMENT  LAYOUT

-------
I6T

-------
         X*TRAX™ - TRANSPORTABLE THERMAL SEPARATOR
                                     for
                   ORGANIC CONTAMINATED SOLIDS
                              Carl Swanstrom
                                Carl Palmer
                    Chemical Waste Management, Inc.
                         Geneva Research Center
                       1950 South Batavia Avenue
                          Geneva, Illinois 60134
                               Presented at:
            SECOND FORUM ON INNOVATIVE HAZARDOUS
                 WASTE TREATMENT TECHNOLOGIES:
                    DOMESTIC AND INTERNATIONAL
                        Philadelphia, Pennsylvania
                             May 15-17, 1990
                                  ABSTRACT
Chemical Waste Management (CWM) has developed a patented system, X*TRAX™, that thermally
separates organics from solids, such as soils, pond sludges and filter cakes, in an indirectly heated
rotary dryer. Vaporized organics and water are transported with a nitrogen carrier gas to a gas
treatment system where they are condensed and collected as a liquid. The gas is then reheated and
recycled to the dryer. To control oxygen, a small portion of the carrier gas is vented to atmosphere
through carbon adsorbers. CWM has constructed a full scale transportable X*TRAX system that has
been contracted for a PCB soils cleanup at a mid-size Superfund site in the Eastern U.S.  The system
is expected to mobilize in mid to late 1990 on that site. In an innovative combination, the condensed
organic liquid will be chemically dechlorinated on site prior to off site disposal. A SITE demonstration
test will be conducted at this site during the performance of the cleanup.  Since early 1988, CWM has
operated both a 5  ton/day  pilot demonstration system, and a 2-4  Ib/hour laboratory system for
treatability studies.  The pilot system has recently completed ten tests using TSCA regulated PCB soils
and is currently in preparation for an extensive testing program using RCRA regulated materials. In
addition to a number of surrogate wastes, the laboratory system has processed over 19 RCRA and
TSCA regulated waste materials.
                                     192

-------
          X*TRAX™ - TRANSPORTABLE THERMAL SEPARATOR
                                           for
                      ORGANIC  CONTAMINATED SOLIDS
1.     INTRODUCTION

The widespread problem of soils and solids that are contaminated with organic chemicals, coupled with
the EPA's increased restrictions on organics in landfills has resulted in the unavoidable fact that millions
of cubic yards of soil and solids will have to be treated  to reduce  or  eliminate  the  organics.
Historically, the most likely treatment alternative has been high temperature incineration, which is
costly, difficult to permit, and requires lengthy mobilization periods for system installation and trial
burns.  Chemical  Waste Management  (CWM) believes that many of these waste streams can be
treated using a thermal separation system; in essence, by drying them.  Wastes such as contaminated
soils, pond or process sludges, filter cakes and others are likely  candidates.  Laboratory and pilot
testing by CWM has shown that at  low temperatures (500-800 °F) many organic compounds including
high boiling compounds  (PCBs) can be successfully separated from solids such as soil, sand, etc.
Thermal separation is now a treatment option with significant advantages in all of the above mentioned
areas for a broad class of waste materials that have relatively low organic concentrations - typically
less than 10%.

Contaminated solids are heated in an indirectly fired rotary dryer to volatilize the organics.  The vapors
are carried to a gas handling system with an inert gas where they are scrubbed for particulate  solids
and cooled to condense  the organics.  The carrier gas  is  reheated and recycled to the  dryer. The
recovered organics  can  be reclaimed,  used on- or off-site  as supplemental fuel or  destroyed by
incineration.  Organic contaminants can range from high  boiling, semi-volatile compounds such as
PCBs, to low boiling, volatile compounds such as RCRA  regulated solvents.1-2 This X'TRAX process
has been granted U.S. Patent No. 4,864,942.

CWM has been actively  developing the X'TRAX process since late 1986. This paper describes the
X*TRAX process and each of the three X'TRAX systems that have been constructed: laboratory, pilot
and full scale.  Test data  are presented for both the lab and pilot systems. The first full  scale unit has
been functionally tested  and will be moved to a Superfund site in mid to late 1990,  depending on
receipt of approvals  from EPA.
2.     SYSTEM DESCRIPTION

The X'TRAX system is a separation process to remove volatile or semi-volatile compounds from a solid
matrix.  Thermal energy is the driving force used to affect the separation.  The process flow diagram
is presented in Figure 1.

Feed material,  which can be either solid or pumpable sludge, is fed into the dryer.  The dryer is an
externally fired rotary kiln.  It is essentially a sealed rotating cylinder with the feed material tumbling
inside and the heat source  (propane burners) on the outside.  Since the dryer is externally fired, the
combustion  products do not contact the waste material (feed) being processed.  The  use of an
externally fired dryer has two distinct advantages.  First, and most important, is that the combustion
gases do not pass through the associated air pollution control (APC) devices.  Propane is a readily
available clean burning fuel.  Air permits for vent stacks from propane combustors are easily obtained,
usually without any required APC devices. This allows the APC devices for X'TRAX to be one tenth
to one hundredth the size of that for an equivalent capacity incinerator. In addition, the small volume
of nitrogen carrier gas discharged makes cleaning  it to very high standards quite  inexpensive.  The
second advantage of external firing is that it makes the X'TRAX system a separation process, not an
incinerator because no organic combustion occurs. It is usually much easier to permit a separation
                                            193

-------
X*

















TRAX: PROCESS FLOW DIAGRAM VENT^^ CARBON
nir I DRUMS 1
^ffo ^—L-^nuwj ^-i—^
1 	 *~ ' '
PRIMARY VWMMRY f
»rJ~i r OIWWR CONDENSER m TFR fv7!

EDUCTOR 	 -™' 	 	 ™ 	 &7^ 	 !,
SCRUI3BER LOW T ^
r . • REHEAT
i, '
UkGANICS — pHASE J CONDENSATE
rrn — 	 ^TDRAlT r?
SLUDGE J±^ MAKEUP 1JSJ «?
\
HIGH T
REHEAT
t I
ROTARY DRYFR ~ ^^r™
— — -- -— •
1
UKY KKUUUU — J
Figure 1
process than a waste incinerator.

The heated solids are discharged from the dryer as a powdered or granular dry material.  For most
applications,  water  will be mixed with the exiting solids to cool them and to prevent dusting.  By
adding reagents at this point, metal containing wastes can be stabilized with the reagent cost being
the only additional expense. The water will normally be condensate from the gas treatment portion
of X*TRAX.

The carrier gas first passes through a liquid scrubber where entrained solid particles are removed and
the gas stream is cooled to its saturation temperature.  The scrubber  also removes a  portion of the
volatilized organics. The recirculated scrubber water continuously passes through a phase separator.
The phase separator collects any condensed light organic from the liquid surface and continuously
discharges a  bottom sludge containing solids, water and organics. The sludge is dewatered using a
filter press.  The dewatered solids are either returned to the feed stream or disposed of.

The scrubbed gas passes to the first heat exchanger where it is cooled  to  10°F above ambient
temperature.  This heat exchanger will produce the bulk of the liquid condensate. The carrier gas now
goes to a second heat exchanger where it is cooled to 40°F.  The liquid condensates from both heat
exchangers are mixed and allowed to gravity separate.  Organics are removed for disposal.  The
condensed water is used to cool and dedust the treated solids exiting  the dryer.

The 40 °F carrier gas now contains some residual moisture and organics that were present in the feed
at levels equal to or less than their equilibrium saturation concentration at 40 °F. The carrier gas then
passes to a  recirculation blower.  After the blower, 5 to 10% of the  carrier gas is vented, and the
remainder is heated to 400-700°F before returning it to the dryer.

The process vent gas stream passes through a particulate filter (2 micron) and then through a carbon
adsorber, where at least 80% of the remaining organics will be removed. Actual practice has shown
removal efficiencies by  the carbon ranging  from 89  to  98%.  This gas is then vented  to the
                                             194

-------
atmosphere.  A 100 ton per day X*TRAX system would release anywhere from 0.25 to 5 pounds per
day of VOCs which is considerably lower than most regulatory constraints.


2.1    Laboratory

Since January, 1988, CWM has operated a laboratory  X*TRAX system  at its Riverdale Technical
Center in Riverdale, Illinois.  This system typically processes 2 to 4 Ib/hour.  It consists of a 4-inch
diameter, 48-inch long electrically heated tube furnace coupled to a small scale gas treatment system
that closely simulates the pilot and full scale systems.

This unit is used for treatability studies and to screen materials for pilot testing and commercial
operations.  To date, 23 separate test runs have been performed, with 19 being on actual RCRA and
TSCA waste  materials. The laboratory system was operated under CWM's TSCA R&D permit for the
Riverdale Center, as well as CWM's Illinois authorization for RCRA treatability studies.

In September of 1989 the system  was transferred to Chem-Nuclear Systems Inc. (CNSI) in  Barnwell,
SC.    CNSI  is using the system to  evaluate the  applicability  of  X'TRAX  to treating mixed
(radioactive/hazardous) wastes, having conducted two tests to date. A second laboratory X*TRAX
system has been constructed and  is operational at the new CWM R&D facility located in Geneva, IL.
A new TSCA R&D permit was  granted on January 30, 1990.


2.2    Pilot System

The pilot X*TRAX system is a mobile unit mounted on two semi trailers;  one containing the dryer and
another containing the gas treatment system. The dryer is 24-inch diameter,  20 feet heated length,
with 10 propane burners. The pilot system has a nominal capacity of 5 tons per day for a  feed
material containing 30% moisture.  Figure 2  is an artist's rendering of the pilot X'TRAX system.

The pilot system was used to provide design data on capacity, material handling, and gas system
performance for the full scale system.  It has been and continues to be used to provide treatability and
emissions data on candidate  waste streams, and is available for the performance of demonstrations.

The pilot system became operational in January, 1988. Since then it has been used to test over 87
tons of materials, including: 59 tons of simulated RCRA wastes, 5 tons of mixed radioactive/hazardous
waste  and  20 tons of TSCA regulated  PCB soils.

The pilot system  is presently installed at CWM's Kettleman Hills Facility in central California.  The
system is operated under a variety of permits at Kettleman.  The most basic of these is an operating
permit from Kings County, allowing CWM to have an air emission source.  CWM also has a variance
from the California Department  of Health Services to treat non-RCRA wastes such as California special
wastes. The testing on PCB materials was conducted under a three month R&D permit from the EPA's
TSCA branch, which expired October 4, 1989.  A 90 day extension was granted starting November
1, 1989.   CWM  has also filed for a  RCRA  RD&D permit to  allow for testing on RCRA regulated
materials.  This permit request is currently under review, and is expected to be granted during the third
quarter of 1990. CWM will perform one year of testing under the RD&D permit, investigating all types
of RCRA wastes but focusing on those that have or will  be restricted from land disposal ("land
banned").
2.3    X*TRAX Model 200 Full Scale Production System

The X*TRAX Model 200 is a full scale production system that was constructed for onsite cleanup of
contaminated soil.  The system is capable of treating 125 tons per day of contaminated soil with a
moisture content of 20%. Like the pilot system, the Model 200 has a rotary dryer and a gas treatment


                                           195

-------
0>
                                                                                                  00

                                                                                                  i-t
                                                                                                  n
                                                                                                   I

                                                                                                  T)

-------
system; however, they are much larger, requiring the use of modular construction techniques. The
Model 200 is fully  transportable, consisting of three semi trailers, one control room trailer, eight
equipment skids and various pieces of removable equipment. Figure 3 is an artist's rendering  of the
Model 200 system.  The area required for the equipment measures 120 ft. by 120 ft.

All of the  equipment has been designed for over the road transport anywhere in the U.S. or Canada.
The  dryer is  the largest of its kind that can be transported over the road.   The components are
mobilized to the project site and assembled using a relatively small 15 ton crane. Approximately three
to four weeks are required to completely install the equipment.  Site preparation involves  grading the
site  level  and providing a  firm base such as compacted gravel. Concrete footings are  not usually
required; however concrete housekeeping  pads may be required. All skids or trailers that normally
contain liquids have integral liquid containment curbs for spill control.

The  system requires three phase, 460 volt electric power, propane storage tanks, and a liquid nitrogen
storage tank.  The electric motors are sized such that the system can be operated from a commercially
available diesel generator if electric power is not available at the site.

Construction of the Model 200 system has been completed and the unit has been functionally tested.
CWM  has a  series  of performance tests planned during  which the  unit will  be operated on non-
regulated  feed materials.
3.     SITE DEMONSTRATION PLANS

In 1988  EPA contracted with CWM to perform a demonstration of X*TRAX under the Superfund
Innovative Technology Evaluation (SITE) program. Under this contract, EPA will perform sampling and
prepare the various reports and CWM will provide the equipment, waste streams, and permits, as well
as operate the equipment.  CWM was to use the five ton/day pilot X*TRAX system on two or three
different  PCB contaminated soils during the TSCA phase of the pilot testing that CWM  conducted at
its Kettleman Hills Facility.  Because of the difficulty of coordinating the schedule for development of
a SITE demonstration test  plan with the short duration permits that CWM  received for the TSCA
testing, EPA elected not to  perform the demonstration as planned.

However, as a result of pilot work done by CWM during the TSCA testing at Kettleman, the company
was awarded a contract to clean up a PCB contaminated site using the X*TRAX Model 200,  full scale
system.  CWM and EPA have agreed to perform a SITE demonstration  during the active part of the
remediation.  This is currently scheduled for late 1990 or early 1991.  A demonstration using the
Model 200 will fully evaluate the operational capabilities of the X*TRAX technology, as well as provide
treatability data.

The  site  of the remediation is primarily contaminated with PCB's,  with some contamination from
chlorinated solvents such as trichloroethene and perchloroethene as well.  The site has a sandy soil
underlain  by clay.  A marshy wetland area  of the  site with high humic content will also require
treatment. The X*TRAX Model 200 will be used to clean the soil, which will  then be returned to the
site as backfill.  A cleanup standard of 25 ppm was established in the Record of Decision (ROD) for
soil returned to the non-wetland portion of the site.

An interesting requirement  of the ROD was that the PCB's be dechlorinated on site as part of the
treatment. In order to perform a chemical dechlorination cost effectively, CWM elected to treat the soil
using X*TRAX, and to dechlorinate only the condensed, recovered oil. The ROD had originally been
developed assuming that the dechlorination would  be performed on bulk contaminated  soil.  By
dechlorinating the condensed oil from X*TRAX, CWM has greatly simplified this step. Also, a much
smaller reactor and less excess chemicals can be used. Furthermore, no chemicals are added to the
soil,  avoiding the  need to establish a secondary treatment standard for residual reagents in the soil
returned to the site.
                                             197

-------
4.     TEST RESULTS

CWM has performed extensive testing as part of the X'TRAX development program. Both simulated
waste materials and actual regulated materials have been tested in the lab and pilot systems.  The
results of most of these tests have been previously reported1'3. The key results to date are summarized
here.  All values for feed and product concentrations are on a dry basis, as is the EPA's convention.
Although the testing program has focused on PCB's to date, this should not be interpreted  as a
restriction on the X*TRAX process from other organic materials.  To the contrary, CWM has pursued
PCB testing as a means to validate the process on one of the most difficult organics to thermally
desorb (because of its very high boiling point).  Processing conditions that are successful with PCB's
will almost certainly work for other volatile and semi-volatile organics.  Also, a large percentage of
Superfund sites that have organics as the primary contaminants, have PCB's as one of the POHCs.


4.1    Lab Testing  Program

The principal use of the laboratory system has been in performing treatability studies for waste streams
or remediation sites that are candidates for the X'TRAX process.  Treatability studies are performed
on a 50-100 Ib sample, and yield results much quicker and at less expense than pilot testing. In all,
23 different test runs have been performed to date, 19 on actual RCRA or TSCA regulated materials.

As a benchmark of the system's capability, a simulated Superfund soil mixture prepared for the EPA
was tested.  Table  1  presents the results.  This material was originally referred to by EPA as the
Synthetic Analytical Reference Matrix, or SARM. It is now called Synthetic Soil Matrix, or SSM. SSM-
1 had high organics concentration and low metals concentration. For both the volatile and semi-volatile
organics, better than 99% removal was achieved.

                                           Table 1

                                  LABORATORY X"TRAX™
                                           SSM-I

Compound
Feed Cone
(ppm)
VOLATILES:
Acetone
Total Xylene
Ethylbenzene
Styrene
Tetrachloroethylene
Chlorobenzene
1,2-Dichloroethane
3,200
2,900
1,900
240
180
130
46
Product Cone
(ppm)
Removal
(%)

16.0
9.50
5.20
< 0.005
0.094
0.180
0.062
99.5
99.7
99.7
> 99.99
99.95
99.86
99.87
SEMI-VOLATILES:
Anthracene
Bis(2-Ethylhexyl)Phthalate
Pentachlorophenol
3,100
3,020
397
12.0
< 0.33
2.8
99.6
> 99.99
99.3
                                            198

-------
To date, eleven lab tests have been performed using TSCA regulated, PCB contaminated soils.  Many
of these tests were developmental, testing different gas system configurations.  The results of this
work have been summarized in Table 2.

                                           Table 2

                                  LABORATORY X'TRAX™
                                   PCB Contaminated Soils
Run #
RS0829
GR0524
GB0504
Matrix
Sandy
Silty Clay
Topsoil
Peed
(ppm)
5,100
962
172
Product
(ppm)
9.7
21
2.8
Removal
(%)
99.3
97.8
98.4
All of these materials had other minor organic contaminants, all of which were generally reduced to
less than detection limits in the treated product (typically in the ppb range).


The most recent test was a series of four tests on soil, pond sludge and mixtures thereof.  These
materials were all from the same site, which is a large remediation project estimated to have in excess
of 500,000 cubic yards of contaminated material. The organic contamination was a complex mixture
of chlorinated semi-volatile organics, aromatics and organic solvents. A summary of the test results
is  presented in Table 3.  Treatment standards for this remediation have not yet been developed.  If
required treatment levels are set on a risk basis, it is very probable that the X'TRAX treated product
would be acceptable.


Presently eight treatability study samples are scheduled for testing in the lab system for this summer.
The agenda is quite full, and continued activity in the area of thermal separation indicates that lab
testing will continue to be in high demand.


4.2    Pilot Testing Program

CWM has performed 57 pilot test runs over the last 2-1/2 years using a variety of simulated, mixed
and TSCA  wastes.  This has resulted in the generation of a wealth of treatability  data as well as
confirmed the operational reliability of the basic process equipment in the X'TRAX system.

Early in the development of the process, a large number of simulated waste streams were tested using
a variety of organic chemicals to spike the feed.  These results  are summarized in Table 4 as an
indication of system performance on various organic chemicals.
                                            199

-------
           Table 3
    LABORATORY X»TRAX™
Non-PCB Soil, Sludges and Mixture
        (Cone = mg/kg)

Run No.
DB0627
Clay Soil




DB0629
Soil/Sludge


DB0706
Sludge






DB0710
Sludge


Parameter
Total Solids (%)
Azobenzene
3,3'-Dichlorobenzidine
Benzidine
2-Chloroanaline
Nitrobenzene
Total Solids (%)
3,3'-Dichlorobenzidine
Azobenzene
Benzidine
Total Solids (%}
Azobenzene
Toluene
3,3'-Dichlorobenzidine
2-Chloroaniline
Benzene
Benzidine
Aniline
Total Solids (%)
3,3'-Dichlorobenzidine
Azobenzene
Concentration
Feed
94.1
3,190
1,820
842
828
45.6
73.1
958
61.0
17.8
52.4
47,900
4,470
3,590
2,100
1,870
1,010
267
47.0
1,070
35.7
Product
100
4.9
<0.66
ND
ND
<0.33
100
<0.66
ND
ND
100
327
<0.42
18.4
47.5
<0.21
3.7
43.3
100
<0.66
ND

Removal (%)
N/A
99.8
>99.96
—
—
>98.6
N/A
>99.0
-
~
N/A
99.3
>99.99
99.5
97.7
>99.99
99.6
83.8
N/A
>99.94
—
           Table 4

       PILOT X*TRAX™
    Surrogate Feed Materials
Compound
Methyl Ethyl Ketone
Tetrachloroethylene
Chlorobenzene
Xylene
1 ,4-Dichlorobenzene
1 ,2-Dichlorobenzene
Hexachlorobenzene
Feed Cone
(ppb)
100,900
91,000
61,810
56,365
78,400
537,000
79,200
Product Cone
(ppb)
< 100.0
15.6
6.5
2.8
1.4
74.1
300.0
Removal
(%)
> 99.90
99.98
99.98
99.99
99.99
99.99
99.62
             200

-------
The pilot system has recently completed a series of tests on ten PCB contaminated soils under a TSCA
R&O permit at CWM's Kettleman Hills Facility.  The last test was completed on January 26, 1990.
Approximately 20 tons of material was tested.

The results of this testing are summarized in Table 5. Nine of the ten materials tested were reduced
to below 25 ppm PCB in the treated product and four were reduced to below 10 ppm. This confirms
at a relatively large scale that X*TRAX can separate PCBs from soil and produce a treated product with
a very low residual PCB concentration.  As with the lab testing, other organic chemicals were present
at lower levels. These were generally reduced to detection limits (typically the ppb range). Oil and
grease (and TPH) were reduced to near or below the 30 ppm detection limit, even when the feed had
45,000 ppm O&G.
                                         Table 5

                                PCB CONTAMINATED SOILS
                                     PILOT X«TRAX™
Run #
0919
0810
1003
0727
0929
Matrix
Clay
Silty Clay
Clay
Sandy
Clay
Feed
(ppm)
5,000
2,800
1,600
1,480
630
Product
(ppm)
24
19
4.8
8.7
17
Removal
(%)
99.5
99.3
99.7
99.1
97.3
During the TSCA testing, the process vent was  continuously monitored  for total hydrocarbon
emissions.  Samples were also taken and analyzed for PCBs. Table 6 summarizes these data.  The
average release rate for hydrocarbons was very low and the PCBs were non-detectable.
                                         Table 6

                                     PILOT X"TRAX™
                              TSCA Testing • Vent Emissions
Run No.
0914
0919
0921
0926
0929
Total Hydrocarbons
(ppm-V)
Before
Carbon
1,320
1,031
530
2,950
2,100
After
Carbon
57
72
35
170
180
Removal
(%)
95.6
93.0
93.3
94.2
91.4
VOC
(Ib/day)
0.02
0.03
0.01
0.07
0.08
PCS'
(mg/m3)
< 0.00056
< 0.00055
< 0.00051
< 0.00058
< 0.00052
       'Note: OSHA permits 0.50 mg/m3 PCB (1254) for 8-hr exposure
                                            201

-------
It should be pointed out that the solids feed system has presented the most problems during operation
of the pilot X'TRAX.  Soils with high sand concentrations presented few problems.  Soils with high
clay content proved  very difficult to convey at a constant rate, and sometimes at any rate. Over the
last two years four  different feed  systems  have been tried.  Each of the  first three were modified
several times before  progressing to the next design. The current feed system has proven itself capable
of metering  and conveying anything from dry sand to damp clay that had to be picked out of an
inverted drum. This design has been incorporated into the Model 200.
5.
DISCUSSION OF RESULTS
In general, results to date  have been quite good for the  X*TRAX  process.  The  basic process
equipment has been reliable, and performance has exceeded expectations, especially  with regard to
the system's capacity.  The area that deserves special attention is that of comparing the residual
organic levels in the product with various applicable treatment standards.

Most of the test data to date on regulated material has been on PCB contaminated soils.  Discussions
with EPA have centered around meeting a treatment standard of 2 ppm. This is the analytical method's
quantitation limit for RGB's  when un-modified procedures are used. As can be seen  from the data,
X*TRAX has not achieved  this standard.   However, there is little or no  basis for  this numerical
standard present in the  regulatory citations.  Rather, it has been implemented as a specific  permit
condition on a case by case basis. The PCB disposal regulations (40 CFR 761.60) simply state that
non-liquid PCB's with concentrations over 50 ppm and less than 500 ppm have to be  disposed in an
approved landfill and that solids over 500 ppm have to be incinerated. The TSCA spill regulations (40
CFR 761.125) give cleanup requirements for soil as  either 25 or 10 ppm, depending  on the level of
restricted access to the site.

Recently, the EPA has issued Superfund LDR #6A4 as guidance to allowable treatment standards for
Superfund soil and debris with respect to the RCRA  land disposal restrictions. This document gives
recommended numerical treatment standards to apply to Superfund soil and debris  when the land
disposal restrictions cannot be met.  For PCBs the treatment standard is given as 0.1 -10 ppm for soils
containing  less than 100 ppm, and 90 - 99.9% removal for soils  greater than  100 ppm.   Similar
standards are given for other classes of organic compounds.

Table 7 is the same treatability data presented in Table 5, with the percent  removal column replaced
with the maximum residual level allowable under LDR guide #6A. Clearly the X*TRAX process meets
the residual levels that could be allowed under these treatability guidelines. It would seem appropriate
to strike a compromise position, somewhere between the LDR Guide's allowable values and the 2 ppm
defacto standard.

                                          Table  7
                                PCB CONTAMINATED SOILS
                                      PILOT X*TRAX™
                               Comparison With LDR Guide #6A
Run #
0919
0810
1003
0727
0929
Matrix
Clay
Silty Clay
Clay
Sandy
Clay
Feed
(ppm)
5,000
2,800
1,600
1,480
630
Product
(ppm)
24
19
4.8
8.7
17
Required Lever
(max)
500
280
160
148
63
        •Based on 90% removal
                                            202

-------
An interesting observation about the residual PCB levels from the X'TRAX process is that the product
concentration is relatively insensitive to the feed concentration.  The product values are generally in
the 5 - 25 ppm range, even when the feed is 5,000 ppm or higher.  The residual levels are however,
strongly dependent on the solids temperature and to a lesser extent the soil matrix. This is a source
of comfort when working on a site where there are "hot spots." If the required treatment level is being
achieved with the average feed material, the treatment of a hot spot should not exceed the standard.

Two of the materials tested in the pilot  system were also tested in the lab system.  Table 8 is a
comparison of the results of these  tests.   Clearly, the residual levels in the product show good
agreement, indicating that the lab system closely simulates the pilot system.
                                          Table 8

                         COMPARISON OF LAB AND PILOT X'TRAX™
                                   PCB Contaminated Soils
Matrix
Sandy
Silty Clay
System
Lab
Pilot
Lab
Pilot
Run ID
RS0829
RS0727
GR0524
GR0810
Amount
(Ib)
19
4,958
31
4,584
Feed
(ppm)
5,100
1,480
962
2,800
Product
(ppm)
9.7
8.7
21
19
6.     COSTS

As is typically the case with waste treatment processes, the operating economics of X'TRAX are
highly dependent on the waste type, specific contaminants and other project specific factors such
as disposal requirements and restrictions, and the amount of material to be processed.  All of this
said, the treatment price for a material that meets the broad scope of acceptable feed for X*TRAX
will usually fall in the $150 - $250 per ton of feed.  This treatment price is in the appropriate range
to make X'TRAX very competitive with both traditional disposal methods and also alternative
treatment  processes.

As with any on-site method, *•  mobilization charges to bring equipment on site need to be spread
over the volume of material re _ jiring treatment. To keep these mobilization charges within reason,
a lower limit of about 5,000 cubic yards is appropriate for the X'TRAX Model 200.
7.
SUMMARY AND CONCLUSIONS
CWM has progressed to a point in the development of the X'TRAX thermal separation process at
which the following statements can be made:

 •     the  X'TRAX thermal separation process has been demonstrated as effective on actual
       waste materials at both the lab and pilot scale

 •     pilot data has shown the process to have very low atmospheric emissions, both for PCBs
       and total organics
                                            203

-------
       very low residual contamination levels have been achieved for the treated product, typically
       in the 5-25 ppm range for PCBs and at the ppb level for most other organics, however, a
       2 ppm treatment standard for PCBs cannot presently be met

       the process is tolerant of wide variation of organic levels in the feed

       the commercial system is easily modified to add stabilization for metal containing wastes

       the process is economically viable, with treatment price in the $150 - $250 per ton; it is
       suitable  for remediations >5,000 yd3
                                        References

1.     Swanstrom,  C.,  Palmer,  C., X*TRA)(rM  Transportable  Thermal  Separator  for  Solids
       Contaminated with Organics, Air & Waste Management Association International Symposium
       on Hazardous Waste Treatment: Treatment of Contaminated Soils, Cincinnati, Ohio, February
       5-8, 1990.

2.     Daley,  P.S.,  Cleaning  up Sites  with On-Site Process Plants,  Environmental  Science  &
       Technology, August,  1989.

3.     Palmer, C.R., Hollenbeck, P.E., Sludge Detoxification Demonstration, Incineration Conference.
       May 1-5, 1989.

4.     USEPA - Office of Solid Waste and Emergency Response, SuperfundLDR Guide #6A: Obtaining
       a Soil and Debris Treatability Variance for Remedial Actions, Directive 9347.3-06FS, July
       1989.
                                            204

-------
 ANAEROBIC PYROLYSIS FOR TREATMENT OF ORGANIC
   CONTAMINANTS IN SOLIDS WASTES AND SLUDGES -

         THE AOSTRA TACIUK PROCESS SYSTEM

                 THERMAL TREATMENT
AUTHOR:
    Robert M. Ritcey, P.Eng.,
    Manager, Demonstration Operations
    UMATAC Industrial Proceses
    A Division of UMA Engineering Ltd.
    Calgary, Alberta
PRESENTATION TO:

     Second Forum on Innovative Treatment Technologies:
     Domestic and International
     Philadelphia, PA
     May 15-17, 1990
                            205

-------
       ANAEROBIC PYROLYSIS FOR TREATMENT OF ORGANIC
          CONTAMINANTS IN SOLID WASTES AND SLUDGES -
                THE AOSTRA TACIUK PROCESS SYSTEM
                                  ABSTRACT


The AOSTRA Taciuk Process (ATP) is a thermal treatment technology which has been
developed and proven effective for separating organic contaminants such as oils and
petrochemicals from solid wastes, soils and sludges. The Process was developed jointly by the
Alberta Oil Sands Technology and Research Authority and UMATAC Industrial Processes for
producing bitumen from Alberta's oil sands deposits as distillate oil product, and for producing
oil from oil shales. The extensive test work in these fields, and recent work directly with waste
materials are the background to the recent commercial application of the ATP System in waste
treatment.

The ATP technology is a continuous flow pyrolysis system which achieves the separation of
organic contaminants from host solids in the wastes, and offers the advantages of minimum air
emissions, ability to recycle the organics, rapid treatment of waste volumes, and low cost. The
ATP System has been extensively tested and its capability demonstrated for separating water
and organic contaminants from soils and sludges.

The plant is a recycling facility for those constituents in the waste which are re-usable. It
renders the solids free of the organic contaminant(s), the liquid products (water and oil) for re-
use, secondary treatment or disposal, and the air emissions meet regulatory criteria applicable
to thermal remediation treatment of wastes.

The first commercial ATP treatment plant has a design capacity of 10 tph and was ready for
service in September, 1989. It will be employed in waste treatment, initially  on  a PCB
separation soil treatment Superfund project in the U.S.A. The plant is portable and will be
used on numerous job sites in North America. Other remediation projects and applications for
the ATP technology  and portable treatment plants of various sizes are being developed.

This paper describes the ATP  System, its scope of applicability in waste treatment, and
considerations for specific applications of portable ATP treatment plants. Performance results
from demonstration  treatment programs and treatment cost data are discussed.
2nd FORUM Mty 15-17,1990

                                      206

-------
       ANAEROBIC PYROLYSIS FOR TREATMENT OF ORGANIC
          CONTAMINANTS IN SOLID WASTES AND SLUDGES -
                 THE AOSTRA TACIUK PROCESS SYSTEM

The AOSTRA Taciuk Process (ATP) is a continuous flow pyrolysis treatment system which is
applicable to  the remediation of various waste soils and  sludges  containing organic
contaminants, such as oils or other hydrocarbon materials and products. This capability
evolved from the extensive development of the ATP technology as a process for producing oil
from oil sands and oil shales which has been conducted since 1976 by UMATAC Industrial
Processes division of UMA Engineering Ltd, under agreement with the Alberta Oil Sands
Technology and Research Authority (AOSTRA).

Over the past five years, the development work has included physical and production testing of
small and large quantities of oily soils from many locations in North America. The work
included tests on contaminated soils and sludges from the oil and petrochemical industries,
wastes from municipal  and urban sites such as old industrial sites or dumps, and from
orphaned or Superfund sites. Contaminants found in these sites include organic and inorganic
materials,  and the ATP technology was tested and demonstrated for separating and recovering
the organics, such as oils, PCBs, coal tars and acid chemicals.

UMATAC offers the ATP technology both in the form of technical expertise and treatment
plant facilities which are  currently available for  commercial application.  The treatment plants
can be designed and provided in a wide range of sizes and configurations to meet the needs of
various situations and clients. The technology is sub-licensed for commercial use in the United
States to SoilTech,  Inc., which is an operating company jointly held by  UMATAC  and
Canonic Environmental Services Corp.

In this remediation work on sites and materials, the need is usually for a treatment facility
which is portable and easily applied to the job, maintains performance capability when treating
wastes of differing characteristics,  and can achieve the treatment at reasonable cost. The ATP
System offers many features to meet these requirements, such as:

      plants in sizes up to 25 tph of feed capacity, or greater as needed.  Larger plants are of
      modularized construction to facilitate moving. Most inquiries are for small,  portable
      plants of 5 tph capacity, or less.
      continuous, stable, safe and efficient operation with widely varying feed materials.
      effective separation of organic contaminants from host soils or other inorganic
      materials,  rendering the soils free of the contaminants (oil, PCB, PAH, etc.).
      recycling of the contaminants which are re-usable, such as oils and petrochemicals.
      ability to operate within, and meet permitting criteria.


The ATP System

The AOSTRA Taciuk Processor and the associated processes of the ATP  System are shown in
Figures  1 and 2.  A technical description of the System is briefly summarized;  more complete
technical descriptions are  available in the references.

The Processor (Figure 1) is a horizontal rotating vessel which can receive feed wastes in
physical form varying from solids to liquids, and which contain varying concentrations of .pa
2nd FORUM Miy 15-17, 1990

                                         207

-------
              RUE MS
       STEAM
                                                           AUOMff
      CLEAN DRY*
     TA1UNGS SAND
                                  __  _             —     	_ MUJ^^^pHHl

                      COOUNG ZONE   |    COMBUSTION ZONE   |**NER
                          ATP PROCESSOR
                     INTERNAL PROCESS FLOWS
                               FIGURE 1
                                             {COMBUSTION
                                     AJR
      STACK
       L
          WATER
FLUE GAS
TREATMENT
L
       Oil
        t
                FUEL
 VAPOUR
TREATMENT
                                         OFF CAS
                  ATP
               PROCESSOR


               OIL FREE
               TAILINGS
                                                    SOURl WATER
                                                   WATER
                                     1
1 * I
p VAPOUR "
OIL
RECOVERY
PRODUCT
(XL l
                   AOSTRA TACIUK PROCESS  (ATP)
                              FIGURE 2
2nd FORUM M«y 15-17, 1990
                                      208

-------
water, solids and organic contaminants.  The usual requirement for treating the waste is to
separate the contaminants from the inorganic solids and render the solids non-toxic so they can
be re-used or disposed directly to landfill.

The Processor effects separation of the solids, water and organics (liquids and solids) by
heating in the first zone (preheat) where the feed is received, and pyrolysis in the reaction
zone. Pyrolysis temperature is usually between 500 and 600°C,  and the operating pressures of
the zones are near atmospheric (-2 to -3 mm).

Heat supply for the Process is generated in the same vessel in a combustion zone which is
situated around the pyrolysis reaction zone as  an annular section. The Processor makes use of
non-vaporizable organic material (coke) which passes with the solids to the combustion zone
from the reaction zone,  and non-condensible hydrocarbon vapor products of the pyrolysis for
process heat fuels. These fuels and the  solids are de-contaminated in the reaction zone before
they enter the combustion zone. Auxiliary industrial fuels such as gas or oil, must be available
for stan up, control and supply to augment any inadequacies in energy  for process heat from
the internally produced fuels.

An important and unique feature of the Processor is the recovery and use of heat from the
solids and combustion gases in the cooling zone, before these products are emitted from the
vessel. This heat is usually sufficient for the requirements of the preheat zone; i.e., to heat the
incoming feed to about 200°C to vaporize the water in the feed  and prepare the oily solids for
the higher temperature retort zone.

The auxiliary process systems are shown in Figure 2 in block flow form, integrated with the
Processor. These systems handle the materials (feed and products), condense the hydrocarbon
vapors and steam to liquid oil and water products, and treat the combustion gases for an
acceptable stack emission.

The products of the ATP System are oil-free solids which are usually suitable for direct
disposal to landfill, and water and oils. The water usually requires secondary treatment such as
steam stripping and biological treatment to remove the organics and reduce the oxygen
demand. Oil from oil industry-type  wastes is usually suitable for addition to refinery
feedstocks; hazardous type oil products can be separately disposed as a relatively small
quantity for incineration by an ATP plant combustor, or by an off site facility. In most cases, it
is important to also establish the levels of non-organic contaminants in the feed, such as heavy
metals or other  hazardous materials, in order to provide the necessary,  secondary treatment for
these in addition to the organics.

The ATP System thus consists of distinct process functions - feed supply, the Processor for
separation of phases,  vapor handling for liquids recovery, solids product handling,  flue gas
treatment and plant control. In the range of process capacities normally required for waste
treatment plants, the System equipment readily lends itself to the modular design needed for
highly mobile, portable plants.

Use in Treating Contaminated Soils and Sludges

The Process has been tested on a wide range of wastes in addition to the oil sands and
shales4*5. The testing during the development and proving phase for the process and mechanical
equipment was  conducted on approximately 14,000 tonnes of oil sands and shales.  Test work
on wastes began in 1986 and has included about 500 tonnes of oily soils and sludges from the
production, refining  and marketing sectors  of the oil industry, waste rubber and  plastics


2nd FORUM M«y 15-17, 1990

                                           209

-------
products, and PCB contaminated soils. In 1989, about 1000 tonnes of oily materials were
treated with the 10 tph commercial plant in the course of plant testing and demonstration.

In addition to these tests on bulk samples using the ATP plants, about 500 small scale batch
tests have been  run on the various wastes. Batch tests simulate the vaporization and pyrolysis
reactions of the ATP System and are used for initial evaluation of the ATP for treating a
waste.

The Processor can treat any mixture of solids and liquids, to an upper limit on solids size of
six to eight inches, so large objects must be separated from the feed by sorting or screening.
Depending on the treatment situation,  sludges can be fed directly as liquid, or pre-mixed with
solids (waste or carrier sand) to enable handling by conventional loader-hopper-conveyor
equipment.

The work and results from four (4) different test programs on oil industry wastes illustrate the
capability and consistency of the ATP System to treat these wastes. Data of analysis
determinations for some of the key contaminants in these wastes are shown in Tables 1 through
5. The program to test PCB contaminated soil  is described as the fifth reference.

The wastes treated in these programs were:
1)     Refinery waste stream sludges  - EPA classifications K051,  K048, and K049. These are
       sludges from API Separators, DAF and slop oil emulsion, respectively.
2)     Oily soil from a decommissioned refinery, labelled "A".
3)     Oily soil from a decpmmissioned refinery, labelled "B".
4)     Drilling  cuttings which contain diesel oil, from use of invert type drilling  mud.
5)     PCB contaminated soils.

1)     Refinery waste stream sludges

UMATAC participated in the 1987 program of the American Petroleum Institute (API) Task
Force which studied technologies applicable and  ready for use in treating the  "K" waste
streams of refinery operations in the  United  States. The API wished to determine treatment
options prior  to the imposition by the EPA of the ban on disposal and treatment of these wastes
in land fills.

The products of the ATP tests, which are oil, water and combusted solids, were analysed by
commercial laboratories for characterization of organic and inorganic constituents. This
procedure is  also followed with most of the feeds and products of tests of this type on other
wastes to assess the environmental characteristics of the materials. The results showed  that the
treated solids were oil  free, and that  most of the metals and salts in the feed remain with the
solids.

The API published a report of its study of the technologies, and reported that pyrolysis (the
AOSTRA Taciuk Process) is the "most efficient in removing" organic contaminants from the
solids and in  achieving low leachability of organic  compounds from the product solids. The
reference is API Report #4465, published in May, 19887.

2,3) Refinery wastes  - from Sites "A"  and "B"

A 35 tonne sample of material from the clean-up of a decommissioned refinery ("A") was
received for  treatability tests and plant test operation in 1986. The sample was a blend of a
number of oily sludges and soils at the site. It was processed in the 4.5 tph pilot plant.  A
2nd FORUM May 15-17, 1990

                                          210

-------
second major sample from another site ("B") was received in  1989.  It was about 150 tonnes,
and was processed in the commercial, 10 tph plant.

The treated solids from these tests contained little or no oil and grease. Leachate data on solids
products showed oil and grease levels of 0.7 mg/1 or less, with metals concentrations being low
or less than detectable.  Additional analyses showed monocyclic and polycyclic aromatic
hydrocarbons at or below detection limits in the leachate from  the solids residue.

The flue gas paniculate discharge was less than required levels, with metals emissions being as
much as 1000 times less than the criteria.  Typical emissions data are presented in Table 5. In
the "B" tests, flue gases were sampled using seven sampling trains.  Each train complied with
the approved methods of appropriate regulatory agencies, which were the sampling and
analysis protocols of EPA MM5, or equal.

The results for particulates, HC1, SO,,  and oxides of nitrogen are well within criteria.  There
is a general lack of information in industry on emissions criteria for PAH,  MAH, and phenols
in the stack gases, but total concentrations of these compounds were 1.5, 130 and 70 ug/m3.
The heavy metals concentrations are generally much less than the emissions criteria. Carbon
monoxide reported a maximum of 190 ppm, to a low of 130 ppm, which meets criteria in some
jurisdictions.

Generally all the water products were contaminated at constituent levels above the emission
criteria, and would  require further treatment for reuse or disposal. Possible treatments  would
include steam stripping, air flotation and activated carbon beds. The sour  water, being a
product of the hydrocarbons cracking, contained high quantities of phenols, and detectable oil
and grease. The preheat waters are not exposed to pyrolysis conditions and are  not expected to
contain phenols.

4) Invert drilling mud cuttings

Invert drilling muds are comprised of large amounts of diesel oil. Most of the mud, with oil,
is recovered from the cuttings and recycled in the drilling operation. The separated cuttings
usually still contain a significant amount of oil, averaging 15% by weight. Nearly  120  tonnes
of these cuttings were obtained for testing at the AOSTRA-UMATAC plant in Calgary.

The cuttings, as received, were quite sloppy and had to be contained by a berm. A  full range
of batch, pilot and commercial unit treatment operations in 1988 and 1989 proved oil removal
capabilities, with more than 60 % of the oil being recovered as butane and heavier liquids. The
treated solids were oil free and suitable for reuse or disposal as landfill.

The flue gas paniculate emission was less  than half the allowable criteria, as presented in
Table 2. The solids effluent leachate data showed low levels of chlorides, sulphates and  metals
as presented in Table 3.

5) PCB Contaminated Soil

Soils contaminated with PCB Arochlor 1242 were treated in bulk sample tests  using the 5 tph
Processor unit in 1988. The PCB concentrations ranged from 200 to 20,000 ppm,  and 27 tons
of soil were treated in 2  test runs.  The results showed the soil to be cleaned; the PCBs are
vaporized in the reaction zone and removed as a vapor. No PCBs are subjected to combustion,
and furans and dioxins were not produced. The DRE of the PCB from the feed met the six
nines requirement.
 2nd FORUM May 15-17, 1990

                                            211

-------
This test work demonstrated the ATP System for acceptance of use at the \\fcukegan Harbour
Superfund Site remediation. Soil treatment operations at this site are scheduled for 1991.


Costs and Sensitivities

The capital supply cost of ATP plants ranges from $ 3 million to $ 8 million for plant sizes of
3 to 20 tph, respectively. Table 6 shows approximate costs and operating data for these plant
sizes in applications on non-hazardous waste such as soils and product streams from refineries.
Higher costs are expected when treating more complex wastes which may require secondary
treatment or costly disposal of separated constituents.  Specific data must be prepared for each
usage situation. The costs are reasonably indicative for a commercial facility, but must be
defined much more accurately by specific examination for each requirement.

If the waste quantities per plant set up are less, the unit costs increase because  most costs are
fixed, such as a permanent operating staff and capital recovery costs,  and the production is
reduced while the plant is moved or waits for feed. The annual operating costs for all plants
larger than 5 tph to 20 tph are much  the same for labor because the cost of the  operating staff
is a large portion of the total. The cost for most operating supplies, especially consumables,
varies with the throughput.

Besides the cost of plant operations, the 2 main factors which influence the total unit cost of
treatment are plant supply cost and utilization. The plant usage is  a function of the relocation
frequency and the quantities of waste for treatment. The costs shown are  for the plant supply
and basic operation, and are only generally indicative of the economics of the ATP System.
They do not include other costs, such as site development, permitting, waste supply and
preparation, products disposal, local  requirements affecting labor costs or special monitoring,
analysis and reporting of products and emissions. These items are all peculiar to each waste
situation.

Other factors which can affect the costs of treatment by the ATP System are:
       contents of moisture and oil in the waste.
       particle size distribution of the solids
       acid content of the wastes
       vaporization of the organic materials
       handling or disposing the condensed hydrocarbon products
       mobility of metals in the solids products.

In  summary, the  total costs of pyrolysis treatment of oily wastes with the ATP  System are
generally in the  range of $100 to $200  per tonne of feed, but can range higher or lower
depending on the situation. Costs of direct plant operation only range  from $40 to $150 per
tonne. The objective is to correctly size the plant for continuous operations, which will enable
minimum cost of treatment. Costs of thermal treatment may not be as  low as  other methods
such as  land treatment, but the remediation considerations must include environmental
acceptability of the method and products, and the time requirements.

Alternatively, the costs and economics  can be considered  in a completely different manner in
the case of a large production plant (oil, petrochemical,  etc,) into which an ATP plant is
incorporated to treat waste products or accumulations, and which is operated as part of the
large plant. For this situation, the mobilization  would occur only once, and the plant supply
cost could be amortized over a longer period, say  15 years, and the plant staff cost could be
 2nd FORUM May 15-17, 1990

                                            212

-------
lower as the crew is part of a larger work force.

ATP SYSTEM ADVANTAGES

The ATP System pyrplysis technology offers a number of unique features for consideration of
its use in treating various oily, organic wastes. They include the following:

      the Process has the intermediate step of extracting  the contaminants from the solids,
      which presents them as available for re-use or refining, rather than destruction.
      treatment plants can be supplied for large throughput capacity, to 20 tph or greater.
      low unit costs of plant operation, particularly for the larger capacity plants.
      all of the treatment process steps for extraction of the contaminants are confined to a
      single process vessel.
      the Process operates at low to intermediate temperatures (to 600°C), which enhances
      safety and indicates low energy requirement.
      water and oil contaminants are recovered as separate products, which makes them
      available for secondary treatment, re-use or separate  disposal.
      solids products are hydrocarbon-free, and can usually meet leachate criteria for  direct
      disposal.
      heat make up uses Process by-products as fuel.
      combustion conditions include low gas velocity, long residence time for the fuels, and
      extensive turbulence and mixing in the combustion zone. This assists combustion
      completion and control of the gas emissions
      the System has been extensively tested and shows capability to operate with stability
      and safety on widely varying wastes.

Much interest  has been expressed by industry and regulatory representatives in the AOSTRA
Taciuk Process System, particularly since the first commercial waste treatment plant became
available and was purchased by SoilTech, Inc. The ATP pyrolysis technology can address a
broad field of applications for treating hazardous and non-hazardous industrial wastes. It  offers
fast clean up capability combined  with the recovery and reuse of the contaminants,  while
operating within the  stringent emissions standards required in today's modern world.


References

1.    R.M. Ritcey, "Anaerobic Pyrolysis  of Solid Wastes and Sludges - The AOSTRA
      Taciuk Process System", HAZTECH Canada Conference, Edmonton, Alberta,  October
      1989.

2.    L.R. Turner, "Treatment of Oilsands and Heavy  Oil Production Wastes Using the
      AOSTRA Taciuk Process", Conference  on Oil Field Production Wastes.  May  1989,
      Calgary, Alberta.

3.    American Petroleum Institute, Health and Environmental Sciences Department,
      Washington,  D.C., Publication No. 4465.  "Evaluation of Treatment Technologies for
      Listed Petroleum Refinery Wastes", May  1988.

 4.    W. Taciuk, R.M. Ritcey, "Taciuk Processor  for Treatment of Contaminated Wastes",
      AOSTRA Conference,  Advances in Petroleum Recovery and Upgrading  Technology
      1987. Edmonton,  Alberta.
2nd FORUM May 15-17, 1990

                                          213

-------
5.    M. Vorum, A.H. Montgomery, "The Taciuk Process Technology For Anaerobic
      Pyrolysis Of Solid Wastes And Sludges", 1989, Canonie Environmental Services
      Corporation. Englewood, Colorado.

6.    SoilTech, Inc., "The Taciuk Process Technology: Thermal Remediation of Solid
      Wastes and Sludges". Technical Information Manual.

7.    American Petroleum Institute, "Evaluation of Treatment Technologies for Listed
      Petroleum Refinery Wastes". API Publication No. 4465, Health and Environmental
      Sciences Department, May 1988. API, 1220 L Street, Washington, D.C. 20005.
2nd FORUM May 15-17, 1990


                                         214

-------
TESTING UNIT
YEAR TESTED
        TABLE 1 - FEED CHARACTERISTICS

                       TEST RESULTS DATA
    AOSTRA TACIUK  PROCESSOR TESTS ON OIL CONTAMINATED SOILS

                            Refinery Refinery   Drill
                      API    Hastes   Hastes     Mud   CRITERIA
                    PROGRAM    A"        B"    Cuttings ENVIRON-
- PLANT              BATCH    Pilot   10 tph    Pilot  MENT
                     1987     1986     1989     1989    CANADA
PROCESSOR FEED:
Oil /Grease Ht %
Hater Ht %
Solids Ht %
HASTE ONLY CONSTITUENTS:
Oil/Grease Ht %
Hater Ht %
Solids Ht %

2.1
17.7
80.2

20 1.8
52 30.0
28 68.2

0.3
10.9
88.8

0.5
14.0
85.5

15.3
7.3
77.4

15.3
7.3
77.4
                   TABLE 2 - FLUE GAS CHARACTERISTICS
PARTICULATES

METALS
     mg/m3

     mg/m3
   Arsenic
    Barium
   Cadmium
  Chromium
    Cobalt
    Copper
      Lead
   Mercury
    Nickel
  Selenium
  Vanadium
      Zinc
979*   17.655
18.9
0.040
30
-0.0001
0.105
0.025
0.425
0.430
0.021
0.150
0.004
0.450
0.045
-0.0031
1.3104
0.0035
0.0071
0.0192
0.0017
0.0070
0.0008
0.9638
-0.0025
0.0037
0.7603
50
                                                                         .2
                                                                          1
                                                                          1
                                                                          5
                                                                          5
                                                                         .2
                                                                          1
                                                                          1
                                                                          5
                                                                          5
* - NO BAGHOUSE IN THE FLUE GAS TREATMENT; HET SCRUBBER ONLY.
2nd FORUM M»y 15-17, 1990
                                     215

-------
TESTING UNIT
YEAR TESTED
  TABLE 2 - FLUE GAS CHARACTERISTICS (cont'd)

                       TEST RESULTS DATA
    AOSTRA TACIUK PROCESSOR TESTS ON OIL CONTAMINATED  SOILS

                            Refinery Refinery   Drill
                      API    Wastes   Wastes     Mud   CRITERIA
                    PROGRAM    A"       B"    Cuttings ENVIRON-
- PLANT               BATCH    Pilot   10 tph    Pilot  MENT
                      1987     1986     1989     1989    CANADA
HCL ug/m3
MAH's ug/m3
Benzene
Ethyl Benzene
Toluene
Styrene
Xylene
PAH's ug/m3
acenaphthene
acenaphthylene
anthracene
benzo ( a ) anthracene
benzo ( a ) pyrene
benzo (b) f luoranthene
benzo (k) f luoranthene
benzo(g,h, i)perylene
chrysene
dibenzo ( a , h ) anthracene
f luorene
fluoranthene
indeno ( 1 , 2 , 3 , c , d ) pyrene
naphthalene
phenanthrene
pyrene

3516
70
8067

2404
7.720
1.520
0.745
0.245
-0.120
0.065
-0.120
0.000
0.125
0.000
0.930
2.595
0.000
6.105
4.570
3.275
6 75000
69.0
2.6
13.1
0.6
18.1
-0.091
0.090
0.110
-0.091
-0.091
-0.091
-0.091
-0.091
-0.091
-0.091
-0.091
0.160
-0.091
0.596
0.537
0.110
NOTE:  (-nn)  indicates  'value  less than detection limit of nn'
       {-)  indicates  'NOT DETECTED'
       {	)  indicates  'not tested for1
       (NA)  indicates  'Not Applicable'
2nd FORUM Miy 15-17, 1990
                                       216

-------
        TABLE 3 - LEACHATE CHARACTERISTICS OF TREATED SOLIDS

                                   TEST RESULTS DATA
                 AOSTRA TACIUK PROCESSOR TESTS ON OIL CONTAMINATED SOILS
TESTING UNIT
YEAR TESTED
- PLANT
        Refinery  Refinery   Drill
  API    Wastes   Wastes     Mud   CRITERIA
PROGRAM    A"        B"    Cuttings ENVIRON-
 BATCH    Pilot   10 tph    Pilot  MENT
 1987     1986      1989     1989    CANADA
Oil/Grease
Specific Cond,

MAJOR CATIONS
METALS
    mg/1
  umho/cm

     mg/1
  Chloride
  Sulphate

     mg/1
   Arsenic
    Barium
     Boron
   Cadmium
  Chromium
      Lead
   Mercury
  Selenium
    Sodium
  Vanadium
      Zinc
    1.0
                                    -0.1
                                    -0.1
NONE
                                              520
                                             2720
           0.026
            0.14
            0.59
          -0.001
          -0.001
            0.02
          0.0003
           0.011

            0.53
           0.064
0.7
                    40.1
                     824
                              180
                   165
                   109
         0.03
         0.12
                                                   -0.0008
                    5
                  100
                  500
                  0.5
                    5
                    5
                  0.1
                    1
                                                                3.5
                                                      0.07
2nd FORUM May 15-17, 1990
                                      217

-------
    TABLE 3 - LEACHATE CHARACTERISTICS OF TREATED SOLIDS (cont'd)

                                  TEST RESULTS DATA
                AOSTRA TACIUK PROCESSOR TESTS ON OIL CONTAMINATED SOILS
TESTING UNIT - PLANT
YEAR TESTED
ORGAN I CS:
MAH's mg/1
Benzene
Ethyl Benzene
Toluene
Styrene
Xylene
PAH's ug/1
acenaphthene
acenaphthylene
anthracene
benzo ( a ) anthracene
benzo ( a ) pyrene
benzo ( b ) f luoranthene
benzo(k) f luoranthene
benzo (g/h, i)perylene
chrysene
dibenzo ( a , h ) anthracene
f luorene
f luoranthene
indeno ( 1 , 2 , 3 , c , d ) pyrene
naphthalene
phenanthrene
pyrene
NOTE: (~nn) indicates 'value
Refinery
API Wastes
PROGRAM A"
BATCH Pilot
1987 1986
XXX XXX
{-) -0.2
-0.2
(-) -0.2

(-) -0.2
XX
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
less than detection
Refinery Drill
Wastes Mud CRITERIA
B" Cuttings ENVIRON-
10 tph Pilot MENT
1989 1989 CANADA
X
-0.002
-0.002
0.004
-0.002
0.006
X
-1.0
-1.0
-1.0
-1.0
-1.0
-1.0
-1.0
-2.0
-1.0
-1.0
-1.0
-1.0
-2.0
-1.0
-1.0
-1.0
limit of nn'
       (-)  indicates 'NOT DETECTED'
       (	) indicates 'not tested for1
   x - ANALYSED BY GC/MS
  xx - ANALYSED BY HPLC (units are ppm)
 xxx - ANALYSED BY GC/FID
2nd FORUM May 15-17, 1990
                                     218

-------
             TABLE 4 - CHARACTERISTICS OF TREATED SOLIDS
                                    TEST RESULTS DATA
                 AOSTRA TACIUK PROCESSOR TESTS ON OIL CONTAMINATED SOILS
TESTING UNIT
YEAR TESTED
- PLANT
        Refinery Refinery
  API    Hastes   Wastes
PROGRAM    A"       B"
 BATCH    Pilot   10 tph
 1987     1986      1989
  Drill
   Mud   CRITERIA
Cuttings ENVIRON-
  Pilot  MENT
  1989    CANADA
Oil/Grease       nig/I
Specific Cond.  umho/cm
MAJOR CATIONS
METALS
     mg/1
  Chloride
  Sulphate

     mg/1
  Aluminum
   Arsenic
    Barium
     Boron
   Cadmium
  Chromium
    Copper
      Lead
    Nickel
  Selenium
  Vanadium
      Zinc
                                NONE
                                              500
                                             2800
                      50
                      88
                     422
      17
    4230
    3290
    2180
13500
5.1
86.5
33.4
-0.1
26.3
195
0.65
0.3
26.00
31.00
38
39.00

122
142
70.00
2nd FORUM M«y 15-17, 1990
                                       219

-------
       TABLE 4 - CHARACTERISTICS OF TREATED SOLIDS - ORGANICS

                                    TEST RESULTS DATA
                 AOSTRA TACXUK PROCESSOR TESTS ON OIL CONTAMINATED SOILS
TESTING UNIT
YEAR TESTED
             - PLANT
        Refinery Refinery
  API    Wastes   Wastes
PROGRAM    A"       B"
 BATCH    Pilot   10 tph
 1987     1986     1989
  Drill
   Mud   CRITERIA
Cuttings ENVIRON-
  Pilot  MENT
  1989    CANADA
ORGANICS:
MAH's
                  mg/1
               Benzene
          Ethyl Benzene
               Toluene
               Styrene
                Xylene
PAH's
                  ug/1
           acenaphthene
         acenaphthylene
             anthracene
     benzo(a)anthracene
         benzo(a)pyrene
   benzo(b)fluoranthene
   benzo(k)fluoranthene
   benzo(g,h,i)perylene
              chrysene
 dibenzo(a,h)anthracene
              fluorene
           fluoranthene
indeno(1,2,3,c,d)pyrene
            naphthalene
           phenanthrene
                pyrene
                                   XXX
                                     -0.1

                                        3

                                      0.8
                                            XXX
                                            XX
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                                            -0.01
                     X
                     -0.1
                     -0.1
                     -0.1

                     -0.1

                     X
                    -0.05
                    -0.05
                    -0.05
                    -0.05
                    -0.05
                    -0.05
                    -0.05
                     -0.1
                    -0.05

                    -0.05
                    -0.05
                     -0.1
                    -0.05
                    -0.05
                    -0.05
NOTE:  (-nn) indicates 'value less  than detection  limit of nn1
       (-) indicates 'NOT DETECTED'
       (	) indicates 'not tested for'
   x - ANALYSED BY GC/MS
  xx - ANALYSED BY HPLC (units are  ppm)
 xxx - ANALYSED BY GC/FID
2nd FORUM May 15-17, 1990
                                      220

-------
                                  TABLE 5

                           FLUE GAS EMISSIONS
                 AVERAGES FOR THREE TESTS - MB" SAMPLE

Particulates
Hydrogen Chloride
Total Specified Metals
Sulphur Dioxide
Oxides of Nitrogen
Total Reduced Sulphur Compounds

Emissions
Criteria
mg/m *
20a
75a

200a
300a


Concentration
mg/m *
17.655
0.006
3.360
1.400
108.6
<3.0
Concentration
ug/m *
Emission
mg/h
0.090
<0.001
0.017
0.007
0.6
<0.016
Emission
mg/h
Total Poly-Aromatic Hydrocarbons


Total Non-Chlorinated Phenols


Total Mono-Aromatic Hydrocarbons
   1.330          6.812


  69.246        344.853


 127.6          645.9
      At 25°C, 760 mm Hg, dry basis, O2 cone. 11.8 % (avg)
      Environment Canada Criteria for Haz. Waste Incinerators and Thermal Treatment
      Facilities, O2 cone. 11 %, 25°C, 760 mmHg.
Mole Percentages  of Total Emission:

Carbon Monoxide


Carbon Dioxide


Hydrogen Sulphide
0.016


7.82


<0.0001
2nd FORUM May 15-17, 1990
                                      221

-------
                                   TABLE 6
               BASIC COSTS OF ATP PYROLYSIS TREATMENT (1)
     ITEM
   PLANT SIZE AND OPERATING ASSUMPTION
ASSUMPTIONS:
PLANT SIZE - TPH
ANNUAL THROUGHPUT (T)
UTILIZATION - TIME %

PLANT CAPITAL ($ x 1000)
      3
  21,000
      80

   3,000
    10
70,000
    80

 5,000
     20
140,000
     80

  8,000
UNIT COSTS - $/T:
PLANT OPERATION
125 to 175     75  to  125     40 to 90
(1)    Costs are for the supply and operation of the ATP only. See text for exclusions.
                                      222
2nd FORUM May 15-17, 1990

-------
          ALGASORB®:  A NEW TECHNOLOGY FOR REMOVAL AND RECOVERY OF
                        METAL IONS FROM GROUNDWATERS

                   By:  Dennis W.  Darnall, Sandy  Svec  and Maria  Alvarez
                        Bio-Recovery  Systems,  Inc.
                        P. O. Box 3982, UPB
                        Las  Cruces, NM  88003
                                 ABSTRACT
       A  new  sorption process for removing toxic metal ions from  water has been
developed.   This  process  is based  upon the natural,  very strong affinity  of
biological materials,  such  as the cell  walls  of plants and  microorganisms,  for
heavy  metal ions.    Biological  materials,  primarily algae,  have  been immobilized in
a  polymer to produce a "biological" ion exchange resin,  called AlgaSORB®.  The
material  has a  remarkable  affinity  for  heavy metal  ions and is  capable  of
concentrating these  ions by a  factor of many  thousand-fold.   Additionally, the
bound  metals can  be stripped  and  recovered from the  algal material  in  a  manner
similar  to  conventional  resins.

      This  new technology has been  demonstrated  to be an effective  method  for
removing  toxic  metals from  groundwaters.   Metal concentrations  can be reduced to
low  pans per  billion (ppb) levels.   An  important characteristic  of  the binding
material  is  that  high concentrations of common  ions such  as  calcium,  magnesium,
sodium, potassium, chloride and sulfate do not  interfere with the binding  of heavy
metals.   Waters containing a total dissolved  solids (TDS) content of several thousand
and  a  hardness of  several hundred  parts  per  million (ppm) can be successfully
treated to remove  and recover  heavy metals.   The  process  has  been  demonstrated
under the SITE  Emerging  Technology  program for  the  effective removal of
mercury  from   a  contaminated  groundwater.

                  This paper  has  been reviewed in accordance  with
                  the U.  S. Environmental  Protection  Agency's  peer
                  and  administrative review policies and   approved
                  for presentation   and publication.
                                         223

-------
                                INTRODUCTION
       In  recent  years  there  has  been increased attention  focused  on pollution of
water  supplies by  heavy metal  ions.   These metals are toxic in rather low
concentrations and  can lead  to acute and  chronic  illness  in  humans  and other
animals.   Past waste  disposal  practices  have  resulted in  serious  contamination  of
the  environment.   The major  sources of heavy  metals are  leachates  from legal and
illegal landfills and  drainage  from old mines.   Several of the sites  on the National
Priorities List are  landfill  facilities  that have problems  caused  by  improper
placement  of  heavy metal  wastes below ground.

       Chemical  treatment using  lime or caustic precipitation  has been  the  most
common method  of removing  dissolved metals  from  leachates, mine  drainages, and
contaminated   aquifers.    Other  treatments  include  reverse  osmosis,  electrodialysis,
and  carbon adsorption.   The energy  and chemical  costs  of these various methods
become  major operating expenses  during  the use of  these  treatments.   A serious
limitation  of  many  of the  current  treatment  technologies  is  the  difficulty  and/or
expense  of treating  waters for  removal  of  heavy metals  to allowable drinking
water  levels.   Alternatives for  economical  recovery  of  dissolved  metals   from
contaminated  waters  are limited.   The development  of rapid,  widely  applicable, low
cost methods  for  the  removal,  recovery and recycling of heavy  metal ions  from
contaminated  waters at Superfund sites  is  a  high  priority.

       For  a  number of years  Bio-Recovery Systems  has been  working with  a  new
sorption  process  for removing  heavy  metal  ions  from water (1-4).   This sorption
process is  based  upon the natural, very  strong  affinity of the cell  walls  of algae for
heavy  metal ions.   Algal  cells  have been immobilized in a  silica  gel  polymer
(AlgaSORB®) and used much  as  an  ion-exchange resin.   The algae  are killed in  the
immobilization process  indicating  that  sorption  does  not  require  a  living
organism, and hence the algal  matrix  can be exposed, with little  or no ill effects,  to
solution  conditions which  would normally kill  living  cells.   The  pores of the
polymer  are  apparently  large  enough  to  allow free diffusion of ions  to  the  algal
cells,  since similar concentrations  of metal  ions  are  bound  by free  and  immobilized
cells.

       AlgaSORB®  functions as  a "biological"  ion  exchange resin and binds  both
metallic  cations and  metallic  oxoanions.   Anions such  as  chloride  or sulfate are
only weakly  bound or not  bound  at  all.    Like  ion-exchange resins,  the  algae-silica
system can  be recycled.   Metal  ions  have  been  sorbed and stripped  over  as  many as
75  cycles  with no noticeable  loss in  efficiency.   In  contrast to current ion
exchange technology,  however,  another real  advantage  of the  algae-silica  matrix
is  that the  components of hard  water  (Ca+2,  Mg+2) or  monovalent cations  (Na*, K+)
do  not significantly  interfere  with the binding of  toxic, heavy  metal  ions (2, 3).
The  binding of Ca+2  and  Mg42  to ion-exchange  resins often limits its usefulness
                                           224

-------
 since  these ions  are  frequently  present  in high  concentrations and  compete  for
 heavy  metal ion  binding.   Frequent  regeneration  of ion-exchange resins  is
 necessary  in order to  remove heavy metal ions  from solution.   Another  advantage
 of AlgaSORB® is that it has a high affinity for metal  ions.   Effluents from AlgaSORB®
 columns  frequently show that  metal ions can  be reduced to  the part per  billion
 level.    AlgaSORB*  can be  used  to remove aluminum, cadmium, chromium,  cobalt,
 copper,  gold,   iron,  lead,  maganese,  mercury,  molybdenum, nickel,  platinum,  silver,
 uranium, vanadium, and zinc.   Many of  these  metals are hazardous wastes  at
 Superfund   sites.

       A major advantage of  AlgaSORB® is  that the  efficiency  of heavy  metal  ion
 removal is  not diminished by  the presence of organic compounds in the  water.   For
 example, waters  containing  copper  ion  and  high  concentrations of  organics  such
 as butyl cellosolve,  alcohols, other  etchers  and  halogenated  hydrocarbons  have
 been successfully treated (4).   Humic and  fulvic acids seem to have little or no
 effect  on metal ion binding  to AlgaSORB®.   The  presence of organic  compounds
 limits  the  usefulness  of ion exchange resins  since  organics often will bind  to these
 resins  and   decrease  their metal  binding  capacity.

       In  1988, Bio-Recovery  Systems was selected to participate in the  Superfund
 Innovative  Technology  Evaluation (SITE)  Emerging Technologies Program.   The
 goal of  the project  was  to test AlgaSORB®  for the removal of mercury from  a
 contaminated   groundwater.
                       DESCRIPTION OF GROUNDWATERS
       A number of years  ago  an  industrial process using mercury resulted  in  soil
contamination  with  elemental  mercury.  The  mercury  subsequently  percolated
through  the  soils and  contaminated groundwater.   At some point  the  mercury  was
oxidized to  the  bivalent oxidation  state and  was found  at  various  concentrations in
the  groundwaters depending upon  the  monitoring  site.    Currently,  the
groundwaters are extracted  from  an  upper perched  groundwater table via  a
drainage gallery.   A facility has  been constructed  to treat  extracted  groundwaters
by  the use  of precipitation with  dithiocarbamates,  followed  by  polishing with
activated carbon  and a  specialty ion exchange  resin.   The  water is pumped  from
the  gallery  at  mercury  concentrations  of  0.1-3.0 ppm and  is  currently treated to
allowable discharge  limits  of 10  ppb  mercury.

       Wells monitoring  the ground water during  the late  1980's showed seasonal
variations in the mercury  concentrations.   It  appears that mercury  levels   decrease
in the dry seasons compared to the rainy seasons.   Chemical  speciation of the
mercury  in   the  groundwaters  was  not rigorously  determined, but  speciation
studies  on  soils  overlying  the  groundwater  indicated the  predominant species  was
oxidized  inorganic  mercury.    The  composition  of  other  elements in  the
groundwater  seems to change with  the  seasons  as  well,  but  an  average  composition
is given  in   Table 1.   Variations in mercury content over a four year monitoring
period in waters  from  two  wells, which are well  separated from one another, are
shown in Table 2.
                                          225

-------
        TABLE 1. AVERAGE COMPOSITION OF MERCURY-CONTAINING
                             GROUNDWATERS
              Constituent
Concentrations   (mg/1)
              Chloride
              Sodium
              Calcium
              Magnesium
              Total  Dissolved  Solids
              pH
          5,800
          2,900
           460
           440
         11,000
            8.0
        TABLE 2. SEASONAL VARIATION OF MERCURY CONCENTRATION IN
                           MONITORING WELLS
Month/Yr
Oct/1
Nov/1
Dec/1
Jan/2
Mar/2
Apr/2
May/2
Sep/2
Dec/2
Feb/3
Sep/3
Dec/3
Apr/4
May/4
Jun/4
Aug/4
Sep/4
Oct/4
Well 1
(mg/1)
9.60
3.35
0.29
5.50
3.80
10.00
4.20
7.70
6.10
6.20
8.50
2.70
4.00
4.00
4.40
5.80
7.70
13.00
Well 2
(mg/1)
0.370
0.293
0.426
0.230
0.390
0.200
0.300
0.370
0.510
0.500
0.240
0.140
-
0.260
0.170
0.180
0.086
0.240
                      EXPERIMENTAL PROCEDURES
      Mercury  analyses  were  performed using the  EPA  Method 245.1 of cold vapor
atomic absorption  spectroscopy  (5).   A Perkin Elmer Model  3030B AAS instrument
was  calibrated  daily  for  mercury, and a  calibration verification  record  was
maintained using data collected  by  the analysis  of EPA certified check standards.
Preparation of  standards  for  mercury  analysis  were  performed  in  accordance with
                                       226

-------
 the specifications in Methods for the Chemical Analysis of Water and  Wastes (5).
 Spiked samples  were analyzed with  each  batch of samples to  determine  if  matrix
 interference  existed, and frequent blanks  were run to  ensure  there  was no
 mercury  carry  over during  analysis.

       Mercury  concentrations  in  groundwaters,  column  effluents  and
 regenerating  solutions  were  determined  by  linear  regression calibration  curves
 generated  from  four point  standard  calibration  analysis  (5).

       Samples collected in the field pilot studies  were split and  sent to  Woodward-
 Clyde Consultants,  EER Technologies  and  Bio-Recovery  Systems for mercury
 analysis.

       Laboratory tests  on the efficiency of mercury  adsorption on AlgaSORB®
 were determined using small glass columns  (1.5 cm i.d. x  20 cm)  which contained
 the  sorbent.  Mercury-containing  groundwaters  were  pumped  through  the
 column at flow  rates which  varied  from 6-20  bed  volumes per hour.    Effluents
 from the  columns  were collected using a  fraction  collector and  mercury  content
 was  determined.   Once  the  columns became  saturated  or leaked  mercury above
 discharge limits  (10 ppb),  the  column  was  stripped with  10-bed volumes  of  a
 selected  stripping  reagent  followed  by  10-bed volumes  of deionized  water.
 Analyses  of stripping  effluents   were  performed  to  verify  stripping.

       Pilot  studies  were conducted  with  a small  portable effluent treatment  system
 which has  two  AlgaSORB®-containing columns in  series and which is  capable of
 treating flows of up  to 0.5 gpm.   On-site pilot testing was  conducted on  August 30 to
 September  1, 1989 and November  6  to  December 1,  1989.


                         RESULTS AND DISCUSSION
       Samples  of groundwater were  collected  at  various times during  1989.   All
samples  were acidified to pH  2 with nitric acid in the  field prior  to  transport for
laboratory  studies. Once  the samples were received  at  Bio-Recovery  Systems, the
solutions  were  neutralized  to  the  original pH with  dilute  sodium  hydroxide.
Laboratory and  field studies were  complicated  by the fact  that over  a 10  month
period,  mercury  concentrations  changed by an  order  of magnitude.   Table 3  shows
mercury  concentration variation  over  the  sampling  period.

       Different  species of  algae  can be  immobilized  to  produce  different
AlgaSORB® resins.   Since different bioploymers comprise the cell walls  of different
algae,  some  species  of algae  behave differently from others with  respect  to  metal
ion binding.    Thus, different  AlgaSORBs containing different algal  species were
tested  for  mercury  removal from  the  groundwaters.    However, these  experiments
were  complicated by the  fact that  consistent  mercury  removal  performance  was
not observed  using  a single immobilized  alga on  waters collected  at different times.
For example, Table  4 shows  mercury contents in effluents  from columns
containing  AlgaSORB 602.  The results shown in  Table  4 can not simply be
explained by  variation in mercury content in  the influent  waters.    For  example,
                                         227

-------
Column  C  shows lower leakage levels  of mercury  than column B in the first 40 bed
volumes  of effluent even  though  the  influent  concentration  of mercury  was
nearly three times higher  in column C  than column B.   This  suggests  that  a
variation  in the  chemical  species  of mercury may  occur  with time.


         TABLE 3. MERCURY CONCENTRATIONS IN GROUNDWATERS


Sample Number
103-13089
176-42089
177-42089
265-070589
343-090189
368-100489
369-100489


PH
8.5
8.0
8.0
7.9
7.8
7.9
7.9
Mercury
Concentration
(Hg/D
150
435
144
1120
620
1550
1550

Date
Collected
01-30-89
04-20-89
04-20-89
07-05-89
08-31-89
10-04-89
10-04-89
       TABLE 4. TEST OF AlgaSORB602 ON MERCURY-CONTAMINATED
                               GROUNDWATERS t

Bed Volumes
of Effluent
1-4
17-20
37-40
54-60
69-72
73-76
93-96
113-116
133-136
149-152
A*
Effluent Hg
(Ug/0
0.5
0.8
1.3
4.0
.
2.2
2.3
3.0
5.0
6.5
B*
Effluent Hg
(Jig/1)
9.9
10.1
21.8
14.8
.
31.0




C*
Effluent Hg
(Hg/0
1.3
3.4
8.1
27.0
72.5





    t  Groundwalers collected at various times  were pumped at a flow rate of 6 bed-volumes per  hour
      through identical columns containing AlgaSORB 602.  Effluents from each column were collected
      in  four bed-volume fractions and analyzed for mercury content.  Influent  flow rate was  six bed
      volumes per hour.

    *  Column A influent water was collected  January 30. 1989, and had an influent mercury
      concentration of 150 ng/1.  Column B influent water was collected April 20, 1989, and had  an
      influent mercury concentration of 435 ng/1.  Column C influent water was collected July 5, 1989,
      and had an influent mercury concentration of 1120  ng/1.
                                           228

-------
       Over  99 percent  of the  mercury was  stripped from  the  columns in Table  4 by
the passage  of  10  bed volumes of 1.0 M sodium  thiosulfate through the column (data
not  shown).

       After examining  several  different AlgaSORB®  preparations  and noting
similar types of variations as shown  in  Table  4, it  was decided to settle on two
different  AlgaSORB® resins for final  testing.  While  these two  resins could have
been  blended into  a single column, they  were placed in  two columns which  were
connected in series and  from  which  effluents  samples could  be taken  from each
column for   mercury  analysis.   Table  5 shows results of  these  experiments.

       Data  in Table  5  shows  that the two  columns  arranged  in  series were effective
in mercury  removal to below  one ppb through passage of 250 bed volumes of
mercury  contaminated  waters  which  contained 1550 ppb mercury.

       On site  pilot testing of the AlgaSORB®  resins  was  performed from  November
6, 1989,  to December 1,  1989.   Two columns  (2.54 cm  i.d.  x 81 cm) were  separately
filled with AlgaSORB 624 and  AlgaSORB 640.  The columns each had a bed volume of
400  ml  and were  connected in  series.    Mercury-contaminated waters were pumped
through the  two columns and  two bed-volume fractions (800  ml) were  collected,
split and  sent  to HER Technologies, Woodward-Clyde and  Bio-Recovery  Systems for
analysis.   Results  of on-site pilot  testing  are  shown in Table  6.


        TABLE 5.  TEST  OF AlgaSORB 624 AND AlgaSORB 640  ON MERCURY-
                      CONTAMINATED GROUNDWATERS*
          Bed Volumes of Effluent                    Effluent  Hg  (u.g/1)
0-12
12-24
24-36
48-60
60-72
84-96
108-112
132-144
168-180
192-204
252-264
288-300
312-324
324-336
0.3
0.2
0.2
0.3
0.5
0.7
0.8
0.9
0.8
0.9
0.6
0.6
2.0
1.9
     * Two columns (1.0 cm i.d. x 37 cm) coupled in series were used.  Groundwaters collected October 4,
      1989, and containing  1550 ng/1  mercury were passed through  the  columns at a rate of six bed-
      volumes per minute.  Ten bed volume fractions were collected and analyzed for mercury.  Data
      shown  above are mercury concentrations in effluents from  the second  column.
                                         229

-------
       By the  time the on-site testing  bad begun in  November the  mercury
concentration  in  the ground waters had  changed  from  about  1500  ppb  (in  October)
to about  700 ppb  (see  Table 6).   With  the  exception of the  first fraction  collected,
the  data  in  Table  6 shows  that  over 500  bed volumes of  mercury-contaminated
waters  were  treated before mercury  in effluents  approached  the   10 ppb discharge
limit.
         TABLE 6. ON-SITE PILOT TESTING FOR MERCURY REMOVAL FROM
                               GROUNDWATERS *
                                           Mercury Concentration  fue/1)
 Bed Volumes of                 Bio-Recovery   Woodward Clyde  EER  Technologies
     Effluent        Influent      Analysis          Analysis           Analysis
7-8
85-86
163-64
229-230
289-290
313-314
343-344
379-380
415-416
449-450
467-468
503-504
533-534
587-588
A'.
f.(

9.5
5.3
2.1
1.4
1.8
1.9
5.5
2.0
1.8
4.9
4.0
5.8
7.7
10.5
l&* fifiO
) \> OOU
)\j / /U
14.2
8.0
3.6
1.4
2.6
2.4
9.3
3.1
3.2
7.8
7.2
9.6
10.3
13.0
780
/ ou

11
<10
<10
<10
<10
<10
10.0
<10
<10
10.0
<10
<10
<10
15
AOO
790
/ 4*\)
     '  A portable water treatment system was equipped with  two columns connected in series.  The first
       column was filled with AlgaSORB 624 and the second was filled with AlgaSORB 640.  Groundwaters
       were pumped through  the system at a  flow rate of 6 bed-volumes per hour.  Effluent samples were
       collected  and sent to  Woodward-Clyde Consultants, EPA  (EER  Technologies Corporation)  and Bio-
       Recovery Systems for analysis.

     *  Influent samples were collected for  analysis just prior  to 436 and 600 bed volume  fractions of
       effluent were  collected.
                                 CONCLUSIONS
       Initial  laboratory  testing  of AlgaSORB9 resins clearly  showed promise  for
mercury recovery  from  contaminated  groundwaters.   Once  the  mercury  was  loaded
on  the  resin,  it  could be  stripped with  sodium thiosulfate.    On-site pilot testing
confirmed  the laboratory studies  and  showed that AlgaSORB* resins treated over
500 bed  volumes  of groundwaters before  mercury  levels  in effluents  exceeded
discharge  limits.
                                            230

-------
                                    JCES
1.  Darnall, D.W.,  Greene, B., Henzl,  M.,  Hosea, M.,  McPherson, R., Sneddon, J.
    and  Alexander,  M.D., Binding and Recovery of Gold(III) and  Other Metal
    Ions from  Aqueous  Solution by Algal Biomass, Environ. Sci. Technol. 20:
    206, 1986.

2.  Darnall, D.W.,  Greene, B., Hosea,  M.,  McPherson,  R.,  Henzl, M., and
    Alexander,  M.D., Recovery of Heavy  Metal Ions  by  Immobilized  Algae, 1m
    R. Thompson  (ed.),  Trace Metal Removal From Aqueous Solution, The Royal
    Society of  Chemistry, London, 1986, p.  1.

3.  Greene, B., McPherson,  R. and Darnall,  D.W., Algal  Sorbents  for Selective
    Metal  Ion  Recovery, In:   J.W.  Patterson and R. Passimo  (eds.), Metals
    Speciaiion,  Separation and Recovery.   Lewis Publishers,  Chicago,  Illinois,
    1987, p. 315.

4.  Darnall, D.W.  and Gabel,  A., A New  Biotechnology for  Recovery  Heavy
    Metal  Ions  from Wastewater, Jn;   Proceedings of the Third National
    Conference  on  New  Frontiers for Hazardous  Waste  Management,  U.S.
    Environmental  Protection  Agency, EPA-600/9-89-072, 1989, p. 217.

5.  Methods  for the  Chemical Analysis  of Water  and Wastes.  EPA-600/4-79-020,
    U.S. Environmental  Protection  Agency, Revised  March   1983  and
    subsequent  EPA-600/4 Technical  Additions  Thereto,  Cincinnati,  Ohio, 1983.
                                    231

-------
      BIOSORFTIOH -  A POTENTIAL MECHANISM FOR THE REMOVAL OF
           RADIOHUCLIDES FROM NUCLEAR EFFLUENT STREAMS

                        Peter Barratt Ph.D

                      Biotreatment Limited
              5 Chiltern Close, Cardiff CF4 5DL, UK
Abstract

Fifteen  fungi  and  bacteria  were  isolated  from  both  metal-
contaminated  and  uncontaminated  environments,  and  tested  for
their tolerance to strontium, ruthenium and cobalt in agar media.
Three fungal  isolates,  Trichoderma viride,  Penicillium expansum
and Aspergillus niger, were  examined  for their ability to remove
these metal  ions  from  solution.   T.  viride  and P.  expansum
removed 100% of cobalt and strontium from solution within 2 days,
                                                              wai
                                                              -1
at concentrations of lOmgl"  ,  although  biosorption of metals was
much  less  effective at  concentrations between, 30  and 200mgl
Composition of the  liquid  medium was found to affect biosorption
of the metals studied.   Additional studies were undertaken on the
biosorption   of    these   metals   by   immobilised  biomass   of
Saccharomyces cerevisiae  and A.  niger in  laboratory  filter  bed
systems.  Removal of >50% of ruthenium in a model effluent stream
was measured in a system containing A. niger.

Introduction

Of the  various stages  involved  in processing and reprocessing in
the nuclear  fuel  industry, many produce  low level wastes in the
form of aqueous effluents which contain both radioactive and non-
active contaminants that can be considered as hazardous to the
                            232

-------
environment  (Schumate et  al,  1978).   These reprocessing effluent
streams vary widely in composition.  Some of the most significant
radioactive isotopes in these streams are fission products, such
as 137Cs, 144Cs, 89Sr, 90Sr and 106Ru, but large quantities of
non-radioactive effluent components, notably nitrates, also occur
during fuel reprocessing.

It  is now  well established  that many  microorganisms have  the
capacity to accumulate metallic cations from the environment,  via
a process generally referred to as biosorption.  In the microbial
cell  this  phenomenon  is  usually  considered to  comprise  two
distinct  phases,   those  of  active  uptake, dependent  upon  the
metabolism  of  the  cell, and  metabolism-independent  cell  surface
binding (Gadd, 1986; Norris and Kelly, 1976).
                                                        i
Cation accumulation by microbial cells is a complex process.  The
degree of  concentration inside the  cell (absorption) or  on  the
cell surface (adsorption) are dependent upon the cationic species
and upon the  properties  of the cell in terms  of  such factors as
cellular charge, metal requirement and tolerance, and competition
for active  binding sites  amongst  different cations, as  well as
environmental factors such as pH,  temperature and the presence of
other  chemicals in solution  (Gadd,  1986; Ross  and  Townsley,
1986).  However, where conditions are conducive it has been found
that  there  is  a   rapid  accumulation  of  cations from  solution,
often  in  the first minutes  following the  addition  of  microbial
biomass (Kiff and Little, 1986)

The uptake of radionuclides by microbial biomass and biomass cell
wall constituents  has been observed  in previous studies,  and the
biosorption of Co  , Sr  , Cs   and Ru   has been demonstrated on
a number of occasions (Gadd, 1986; Norris  and Kelly, 1977; Ross
and Townsley, 1986).  On the basis of these observations there
                            233

-------
appears to be  some  potential for the use of microbial biomass as
a  means  of  removing  radionuclides  from  contaminated  effluent
streams.

Many of the  treatment  processes currently in use for the removal
of  active contamination  from effluent  streams  demonstrate high
efficiency where  substantial concentrations of radionuclides are
present.  However, where effluent metal  concentrations lie in the
lower range  of 1  to lOOmgl" , and amelioration is still required
before  discharge,  some of  the  classical  treatments become less
economical and even ineffective.  It  is in this area that novel
microbial-based  techniques  may be  able  to  offer  considerable
benefits in the future (Schumate et al,  1978).

3. Experimental methods

3.1. Biomass screening and  selection

A  number  of  microorganisms  were  isolated  from  the  natural
environment  using standard  agar plate isolation techniques, and
maintained  in  axenic  culture.   These organisms included fungi,
bacteria  and yeasts from a variety  of sources, including metal-
contaminated   soil   from   a   former   waste   lagoon   (Luton,
Bedfordshire,  UK),   contaminated leaf  surfaces near  a  smelting
works  (Avonmouth,  UK), and uncontaminated agricultural sources,
notably silage and hay.

As  an  initial  indication of the organisms' tolerance to elevated
metal  concentrations,  each isolate  was  cultured  on agar plates
containing a range of  concentrations of  the non-radioactive metal
species ruthenium (101Ru),  strontium  (88Sr),  cesium (   Cs) and
cobalt   (59Co).     These    metals    were   selected  as   being
representative  of   four   important   cationic  constituents
effluent  streams  from  the UK nuclear reprocessing  industry.

-------
Agar plates were prepared containing 0,  5,  10,  50,  100,  200,  400
and  500mgl~   of  each metal.    Fungal  and  yeast isolates  were
streaked onto malt extract agar  (Oxoid,  UK), and bacteria  onto
nutrient  agar  (Oxoid,  UK),   each  containing  the  appropriate
concentration of   metals  in  the  form   of  ammoniated  ruthenium
oxychloride   (CltH.-N-.O-Ru.,),   and   the   chloride  salts   of
strontium,  cesium  and cobalt.   All plates  were  incubated over a
period of 5 days at  25°C, after  which the presence  or absence of
microbial growth was recorded.

Some  of  the  microorganisms obtained  were  later identified to
species level.

3.2. Biosorption studies in liquid culture

The adsorption of cobalt and strontium by living cell cultures of
four of  the  microorganisms  isolated in  3.1  was  tested in liquid
culture.

Three  fungal and  one bacterial  isolate were  tested for  their
growth and metal uptake in  pure  culture.   Fungal isolates A3, A6
and A8 were  grown in two  liquid media; malt extract (ME)  broth
(Oxoid, UK) to simulate an organic-rich environment, and Czapeks-
Dox  (CD) broth  (Oxoid, UK), a  synthetic inorganic  medium.   The
bacterial isolate, B4, was cultured in nutrient broth (NB; Oxoid,
UK).

Conical  flasks  containing  100ml of the selected  sterile  media
were prepared.  A range of concentrations of strontium and cobalt
had  been  established  in  these  liquid  preparations,   by  the
addition of SrCl- and CoCl0, to give concentrations of each metal
                                               -l
as  follows:   0,  10,  20,   50,  100  and  200mgl   .    Flasks  were
inoculated with the  relevant  cell culture  to give  the following
treatments at each metal concentration:
                            235

-------
              A3 in ME           A3 in CD
              A6 in ME           A6 in CD
              A8 in ME           A8 in CD
              B4 in MRD

Each treatment flask was duplicated to give  a  total of  84 liquid
samples.    After   inoculation  of  the  flasks   under  sterile
conditions, each was  plugged  and  incubated  at 25°C on a rotary
shaker (ISOrpm) for 48h.

Following  incubation  cell  culture  media were  filtered  through a
packed glass  bead  column  and  centrifuged at  6000rpm for 15min.
The supernatant from  each  treatment was  collected for  cobalt and
strontium  analysis by  atomic  absorption  spectrophotometry (AAS).
Uninoculated control samples for each liquid medium at each metal
concentration were also analysed for the two metals.

3.3. Metal analysis

All metal  analyses were performed using a Perkin Elmer AAS (model
2380),  with  the  appropriate  cathode  lamp  for   each element.
Cobalt  was analysed  at a wavelength of 240.7nm,  ruthenium at
349.9nm, and strontium at  460.7nm.  All three metals were ionised
in  a  nitrogen/acetylene flame.  Lanathanuro  chloride was used to
enhance  the   sensitivity   of  the  machine  to  ruthenium  and
strontium.

Metals  in  the samples were measured  against standard  analytical
solutions.
                            236

-------
3.4. Biosorption studies in a model effluent stream

Using information  from various sources  in the UK  nuclear  fuels
industry, a model 'low level' effluent stream was prepared in the
laboratory.     This  comprised  the   following   non-radioactive
components:

              Ruthenium (as RuCl.s)              5mg
              Cesium (as CsCl2)                5mg
              Strontium (as SrCl2)              5mg
              Cobalt (as CoCl2)                5mg
              Sodium nitrate               20000mg
              Sodium hydroxide               SOOmg
              pH                               8
              Deionised water               1000ml

Model filter beds  comprising approximately 350g  of a mixture of
washed sand and gravel (3:5  wt/wt),  supported on a woven polymer
mat, were  constructed in Buchner  funnels in the  laboratory.   A
header tank  (1  litre  capacity)  with  a  perforated base, of the
same diameter  as  the  Buchner funnel, was suspended  above  each
filter bed,  to allow  even  distribution  of liquid  from  the  tank
across the surface of the bed as it percolated through.  Control
beds  contained  sand  and  gravel  only,   and  biomass  treatments
incorporated lOg dry weight  of either S. cerevisiae or A. niger,
evenly distributed amongs- the gravel base.

Biomass  had  been  collected as waste  products from  the brewing
industry (Bass Brewing, UK), and a citric acid fermentation plant
(Rhone-Poulenc, UK) respectively, and had been air-dried prior to
incorporation into the gravel filters.  Each treatment system was
duplicated, giving a total of six filter systems, and each system
was thoroughly rinsed through with distilled/deionised water, and
allowed to drain.
                            237

-------
Effluent stream  (1  litre)  was pumped into each header  tank at a
known rate, via a series of  peristaltic  pumps,  and as the stream
passed  through  the  filter  bed,  samples  were  collected  for
analysis.  For each  filtration  system,  4 x 200ml treated samples
were  collected  at  200ml  intervals,  so  that at  the end  of  the
experiment,  each   system   had  collected  one   sample  of  the
following: 0-200, 200-400,  400-600 and 600-800ml.

Collected  samples  were   analysed   for   strontium,  cobalt  and
ruthenium by AAS.

4. Results and Discussion

4.1. Metal tolerance screening

Table 1 gives the results  of the  metal  tolerance tests.  In all,
15 organisms were screened in this way,  and tolerance appeared to
vary widely between isolates.  Seven organisms were identified to
species,  and  these  were:  Saccharomyces  cerevijsiae,  Penicillium
spinulosum, Penicillium expansum,  Aspergillus niger, Trichoderma
viride,   Zoogloea ramigera  and Bacillus subtilis.   The results
show  that  only  3 of these  (Al,  A3 and A6) were  able to grow on
agar  media containing  the  highest  concentration  of   the  four
metals  (SOOmgl   ).   These  three organisms were all identified as
fungi or  yeasts.   Isolate  Al was S.  cerevisiae,  A3 P.  expansum,
and  A6  T.  viride.   The  bacterial  isolate,  B4, which showed
limited growth at a concentration  of 400mgl~ ,  was B.  subtilis.
These four microbial isolates were selected for further work.  In
addition, fungal isolate A8,  (A. niger), was selected.

The concentrations  of metals in the media used  in this stage of
the study  were  for the  purpose of  organism  selection only.  Due
to  the  metal  binding properties of  some organic molecules it is
unlikely that the concentrations specified represented  the
                            238

-------
available  metal  ion  concentrations  present  in  the  agar  media.
The  interaction  of  metals  with  non-biomass organic  materials
should always be borne in mind during experiments of this kind.

4.2. Metal uptake from liquid media

Figures 1-4 illustrate the results of microbial treatment of Co2+
ions in metal-supplemented liquid media at concentrations of 200,
              3 On
              .2+
100,  50  and  30mgl~   respectively.    Figures  5-8  give  similar
results for Sr
Cobalt concentrations in  CD  medium treated with P.  expansum were
considerably  less  than any  other treatment  after  incubation of
the  200mgl~   solutions.     This  treatment  demonstrated  a  22%
decrease when compared with the control samples with no inoculum.
A  similar  pattern  was  detected  in  media  containing  an  initial
concentration of lOOmgCol" ,  where the treatment inoculated with
P.   expansum  reduced   Co    concentrations  by  26%  after  48h
incubation (Fig. 2).

At 200mgl  , substantial Sr   accumulation  in biomass occurred in
both P.  expansum and T.  viride cultures  (Fig.  5).   Differences
between  Sr    uptake from  the two media were  not well defined at
100  and 200mgSrl~  ,  although  Co    uptake by  P.   expansum  was
clearly  greater  in  CD.   This could have  been due to the binding
of  metal  ions  to  organic   molecules  in ' ME  broth  at  higher
concentrations,  which  may  consequently   neutralise  the  ionic
charge  and  reduce  the  overall  affinity  of  the metal  for  cell
surfaces  (adsorption).    The   results  for   Sr    at  100  and
200mgSrl   show no such difference between  biosorption in the two
mycological media,  which  suggests that  Co*""1"  may have  a greater
affinity than Sr   for components of the ME broth.
                             239

-------
The pattern  of  increased biomass affinity for  Co    in CD medium
is consistent for P.  expansum at all metal concentrations
                                                         •

At  lOmgl    P.  expansum  and T.  viride  both  demonstrated  100%
removal  of   Co    from CD  broth  (Fig.  9).   The  data  for  Sr2"*"
accumulation in these  two  organisms  were  very similar to that of
Co  , with complete  removal of lOmgSrl    taking place from both
media  after  incubation  with  P.  expansum,  and  85%  reduction
occurring in the T.viride treatment in CD broth (Fig. 10).

Removal  of  Co and Sr by  T.  viride, although  almost completely
effective in CD, was poor in ME broth.

In general,  incubation  in  the  presence  of  A.  niger  had little
effect  upon  the  concentration  of metal  ions  in  solution.   At
     _i
lOmgl    only minimal reduction in  Co occurred, and  the maximum
reduction observed in Sr was only 17% of the untreated control.

4.3. Metal uptake from effluent stream

Figures  11 and  12 illustrate the  concentrations of metals in the
effluent stream after  filtering through the  sand and gravel beds
containing A. niger and S. cerevisiae respectively.

Both  biomass treatments  resulted in a  marked decrease  in the
metal  loading of the  stream  in  the first  200ml  treated when
compared with the non-biomass controls.   Following  this initial
decrease   in   the   concentrations,   metal    levels    rose   by
approximately 50% in all  treatments between  200  and  400ml, and
generally these   levels,  all markedly  lower  than the controls,
were maintained between 200 and 800ml.   Differences between the
three samples taken from 200-800ml were minimal for all metals in
either treatment.
                            240

-------
Although the  trends  discussed  occur  in nearly all the data sets,
the results  for  ruthenium in  the effluent  treated with A. niger
differ,  in that,  between the  samples taken  from 0  and  800ml,
there  were   no  significant   differences,   and  ruthenium  was
maintained  at  <50%  of  that   in either   the  influent  or  the
controls.  In this way  Ru   adsorption by  A.  niger appears to be
the most  efficient of  the  treatments applied to  the filter bed
systems.

Results  from  the   biomass   filter   systems  demonstrated  some
biosorption of  Co,  Sr and Ru  from  an acid effluent  by A. niger
and S. cerevisiae.  The positive results obtained for dried
A.  niger  biomass,  compared with those  from  the cell  culture
experiments,  indicates  the  influence  of  both  the form  of  the
bioroass  and the  environmental  conditions  prevalent  within the
liquid on cation accumulation by microbial cells.

Conclusions

The  biosorption of  Sr and  Co  from  axenic  liquid  cultures  of
PeniciIlium   expansum    and    Trichoderma   viride   has   been
demonstrated.   Complete biosorption  of  both Sr and  Co occurred
where the  initial  concentration was  10mgl~  .   At concentrations
above  30mgl~   accumulation  of  metal  ions  was  less  efficient
during 48h incubation.  There was no clear pattern of biosorption
at metal concentrations increasing between 30 and 200mgl   .

The uptake  of Sr and  Co  by a culture of  Aspergillus  niger was
generally   poor   in   liquid   media,   even   at   low   initial
concentrations (10mgl~  ).

There  appear   to  be  significant   effects  of   the  chemical
environment on biosorption by  fungal biomass in a liquid system.
Other organic components in the liquid phase may inhibit
                            241

-------
biosorption of Co and Sr.  It seems likely that there are complex
interactions  in  operation during  the  biosorption process  which
influence   its   effectiveness.     Such   factors  may   include
competition  between  cation  species,  metal  sequestration  with
organic molecules  in solution, pH  and  the physical form of  the
biosorbant matrix.

The use of  living microbial cells  in continuous  culture systems
for the removal  of metal ions from nuclear  effluent  streams  may
be a feasible proposition where metal concentrations are low,  and
effluents can be  held in treatment tanks  or  holding  ponds  prior
to treatment, although  further verification of the  processes is
required  both in  stable  and  radioactive  streams.    Subsequent
extraction  of  bioaccumulated  metals  into a  small volume  of an
inorganic matrix is also likely to be a prerequisite for a viable
treatment process,  as biosorption  treatment  has the potential to
give rise to higher  level waste  from  low level streams,  and such
waste  is  likely   to  require  encapsulation.    Encapsulation  of
organic  wastes  is  not  a  favoured  practice, in  the  nuclear
industry,  and although  the elution of  active metal species prior
to disposal  is  clearly  an option  for such  a treatment process,
this may  be more difficult where  ions are accumulated inside as
well as outside the cell (Gams, 1986).

References

Gadd.  G.  M.   (1986) The  uptake  of  heavy  metals by  fungi  and
yeasts:   The  chemistry  and  physiology  of   the  process  and
applications  for  biotechnology.    In:  Immobilization  of  ions by
bio-sorption  (eds.  H.  Eccles  and  S. Hunt).    Published  for  the
Soc. Chem. Ind.  by Ellis Horwood, U.K.
                            242

-------
Kiff R.  J.  and D.  R. Little.  (1986) Biosorption  of  heavy metals
by  immobilised fungal biomass.   In   Immobilization of  ions  by
bio-sorption  (eds.  H. Eccles  and S. Hunt).   Published  for  the
Soc. Chem. Ind. by Ellis Horwood, U.K.

Norris P. R.  and D.  P.  Kelly.  (1977) Accumulation of cadmium and
cobalt by Saccharomyces cerevisiae.  J.  Gen. Micro. 99:317-324.

Ross I.  S.  and C.  C. Townsley (1986) The uptake  of  heavy metals
by filamentous  fungi.   In Immobilization of ions  by bio-sorption
(eds. H. Eccles and  S. Hunt).  Published for  the  Soc.  Chem.  Ind.
by Ellis Horwood, U.K.

Schumate  II  S. E.,  C.  W. Rancher,  G.  W.  Strandberg  and C.  D.
Scott.  (1978)  Biological processes for  environmental  control  of
effluent streams in  the  nuclear  fuel cycle.  In Waste  Management
and  Fuel Cycles  1978.    Proc.  Symp.  Waste  Management,  Tucson,
Arizona.

Acknowledgeinents

Grateful  thanks to  Deirdre  O'Toole  for  her technical  support
during the  course  of this project, and  to  the  Department of  the
Environment (UK) for funding the work.
                            243

-------
   TABLE  1
ORGANISM VIABILITY AT 25°C
TN

ORGANISM

YEASTS AND FUNGI
Sacc. cerevisiae (Al)
Pen. spinulosum (A2)
Pen. expansum (A3)
Fungus A 4
Fungus A 5
Tr. viride (A6)
Fungus A7
Asp. niger (A8)
BACTERIA
Bacterium Bl
Bacterium B2
Bacterium B3
Bacillus subtilis (B4)
Bacterium B5
Bacterium B6
Zoo. ramigera (B7)
THE PRESENCE OF VARYING
CONCENTRATIONS OF Ru.
Sr, Cs and Co
4- : growth
- : no growth
na : no result available
METALS CONCENTRATION (mg.l'1;
0 5 10 50 100 200 400

+ na + + + 4- 4-
+ na 4- +
4- na 4- 4- 4- 4- +
+ 4- 4- 4-
4- 4- 4- + 4- 4-
+ + + 4- 4- 4- 4-
+ + + 4- +' - -
+ + 4- 4- + - -
+ + + + 4-
+ + + 4- + 4- -
+ 4- 4-
+ + + + + + +
+ + 4-
+ + 4-
+ + + ----


500

+
+
-
4-

-
-
-
-
-
-
—
244

-------
            COBALT CONCENTRATION (mg/1)
                                                                                      COBALT CONCENTRATION (mg/I)
               w
               o
                  6
                  o
ro
-P>
en
    i
z
O   H


E   *
w   B
M   "
Ol
o
I
(O
§
M
Ol
O
                                             O
                                             O
                                             to
                                           003
                                           II
                                            oo
                                            z«
                                             i

                                            1
                                             31
                                             m
                                              O
                                                 w
0
§
w
>
&
m
OB
>-
wO
o<
o>
3r
oO
b«
°U
-i
op
c>
3ir
^
°5
is
§>
v
2«
FROM
CUBATI
oo
Z?
^
M
O
c
»
w

1— *




















-------
COBALT CONCENTRATION (mg/1)
           o
           j
Ol
o
1
o
I
o
I
                          9
                          O
o
o
03
                                  H
                               t  s

                             §<
                             3r
                             -CO
                             zo>

                             §5

                             a
                               H

                               31
                               m
                                                          COBALT CONCENTRATION (mg/l)
                                                                 M

                                                                 •0
                                                                 M
                                                                 O
                                                                                   B

                                                                                   r
                                                                                                        o

                                                                                                       3r
                                                                                        §S
                                                                                                      §?
                                                                                                        m
                                                                                                            w

                                                                                                            LJ

-------
         FIGURE  5

         STRONTIUM REMOVAL BY FUNGAL BIOMA88 FROM CULTURE
                  MEDIA (200 mg.Sr/1) DURNG 48h INCUBATION
        860

        20O-
         160-
         100-
         60-
               12 CONTROL
                        B CZAPBKS-DOX   (B MALT EXTRACT
                    A.NK3EH
                            T.VRCE


                        FUNGAL SPECIES
P.EXPAN8UM
         FIGURE  6

         QTHONTIUM REMOVAL BY FUNGAL  BIOMASS  FROM CULTURE
         STRONTMEDIA  (100 mo-Sr/l) DURING 48h INCUBATION
1
o

8
        260
        20O-
16O-
         100-
                                S CZAPEKS-DOX   B MALT EXTRACT
         60-

-------
^>



?
FIGURE 7

STRONTIUM REMOVAL BY  FUNGAL  BIO MA S3 FROM CULTURE
        MEDIA (60  mg.Sr/l) DURING 48h INCUBATION
80-
70-


eo-


60-


40-


30-


20-


10-


 0-
      C3 CONTROL
           A.NK3ER
                               12 CZAPEKS-DOX   EB MALT EXTRACT
                                    T.VRDE
                       FUNGAL SPECIES
P.EXPAN8UM
 o

 g

 E-
 FIGURE  8

 STRONTIUM REMOVAL BY FUNGAL BIOMASS FROM  CULTURE
          MEDIA (30 mg.Sr/l)  DURING 48h INCUBATION
 80-
 7O-


 60-


 60-


 40-


 30-


 20-


  10-
       (2 CONTROL
            A.NUER
                                 12 CZAPEKS-DOX   B MALT EXTRACT
                                      T.VRDE


                                  FUNGAL SPECIBS
 P.EXPAN8UM
                         248

-------
              STRONTIUM CONCENTRATION (mg/1)
                                                                                             COBALT CONCENTRATION (mg/l)
ro
       c
       z
       o
       M

       •D
       O
a-
o
m
               I   I   I  I  I   I  I   I
                                   « o  2
                                   I   I  I
                                     (0
                                     H
                                     3)

                                     O
                                                    cn
                                                    c
                                                    »
                                                    M
                                              is
  11




30 Q
                                               OoQ
                                               do
                                               02

                                               20

                                                5


                                                33
                                                m
                                                                         c

                                                                         o
                                                                         •o
                                                                         M
                                                                         o

                                                                         9
                                         a
                                         o
                                         m
                                                                               O

                                                                               O
                                                                               ID
                                                                                                                 8
                                                                               *
                                                                               m
                                                                                                                      §
                                                                                                                      VŁ>

                                                                                                                            DS
                                                                                                                 03 3D


                                                                                                                 51

                                                                                                                 ip

                                                                                                                  H

                                                                                                                  V
                                                                                                                  m

-------
             FIGURE 11

             METAL CONCENTRATIONS IN EFFLUENT STREAM
              AFTER TREATMENT WITH ASPERQILLUS  NIGER
                          m CONTROL
                          E3 300-400ml
                            600-800ml
            ED 0-200ml
            OH 400-600ml
                 COBALT
 STRONTIUM
RUTHENIUM
                             METAL SPECIES
          FIGURE 12

          METAL CONCENTRATIONS IN  EFFLUENT STREAM  AFTER
             TREATMENT WITH SACCHAROMYCES CEREVISIAE
                            CONTROL
                            200-4OOml
                            600-aOOml
             S3 0-200ml
             IE 400-600ml
63
a
                 COBALT
  8TRONTWM

METAL SPECIES
RUTHENIUM
                             250

-------
  BlOTRCX...
                       •                    ... for environmental solutions, naturally!
                Biological Treatment of Wastewaters

                                by

                        Thomas J. Chresand
                        Dennis D. Chilcote

                           BioTrol,  Inc.


                           Presented at
               U.S. Environmental Protection Agency
       Second Forum on Innovative Hazardous Waste Treatment
             Technologies: Domestic and International

                    Philadelphia, Pennsylvania
                          May  15-17,  1990
                             ABSTRACT

     The BioTrol Aqueous Treatment System  (BATS) was  demonstrated
as part of  the Superfund Innovative Technology Evaluation  (SITE)
program   for   treatment   of   groundwater   contaminated    with
pentachlorophenol   (PCP).      The   system   employs   indigenous
microorganisms; however, it  is also amended with a specific  PCP-
degrading bacterium.  A mobile trailer-mounted system was used for
the demonstration.  Three flow rates were tested, corresponding to
residence times of 9,  3 and  1.8 hours.   PCP removal ranged  from
97.6 to 99.8 percent, with average effluent concentrations  as low
as 0.13 ppm.   It  was  shown  that biological  degradation was  the
predominant   removal   mechanism   while   air   stripping   and
bioaccumulation were negligible. Acute biomonitoring with minnows
and  water  fleas   showed complete  removal  of  toxicity by  the
treatment system.

     The  BATS  has  also been shown to  be highly  effective  for
treatment of a variety of wastewaters, including process and  lagoon
waters, and for.  a  wide  range of  contaminants..    Results  are
presented  on  treatment  of  gasoline,  phenolic,  and   solvent-
contaminated wastewaters.
         BioTrol, Inc -11 Peavey Road • Chaska, Minnesota 55318 • 612/448-2515 • FAX 612/448-6050

                                   251

-------
                      TECHNOLOGY DESCRIPTION

Reactor  Design

     Reactor  design   is   a   critical  aspect   to   successful
implementation of this technology.  BioTrol  employs a multi-stage,
submerged, aerated, fixed-film reactor that provides a high biomass
concentration  and, therefore, reduced reactor volume.  The reactor
and a schematic of the trailer-mounted system are shown  together in
Figure  1.    The  multiple  stages create a dispersed plug  flow  of
wastewater  through the reactor.  The plug  flow configuration  is
important in that contaminant concentrations typically  fall in the
range  of first  order removal  kinetics.    That  is, the rate  of
removal  is  proportional to the concentration  of contaminant.  A
completely   mixed   tank  system  operating   at  a   low   effluent
concentration  will  experience low removal  rates.   A system  with
plug flow characteristics, on the other hand, will  experience the
same low rates  in  the  effluent sections;  however, the  upstream
sections will  operate at  very high rates,   thus yielding higher
overall removal  rates.

     The  use  of a  fixed-film system  allows  for  a   long   cell
retention  time  and,  therefore,  lowered   production   of excess
sloughed biomass.  Moreover, the fixed-film system  eliminates the
often  problematic biomass  separation step  which  is  crucial to
successful operation of an activated sludge system.


Bacterial Amendment

     Many   of  the  priority  pollutants   can   be  degraded  by
microorganisms  indigenous  to  a  given  wastewater.    For these
compounds,   treatment can  be  accomplished  by  simply  adding the
appropriate inorganic nutrients and  allowing time for acclimation.
However, in cases where a highly toxic or recalcitrant compound is
to be treated, the appropriate microorganisms may not be  present.
In these cases, treatment can be accomplished by adding organisms
with the appropriate degradative  capabilities.   This  technique,
called  microbial   amendment,   is  finding   increasing  use  as
microbiologists continue to isolate  organisms with novel metabolic
pathways.

     As an example of microbial amendment, a  Flavobacterium species
is used by BioTrol  for treatment of pentachlorophenol-contaminated
wastewaters.  This microorganism can perform rapid mineralization
of pentachlorophenol at concentrations up to 200 ppm.
                                  252

-------
                          PREVIOUS WORK

     In the fall of 1986, a nine-month pilot study was performed to
demonstrate  the feasibility of  groundwater treatment  at a wood
preserving  facility.   The study  was  funded by  a grant from the
United States Geological Survey.  A single-stage packed bed reactor
(30 gal)  was activated with indigenous microflora and later amended
with inoculations of a Flavobacterium species.  The target compound
in this study was PCP, present in the groundwater at concentrations
from 60 to 100 ppm.  A typical monthly performance chart is shown
in Figure 2.  As shown in Figure 2, the flow rate  for most of the
month was 25 gph, corresponding to a residence time of  1.2 hours.

     The study  proved  the effectiveness of  amending a microbial
consortium  with  a unique   organism  to  treat   a highly  toxic
groundwater.  The specific rate of PCP  degradation demonstrated was
as high as  70 mg  PGP/liter reactor volume/hour - or 105 Ib PCP/1000
cf/day - a loading rate which would be high even  for treatment of
the  BOD  of  municipal  sewage.   Moreover,  the  nine-month  study
demonstrated the stability of such a microbial system.
                           BATS Performance
 Figure 2
                                  253

-------
                   RESULTS OF SITE DEMONSTRATION

In July and August of 1989 a demonstration of the  BATS was  carried
out  under the U.S.  EPA SITE program.   A wood treating site  in
Minnesota  was -chosen  as  the  location  for  the demonstration.
Groundwater  at the site was contaminated with PCP, creosote, and
heavy metals (copper,  chromium, and arsenic).  The  goals of the
demonstration were to:


     1) Reliably measure the removal  of contaminants  across the
     reactor at  a  series of different  flow rates

     2)  Carefully  monitor  the fate  of  PCP  to determine the
     predominant removal mechanisms.   In particular, to  determine
     complete  or  partial  degradation  and   the   extent  of air
     stripping,  and/or bioaccumulation

     3) Determine the decrease in toxicity of  the groundwater  after
     biological  treatment

     4)  Determine  operating   costs   for  treatment  of  similar
     wastewaters


A mobile trailer-mounted system with a 5  gpm  capacity was used for
the  demonstration.   The test  was  carried  out  over  a six-week
period, and consisted of treatment at three flow rates for  2  weeks
each.  The flow rates investigated were 1, 3,  and 5 gpm.   Since the
packed volume  of the reactor was  540 gallons, these  flow  rates
corresponded to residence times  of 9, 3 and 1.8 hours respectively.


PCP Degradation

     Table 1 shows the  results  of PCP removal  by the  BATS.   The
data are  presented as averages over the  two-week period of each
flow rate.   Two  important points are  illustrated by  the data  in
Table 1.   First,  that treatment  of PCP was effective up to  42 ppm,
and second, that even at the highest loading  rate  (5 gpm) effluent
concentrations below 1 ppm were achieved.

     To  determine  the   mechanism  of  PCP removal,  analyses   of
chloride ions  in the influent and  effluent  were  performed.   The
complete destruction, or mineralization, of PCP will yield chloride
ions in a 5:1 molar ratio (C1~:PCP).  Thus, if PCP is disappearing
based on chromatographic analysis,  and Cl~ is not appearing in a
stoichiometric ratio, it is likely  that the  PCP  is  being either
partially  degraded or  transformed.    Likewise,  analyses  of air
exhaust samples  were performed to determine whether air  stripping
was  a significant  removal  mechanism.    The  results   of these
measurements are given in Table 2.
                                    254

-------
      The data in Table 2 show that more Cl~ was produced than would
 be expected by stoichiometric degradation of PCP alone.  It is not
 clear why  the  amount  is  higher  than  expected;  however,  the
 hypothesis   of   PCP  mineralization  is  nonetheless   supported.
 Likewise, the  air stripping results indicate that  biodegradation
 was the predominant removal  mechanism.

      Concentrations of the various PAHs in the incoming well water
 were  lower than  expected.   In many cases the concentrations were
 below detection limits.  Thus, no conclusions on biodegradation of
 PAHs  were drawn from the study.  BioTrol has measured  significant
 removals of  2  to  4  ring PAHs  in  other studies.   Low levels of
 arsenic and various other  heavy metals were found in  the  system.
 There was little or no change in concentration upon passage  through
 the system.

 Bioassavs

      It was  anticipated that the  groundwater would be toxic to
 aquatic organisms since LC50 values for PCP tend to be  quite  low.
 Thus,  biomonitoring was performed to determine the system's  ability
 to  detoxify the  groundwater.  Fathead minnows  (P.  promelas) and
water fleas fD.  Macma)  were used for the bioassays.  Table  3 shows
the results of these bioassays.


%

Table 3
Bioassay Results
of wastewater in test
Minnow
Week Influent Effluent
1 0.35
2 0.84
3 0.26
4 0.54
5 0.61
6 0.66
>100
>100
>100
>100
>100
>100


water
Water
Influent
0.3
1.1
0.43
0.3
0.2
0.2



Flea
Effluent
>100
>100
>100
35
>100
>100

    The values in the table indicate the percentage of influent or
                                 255

-------
                 TREATMENT OF OTHER CONTAMINANTS

     The  technology demonstrated  in the  SITE program has  vide
applicability for  the  treatment of wastewaters contaminated  with
various toxic organics.  In fact, treatment of PCP  demonstrated a
"worst case" in that PCP is one the most highly toxic of the common
priority pollutants. BioTrol has had success treating a variety of
other wastewaters with the BATS approach.   These waters range  from
very  high  strength process water with  high  concentrations  of
substituted phenols to low strength groundwater contaminated  with
trace concentrations of benzene.

     Figure  3  shows the performance of  a full-scale BATS  unit
treating  gasoline-contaminated groundwater  (benzene  as  primary
contaminant) at 15  gpm.  This system has  consistently performed >99
percent removal since installation,  and  the effluent gualifies for
discharge without a polishing step.
                        BATS Performance
                         Benzene Treatment
             5000
             4000
             3000
             2000
             1000 -
                 influent, ppb
     Effluent, ppb
              40
                                                 30
                                                 20
                                                 10
                                                 0
                             10
15
20
25
                              Days
                                256

-------
                         BATS Performance

                 Treatment of Solvent-Contaminated
                           Process Water
                         Influent       Effluent
                           (mg/L)          (mg/L)        % Removal

Methylethylketone          43.0          <0.005          >99.9

Total BTEX                  1.3          <0.01           >99

Tetrahydrofuran             5.7           0.014          >99.7

Total Unknown Peaks         5.0          <0.05           >99
     In summary, the BATS  technology  demonstrated under the SITE
program has wide applicability beyond treatment of PCP-contaminated
groundwaters.  Many contaminants are amenable to biodegradation by
indigenous microorganisms,  and  high removal  efficiencies  can be
achieved at relatively high loading rates.  Contaminants amenable
to  treatment  include  gasoline   components,  solvents,  PAHs,  and
substituted phenols.
                                257

-------
     A-S Biotekniskjordrens
BIOTECHNICAL SOIL  PURIFICATION OF SOIL POLLUTED

BY OIL/CHEMICALS
 Presented at:

 "Second   Forum   on   innoyative   Harzadous
 Waste   Treatment Technologies -
 Domestic and  International"
Hay 15 - 17,  1990
Philadelphia,  Pennsylvania.
                            SUSANNE SGHI0TZ HANSEN
                            H.Sc.
      Administration &Anlaeg       Telefon         Telefax         Giro

      Maglehojvej 10            53504668       53504490       8465800
      4400 Kalundborg          33156872                    A/SReg.nr. 151387

                               258

-------
BIOTECNICAL SOIL PURIFICATION OF SOIL POLLUTED BY OIL/
By Maniging director H.sc. Susanna S. Hansen,
                   Biotecnical Soilclean Ltd.
SUMMARY;

Soil polluted by light oils and by some easily decomposable
cemicals such as esters, katones and alcohols may be cleaned
mikrobioally in about 2-8 months. Heavier pollution will
require a processing time of 8 - 16 months.

The processing is one of oxidizing, watering, and adding
nutrients and bacteria to the soil!                ««x"y

Soil polluted by organic solvents may be cleaned by gas
extraction, with subsequent cleaning of the air in a biofilter
and/or coal filter, in 1 - 4 weeks (MIBK, chlorbenzene, toluene,
6tC•J•
BACKGROUND - MICROBIOLOGY

Oil and many chemicals are organic substances - and may thus be
decomposed microbially.
60% of a surface garden soil contents of fugi and bacteria being
oil decomposing.
Descreibed  for the first time in 1895 when  Miyoshi  deschreibed
how the mould penetrate parafine.
More  than 100 species of.microorganisms belonging  to  bacteria,
mould fungi and yeast fungi are said to be able to decompose oil.
DEGREPATION  - CONDITIONS

First and foremost OXYGEN but also soil moisture (appr.60%),
supplementary  nutrients,  and  right pH  (7-9  )   are  elements
basic to how the oil/chem. decomposing bacteria thrive.
The  decomposition works well down to 8 - 5 degrees of C,  but  a
higher temp, will promote the decomposition.

SLUDGE FARMING

The pricipeles of "using11 natures own forces has been applied for
some  years  to get ridof oil sludge from refineries  where  this
sludge  has been scattered on agricultural areas and  mixed  with
the surface soil by harrowing.
Small guatities of oil (1%) would seem not to harm the plants
(but  it may be a danger to the groyundwater). On  the  contrary,
one may notice that grass the year after will grow better on  the
oilspot than outside.


                               259

-------
SOME PROBLEMS -
                                            ray grass
                                            •'  stress"
                                               that  AROMATE
It  has been shown that at 2% radishes and
growth inhibition.  The radish root s show
way of short, corkscrew curled roots.
Flourescence   microscope  photos  indicate
penetrate.
The oil are able to percolate to the groundwater(  few dops
1000 1 of drinkingwater).
Heavevy  degredable  parts  of  the oil  and  heavy  metals
accumulate.
                                                               will
 CONTROLED  "SLUDGE FARMING"

 A  place  where  the oil  soil could be taken  in;  Created  in  a
 controled way; without any waste to the enwiromental;  and put back
 into use; seemed to be a  useful solution to the problems  related
 to sludge farming but using  the positive elements such as:

 * it is a cheap cleaning  method
 * it is not energy consuming

 ** S!sSmeŁhoSywS!lUno? gfne^an^Iecondary pollution (smoke
   or polluted water:

 The drawbacks of this method are:

 * it is  fairly  slow

 i it doe^not dTSnySing about the inorganic pollution elements
   such as heavy metals:


 BIOTECNICAL SOILCLEAN  LTD.

 the plants which has been in function for three years; are  build
 after  a system; which has been arranged in giving high  priority
 to  the  safty  to the envinment; as would appear  from  the  plant
 sketch  below:  two  sets  of drainage  systems   (nos:  2 and  4)
 diaphragm  (3). Fibertex protect the diaphragm. The  top-layer  is
 jointed SF-brlcks. A specially close-vibrated concrete brick.

 Rainwater is recycled  from the percolate bassin,  which is airedso
 it may be sprinkled back over the stacks.
 Thus generating propagation of bacteria.
 Nutrients and tentatively bacteria are added to the   bassin.
               RAMPART KITH

               PLANTATION
 Fig.  i.

 Outline of plant.
           Watercanal
           for surface
           outflow
           to reservoir
                                                         Pile
                                                 .6 cm SF-stones
                                                 .Stable gravel


                                                 •Primary drainUI)


                                                 Protective matta)

                                                 Membrane tl)


                                                 Secondary
                                  260
                                                 Aeration _inotallnt ion

                                                 Reservoir for oer-
                                                 colator-and surface-
                                                 water

-------
 OPERATION AND OPERATIONS EXPERIENCE

SOIL DESCRIPTION:
The  way this place is run requires rather strict control
soil  received: we need to be sure that the soil  may be  re1
at  no risk, and without running too much of finanslal  ris)
should  be  able to clean the soil within the  time  allowed
financially, the ideal thing would, of course,  be if the soil  had
been  through  analysed  before arriving -  but  this  is  rather
unrealistic.
Analyses  provide  answers  merely to what was  asked  -  and  the
answers being greatly dependent on the analysis method, too.

The method seemed by us to be the most acceptable is: interwevs
of  the  landlords  and GC/FID (capillary  column)   to  chek  the
content of  extractable contents in the soil.
 Fig.  2.   Gascromatogrammes showing the oilcontents in the
          same soil  shown as funktion  of the time.
 On the site,  appr.  100.000 tons of soil  has  been treated,  divided
 on appr.  630  cases.

 The  major part  of this soil volume has  been  polluted  by  oil
 products (diesel, light fuel oil,  turpentine, petrol,  heavy  fuel
 oil)  minor  quantities(   10%) having  been polluted  by  MIBK,
 acetone ect).
 Degredation times ranged:  for the light fractions,  it takes
 two  to  eight  , and for the heavy  fractions,   from  eight
 sixteen months.
from
 to
                            261

-------
Fig.  3.  Airphoto of the plant.
The  acceptance criteration for soil for recycling is 100 ppm  so
far. For the light fractions, we normally do not recycle the soil
until the content has come down to appr. 50 ppm.

As  most  of  the polluted soil is  raw-soil,  there  is  certain
recycling limitations. It will generally be used as filler soil.


ORGANIC SOLVENTS


It is not acceptable, due to the working enviroment and for
enviromental reasons, to clean soil polluted by organic solvents
by this method. Due to this fact,  Biotecnical Soilclean Ltd.
cleans the soil for this type of pollution by gas extraction with
subsequent cleaning of the air in a biofilter and/or coal filter.
                           262

-------
Description of Gas-extraction Tent.

For the  tine being, we have two,  one stationary at the processing
site at  Kalundborg, and  one mobile which may  easily be moved  from
one site to an other, thus conducting "on site" purification.


The basic idea of the two is the same, that of
     1 a dense bottom
     2 one layer of specially
       developed concrete brick,
       being perneable (MULTI BLOCK
     3 a conpreaaed air-unit.
     4 an air - tight tent of an
       organic solvent proof diejhrag
       diapragn
     5 a blover, providing sub-
       pressure in the tent
     6 a blofilter
     7 a coal filter
      § test tubes
Fig. 4.   GAS-EXTRACTION TENT.
The mobile tent has been built up over a slightly enlarged open-
top container,  enabeling the  whole system to be pulled up on a
truck and moved.

TESTS.
In two rounds,  Institute  of Tecnology, Denmark,  has conducted
measurings  in order to evaluate how much would be absorbed by the
biofilter,  and how much would have to be absorbed by the coal filter,

We were aiming at having  as little as possible land up in the
coal filter.

     1.   Measuring series  (table 1) showed in the first round
          that 30 - 90 %  of what was blown out  (toluene) was
          absorbed by the biofilter.
                             263

-------
        TABLE It
        MEASUREMENTS OF THE REDUCTION OF TOLUEN IN THE MR WHEN CONDUCTED
        THROUGH A COMPOST FILTER AS WELL AS A CHARCOAL FILTER.
POINTS OF
SAMPLING:

j>
C


C

I:
C
(ppm)
8.6
1
0
1.4
0.7
0
0.9
0.6
0
TOLUEN
(mg/m*)
33.2
3.8
0.1
5.1
2.0
0
3.5
2.3
0
REDUCTION

B9
100

45
100

34
100
                          0.3
                                    1.3
                                                     62
         Airsaroples points:
               A:  Compost filter inAtake.
               B:  Compost filter exit.
               C:  Charcoal filter exit.
         Method:
               Airsamples in charcoal tubes.
               Analysis on G.C.
        Comments.

        The reasons why the figures were no higher nay be that the
        bacteria on the graftet filter had been starved for some
        months before we got the tests going, an that the
        concentration of the rather large air volumes blown trough
        were too low to keep a dense bacterial strain alive.
Various tests have been run  in the permanent site  tent.

As these tests  were required to be realistic, the  series of tests
was controlled  by what emerged in the  way of current cases.

I.e.,  the pollution cases cropping up  should also  be
      *

      *

      *

      *
of a  suitable size,  i.e. it  should be possible to
 ?rocess it in max  two rounds in the tent,
 t would have to be  a case polluted by organic solvents
only,  and not, for instance,  a nixed pollution with oil,
they  were to await our notifying the county of the test
and our recieving  their reply,
and the pollution  were not to include chlorinated
compounds which are  not handled by us on the site.
                              264

-------
To achive,  in a single test,  concentrations appropriately h,
the air stage to give clear measuringresults, we designed an'
accident by pouring five  liters of toluene out  in the soil in the
tent.

This will undoubedly produce  the best blow-uot  as the toluene
will not have time to penetrate into the micropores of the soil.
The measuring is just taking  place.

With these  tests, for which no mass balance have been made
 (beyond the artificially  generated pollution),  it may prove  hard
to 5udge haw much pollution has been blown off, and how much has
been decomposed by pumping oxygen into the soil.
Measuring results.
     TABLE 2;
     SUMMARY OF THE VARIOUS SOILS CLEANSED IN THE BLOW-OUT TENT.
CASE
NUMBER
470/89
515/89
Mixture
507/89
(104)
POLLUTION
ether
dibutyl-
phtahalat
toluen
clorbenzeen
CONSEHTRATION
(ppm in soil)
START FINISH
loot
26
1500
170
to
0. 5t *
(n.d.)
n.d.
n.d.
n.d.
/•»•*
TIME
(days)
9
36
10
49
2e
                                                        2O tummtr
         This eerie has been measured as reduction of the total area
         of the gascromatogramrae.
 CASE.

 The Mobile Tent case has been working since December on a
 pollution caused by chloride benzene.

 The pollution had been excavated and wal left in containers.
 In this case the biofilter was left out, as chlor benzene is
 toxic forbacteria, thus the filter would have been useless  (apart
 from the fact that some amount of absorbtion would probably have
 taken place).

 This cleaning was made under the worst imaginable conditions.

      *    In each container, there was about 25% more soil than
           there should be in the tent. For economical reasons, a
           whole container was emptied at a time.
      *    The soil was clayey and wet.
      *    The cleaning was effected in the coldest months of
           December, January and February.
      *    The blow-out could only take place between 1500 and 1700
           hours, as demanded by the building workers on the site.
      *    For periods of time, site electricity was switched off.
 Requirements were made by the County jof Frederiksborg:

      *    The soil was to be cleaned down to 0.1 ppm, for later
           deposit on a local dump!


                                  365

-------
        ppm
50

40
30
20
10

I

\
\
\
v..._
****f

3 10 20
                                                     DAYS
                                                   30
Fig. 5.  REMOVAL OF CHLORBENZEZE FROM  SOIL BY GASEXTRACTION.
A less expensive way of heating the air would undoubtedly
improve upon this method during the winter half  of  the year.
                            266

-------
In Situ Physical and Biological Treatment
of Volatile Organic Contamination:
A Case Study Through Closure

presented to:
United States Environmental Protection
Agency

SECOND FORUM ON INNOVATIVE
HAZARDOUS WASTE TREATMENT
TECHNOLOGIES:
DOMESTIC AND INTERNATIONAL
May 15-17,1990
Wyndham Franklin Plaza
Philadelphia, PA

presented by:
Richard Brown, Ph.D.
Groundwater Technology Canada, Inc.

Richard Tribe, P.E.
Ultramar Canada, Inc.
       GROUNDWATER
       TECHNOLOGY, INC.
                  267

-------
Introduction

      Even with aggressive policies of preventing product losses from occurring or from
having substantial impact, incidents can and will happen.  Additionally with the transfer of
properties, problems can be inherited, either knowingly or unknowingly.  In this context,
it is important to have a systematic approach to addressing environmental incidents.

      Ultramar Canada, Inc., a refining and marketing company operating in Eastern
Canada, has been a proponent and user of a systematic approach to addressing the loss
of petroleum products.  The recently completed clean-up of an Ultramar site in a suburb
of Montreal illustrates the application of this approach.

      There  are  three basic elements that Ultramar addresses  in systematically
responding to any leak or spill.  The first is the safety of the general public, employees
and facilities.  Once a site has been secured,  and all immediate danger has been
addressed, the second element becomes central to an effective response. This element
is an assessment of the situation to define  the impact of the problem - short and long
term, the context of the problem - its geological, regulatory and public settings, and the
extent of the problem - the amount(s) and distribution of the product loss.  When the
situation is defined, an appropriate remedial  action, the third element of response, can be
adopted. Selection of a remedial action involves setting the clean-up criteria and then
choosing the technology (s) best suited to achieving those criteria.

      The case study that will be reviewed in this paper involves a service  station in
suburban Montreal (Figure 1).  The product loss was the result of leakage over a long
period.  Eventually, the product migrated  across a busy intersection to a commercial
facility that served both  as a retail store and a warehouse/distribution facility.  Petroleum
vapors accumulated in the subgrade warehouse facility resulting in a potentially explosive
environment. The remediation project that will be discussed  in this paper, was initiated
in response to safety issues created by the vapor accumulation.
•Copyright Groundwater Technology, Inc. 1990
 All Rights Reserved.
                                         268
                      GROUNDWATER TECHNOLOGY, INC.

-------
Safety
      When a hazardous situation exists or may potentially result from a product loss, the
number one priority must be safety. This means protecting the immediate and long term
well being of the general public, employees at the impacted facility(s), and the facilities
themselves.

      The most immediate hazard which can result from a product loss is the migration
and accumulation of vapors emanating from contaminated soils and/or free  phase
products.   These vapors may represent  an acute hazard  in the form of explosive
environments.  Vapors can accumulate in basements, utility lines and  buildings with
subgrade construction.

      A second, longer term hazard due to  a product loss is the potential  health impact.
This may be due to the inhalation of vapors or the ingestion of contaminated groundwater.

      Where a safety issue  exists, the immediate response must be to secure the
impacted facility.  This  means stopping the source and removing the acute hazards:
Once the situation is secured the safety focus must be to address the longer term safety
issue such as health impacts.

      The safety issues that were created at the study site were a result of the long term
product loss, and were the accumulation of vapors in the basement of the commercial
building across the street from the site. The vapors entered the basement through small
cracks in the southern wall and floor (facing the service station).  At times, these vapors
approached explosive  levels.   For the  most part,  the  vapors were  present  at
concentrations well below explosive levels (Figure 2). They were however, detectable by
smell and created an odor nuisance and potential health risk for workers in the basement
warehouse.

      Ultramar's response to the  safety concerns was to  install a vapor abatement
system, consisting of a vacuum extraction and air intake system keyed into the crushed
                                        269
               JDD GROUNDWATER TECHNOLOGY, INC.

-------
stone fill  below the floor. This system had an immediate impact on the vapor levels
(Figure 3).

      Ultramar's second response was to install an interceptor well between the impacted
commercial building and the service station. This well served to intercept any potential
dissolved or free phase product migrating from the station (Figure 4).

      It  should be  noted that Ultramar undertook these  steps without it  being fully
established that a significant loss had occurred at Ultramar's facility or that the loss was
the source of the vapors in the commercial building. There were other underground tanks
to the west and south of the impacted building which could also represent potential
sources.  However, being the closest service station and because of the immediate safety
issues, Ultramar dealt with the safety problems at hand.

      With the safety  issues in hand, the next phase  of response was to assess the
source and extent of the problem.

Assessment

      Assessment involves three factors.  The first is identifying the  sources of the
contamination  - i.e. line leaks, tank leaks, overfills etc.  The purpose of this identification
is to stop the source.  The second  factor involved in assessment is to identify the
migration route(s) of the contamination. The purpose of identifying routes is to assess the
potential  for impact and thus the near and long term liability.  The third factor addressed
in an assessment is to determine the extent of the contamination both areally and degree.
Knowing  the extent  is  important to assessing, again, the potential impact and/or also
important to deciding on the appropriate course of action.

      During an assessment, the geology, hydrogeology and contaminant distribution
(adsorbed, dissolved, and free phase) need to be determined. This is accomplished by
the installation of monitoring wells and by obtaining soil borings.
                                         270
                      GROUNDWATER TECHNOLOGY, INC.

-------
      At the study site, a series of monitoring wells were placed around the service
station and on adjacent properties by Qroundwater Technology, Inc.  The wells were
gauged and sampled.  As shown in Figure 5, the Ultramar station did appear to be a
source of contamination and did appear to be impacting the commercial building. Wells
in the vicinity of the tank pits had several inches of free phase product indicating that the
probable source was a  tank leak. The adjacent station also appeared to be a potential
source of contamination, based on elevated dissolved VOC levels at that station.

      The groundwater contours from the site show that product loss at the station could
be  drawn into the store basement,  Figure 6.   As can be seen, there  is a natural
groundwater low at the  store.  This low draws groundwater radially into the area of the
store.

      The well logs were used to develop two  geological cross sections of the  site
(Figures 7 & 8). The reason for this groundwater low is evident from the geological cross
section. The basement  of the store was blasted into the bedrock underlying the general
area.   The bottom  of  the basement, being below the water table necessitated  the
operation of a french drain and surnp to keep the basement dry.  This operation caused
a draining of the groundwater from the area into the  basement sump. The drainage
pattern is basically radial. Any  contamination in the area would be drawn to the store
basement.   This  is what  appears to have happened  with the Ultramar site.   The
contamination resulting from long term losses at the Ultramar site was drawn across the
roadway  by the continual operation of the basement sump.

      One of the potential complications from this site geology is the ability to access the
adsorbed phase contamination.  In fractured bedrock, it is not unusual to have small
pockets,  "hot spots", of contamination trapped in  small fractures.  These fractures  are
often difficult to remediate and their presence can  result in the long term persistence of
low levels of dissolved VOC.

      A second focus of a site assessment is to define the phase distribution of  the
petroleum product that has resulted from the loss. There are basically four  phases of
                                        271
               in GROUNDWATER TECHNOLOGY, INC.

-------
 contamination - free phase, adsorbed, dissolved, and vapors. The actual distribution is
 a function of the site geology, the amount lost, and the type of product.  For gasoline, the
 relative percentages are as follow:  free phase 20-70%, adsorbed 20-70%, dissolved 0.5-
 3.0%, and vapors 0.1-1.0%.  In fractured bedrock, the amount of adsorbed contamination
 can be elevated due to physical retention in the fractures.

       These contaminant phases are in equilibrium with each  other.  Because of this
 equilibrium, addressing  one  phase often necessitates  addressing other phases,
 particularly when the focus is either vapors or dissolved contaminants.  At the study site
 the initial concern was the appearance of vapors in the store basement and the drawing
 of dissolved contamination into the basement sump. As the source of  the vapors is the
 free phase and/or adsorbed  contamination in  the  bedrock and fill.  Solving  these
 problems necessitates addressing the presence of adsorbed/free phase at the station and
 in the fill  underlying  the  basement.   Figure 9  shows  the  approximate  extent  of
 contamination at the site as determined during  the assessment. A small amount of free
 phase product appeared in monitoring  wells downgradient of the tank pit.  Adsorbed
 phase contamination extended across the street and into the basement fill. The adsorbed
 phase extends through the fill into the surface of the bedrock.

 Remedial Action

       Remediation of  a  product  loss involves not just the successful application  of
 technology, it  also entails identifying  and satisfying the stakeholders, those that have an
 interest in  the problem, and determining and  meeting the criteria that will be used  to
 regulate the cleanup effort.  For a remediation effort to be successful, it must reasonably
 satisfy the  stakeholders and meet the predetermined  cleanup  criteria.   Thus the
 stakeholders and the cleanup criteria become the context for evaluating and choosing a
 remedial plan.

       The stakeholders are persons who are affected by the loss and/or by the operation
• of the remedial system, and persons who have to give the necessary approvals for any
 cleanup program.  The list  of stakeholders can be quite long.  Basically there are four
                                          272
                inn GROUNDWATER TECHNOLOGY, INC.

-------
groups. Company personnel are an important group of stakeholders. These include not
only the operators of the impacted facility but also the insurance, marketing, environmental
and legal departments. A second group are the regulators that must be dealt with. These
include both local and provincial agencies.  Local officials who may be involved are
municipal officials such as the health department, the fire department and police. The
provincial agency was the ministry of the environment.  A third group of stakeholders is
the general public, both those immediately impacted as well as those in the immediate
area who may be "concerned" about activities on the site. The final group of stakeholders
are the contractors hired to address the problem.

      With the stakeholders there are two items of concern.  The first is the protection
of their well being, that is the health and safety of impacted parties and the economic well
being of the company.   The  second is the  understanding of stakeholders of the
effectiveness/appropriateness of the  remedial action(s) chosen. If either concern is not
addressed there  may be significant resistance to the remedial program.  Education is,
therefore, key to obtaining the necessary approvals. Education entails both an accurate
assessment of the problem and documentation of remedial options.

      The cleanup criteria chosen for the site must address the interests of the different
stakeholder groups.  Obviously the priorities and relative weights of these interests need
to be considered. At the study site the cleanup criteria chosen were as follow:
      o     Restoration and  maintenance of satisfactory working conditions in the
            warehouse structure.

      o     Minimum disruption of operations in warehouse facility and Ultramar Station.

      o     Work be carried out in a reasonable time frame.
                                         273
                 LQ GROUNDWATER TECHNOLOGY, INC.

-------
      o     The solution is permanent.

      o     The solution is cost effective.

Within this context of identified stakeholders and cleanup criteria, a number of different
cleanup options were evaluated.  This evaluation was conducted in two stages. In the first
stage all possible options were listed in a "brain-storming" session without regard to cost
or practicality.  In the second  stage the  options were  evaluated  in relation  to the
established cleanup criteria.

      The treatment options that were considered consisted of the following:

            Purchasing impacted business
            Excavating soil under basement
            Pump and treat
            In situ bioreclamation.

The first three were rejected.  Purchasing the building was neither a permanent solution
nor cost effective. Excavating the basement did not address contamination outside the
building,  would be extremely disruptive and very costly.  Pump and treat would not
effectively restore working conditions and is not effective  against  adsorbed phase
contaminants.

      In situ bioreclamation was chosen as the treatment option  that  most fully satisfied
the cleanup criteria.  Bioreclamation addresses all phases of contamination. It destroys
the contaminant, so it is a permanent solution. It can be completed in  a reasonable time,
2-3 years, and is, after installation, not disruptive of routine operations. Having a system
which could operate without  inconveniencing the  workers in the warehouse was an
important factor in  choosing bioreclamation.   While expensive compared to  non-
permanent solutions such as pump and treat, bioreclamation was  a cost effective solution
for this particular location in that it brought a permanent resolution to the contamination.
                                          274
                nm GROUNDWATER TECHNOLOGY, INC.

-------
      Even with the selection of a remedial operation, there were still several tasks that
needed to be accomplished before the system could be installed and operated.

      One such task was to select a contractor. The most important criteria for choosing
a contractor is a proven track record. Documented case histories and operating sites that
can be visited are invaluable in obtaining a first hand evaluation. Based on these criteria,
Ultramar chose Groundwater Technology, Inc. as the remedial contractor. Groundwater
Technology  had successfully completed  and  was  currently  operating a  number of
bioreclamation projects in the United States.

      The second pre-installation task was to gain  regulatory approval.  Many of the
regulatory questions  about impact  migration, disruption  to the business, safety,
performance were addressed in a report.  Other questions were handled in a face-to-
face meeting.  Key to  Ultramar's gaining regulatory approval was the "homework" that
Ultramar undertook before meeting with the ministry. Ultramar visited bioreclamation sites
in the United States to  see the operations and to talk with effected parties. This enabled
Ultramar to more knowledgeably answer the ministries' questions.  Having satisfied the
ministries' concerns, Ultramar obtained approval to proceed.

Principles of Bioreclamation

      Bioreclamation,  in situ aerobic  biological  treatment, is one  of the most versatile
remediation processes, dealing with a wide range of organic compounds in a number of
different hydrogeological conditions. In situ enhanced bioreclamation is a proven method
for remediating groundwater aquifers  contaminated  with petroleum hydrocarbons and
many organic chemicals. It is simply the use of common aerobic soil bacteria to degrade
organic contaminants.  It involves the stimulation  of indigenous bacterial through the
addition of essential nutrients.  The bacteria used in the process are already there.
                                         275
                 m GROUNDWATER TECHNOLOGY, INC.

-------
      The process can be simply viewed as a step-wise degradation as pictured below:

                      Biodegradation of Organic Chemicals:
                             Step-wise Metabolism

CONTAMINANT                CELL MATERIAL               MINERALIZATION
               Bacteria                           Bacteria
        (C,H)                  (C,H,N,P,0)
Bacteria which can feed directly on the contaminant use it to grow, producing primarily
cell material. Looking at this first step one can see what is required for stimulation. The
contaminant is  primarily carbon  and hydrogen; cell  consists  of carbon,  hydrogen,
nitrogen, phosphorus, and  oxygen (with minute amounts of minerals).  The  key to
accelerating this natural degradation process is to add sufficient nitrogen, phosphorous,
and oxygen (N,P,O) to balance the available carbon and hydrogen (the contaminant). On
a contaminated site, biodegradation is already occurring but is very slow because the
bacteria quickly expend the naturally occurring nutrients and oxygen.

   Once the specific degraders convert the contaminant to cell material they die and serve
as a readily utilizable food source for other bacteria. This secondary metabolism also
requires continued addition of oxygen and nutrients to sustain high biological activity. The
end result of the overall process is the conversion of the  contaminant into carbon dioxide
and water.

      The  key  to  in situ bioreclamation is getting the  nutrients and oxygen to the
contaminated area.  The first stage in this process is to hydrogeologically create  a
reaction vessel by pumping  and  injecting  groundwater.   This accomplishes two things.
First, the treatment area is confined; second, the contaminant and nutrients (N,P,O), the
reactants, are mixed. Since normal groundwater flow ranges from 10 to 200 feet per year,
it could take years for nutrients and oxygen under natural conditions to traverse even a
                                         276
               JLD GROUNDWATER TECHNOLOGY, INC.

-------
small site.   By increasing the gradient through  pumping  (draw-down) and injection
(mounding), transport times are reduced from years to months or days.  Since the
availability of nutrients and oxygen limit the process, the more quickly they can be added,
the faster the remediation.  Compared to the rate of nutrient transport, the biological
processes are essentially instantaneous.

    Once the groundwater system is properly installed, maintaining a proper nutrient (N,P)
balance is relatively straight forward.  The nutrients are generally soluble ammonium and
phosphate salts. They can be added at quite high levels and are eventually recycled by
the biological process.  Thus nutrient supply is easily managed.

    With nutrients, the key management principle is maintaining a threshold concentration.
Beyond this threshold level additional  nutrients do not benefit the biodegradation process.
The threshold level is a function of two factors: the basic requirements of the metabolic
cycle, and the degree of sorbtion and  retention of the nutrients by the soil matrix. The first
factor, the metabolic requirements is  quite small.  The ideal metabolic ratio of carbon to
nitrogen is 10:1, and  carbon to phosphorus, 30:1. Soil retention of  nutrients, however,
can be quite high on the order of 10's to 100's of ppm. It is this retention factor that drives
the nutrient concentration in applying  bioreclamation.  For example, the minimum nutrient
concentrations in an in situ bioreclamation system are typically 100 ppm versus 1-2 ppm
in a typical aqueous bioreactor. The difference is the impact of soil retention.

    With oxygen, however, the supply/demand situation is quite  different.  Maintaining a
high oxygen tension is critical to biodegradation.  First, considerably more oxygen than
nitrogen or phosphorus is required for biodegradation.  Each kilogram of hydrocarbon
that is metabolized requires  approximately 3.5 kilograms of oxygen to convert it to
and water:
            Reaction:   (Chy  +1.5 Ob  — > COb  +
            Weights:   14Kg    48Kg     44Kg   18kg
               IDE                    277
               ~TJJ GROUNDWATER TECHNOLOGY, INC.

-------
   In unsaturated soils (above the water table) oxygen supply is relatively easy, requiring
only  active aeration.  Each 1000 cubic meters of air supplies  approximately 250 kg of
oxygen. A soil vent system consisting of vacuum blowers, such as was employed under
the commercial building, is an effective means of aeration.

   In saturated soils (below the water table) oxygen supply is much more  difficult.
Natural dissolved oxygen  levels in groundwater are 4-5 ppm;  the maximum is 8 ppm.
Thus, if oxygen in a saturated system is supplied by air diffusers, a common method of
aerating water, approximately 125,000 liters of water would be needed to supply one
kilogram of oxygen. A more effective means of oxygen supply in a saturated system is
to use a chemical oxygen  source, hydrogen peroxide.

   Hydrogen peroxide decomposes by a  variety of mechanisms to give oxygen and
water:
            2HA  —> 2^0  +  Oj

During decomposition, each part of hydrogen peroxide supplies one half part of oxygen.
Decomposition is mediated by heavy metal catalysis, surface area effects, and biological
catalysis. The two most common decomposition catalysts are iron and enzymes such
as catalase.  Most common aerobic bacteria contain catalase  and therefore are able to
"use" hydrogen peroxide as an oxygen source.

   An advantage of using hydrogen peroxide as an oxygen source is that it is  miscible
in water and may, therefore, be theoretically added in any concentration.  However,
hydrogen peroxide is also a known biocide, as evidenced by its  use as a topical antiseptic
in many homes.  Thus, it must be added at levels that are not biotoxic but are  still able
to maintain high oxygen availability.  The  practical limit in  peroxide concentration  is
-2,000 ppm. At this level, 1,000 liters of peroxide amended water supplies close to one
kilogram of oxygen. This is greater than a two order of magnitude increase in oxygen
availability over air  diffusers.  Because of its ability to supply significantly more oxygen,
and  because the low permeable soils limited water circulation, hydrogen peroxide was a
critical factor in the use of bioreclamation at the Ultramar site.
               ,-H-~-                    278
               mi:
                  ID GROUNDWATER TECHNOLOGY, INC.

-------
   In choosing Groundwater Technology, Inc. (GTI) as the remediation contractor, one
of the factors that Ultramar considered in particular was GTI's experience with the use of
hydrogen peroxide. GTI has been actively employing hydrogen peroxide in bioreclamation
systems since 1983.

System Installation and Operation

      The bioreclamation system was installed at both the Ultramar service station and
the commercial building.  In both cases, the system was designed so that oxygen and
nutrients could be swept through the  areas of contamination effectively stimulating
biological activity.

      The service station system is pictured in Figures 10 and 11. Figure 10 shows the
injection recovery system, which consisted of two recharge galleries and an interceptor
trench keyed into a recovery well (RW-2). The first gallery was placed between monitoring
wells MW-1 and MW-2 to treat the area around  the southern most pump island.  The
second  gallery was  positioned to  sweep the front of the  station and treat the
contamination resulting from the tank-pit. The interceptor trench stretched across the full
front of the station and was positioned to prevent the migration of contamination towards
the commercial building. The process equipment for the service station system is pictured
in Figure 11.  Extracted groundwater was air stripped to remove dissolved hydrocarbons.
A dilute peroxide solution (10%) was continuously metered into the stripped groundwater
at about 2,000 mg/L.  On a batch basis,  a nutrient solution consisting of ammonium and
phosphate salts was also added to the groundwater. This was accomplished by diverting
groundwater to a 300 gallon mix tank to make a  nutrient concentrate.  The concentrate
was  then  metered into the water stream  over a 8-10 hour  period.  The amended
groundwater was then discharged into the recharge galleries. The injected water flow rate
was adjusted to maintain hydrogeological balance. Excess water was disposed of to the
municipal sewer.  Total groundwater flow  ranged 3-5 gpm.  At this rate the  system
supplied approximately 22 Kg of oxygen a day.
                                         279
               Em GROUNDWATER TECHNOLOGY, INC.

-------
      Figures 12 and 13 show the bioreclamation system as installed and operated in the
commercial building. The injection/recovery system is pictured in Figure 12. The injection
system consisted of six injection points that were drilled through the basement floor and
screened in the gravel fill. Recovery of groundwater was done through recovery well RW-
1 and the  french  drain sump.   Figure  13  shows the  process  equipment for  the
bioreclamation system.  Groundwater from the sump and  RW-1 was amended with
hydrogen peroxide at -2,000 ppm and nutrients (on a batch basis) and reinjected into the
injection ports.  A hydrogeological water balance was maintained by disposing of excess
water to the municipal sewer.  Total groundwater flow rate was 3-5 gpm.

Results

      The implementation and performance of the bioreclamation system varied by area
on the site.  As shown in Figure 14, there were four focus  areas in the site remediation
program.

      Area "A", the area surrounding the southern most pump island generally evidenced
low initial levels of contamination. It was the first area treated and rapidly responded.

      Area "B" was the area on the station around the  tank pit  (the source of  the
problem).  It was the most heavily contaminated with  both free phase and adsorbed
phase contamination being present. This area was the second area treated and was the
area treated for the longest time.

      Area "C" was the area between the service station and the commercial building.
While having some contamination present, this area was not specifically treated because
of accessibility. (Area "C" was under a heavily traveled roadway.) While not specifically
treated, contamination in Area "C", being downgradient, was none-the-less impacted by
the treatment systems in Areas A & B.

      The final area, Area "D", was  the basement warehouse. This area had both
dissolved and adsorbed phase contamination resulting in a vapor contamination problem.
               DTK                    28°
               LED GROUNDWATER TECHNOLOGY, INC.

-------
 Treatment consisted of venting and bioreclamation. This area was the last area to be
 bioreclaimed.  Response to bioreclamation was faster than observed in Area "B" due to
 the lower level of contamination.  The Area "D" system was terminated before the Area "B"
 system.

       The specific results for the bioreclamation systems are pictured in Figures 15-20.
 Results are given for the individual focus areas.

       Figure 15 shows the bacterial response to the addition of oxygen and nutrients for
 the bioreclamation system on the service station and in the warehouse. Prior to remedial
 activity the  counts  of hydrocarbon utilizing bacteria were  10* colony forming units
 (CFU)/ml in both areas.  The first  phase of remediation was the circulation of treated
 (aerated) water. This resulted in slightly higher counts in the treated wells relative to
 background wells.   Addition of nutrients with the aerated water resulted  in a 2-3  fold
 increase in hydrocarbon  utilizing bacteria.  Administering hydrogen peroxide to boost
 oxygen availability resulted  in an order of magnitude increase in hydrocarbon utilizing
 bacteria.  These results show that biodegradation requires the supply of nutrients  and
 high levels of oxygen.

       The results for Area  "A"  are depicted in  Figure 16.   Prior  to remediation  the
 maximum dissolved level was 15,000 ppb total VOC (volatile organic chemicals). With
 simple pump and treat, this  level was dropped to ~ 1,300 ppb at the beginning of the
 bioreclamation  phase.   Continued reduction in dissolved VOC's was rapid with  the
 subsequent  injection of nutrients and oxygen in recharge gallery #1.  The VOC level
 dropped from the 1,300 ppb to < 100 ppb  in six months of treatment.  At this point
 injection through recharge gallery #1 was terminated.  The post treatment levels of VOC
 varied  between 100  and  250 ppb over the next two years indicating that treatment had
 been effective.

      In Area "B", the results of treatment do not show the simple response as did those
for Area "A". This is due to the fact that the degree of contamination was much greater,
and the source was different, being a tank leak as opposed to surface spillage (during
                                         281
                ED GROUNDWATER TECHNOLOGY, INC.

-------
fueling of autos).  Figure 17 shows that the initial pump and treat operation (recovery
through RW-2) had a substantial impact on the contamination reducing the dissolved VOC
level from the appearance of free phase to VOC levels < 10 ppm. With the implementation
of the bioreclamation system the VOC levels generally remained low, with the exception
of two major concentration spikes at  16-18 months and 28-30 months of total remedial
operation. The reason for these spikes is evident in the overlying plot of depth to water.
As can be seen, the increase in VOC levels coincides  with a drop in the water table.
There are pockets of deep contamination which are  normally "locked" in the bedrock.
When the water table drops significantly, these are "exposed" and result in an increase
in dissolved VOC. It should be noted that the second  spike was much less than the first,
indicating that bioreclamation had effectively addressed even the deep contamination.
The residual contamination in Area "B" is now confined  to an isolated hot spot where there
is still minor bedrock contamination. Because the groundwater VOC levels in this formerly
highly contaminated area have stabilized at <10 ppm, the bioreclamation system was
terminated.

      Figure 18 shows the impact of the bioreclamation process on Area "C".  The results
parallel those for Area "B". Generally, there was a reduction in dissolved VOC levels.
There is,  however, a depth-dependent spike in VOC level. This, again, indicates that
there  are pockets of bedrock contamination that impact dissolved levels at low water.
What  is interesting in the results for Area "C" is that the concentration lags the depth to
water.  This is due to the fact that Area "C" is downgradient. The lag time is due to the
transport time for the dissolved VOCs to migrate from Area "B" to Area "C".  The results
for Area "C" show that general site contamination has been reduced. With the second low
water event (months 28-30) there was not a corresponding  increase in VOC levels as was
seen after the low water event at months 18-20.  This indicates that the bioreclamation
system is also addressing the trapped contamination  in the bedrock.

      Figures 19 and 20 depict the results for Area "D".  The initial phase of operation in
Area "D" was the vapor extraction and interceptor well system.  During this period the
dissolved VOC levels were reduced from 15,000 ppb to 3,200 ppb indicating that the
contaminant load was being reduced.  With the implementation of the bioreclamation
                                         282
                      GROUNDWATER TECHNOLOGY, INC.

-------
system the levels were reduced to <200 ppb in 10 months of treatment. During treatment
there were considerable fluctuations in VOC level due to changes in water table and the
activity of the bacteria.  However, the trend was  generally downward.  At the end of
treatment the residual contamination was confined to an isolated area at the front of the
store (near RW-1). In this area the subfloor consisted of a tight clay fill which trapped the
contamination. The impact of this hot spot was negligible on the final results as shown
in Figure 20. At the end of treatment,  the general air VOC concentration was under 500
ppb and the dissolved level <200 ppb. These were well below the target cleanup levels.
The bioreclamation/venting system  has been removed.
Costs
      Total project costs for the site have been approximately $850,000 direct costs. The
breakdown of costs are as follow:

      Emergency response (vent system & interceptor well)            $ 50K
      Site investigation                                            $ 80K
      System design                                              $ 50K
      System installation                                           $100K
      Capital equipment                                           $200K
      Operations (3 years)                                         $330K
      Post closure monitoring                                      $ 40K

      The equipment was designed and installed as self contained, trailer mounted units
so that it would be reusable. With the termination of the bioreclamation program, the units
from this site have been removed and installed at other Ultramar sites.  Thus the capital
expenditures were, in fact, a long term investment for Ultramar.

Conclusion

      The loss of petroleum products from retail operations is a situation that should be
prevented to the fullest extent possible.  However, such events can and do occur.
                                        283
               PZR
                     GROUNDWATER TECHNOLOGY, INC.

-------
      When product losses do occur having a systematic approach, makes managing
the costs, liability and exposure a controllable process. The approach that Ultramar has
developed and effectively employs is an orderly process consisting of:

      o     Addressing the immediate safety concerns

      o     Defining the extent and impact of the loss

      o     Identifying driving forces for cleanup, i.e., stakeholders, regulatory factors,
            and cleanup criteria

      o     Choosing appropriate remedial action (s)

      o     Installing and operating the remedial system

      o     Attaining closure.

      The above general process can be used to deal with  a wide range of loss
situations.  When site specific factors necessitate aggressive action, the use of innovative
techniques such as  bioreclamation can be used as part of this process to significantly
reduce contaminant levels and achieve a permanent solution. This paper documents the
successful application of Ultramar's response system to a site involving long term losses
of product resulting in  the vapor contamination of an adjacent commercial building.
Bioreclamation  was chosen, in this case, to eliminate persistent adsorbed phase
contamination.

      Activity at the site was initiated by a safety concern - the accumulation of vapors
in a basement warehouse. Ultramars response was to address these immediate concerns
and then assess the extent of the problem and its long term implication. Because of the
present and potential for continued off-site impact, Ultramar chose an aggressive remedial
approach, bioreclamation, which could bring the site to closure.
               HTW                    284
               :TD GROUNDWATER TECHNOLOGY, INC.

-------
      The application of the bioreclamation system was conducted in two parts. One part
dealt with the contamination at the Ultramar facility.  This was to treat the upgradient
source of the problem. The second part dealt with contamination specifically in and under
the commercial building.  This was to restore the impacted building.

      The total remediation program was conducted over a three year period. The
bioreclamation systems were operated over two and one half years.  The implementation
of the bioreclamation was a phased program with different areas being treated for different
periods of time. The overall remedial program successfully attained its cleanup goals and
the site is currently being monitored.

      In summary, dealing with product losses is a manageable process. It entails
developing a systematic process and using the appropriate tools.  Diligence  to both
technical and political (regulatory, social) issues makes the system work smoothly. A part
of the success of the remedial program at this site was the attention  to this non-technical
issue. The bioreclamation system caused minimum inconvenience to the workers and
operations of the warehouse and therefore was not an added "burden". Second, the
professionalism of the contractor and the Ultramar personnel involved in the remediation
prevented any conflicts between environmental activities and normal site operations, and
kept all the stakeholders  informed and satisfied.
                                        285
                      GROUNDWATER TECHNOLOGY, INC.

-------
                            BIOGRAPHIC SKETCH

                            RICHARD A. BROWN

      Dr. Richard Brown is currently the Director of Chemical Technology at Groundwater
Technology, Inc.  His responsibilities include project management, system design, and
business development for in situ and on-site treatment of hazardous wastes.  He has
developed and implemented remediation systems using biological and vapor extraction
technologies.   Dr. Brown is responsible for development of new  physical/chemical
treatment processes  including advanced oxidation and metal recovery.

      Previously Dr. Brown was the Director of Business Development of Bioremediation
Systems, a Division of Cambridge Analytical Associates, (CAA). Prior to joining CAA, Dr.
Brown founded FMC Aquifer Remediation Systems Division.

      Dr. Brown holds a B.A. degree in Chemistry from Harvard University in Boston, a
M.A. degree and a Ph.D. in Inorganic Chemistry from Cornell University in Ithaca, New
York. He has authored  15 patents and is involved with many professional scientific
societies such as The American Chemical Society of Petroleum Engineers, The Water
Pollution Control Federation, and The American Chemical Society.

      For more information on groundwater and soil remediation using bioremediation
technology, Dr. Brown may be contacted at Groundwater Technology, Inc.  located at 100
Youngs  Road,  Mercerville, NJ 08619,  phone number (609) 587-0300.
                                        286
                     GROUNDWATER TECHNOLOGY, INC.

-------
J
©
•/
J>
MM
-\s
— •, — ArmOKHlATC UXATnUOF
\ UNOCHCHCMIO TANk
f
_J
-l—j


-K>
k >
r
i 	 ••*
RMNC« UHDtRCROUNO TANKS— <^~^-ZIJl'
IISS


IIT»
MW-9
s.v * y /

Ss ~^\\
rvntn ISLANDS 	 ^_^
^°
A-l
r~!' '
1 I1 1
1 l< '
— L.J'.-1_
MW-9
• i l^
* — RfCM**»CC 6*
*
'
/
	 ,< 	 .
LLCHV NO 1 -^
ft
J
nkitKiw;
ONUM-.
— i rtuci . 1
CMCIOSUNC 	 v ' ]
""" ^^
r- Iti
OHAUST moit
BASCMCNT VCNTIlATKn
MW-S
•
CW-4
0
MW-IZ
• CW-3
0
MW-4 • A RW-I
^

•M
FIGURE 1
SITE MAP
PROJECT:
LOCATION
DRAWING NO.'
MONITORING DATE'
A RECOVERY WELL
• MONITORING WELL
• UNDERGROUND TANKS
weLl lOtNTWlCAtlON
ex • HW- i - veil, to
- w»it« T«gie f icv.ON rtm
© CORE WELL
0 «O
SCAII IN net
Mte "*< w-«r
» II
MW-U
•
' ,- INTCICCrTOO '»eNCM
amr^ MWH(<5 ^X^ IRCCHMCC SAueitT »OJI
Y"^"^"*"' _ ~~L— -^^ — Pours m.*NOS
! jj ~jj |! 1 ^ 	 »eCHAH«€ CAlXtHT HO.*
n^^ "«°
1 r sowici
U STATION
U0H.W-I4 onua^ri-,
Wi-J^«' J
*1
-------
CO
                                            BASEMENT LEVEL 1135
        FIGURE 2
VAPOR  CONCENTRATIONS
      IN  BASEMENT
      (Total VOC ppm)
PROJECf
LOCATON'
DRAWING NO.' 19-8167 2

O VAPOUR PONT

• AIR SAMPLING LOCATION

© INJECTION PORT

• CORE WELL
                                                                                                                                                OATC ••*.«.(.?
                                                                                                                                    LCGEND
                                                                                                                                    • MONITORING WELL

                                                                                                                            MW-4    • "ECOVERY WELL
                                                                                                                          VUVC f OK
                                                                                                                          VACUUM cannot.
                                                                                                                                            GROUNOWATER (.
                                                                                                                                            OL RECOVERY
                                                                                                                                            SYSTEMS. WC

-------
ro
CO
                                                                                                                              FMMIMC a
                                                                                                                         VENTILATION SYSTEM
                                                                                                                       WITH VAPOR CONTOURS
BASEMENT LEVEL 1135
                                                                                                                       O VAPOU* POUT


                                                                                                                       • AIK SAMPLING LOCATIOM


                                                                                                                       0 INJECTION POUT



                                                                                                                       • CORC WELL
                             I	I
                                                                                                                                   GROUNOWATER&
                                                                                                                                   OL RECOVERY
                                                                                                                                   SYSTEMS. WC

-------
PO
VO
o
                                                                                                                                               HGUME 4
                                                                                                                                    GROUNOWATER CONTROL WITH
                                                                                                                                         INTERCEPTOR WELL
                                                                                                                                     PROJECT
                                                                                                                                     LOCATION
                                                                                                                                     DRAWING NO.' 15-8167-23
                                                                                                                                     MONITORING DATE MARCH 11-67
                                                                                                                                     A RECOVERY WELL
                                                                                                                                     • MONITORING WELL
                                                                                                                                     !_ ___ j UNDERGROUND TANKS
                                                                                                                                       •III IDCNTiriCATION
                                                                                                                                       in  MW  I -WC II 10
                                                                                                                                           9O.ll - •»?!• 1»»U (1C VOX FCC n

-------
ro
vo

©
J<
MW
•
S>
- •^X""' *"*WlllilT€ IOCATWW OF
* UNOC ffdiouMO TANK
i
h ~"-i

j
-»
L ,
A
__ r 	 1
j i
romMft tmoCKiiouNO T«N«$ 	 «C-i-"^_~ " " '
II 5»


IITf
XMW-9
V /**&
- 	 /_ — ^
//-x
/* / 0
( /T~i
l &
KHtn taunt 	 »>_ I O J
^V^ / « MW-7 /
•\ 1 " •30-°00/
^ 1 \ y
\r_.l\ir-| X
« !' f**-*
V i* !i i ^
Nj- • i _-/
., - /^-^
\> • >
*"
u
/
x^"
— ,.<
J
MmiMC
	 ^— (IIUtnT (BOM
\ (AMUCNT VHHItHlOM
\ MW-6
\«
1131
MW-5 	 	
• >r ^ -.
/ STORE \
/X ^\ \
/7' \ \
FIGURE 5
PETROLEUM DISTRIR'JTION MAP
PROJECT.
LOCATION'
DRAWING NO • K- 0167 - 4
MONITORING DATE' FEB. 9-87
A RECOVERY WELL
• MONITORING WELL
1 	 J UNDERGROUND TANKS
Will IOCNlirK>T>ON
ft «»•! - WILL 10
- »»1IH HBll HCVIINfCCfl
O «O
SCAll W Ff < t
OATt «M ;o-«*
MOIt CONTOU* U«C IMOICAKS MtOKAClC
OTCHD OF VITHOLCUM COMtlMINMHM
/ MW-4 • A RW-I 1 1
' Z4.000 > 1
\ / 1 _»I«»-H
l«^ H«0 |
|t^«._^ tt«wet I CAiwet
LJ Simon 1
.--1 — s-±
MW 2 * '
* l>lo°

-------
ro
(O
ro
          FIGURE 6
GROUNDWATER CONTOURS
PRO JET-
ENOROIT
DC SSI N NO 'IS B063- 3
DATE D'OBSERVATIONi M-ZS-86
• PUITS OBSERVATION
A PUITS OE RECUPERATION
                                                  ANOCNS IKSCWWIIS CNfOUIS
                                                                                                                            MfHT
                                                                                                                      DC LA VI NTH ATION
                                                                                                                      CHI >OUt-5at
                                                                                                                                              KM NTiriCATtON DUfUITS
                                                                                                                                               II MW-I IOCNT OUPUITS
                                                                                                                                                  > NM>n PWWAtKMX
                                                                                                                                               RESERVOIRS ENFOUIS
       CROUNOWATER  FLOW
            DIRECTION
                                                                                                                                                         GROUNOWATER
                                                                                                                                                         OIL RECOVERY
                                                                                                                                                         SYSTEMS, INC

-------
ro
VO
CO
                                                                                                                                                          FIGURC 7


                                                                                                                                                        GEOLOGICAL

                                                                                                                                                     CROSS  SECTIONS


                                                                                                                                                PROJECT-

                                                                                                                                                LOCATION

                                                                                                                                                DRAWING HO ' 9«l -130 -8167-3

                                                                                                                                                MONITORING DATE


                                                                                                                                                A RECOVERY WELL

                                                                                                                                                • MONITORING WELL

                                                                                                                                                '_	j  UNDERGROUND TANKS

                                                                                                                                                  WELL IDC)*TI*(CATION

                                                                                                                                                  rx • MW i - we it 10
                                                                                                                                                          - WATCH TABie ClfVIINfffT)

                                                                                                                                                © CORE WELL

                                                                                                                                                         0
                                                                                                                                             IttTCKCCPTO* TftCMCH
                                                                                                                                             IKCCHAKCt CALLCRT HO it

                                                                                                                                                  ram ISIAMOS
                                                                                                                                              HCCMAMf CALICUT HO I

-------
\o
                         Figure 8    GEOLOGICAL  CROSS   SECTIONS
           A
          MW-2
SERVICE
STATION
           B'

           MW-I
             I	
          MW-3 RW 2
            '   '
MW-II MW-4  MW-12
  '     ' .   '
                                      ORGANIC FILL
                                     	
                                   SILTY SAND GRAVEL

                                  WEATHERED LIMESTONE
                              i   1,
                                   LIMESTONE BEDROCK
                        MW-5
                       	1	
                                 A1
                                MW-6
                                  J
                                                          STORE BASEMENT
                                                B
       MW-2
MW-7
  I
MW-8
  I
                                SILTY SAND GRAVEL
                                WEATHERED BEDROCK
                                	—	
                                LIMESTONE BEDROCK
                                                                   •50-
                            Horizontal Seal*
                                                                                  Vertical Seal*
                                                                                     LA COMPACNIE G. T. CANADA
                                                                                     CROUNUWATER TecH.vH.oct INC

-------
ro
vo
tn
                                                                                                                                                            FIGURE 9

                                                                                                                                                     PHASE  DISTRIBUTION
• UNMKCROUM) 1«M«S
                                                                                                                                                 PROJECT-
                                                                                                                                                 LOCATION'
                                                                                                                                                 DRAWING NO.'94I-I3O-8I«T-S
                                                                                                                                                 MONITORING DATE-
                                                                                                                                                 A RECOVERY WELL
                                                                                                                                                 • MONITORING WELL
                                                                                                                                                        UNDERGROUND TANKS
                                                                                                                         OHAUST FROM
                                                                                                                         BASCMCHT VINTIIMION
                                                                                                                                                    wfLL IMMtlf ICtTIOK
                                                                                                                                                    [»,. UW-I -WJlt 10
                                                                                                                                                            - W<7€» T«*lf (ICV (IN F«T>
                                                                                                                                          ADSORBED
                                                                                                                                              INItHCtfTOK TIICNCH
                                                                                                                                              IBtCM««Ct CAUCKT NO SI
                                                                                                                                                   rum niiuios
                                                                                                                                               IWCNJIIWC •Au.or no t
                                                                                                                                 n«o
                                                                                                                                 scnvict
                                                                                                                                 SIAtKW
                                                                                                                                                      ItlltlMHAHM HI N« llltWMX (JIMS l*«

-------
IV)
ID
CT>
                   VJ
1
v_x


	 Jf
^;
C^
                                                          V-iowwts tm
                                                                                                        - «*TI(I OISCHA««C LINC
                                                                                                                                                   ricunc 10
                                                                                                                                        BIORECLAMATION  SYSTEM
                                                                                                                                           SERVICE  STATION
                                                                                                                                         PHOJCCT UlTRAMAN CANADA INC
                                                                                                                                         ioc« r ION-i UNGur utL.oueecc
                                                                                                                                         ONAWING NO  I5-07O2-4
                                                                                                                                         A «ICOVt«Y WtLL
                                                                                                                                         • MONITORING WCLL
                                                                                                                                         f) MONITORING PORT
                                                                                                                                                  SC*l I t«i n f T
                                                                                                                                      •«OTf  U«  O«G'. «^ 87O7 5 * C f
                                                                                                                                           O< ICCTIOM^ A « TO  I f

-------
rvi
10
                                                                                                                    SERVICE STATION

                                                                                                               NUTRIENT INJECTION SYSTEM

-------
10
00
                                            BASEMENT LEVEL
                                      OXYGENATED WATER LINE 2*« PVC
        ricuftc 12

 STORE BIOREMEDIATOR
        SYSTEM

PROJECT
LOCATON'
DRAWING NO.' I5-BI67 2

 O VAPOUR  POUT

 • MX SAMPLING LOCATION


 © INJECTION PORT


 • COM WCLL

        o         >o*

        sc»ic i
                                                                                                                                              GROUNOWATEft t.
                                                                                                                                              OIL RECOVERY
                                                                                                                                              SYSTEMS. WC

-------
                                                                                                                            STORE
                                                                                                                 NUTRIENT INJECTION SYSTEM
                                                                                                                   — (XJ— THREE- WAT VALVE

                                                                                                                         SOLENOID VALVE
	Ji
        	Tc.1
        l             L—'
                                                                                                                I  |l  f^l CIHM MiWIIIHl UN KlltRIW MSTIHS I
                                                                                                                L "  ™^ I ««« \K 14 HlllllHMkK

-------
                     v:
o
o
                                    MW-IO
             UMXMIIOUia TANK
                                                                  mm ISLAMOS
                                                                                                 I  t<
                                                                                              ,  1..1I-
                                                                                              'Mw-e
V


MW-9
• y

^
— t
MW-7
l l
ij
i
INTERCEPTOR
TRENCH 	 ._
J ' 1
mtRKING
ENCLOSUHt 	 -^ 71
MW-« W i
EHHAUST EROM 1
DASEMENT VENTILATION j
II S5
WMW"S STORE
AREA'O"
1
CW-4 j
©
MW 12 ;
• CW-3 I
0 1
FIGURE 11
TREATMENT FOCUS AREAS
PROJECT
LOCATION-
DRAWING NO = 941 -ISO -8167-3
MONITORING DATE'
A RECOVERY WELL
• MONITORING WELL
'• ' '• UNDERGROUND TANKS
WELL lOENTiriCATION
f> M* 1 -*ELL ID
- WATER I««LEElEV(mFEET)
© CORE WELL
0 «0
SCALE W FEET
DATE AM 2O-C

MW-4 • ARW-I
I— -f '"""
^ 	 1 	
AREA-C"
MW-ISl
• f
^ AREA'B"
RW-Z"! 	 MW-<6 f
\^T^]^^

INTERCEPTOR TRENCH
IRECHARCE CALIERT no.5)
_, 	 PUN'S ISLANDS
- RCCMAMC CAILERT NO t
Ir^ JERVCt I CARME
LJ STATION .
* AREAVMTi"^ ~H'^«"^!H" |
u^tt-2.^- 	 r' 	 1 '
r«" 	 	 — 1 t'ttf-l^

s '

                                                                                                          «*CCMM>OC CALICMV NO t -

-------
co
o
           E
           v.
           3
           c
           UJ
           m
                 lo'-,
                 io9-
                 I04-
                 IOS-
                      Figure: 15      BACTERIA    COUNTS



1 — 1
A
I

—
B
\

—
C
!
n
A) CHANGE IN HYDROCARBON UTILIZING  BACTERIA COUNTS
  DURING BIORECLAMATION - STATION SIDE AREA A8B
o I0 1
z
N
^ 10'-
o IO4-
CD
OC
O
g io5-
o
X
V
c V/
B
A










\
\






$
\
1







7-
y,
\
\
u*
A
B
C

                    B) CHANGE IN HYDROCARBON UTILIZING BACTERIA COUNTS
                      DURING BIORECLAMATION-STORE SIDE AREA D
BACKGROUND WELLS



TREATED WELLS

(Nutrient* 8 oxygen added)

AERATED WATER

AERATED WATER + NUTRIENTS

H202+ NUTRIENTS
                                                                         e
                                                                      T CANADA
                                                            GROUNUWATCR TECHNOLOGY INC

-------
                               zoi;
                       §
s
o
a
8
     5-
  o
  S SH
     SH
     8
§0
     w
               TOTAL  DISSOLVED HYDROCARBON (PPB)
       - BEGIN BIORECLAMATION-
       -• STOP BIORECLAMATION —
                m

                                                        m
                                                          «•
                                                          •i
                                                        >
                         o
                         TI
                         CD
                         O

                         O

                         >

-------
           (FREE PHASE)
       I2O
                     Figure 17  AREAHBM  IMPACT  OF BIORECLAMATION
CO
o
CO
                                                    AREA'S'
                                                    DEPTH TO
                                                     WATER
                                                              30   32   34  36
                                MONTHS OF  OPERATION
                                                                   LA CONPACNie G. T. CANADA
                                                                   GROUNOWATF.R TECHNOLOGY INC

-------
                                       111
      60-
                    Figure 18  AREA       IMPACT OF BIORECLAMATION
      50-
00
o
                              12  *   16   18  20   22   24


                                MONTHS  OF OPERATION
                                                           2»
90
                                                                           8.0
                                                                   LA COMPACNIC C.T CANADA
                                                                   GROUNUWATF.R Tr.aiNot.ocv INC

-------
     26OO-I
                    Figure 19  AREA"D"  IMPACT OF BIORECLAMATION
     2OOO-
     1400 •
co
o
tn
      eoo-
                              MONTHS  OF OPERATION
                                                                      96
                                                                LA COMPAGNII: G. T. CANADA
                                                                GROUNIJWATKR TECHNOLOGY

-------
Co
           \
                 \
     m«>i«^«»* »•»
BASEMENT LEVEL
                                                                                                     HOT SPOT ARE*
                                                                                                     (tfiMolvwt <2n>m)
                                                                     IhrMhold)

                                                                    <200ppb  DISSOLVED HYDROCARBON
         FIGURE 20



POST CLOSURE  STATUS


PROJECT

LOCATONt

DRAWING NO.- 13-8167-2


O VAPOUR  PONT


• AIM SAMPLING LOCATION



& INJECTION PORT



• CORE WELL



       0         10'


        SCALC IN rCCT

                   D«1C MAY tMf-
                                                                                                                                                    GROUNtMMTER 6
                                                                                                                                                    ON. RECOVERY
                                                                                                                                                    SYSTEMS, me

-------
         GEODUR PAPER
            FOR

SECOND FORUM ON INNOVATIVE
     HAZARDOUS WASTE
 TREATMENT TECHNOLOGIES:

DOMESTIC AND INTERNATIONAL
     UNITED STATES
     ENVIRONMENTAL
       PROTECTION
         AGENCY
         May 15-17, 1990
       Philadelphia, Pennsylvania
            USA
            307

-------
               Lecture by Svend Mortensen GEODUR A/S
                and introduction to The GEODUR System
THE HISTORY OF GEODUR:

GEODUR is a product made of chemicals in the molecular stabilization technology. The product
is the result of concentrated scientific research and practical experiments.

The chemical process as a result from using GEODUR makes it possible to stabilize loose organic
matters and inorganic materials to a product which is very much alike concrete.

These loose organic and inorganic materials are e.g. mould, desert sand, fly-ash and mud, as
well as slag from combustion
plants and power stations, paper, waste,  etc.

WHAT IS GEODUR:

The name "GEODUR" gradually has become several different things. The following list explains
the word:

The GEODUR Concentrate:  A product (The GEODUR Product) produced of chemicals.

GEODUR A/S:             The company which is producing and marketing the GEODUR
                         Concentrate.

The GEODUR Solution:     A compound of the GEODUR  Concentrate and water  mixed in
                         the ratio 1:100 (1 kg of concentrate for 100 Itr. of water).

The GEODUR Process:     The chemical process, which makes stabilization  possible.

The GEODUR System:      The method of stabilizing which binds heavy metals in soil, fly-
                         ash or other waste products, and the method of producing soil
                         concrete.
The GEODUR Soil
Concrete:                 A type of concrete made of soil, cement and the GEODUR Solution.

The GEODUR Treatment:   The treatment itself by means of the GEODUR Process


MORE ABOUT GEODUR:

When producing GEODUR Soil Concrete almost all kind of available local materials are being
used as aggregates instead of the ordinarily known aggregates. These available local aggregates
possibly can be the soil excavated e.g. the top soil such as mould,  clay, sand etc.

The GEODUR Solution and a small amount of cement (typically 4-8%) are to be mixed into
the soil.

The quite new aspect of making soil concrete is the fact that the GEODUR Process makes it
possible to bind the organic matters
which are  in almost all kind of soils and in unsorted and unwashed types of sand.

                                     308

-------
It has always been impossible to produce a type of concrete of ordinary soil or of polluting waste
products.

The GEODUR Process makes these things possible.

                                       oo 00 oo

The GEODUR Process makes it possible to bind heavy metals in different materials, so they
cannot be set free again. Typically it is heavy metals in polluted soil, fly-ash etc.

Various tests and analysis have been made, e.g. an analysis of
leaching has shown that the leaching of heavy metals is so poor that the heavy metals cannot
be  measured in the samples stabilized by GEODUR.

The following kinds of soils and solid wastes have until now been solidified:

        Garden topsoil, clay and clayey humus, sand and sandy humus, silt, desert sand (salty),
        coast sand (salty), laterite, volcanic lava, mud (sediments with blue clay), municipal sludge
        (heavy metals), hazardous chemicals and oil contaminated
        soils, fly ashes, stabilized and raw sulphite powder.

Among  the test results the following should be mentioned:


                                       FLY ASH

Variable              Content in untreated      Content leached
                           sample              out of sample
                            mg/l                   mg/l
Arsenic
Barium
Cadmiumb
Chromium
Copper
Mercury
Lead
Zinc
(As)
(Ba)
(Cd)
(Cr)
(Cu)
(Hg)
(Pb)
(Zn)
48
449
<5
56
135
0.11
145
249
<0.02
<1
<0.02
0.04
<0.02
<0.001
<0.02
<0.00
Apart from the teachability tests of contaminants also the chemical and ecotoxicological properties
of the leachate are studied together with the mechanical strength and durability of the solidified
materials. Also petrographic analysis have been carried out.

It should be mentioned that a special ecotoxicological fouling test is carried out on different
solidified materials exposed in-situ in receiving waters.

The GEODUR Concentrate is a non-poisonous product (see certificate from the Danish Toxicology
Centre).  Therefore  it does not harm  neither our environment nor the persons being using
GEODUR.
                                           309

-------
FIELDS OF APPLICATION:


1.     GEODUR  and the Environment.

       GEODUR can be used to improve our environment, both

       a)        preventively for infrastructure purposes in the fight against further pollution,
                and

       b)        in the treatment of soil already polluted.

                                   oo OO oo

       a)        GEODUR together with e.g. fly-ash is being used for infrastructure purposes,
                e.g. for road construction. The fly-ash being stabilized by mixing it into cement
                and the GEODUR Solution  is applied on to the ground section excavated
                as the stabilizing base under asphalt. (See Road Construction Project Report).

       b)        Polluted soil is being stabilized by mixing it into the GEODUR Solution and
                cement (only a few %). Then the GEODUR Process makes the soil "concrete
                like" to be casted in any form required easy to be kept on site, or to be re-
                used e.g. as embankments, coast protection,  sound banks, roads, etc.

       Furthermore GEODUR is being tested to solve storage problems of nuclear waste
       materials to prevent these materials  from being spread by dust formation.


2. GEODUR  and Construction.

       As mentioned it is possible to make soil concrete by means of the GEODUR Process
       by replacing the ordinary aggregates by the top soil of the site.

       Thus the purchase and the transportation costs of the aggregates are being eliminated.

       By applying the very special concrete "ANTI-FIRE" on to the GEODUR Soil Concrete
       the strength is being increased  and the concrete appears as ordinary concrete.

       GEODUR  Soil Concrete  is Composed of:

                          Soil from  the site.
                          GEODUR Concentrate.
                          Portland Cement.
                          Water.
                                      310

-------
        How to Produce GEODUR Soil Concrete on the Site:

                           1-2%  GEODUR solution  (Calculated  as percentage  of the
                           soilweight).

                           Dry mixing the soil.

                           Dry mixing the cement required (typically 4-8% of the soilweight
                           which is far less than the amount of cement when producing ordinary
                           concrete)

                           After repeated mixing, water is to be added until the compound
                           reaches a proper consistency for compressing (roads) or for casting
                           (building materials).

       The GEODUR Process activates chemically all the natural binding properties of the soil
       by changing the surface tension. This process of agglomeration produces an ultimate
       stabilizing effect on the soil with high density and strength.


       The PH-factor of the soil  does not have any influence on the final strength of the
       GEODUR Concrete, because it is chemically controlled by the GEODUR Solution.


       Different Types of Soil Can be Used for GEODUR Soil Concrete:

                          Black vegetable mould (garden top soil).
                          Clay and clayey mould.
                          Sand and sandy mould.
                          Mud.
                          Desert  sand (salty).
                          Coast sand (salty).
                          Volcanic lava.
TESTS AND ANALYSISES:

Following tests and analysises have been carried out:

       1.         Ecotoxicological Fouling Test in Sea Wate by:
                      The Danish Water Quality Institute VKI
       2.         Leaching Test by:
                         Steins Laboratory, Denmark and
                     RIS0 The Danish Atomic Research Center
       3.         Petrographical Analysis by:
                         The Danish Technogical Institute
       4.         Mechanical Strength  and Durability Test by:
                 ELKRAFT, The Danish Power Station Association.
       5.         Full Scale Test of Road Construction by:
                 ELKRAFT, The Danish Power Station Association.
       6.         Permeability Test by:
                          Geological Institute, Denmark
       7.         Certificate by:

                           Danish  Toxicology Centre

                                      311

-------
 TEST IN SEA WATER

 The field tests are designed on the basis of the results of a preliminary  investigation in 1989. With
 the objective to get an indication of the stability and leaching properties of different GEODUR stabilized
 waste materials and concrete, test blocks were exposed  in the sea during a period of four months.

 Colonization of fouling  organisms  has been followed with regular intervals, and the accumulation of
 selected trace elements in the biota was examined at the end of the exposure period.

 MATERIALS AND  METHODS

 Five GEODUR stabilized materials have been exposed in the sea together with blocks of commercially
 manufactured concrete as a reference material (table 1).

 Table 1   Materials exposed in  the sea
       NO.     I  CODE    I           DESCRIPTION

1

2
3
A

5
6

106

98
136
U9

110


Fly ash with stones i
i
Polluted soil from Frederikssund i
Sewage sludge from Clostrup Purification Plant
i
Incinerator ash from Refshalec
i
Unpolluted soil from Canlcse !
Concrete (reference)
Test blocks sized 10 cm x 10 cm x 10 cm were delivered by GEODUR. At the end of July six blocks
of each type were suspended one metre below the surface on a raft in the Innerbroad of the Isefjord,
Zealand. The Isefjord is a moderately eutrophicated area with a salinity  around 20 o/oo.

After 5, 10, 12, and  16 weeks  of exposure one or two blocks of each material were collected and
photographed. The species composition and the biomass (60°C constant weight) were described.

At the end of the experiment samples of the dominating fouling organisms (hydroids) were collected
from material Nos. 3,4, and 6, freeze-dried and analysed for cadmium and lead. Samples have been
digested in a teflon bomb system using quartz destined nitric acid. Analysis has been performed by
heated graphite furnace AAS. Background correction and standard addition technique was used.

The blocks were weighed  before the exposure in the sea and after 2 months at room temperature.

RESULTS AND DISCUSSION


Physical Stability

The loss of weight during the exposure in the sea was approximately 1% in concrete (table 2). The
greatest loss of weight was found in incinerator ash with values between 9-16%. This is 3-4  times
higher than the other stabilized materials, where the mean toss of weight varied between 3.5-4.1%.
There was no relationship between the  loss of weight and the length of the period of exposure.

                                           312

-------
Table 2. - Loss of weight In percentage after exposure in the sea.
NO.
1
2
3
4
5
6
MATERIALS
riy «h
Polluted foil
Sewage sludge
Incinerator ash
Unpolluted coll
Concrete
PERIOD OF EXPOSURE
5 weeks
5.6-5.7
4.2-4.S
3.9-4.6
12.1-16.0
2.5-2.8
0.6-1.6
10 week*
3.2-4.2*
0-3.4
2.8-3.7
5.2-15.4
2.5-2.6
0.9-0.9
12 weeks

7.7
4.2
11.7
4.4
0.6
16 weeke

5.2
4.2
9.2
6.3
1.1
* drifted ashore
Colonization, Species Composition and Biomass

After 5 weeks of exposure there was only a slight 'hairy' fouling on the blocks and quite a numerous
fauna of free living crustaceans, especially amphipods. After 10 weeks the blocks were covered by
a dense growth of hydroids (Laomedea longissima) (figure 1.).

A further development of the fouling community was seen after 16 weeks (table 3).


Table 3. - Species composition of the fouling community after 16 weeks of exposure.


COELENTERATES
CRUSTACEANS
Microdeutopus gryllotalpa
Corophium Insidiosun
Idotea baltica
Dexanine spinosa
Phtisica marina
Ostracods
POLYCKAETES
Nereis sp.
Harmothoe sp.
Blue mussels
Snails
Sea squirts
Number of species
MATERIAL NO.
2


444
44
44
4

444

4
4



9
3


+ 44
4-4
4

4
+ +





*
e
4


444
44
+
4
4
4



4


9
5


+ 44
+
4
4



4
+



e
6


+++
++
++
+
++


+


+

9
                    < 5 individuals
                    5-25 individuals
                    > 25 Individuals
During the period of exposure there was a slight increase in the richness of the species, and the same
species and the same number of species developed on the different sorts of stabilized materials and
on concrete.

The biomass of the fouling organisms (totally dominated by hydroids) increased rapidly after 10 weeks
of exposure (figure 2 and table 4).

The biomass developed in the same way on blocks of concrete, sewage sludge, and unpolluted soil.
A higher biomass was found in  incinerator ash, and until 12 weeks also on blocks of polluted soil.

                                         313

-------
However, during the last four weeks of the experiment there was a staanation in the biomass on the
blocks of polluted soil.
                                     BIOMASS
                                                             .. 4
                                                              16
Figure 2 - Fouling biomass on test blocks after 10,12, and 16 weeks of exposure in the sea.
Table 4 - Fouling biomass on test blocks after 10,12, and 16 weeks of exposure in the sea.
NO.
2
3
4
5
6
MATERIAL
Polluted soil
Sewage sludge
Incinerator «»h
Unpolluted soil
Concrete
10 WEEKS
g DW
6.72
2.36
3.86
2.30
2.65
*
253
89
145
87
100
12 WEEKS
g DW
8.87
3.59
9.43
5.17
3.27
*
271
110
288
158
100
16 WEEKS
g DW
8.66
11.18
13.58
10.64
10.84
%
BO
103
125
98
100
Trace Elements

The concentration of cadmium and lead is the same in hydroids grown on blocks of sewage sludge
as on blocks of concrete and the level is almost the same as in blue mussels from unpolluted areas
(table 5). In contrast, slightly elevated levels of cadmium and very high values of lead are measured
in hydroids grown on blocks of incinerator ash. Compared to sewage sludge, which has a higher content
of lead, the stabilization and binding of lead in incinerator ash has been unsuccesfull. The blocks of
incinerator ash were porous and had the  greatest  loss of weight (table 1).
In spite of the  different physical and chemical properties the highest biomass was found on blocks
of incinerator ash (table 4).

                                           314

-------
Table 5 - The concentration of cadmium and lead in hydroids grown on stabilized test blocks
          of sewage sludge, incinerator ash and concrete and the concentrations in unstabllized
          material. All values in mg/kg DW.
NO.
3
4
6
MATERIAL
Sewage sludge
Incinerator aah
Concrete
HYDROIDS
Cadmium Lead
0.32
0.46
0.35
3.1
79
2.7
UNSTABILIZED
MATERIAL
Cadmium Lead
51
70

6410
2500

CONCLUSIONS

Within an exposure period of four months in the sea test blocks of concrete had a loss of weight around
1%. The mean loss of weight of fly ash, polluted soil, sewage sludge and unpolluted soil varied between
3.5 - 4.1%. The loss of weight in incinerator ash varied between 9-16% with a mean value of 12%.

During the period of exposure a fouling community, quantitatively dominated by hydrojds (Laomedea
longissima) developed on all materials, and there was no differences in species composition and species
richness. Compared to biomass on blocks of concrete, the biomass was the same on blocks of sewage
sludge and unpolluted soil and higher on blocks of incinerator ash and polluted soil (up to 12 weeks).

The concentrations of cadmium and lead is the same in hydroids grown on blocks of sewage sludge
and concrete. The level in hydroids is  comparable to blue mussels from unpolluted areas.

In hydroids grown on blocks of incinerator ash the concentration  of cadmium is almost the same in
hydroids grown on concrete, but the concentration of lead is about 30 times higher.

It is therefore recommended to improve the stabilization of incinerator ash. Renewed fouling tests and
physical and chemical laboratory tests are planned in 1990  on incinerator ash and  soil polluted by
trace elements.
                                           315

-------
                          THE  SITE PROGRAM:
                  HEAVY  METAL  FIXATION  IN  SOIL
                                Prepared by:
                            Philip N. Baldwin, Jr.
                        Chemfix  Environmental Services
                       2424 Edenborn Avenue,  Suite  620
                            Metairie, LA   70001
      CHEMFIX TECHNOLOGIES,  INC. (CTI) 1s a publicly owned company based in the
New Orleans, Louisiana area with historical  roots  going  back  to  CHEMFIX,  INC.
founded in Pennsylvania in  the  early  1970's.
      CTI specializes in the stabilization of hazardous, industrial and municipal
wastes.  CTI operates fixed facilities on the West,  Gulf and East Coasts as  well
as a couple to be operational in the near future  in  the  Midwest  region  of the
country.  CTI  also  has an  active licensee in the far East.   In addition  to
stabilization  activities,  CTI   owns  two  EPA  contract analytical  laboratory
companies:  One in  Louisiana  (E.I.R.A.,  Inc.)  and  one in  California  (BTC
Associates).
      CTI has  historically  based  its stabilization  business  on the  patented
CHEMFIX* Process  which involves the  use  of  calcium  based setting agents  and
liquid anionic  silicates. In the late 1980's, the basic  CHEMFIX® Process evolved,
when  treating  hazardous wastes,  into  the CHEMSET~/CHEMFIX*  Process.    This
modification  has  been  brought about  by  the requirements  of  the  landban
regulations   and   Best   Demonstrated  Available  Technology   (BOAT)   treatment
standards.  The CHEMSET~ Process modifications to the historical CHEMFIX* Process
were also brought about by the need to treat wastes at  a higher solids and higher
                                        316

-------
mobile analytes content than had been typical  in  the  1970's and early to middle
1980's. These CHEMSET™ modifications are many  and they add substantially to the
depth of stabilization capability established  by  the  CHEMFIX* patented Process.
      The addition  of CHEMSET1" molecular bonding enhancement  reagents  to the
CHEMFIX* Process allows  improved interactions with analyte anions  and cations as
well as  non-ionic  organic molecules. The results are decreased  molecular and
physical diffusivities  of the analytes  of  concern.   Fixation is  affected by
several  types  of chemical  bonds:  (1)  Ionic,  (2)  Co-valent and  (3)  Electro-
static.

                         MOLECULAR  BONDING ENHANCEMENT
Types of Bondings
      Ionic               Na:Cl     Sodium chloride (Table salt)
      Co-valent           Si::0     Silicon Dioxide (sand)
                          0*
      Electrostatic       H-0       Mater and their hydrates
The traditional CHEMFIX*  Process,  as  it has been applied  to  industrial  metal
contaminated wastes, was geared toward the reduction in the mobility of cation
analytes. The CHEMSET™ addition to the CHEMFIX* Process substantially broadens
the capability of the fixation technology.
      In 1988 when CTI and the EPA agreed to a joint participation in the 1989
SITE PROGRAM, CTI  was looking for a site that could  utilize some of the newest
expressions of stabilization  technology.   The plan  was to tackle  a  very high
level  of a heavy metal in a high solids matrix.  It  should be noted that since
                                          317

-------
late 1988, the concept of CHEMSET"1  has moved forward substantially and though
the chemistries used at Clakamus,  Oregon  in March 1989 were successful overall,
they would not  be  repeated today on similar wastes  without  using the updated
refinements.   Chemical  fixation  has become like the microchip business:  Good
today, better tomorrow.  Regardless of the more  advanced  stage of the technology
today in 1990 vs.  late 1988,  the  results  of the CTI/EPA  1989 SITE PROGRAM stand
on  their  own  as  impressive  indication  that  the  family  of  CHEMFIX  Process
technologies are viable when  applied to very high concentrations of heavy metal
in a high sol Ids waste matrix.
      The high  solids  processing equipment taken  to   the  Clakamus  site  has
demonstrated  an ability  not  only to  successfully treat high  solids,  but
demonstrated the capacity to treat an  average of 400 tons of contaminated soil
per one shift day.  Peak production rates exceeded 600 tons per day during a late
1989 commercial  job.
      During  the actual  SITE  program, four individual areas designated as A, C,
E, and F by the EPA, were to  be  treated.  Only a total  of 40  cubic  yards  were
treated in the field, which posed  an unexpected control  problem  at  such a low
volume output.   The  Pre-Production laboratory data for the key analytes is shown
below.
                                         318

-------
                        PRE-PRODUCTION TOTALS ANALYSIS
                                   RAW WASTE
                                    TOTALS
                                    mg/kg
                                    mean
                  AREAS:    AC          E        F

                  LEAD:     21,000   117,000   88,000   47,000

                  COPPER:   18,000     8,500  110,000   48,000
                         PRE-PRODUCTION TCLP ANALYSIS
AREAS:

LEAD:
COPPER:
                  RAW WASTE
                  mg/1
                  mean
A
650
78
C
590
23
E
680
4.5
F
220
110
      NATURFIL*
      mg/1
      mean


A     C      E      F

0.12  0.003  0.23  0.03

1.2   1.8    0.60  0.93
                                          319

-------
*»****»*»**************«**»»*******************************»************«*»>»*


                          PRODUCTION TOTALS ANALYSIS

                                   RAW WASTE
                  AREAS:
                                    TOTALS
                                    mg/kg
                                    mean
                  LEAD:    21,000   140,000   92.000   11,000

                  COPPER:  18,000     18,000  74,000   33,000

******************************************************************************
                           PRODUCTION TCLP ANALYSIS
AREAS:

LEAD:
   RAW WASTE
   mg/1
   mean

A    C    E    F

610  880  740  390
COPPER:   45   12   81  120
              NATURFIL*
              mg/1
              minimum/mean

A        C        E         F

.05/.05  .05/2.5  .05/47    .05/.10

.3S/.57   .43/.S4  .16/.65   .49/.60
% Deviation of Liquid                     -8.0%     -8.2%    -23.0%    -17.4%
****************************************************************
                                         320

-------
      This data Indicates that a very high metal  content in a high solids waste,
both  total  and mobile,  can be handled  by the  CHEMSET~/CHEMFIX*  technology.
MURPHY, however,  came  along with us on  our Clakamus project.   A  malfunction
occurred in our computer control  panel  and for a series of reasons  we had to
complete the brief demonstration  with only semi-automatic  control  conditions.
This situation led to a greater than a +/-  5% reagent variation and as a result
we were not able  to  completely reproduce  all of  the  Pre-Product ion  treatment
results.  In spite of  this  deficiency, the mean  TCLP  treated  waste values for
Area E still represented  a  94% reduction  in lead  mobility  and a 99% reduction
for Copper.  The results from Area C represented a 99% TCLP mobility reduction
for lead and 95%  for Copper.  Less contaminated Areas  A &  F  reached  well  over
99% in mobility reduction for both metals.
      What was  re-learned from the problems and successes CTI had at the Clakamus
site, is that  the chemistries  designed  into the  fixation  technologies by CTI
are not random but are waste specific.   Any particular chemistry design can
address a wide variety  of cations, anions,  mixtures of each  and more recently
non-ionic organic compounds, but  the  chemistries must  be monitored  and mixed
properly to get the consistent and  dramatic TCLP mobility reductions that we have
seen are possible. There is a great deal  yet to learn but the CHEMSETYCHEMFIX*
Processes are an  excellent  format for unlocking waste  stabilization  problems.
CTI also continues to  work  on improved  chemical  delivery  and  mixing systems.
      In closing  I would  like to  make a  few final overview comments  about our
SITE PROGRAM experience.
            The PC8 levels at the  SITE were too  low  to make any hard  decisions
            and  hence   I have  left  any   of  that  discussion  out  of  this
            presentation.
                                         321

-------
            The leachate tests applied by the EPA to the stabilized products were
            the TCLP, Multiple Extraction Procedure (MEP), ANS 16.1 and the Batch
            Extraction Test.  Through field dellstlngs for AMOCO, GENERAL MOTORS
            and US  STEEL,  CTI  has a  track record of  achieving excellent MEP
            results.   The  MEP  data produced  from the Clakamus site was  only
            performed on location C. Although  the chemical deliveries were only
            slightly off of prediction,  site  C  had  the  highest mean  lead
            concentration- nearly 14%  lead. These two factors together produced
            erratic test results.

            The  resultant  product  of  the CHEMFIX*  and/or CHEMFIX*/CHEMSET™
            processes (NATURFIL*) is a clay like  material,  not a monolith. The
            American Nuclear Society 16.1 extraction test was mistakenly applied
            to the NATURFIL* produced at the SITE.  This  test  is to be applied to
            monoliths,   not  to  clay like materials. Any  material  that  might
            physically sluff off of a  clay  like substance would be picked up as
            leached product thereby improperly slanting the results.  In  spite
            of this problem,  the NATURFIL*  still  produced     ANS   16.1   test
            results that EXCEEDED the  NCR criteria by  6  orders of  magnitude.

      CTI was  pleased to participate in the SITE PROGRAM, but  with the experience
of having participated, we have made several  suggestions  to the EPA that  might
help future programs:
      1.    There should be better co-ordination of the types  of and difficulties
            of wastes treated during a SITE PROGRAM YEAR. CTI treated soils with
            up to 14% lead  and  11% copper with combined TCLP mobilities  of over
                                         322

-------
            600  ppm while  most  other contractors  dealt  with mobile  metal
            concentration less than 50 with totals less than 1.0%. We chose this
            challenge  but  in  so  doing  made  any  real  comparisons  between
            technologies more difficult for the public.
      2.    Closer oversight is needed by the EPA R&D department regarding the
            type of tests that are applied to the treated wastes.
      3.    More interface  is  needed  between the EPA and  the  treatment  firms
            regarding the amount of waste that should be treated as that relates
            to the equipment available from the treatment firm.
      4.    Treatment is both engineering capability,  equipment and chemistry.
            The EPA only evaluated chemistry AFTER the fact. This  shortchanges
            the concept of waste treatment.

      CTI believes that the  days of stabilization as an art are passing quickly
and that the days of stabilization as a science are upon us.  CTI is pleased to
be apart of that science!
                                          323

-------
 Evaluation of Three
 Leading Technologies


 For Second Forum on Innovative Hazardous
 Waste Treatment Technologies


 By
 Armand A. Balasco
   Case Study on Innovative Technology
 Development for Rocky Mountain Arsenal
                   Contracting Agency: U.S. Army
                   Toxic and Hazardous Materials
                   Agency (USATHAMA)

                   Performance Period: July 1086-1988
ArthirD Little

-------
          DOD Hazardous Waste Sites on
          Super-fund National Priority List
ArtlurD Little
 Rocky Mountain Arsenal
 (RMA) profile

 • Located northeast of Denver, Colorado
 • Occupies more than 17,000 acres
 • Began operations in 1942
ArthirD Little
                     325

-------
     Rocky Mountain Arsenal
     (RMA) profile
     • On-site operations (1942-1982):
       - Chemical/incendiary manufacture
       - Chemical munitions demilitarization
       - Insecticide/pesticide manufacture
    ArthirD Little
General Map
   of RMA
North Boundvy
           IrondaJe
       Ran CtaMifkaUon
          and
       Warehouse Are*-
            StopMonM'l
             AJrport
    ArthirD Little
                 N
                 I
                           326

-------
 RMA Basin F profile


 • Asphalt-lined evaporation basin
 • Period in use: 1956-1982
 • Average depth: 10 feet
 • Surface area: 90 acres
 • Capacity: 243 million gallons
ArtfurD Little
  RMA Basin F profile
  • Current amount of hazardous material:
   - Soils (> 400,000 cubic yards)
   - Liquids (4 million gallons)
  • Major contaminants:
   - Organics
   - Inorganics
   - Metals
AtthirD Little

                       327

-------
 RMA Basin F profile
   Basin F material
   contaminant
   Aldrin
   Isodrin
   Endrin
   Sulfoxide
   Sulfone
   Mercury
   Arsenic
Contaminant
cone, (ppm)
  100-1,100
  50-1,300
    65-180
   130-300
  300-700

       <1
       20
ArthirD Little
 Situation at RMA


 • Industrial operations generated variety
   of wastewaters
 • Wastewaters stored in on-site lagoons
 • Lagoons pose threat to environment
 • Environmentally acceptable and reasonable-cost
   clean-up needed
ArthirD Little
                      328

-------
   RMA case study objectives

   • Review treatment technology databases
   • Evaluate/rank most applicable technologies
   • Bench-scale test most promising technologies
   • Develop full-scale designs/cost estimates
   • Select technology for on-site (RMA) pilot-
     scale testing
  ArthirD Little
 Technologies to  be evaluated/ranked

 • Encapsulation
 • Fluidized/circulating  bed combustion
 • Classification
 • In-situ glassification
 • Soil washing
 • Solvent extraction
 • Wet-air oxidation
ArthirD Little

                        329

-------
        Evaluation criteria employed
       (Level of importance)
       Highest
       Safety
       Treatment of organ ics
       Throughput rate
       Vendor test facilities
       - System safety
       - Permitting
       - Availability
Average
Treatment of metals
Treatment of inorganics
Capital cost
Operating cost
Reliability
Proprietary status
Vendor test facilities
- Emissions monitoring
Lowest
Associated equipment
Equipment complexity
Maintenance
Ease of operation
Barriers
      ArthirD Little
       Relative evaluation of leading
       technologies for  Basin  F
Technology
Rotary kiln
 (base case)
Encapsulation

Fluidized/circutating
 bed combustion

Glassification

In-situ vitrification

Soil washing
Engineering
design/ Capital Operating System
status cost cost complexity
























Perceived effectiveness
Organics Inorganics












      ArthirD Little
                                     330

-------
 Technologies selected for laboratory/
 bench-scale testing

 • Fluidized/circulating bed combustion
 • Glassification
 • Soil washing
ArthirD Little
 Fluidized/circulating bed combustion

 • Vendor test facility permitting problems
 • USATHAMA schedule constraints
 • Elimination of technology from
   further consideration
ArthirD Little

                      331

-------
                 Classification
        ; Basin F
         Material
    • Glass >•
    Product
                                      .Scrubber
                                       Effluent
     ArthirD Little
                       Cleaned
GLASSFICATION BENCH-SCALE
TEST SYSTEM
        OH-Gas
       Samping Port   Pitot
                 Tube
        Jouto-Heated
      Ceramic-Uned MeRer
                                                  Stack
                                               Blower
                        Scrubber
                Sampbtg Port  Samping Port
^Continuous
   Analyzers
      ArthirD Little
                            332

-------
 Glassification equipment/
 test conditions

 • Glass melter volume: 8.2 liters
 • Glass melter operating temperature: 2,200°F
 • Glass melter operating pressure: -3 to -5 in.
ArthirD Little
  Classification equipment/
  test conditions
   Glass melter off-gas treatment system:
   - Venturi scrubber
   - Heat exchanger
   - Demister
   - HEPA filter
   - Activated carbon bed
 ArthirD Little


                        333

-------
 Classification equipment/
 test conditions
 • Glass melter feed system:
   - Submerged drop tube (Basin F material)
   - Above glass melt surface (glass formers)
 • Basin F/glass formers
   - Basin F material
   - Sodium carbonate
   - Sodium borate
ArthirD Little
 Classification equipment/
 test conditions


 • Basin F feed rate: 3.3 Ibs/hr
 • Glass formers feed rate: 0.5 Ibs every
   15 minutes
 • Glass draw-off rate: 9.9 Ibs every 2 hrs
ArthirD Little

                       334

-------
 Classification equipment/
 test conditions
 • Gas sampling performed:
   - Modified Method 5 (6.2 hrs)
   - Method 101A (2.2 hrs)
 • Continuous gas analyzer:
   - CO    - SO2
   - C02    - NOx
   -O2
ArthirD Little
 Classification test results

 Advantages
 • Organochlorine compounds: > 99.99% reduction
 • Organophosphorus compounds: > 99.76%
   reduction
 • Organosulfur compounds: > 99.60% reduction
ArthirD Little

                       335

-------
 Classification test results

 Advantages
 (continued)
 • TCLP leachate from glass product
   - No target contaminants detected
   - Endrin, arsenic, and mercury levels below
    regulatory limits
JirthirD Little
 Classification test results

 Disadvantages
 • Volatile metals (e.g., mercury, arsenic)
   in off-gas
 • Acid gases (e.g., HCI, H2SO4, H3PO4)
   in off-gas
 • High dust carryout in off-gas
ArthirD Little


                        336

-------
                 Air
                Emissions
                               Soil Washing
      ArthirD Little
SOIL WASHING LABORATORY-SCALE

TEST SYSTEM
                                 Add
              Add
               and
             FbccUant Rocotant
      •I Organic
      Decant
6th Organic  Froth
 Decant   Slurry
Decant  1st Add 2nd Add
     Wash   Wash
3rd Add Washed
 Wash  Material
       ArthirD Little
                           337

-------
  Soil washing equipment/
  test conditions
    Organic washing (five stages):
    - 700 grams of basin F material
    - Toluene/kerosene/octanol (TKO) addition
    - Heating (120°F)
    - Mixing
    - Organic layer decanting
    - Slurry removal  to flotation stage
 flrthirD Little
  Soil washing equipment/
  test conditions
   Froth flotation (1,000 gram cell):
   - Slurry from organic washing
   - Reagents (caustic, sodium silicate
    and surfactant) addition
   - Heating (120°F)
   - Aeration
   - Mixing
A rthirD Little


                        338

-------
 Soil washing equipment/
 test conditions
 • Froth flotation (continued)
   - Froth removal
   - Flocculant addition
   - Clean slurry setting
   - Water decanting
   - Settled solids removed to acid washing
ArthirD Little
  Soil washing equipment/
  test conditions
   Acid washing (three stages):
   - Settled solids from froth flotation
   - Acid (HCI) addition
   - Clean solids settling
   - Water decanting
   - Clean solids filtering
   - Filtrate removal
   - Clean solids removal for analysis
 ArthirD Little

                        339

-------
  Soil washing test results


  Advantages

  • Organochlorine compounds: >98.9% reduction
  • Organosulfur compounds: > 78% reduction
  • Metals: < 55% reduction
ArtlurD Little
 Soil washing test results


 Advantages
 (continued)
 • TCLP leachate from washed material
   - Some target contaminants detected
   - Endrin, arsenic, and mercury levels below
    regulatory limits
 • Acid washing may not be required
ArtlurD Little


                       340

-------
  Soil washing test results

  Disadvantages
  • Organic extractants required
  • Minimum three stages organic washing required
  • Necessity for stream recycle
ArthirD Little
 Soil washing test results

 Disadvantages
 (continued)
 • Non-pesticide organic contaminants do not
   follow pesticides in spent organic extractant
   - Carbon adsorption required
 • Technology requires many operational steps
ArthirD Little

                        341

-------
 Results - RMA


 • Fluidized bed combustion, glassification, and
   soil washing selected for testing
 • Fluidized bed combustion not tested; permits
   could not be obtained
 • Glassification and soil washing both
   technically attractive
 • Glassification selected for pilot testing
   at RMA
ArthirD Little
                          342

-------
                   Soil Vapor Extraction and Treatment of VOCs
                             At a Superfund Site in Michigan
                           Joseph P. Danko, Michael J. McCann, and William O. Byers
                                                CH2M HILL
                                              Corvallis, Oregon
Abstract
     The Verona Well Field supplies potable water to the City
   of Battle Creek, Michigan, three townships, and another
   small city. The combined service area equates  to
   approximately 50,000 people. In 1981, the well field and
   surrounding area were found to be contaminated with
   chlorinated solvents and other volatile organic compounds
   (VOCs). The contaminant plume extended throughout an
   area of approximately 1 mile by one-half mile. Two facilities
   operated by a local solvent wholesaler/distributor were
   identified as the primary sources of the contamination.
   VOC concentrations as  high as 1,000 parts per million
   (ppm) were found in the groundwater and soil at the
   wholesaler's primary facility.
     EPA selected enhanced volatilization of VOCs using soil
   vapor extraction (SVE) to clean up the contaminated soils
   at the primary contaminant source area. A pilot SVE system
   was installed in late 1987 and operated for about 69 hours,
   removing approximately 3,000 pounds of VOCs from the
   vadose zone. Using the results of the pilot phase, a full-
   scale SVE system was designed and installed. It began
   operation in March 1988 and has since removed more than
   40,000pounds of VOCs.
     From system startup until January 1990, extracted vapors
   containing VOCs were passed through vapor-phase activated
   carbon for treatment prior to discharge. In January 1990,
   the carbon system was replaced by a catalytic oxidation
   system capable ofonsite destruction of VOCs. The catalytic
   oxidation system is expected to provide more continuous
   operation and a cost savings over the life of the project.
Introduction
   Soil vapor extraction (SVE) is rapidly becoming a
common and preferred treatment method for removing
volatile organic compounds (VOCs) from contaminated
soil. This paper presents the case history of an operating
SVE system at the primary contamination source area
in the Verona Well Field Superfund site in Battle Creek,
Michigan. The paper includes a site description, a
description of the SVE system and the associated offgas
treatment system, and a summary of SVE and offgas
treatment performance to date.

Background
   The Verona Well Field (Figure 1) provides potable
water to approximately 50,000 residents in and around
Battle Creek, Michigan. In  August  1981,  VOC
contamination was detected in approximately one-
third of the city's production wells and in numerous
private wells outside the well field.
   Initial investigations of the site by the U.S. EPA
Technical Assistance Team identified three primary
potential source areas and a contaminant plume with
VOC concentrations of up to 346 micrograms per liter
(Hg/l) in the well field. These three areas are also
identified in Figure 1.
                                                                            FIGURE 1
                                                                          VICINITY MAP
   A solvent distribution facility run by the Thomas
Solvent Company was found  to be  the most
contaminated of the source areas. This facility, referred
to as Thomas Solvent Raymond Road (TSRR), was used
for the storage, transfer, and packaging of chlorinated
and nonchlorinated solvents from the mid-1960s to
1984. Contamination of the soil and groundwater at
TSRR is thought to have resulted from underground
tank leakage and surface spills in the tank truck loading/
unloading area and the warehouse (now demolished).
Leak tests showed nine of 21 tanks on the site to be
leaking (Figure 2). Vadose zone contamination covered
more than an acre.


Remedial Measures
   In May 1984,  EPA signed a Record of Decision
(ROD) to implement an Initial Remedial Measure (IRM)
at the Verona Well Field site in order to protect the
                                                                                                  Ob-
                                                      343

-------
city's water supply. The IRM included converting a
number of the contaminated production wells to
blocking wells to prevent further  migration of the
contaminant plume. An  air stripping system  was
installed to  treat the extracted groundwater. Three
additional production wells were also installed as part
of the  IRM to supplement the city's existing water
supply system.
     A OROUNONOEREXnvieiKINWaL
                   FIGURE 2
                 LOCATION OF
             UNDERGROUND TANKS
   In the fall of 1985, EPA signed a second ROD that
 addressed the TSRR facility. The ROD  specified a
 corrective  action that included a network of
 groundwater extraction wells to remove contaminated
 groundwater, the treatment of extracted groundwater
 by air stripping,  and an SVE system to remove VOCs
 from the unsaturated zone.


 Groundwater Extraction
   The installation of nine groundwater extraction
 wells at the TSRR facility began in the fall of 1986. The
 groundwater extraction system  started up in March
 1987. Since that time, more than 11,000 pounds of
 VOCs have been  extracted with the groundwater.
 Groundwater concentrations initially were as high as
 19,000 Jig/l total VOCs; as of October 1989, the
 concentrations had dropped to about 1,500 ng/l.


 Soil Vapor Extraction
   Contract documents for the construction, operation,
 and maintenance of an SVE system at the TSRR facility
 were issued for bid in June 1987. The contract called
 for the following performance objectives:
      •   None of the soil samples analyzed could
         have VOC concentrations above 10
         mg/kg of dry soil.
      •   No more than 15 percent of the soil
         samples analyzed could have VOC
         concentrations above 1 mg/kg of dry soil.

  The contract also included the removal of the 21
underground storage tanks. The underground storage
tanks were surrounded by heavily contaminated soil,
and direct excavation and removal would significantly
violate the State of Michigan's air quality criteria. This
problem has been avoided by using the SVE system to
remove the majority of VOCs  before removing the
underground tanks.
  The contract was awarded to Terra Vac, Inc. of San
Juan, Puerto Rico in October 1987.


Site Characteristics
  The lithology at the site consists of fine- to coarse-
grained  sand with trace day, silt, and pebbles. The
water table is approximately 25 feet belowgrade and
the hydraulic gradient is to the northwest.
   In addition to the vadose zone contamination, there
was also a floating nonaqueous phase liquid (NAPL)
layer in the vidnity of Extraction Well 8, whose location
is shown in  Figure  2.  This well was installed as  a
product recovery well,  combining  groundwater
extraction with  intermittent removal  of the  NAPL
product as it accumulated.


Soil Vapor Extraction (SVE) System
   The SVE system includes vapor extraction wells, an
air/water separator, offgas treatment equipment, and a
blower. Figure 3 presents a simplified schematic of the
system. The system  was installed following  a
preconstruction investigation designed to help place
SVE wells and to  determine  the pre-operation
magnitude of contamination.
   The extraction network  consists of 23 2- and 4-
inch-diameter PVC  wells with slotted screens from
approximately 5 feet belowgrade to 3 feet below the
water table.  The wells are packed with silica  sand,
sealed at the screen/casing interface with bentonite,
and grouted to the  existing grade to  prevent short
circuiting. The wells  are  connected by  a surface
collection manifold. Each well head has a throttling
valve, sample port,  and vacuum pressure gauge. The
 surface manifold is  connected to a centrifugal air/
water separator which is in turn connected to a blower
 and offgas treatment system.


 Offgas Treatment Systems
   The  original  SVE system at the TSRR site  used a
 carbon adsorption  system  to  remove contaminants
 from the vapor stream. That system was replaced in
January 1990 with a catalytic oxidation unit for offgas
 treatment (both systems are described below). Following
                                                       344

-------
                     TYnCM.MI.WOfl
                     emuenoHWEu.
                                               FIGURES
                                       SCHEMATIC OF SOIL VAPOR
                                         EXTRACTION SYSTEM
offgas treatment, the air is discharged to the atmosphere
through a 30-foot stack.

Carbon Adsorption System
  The carbon adsorption system consisted of two sets
of four stainless steel carbon vessels connected in series.
A schematic of the system is presented in Figure 4.
  The first (primary) set of vessels was used for the
majority of VOC adsorption, while  the second set
acted as backup  in the event  of contaminant
breakthrough in the primary set. The system was
installed under negative pressure to prevent the leaking
of VOCs. Each vessel held 1,000 pounds of  vapor-
               phase carbon. The vessels were connected to header
               piping  with flexible  hoses and quick disconnect
               couplings. Sample ports, vacuum pressure gauges, and
               temperature probes were installed upstream, between,
               and downstream of the two sets of vessels. A carbon
               monoxide meter was also installed between  the sets
               of carbon vessels for early detection of combustion in
               the primary carbon units.
                 During system operation, an in-line organic vapor
               analyzer (HNu) between the primary and  secondary
               carbon units was  used to monitor contaminant
               breakthrough. When  breakthrough occurred, the
               primary  carbon  unit was changed out and the
               secondary carbon was placed In the primary position.
                                                      VACUUM EXTRACTION UNIT
                        PRESSURE INDICATOR

                        TEMPERATURE INDICATOR
 (S)  SAMPLING PORT

 Q FLOWMETER

       FIGURE 4
SCHEMATIC OF CARBON
 ADSORPTION SYSTEM
                                                    345

-------
This minimized the chances of breakthrough of the
secondary system.


Catalytic Oxidation System
   The catalytic oxidation system, designed to destroy
contaminants onsite, is essentially a thermal reactor
used for the  destruction of  chlorinated  and
nonchlorinated VOCs. A schematic of the system is
presented in Figure 5. The SVE air stream flows from
the  positive  displacement blower through a  heat
exchanger, where  it is preheated to  approximately
430°F. The preheated gas then enters the burner, where
it is heated to approximately 800°F by a natural gas
burner. The air stream is then passed through a catalyst
bed, which causes VOC oxidation. The catalyst, without
being altered, accelerates the oxidation reaction by
adsorbing the reactants  (oxygen and  VOC) on
catalytically active sites1. This greatly  accelerates the
reaction rate converting chlorinated VOCs to carbon
dioxide, water vapor, and hydrochloric acid gas. The
catalyst also allows the oxidation reaction to occur at
much lower temperatures than would conventional
thermal incineration.
   The gas stream  exits the reactor at  approximately
820°F and enters the shell side of the heat exchanger.
In the heat exchanger, the heat produced from the
natural gas burner and from combustion of VOCs in
the catalyst bed is used to heat the inlet gas and reduce
the  fuel requirements  of the preheat  burner. In the
heat exchanger, exhaust gas is cooled  to about 550°F
before being vented to the atmosphere through a 12-
inch-diameter stack.
   Because the catalytic oxidation system uses natural
gas as fuel and is designed for remote operation, the
following automatic shut-down controls have been
included:
         •  Cutoffs for low and high temperatures
           (750°F and 1,300°F)
         •  Ultraviolet flame sensor cutoff (if flame
           goes out)
         •  Blower default cutoff (if the SVE blower
           fails)
     The catalytic oxidation system  was selected  to
  replace the vapor-phase carbon system at the TSRR site
  because:
         •  The catalytic oxidation system does not
           require the intermittent downtimes needed
           by the carbon adsorption system. Site
           cleanup could therefore proceed faster.
         •  VOCs are destroyed onsite with the
           catalytic oxidation system  rather than
           transferred onto carbon and hauled several
           hundred miles for thermal destruction. The
           National Contingency Plan favors onsite
           destruction of wastes.
         •  Use of the catalytic oxidation system is
           expected to result in  an overall cost savings
           for the project based on the estimated
           remaining mass of contaminants.

  Performance of the SVE and
  Offgas Treatment Systems


  Pilot Phase SVE
     Pilot-phase operation of the SVE system was initiated
  in November 1987. Figure 6 shows the location of the
  SVE wells and the piping layout at the TSRR site. Prior
  to complete system operation, individual wells were
  operated to determine their radius of influence, flow
  rate, and initial extraction rate. All gas stream analyses
  were done using an  onsite gas chromatograph.
                      PARTKULATE
                        ALTER
                                                 EXHAUST TO
                                                 ATMOSPHERE
HE AT EXCHANGER ^- BURNER
                                                                                  CATALYST BED
       EXHAUST FROM SOIL
       VAPOR EXTRACTION
       SYSTEM
                                                 FIGURES
                                   SCHEMATIC OF CATALYTIC OXIDATION
                                       SYSTEM WITH HEAT RECOVERY
 •Hardlson, LC. and Ej. Dowd. August 1977. Emission Control Via Fluidized Bed Oxidation. Central Engineeriny Progress.
 Radian Corp. July 1984. Control of Industrial VOC Emissions by Catalytic Incineration. Prepared for the Industrial Environmental Research Lab.
                                                         346

-------
  Total VOC concentrations ranged from a low of 2
mg/1 at VE-6 to 204 mg/1 at VE-2. The radius of influence
of individual wells was measured by recording the
vacuum in nearby vapor  extraction wells and in
vacuum piezometers. A 1-1/4-inch water vacuum was
recorded  60 feet from an  extraction well. Vacuum
piezometers located between tanks to investigate tank
shielding showed at least a 2-inch water vacuum.
  Pilot-phase operation continued intermittently over
IS days with a total operation time  of 69 hours.
Approximately 3,000 pounds of contaminants were
removed during the pilot test, as  determined by gas
stream and carbon loading analyses. The average VOC
stack cc  centration was 0.0666 mg/1. At an average
flow rate of 500  cubic feet  per  minute (cfm),
approximately 4.6  pounds of VOCs were released
during the pilot phase,  indicating  an approximate
removal efficiency of 99.8 percent.
   A
   EW2
   A
  BM
                                          eo
             aoLwpon EXTRACTION (SKI w^i.

           A QROUNDWATER EXTRACTION WCU.
                    FIGURE 6
        SOIL VAPOR EXTRACTION SYSTEM AT
      THOMAS SOLVENT RAYMOND ROAD SITE
Full-Scale SVE
   The full-scale SVE system began operation in March
1988 with 23 vapor extraction wells. Operation has
generally been limited to under 14 wells at any one
time, with well head concentrations determining the
wells selected for operation. This is done to maximize
contaminant loading to the offgas system. The typical
flow rate from individual wells is 100 scfm, with the
combined  flow  between  1,400 and  1,600 scfm.
Wellhead pressure in the extraction wells is generally 2
to 3 inches of mercury.
  As of May 1,1990, approximately 43,000 pounds of
VOCs have been removed by the system in 252 days of
operation. The total VOC loading rate has dropped
from an initial high of approximately 1,080 pounds of
VOCs per day to under 100 pounds per day. The NAPL
layer originally observed under the site has not been
present since October 1988.


Offgas Treatment Systems

Carbon Adsorption
  Between November 1987 and January 1990, VOC
contaminants were removed from the vapor stream
using activated carbon adsorption. During that time,
approximately 266,400 pounds of carbon were loaded
and shipped offsite for regeneration. The average carbon
loading rate during this period was 16 percent.


Catalytic Oxidation
  Since January 1990,  VOCs have been destroyed
onsite  using  the catalytic oxidation system. The
catalytic  oxidation system was performance tested
before being put into operation. Process operating
data from the performance tests are shown in Table 1.
The  primary difference between the  four runs was
reactor operating temperature, which was varied to
determine  the effect of operating  temperature on
destruction removal efficiencies.
  Table 2 presents the results from discharge stack air
sampling  during  the performance  tests. All
contaminants  except tetrachloroethene  were
significantly less than the  allowable state discharge
limits  at each  of the operating temperatures.
Tetrachloroethene at 780°F had an average discharge
concentration of 204 ppb, approximately 60 percent
of the discharge limit. At 860°F, the stack concentration
of tetrachloroethene decreased to 35 ppb, 10 percent
of the  allowable  limit. The highest detected
concentration of hydrochloric acid, a byproduct of the
oxidation, was 21 ppm, well below the discharge limit
of 132 ppm.
  Since the performance test,  the catalytic oxidation
system has operated at  800°F without any operating
difficulties.
Acknowledgements
   The soil vapor extraction system is part of a U.S.
EPA Superfund Operable Unit Remedial Action under
Contract No. 68-01-7251. This paper has not been
subjected  to the Agency's peer and administrative
review. It  therefore does not  necessarily reflect the
views of the Agency, and no official  endorsement
should be inferred. Similarly, any use of specific names
in the paper should not be viewed as an endorsement.
   The authors would like to thank Margaret Guerriero
of the U.S. EPA and Terra Vac, Inc., the SVE contractor,
for their cooperation in making this paper possible.
                                                     347

-------
                                             Table 1
                                  Catalytic Oxidation System
                             Performance Testing Operation Data
Parameter
Temperatures
Heat Exchanger (in)
Heat Exchanger (out)
Catalyst (in)
Catalyst (out)
Stack
SVE Process Gas
Pressure drop (flow rate)
Temperature
Pressure
Flow rate
Natural Gas
Flow rate
Mass rate
Miscellaneous Data
Stack velocity
Stack flow rate
NG valve position
DA valve position
Units

oF
°F
op
oF
oF

in. Hp
op
in.Hg
scfm

scfm
pph

fpm
scfm
%
%
Run A
(1/9/90)

98
430
780
792
468

3.4
41
4.2
1,579

13.4
37.0

3,740
1,636
43
13
RunB
(1/10/90)

93
446
820
838
482

3.4
38
4.4
1,577

_
—

3,290
1,626
46.7
13
RunC
(1/10/90)

95
470
860
875
510

3.4
37
4.3
1,582

— —
—

3,321
1,418
49.5
13
RunD
(1/30/90)

91
431
800
822
—

3.4
39
4.5
1,572

13.7
37.8

—
—
—
—
     Note: — means not measured.
                                              Table 2
                                 Discharge Stack Air Sampling
                        Catalytic Oxidation System Performance Tests







Component
Chloromethane1
Tetrachloroethene1
Tetrachloroethene
Trichloroethene
Carbon Tetrachloride
1,1,1-trlchloroethane
Chloroform
Methylene Chloride
Benzene
Vinyl Chloride
Acetone2
2-Butanone2
Toluene
Hydrogen Chloride2

IlniK
Ulllld
ppbv
ppbv
ppbv
ppbv
pptv
ppbv
pptv
ppbv
ppbv
ppbv
ppbv
ppbv
ppbv
ppmv
MDNR
Limits
• -•••HT"*j
_4
347
347
1,337
249,000
_«
161,000
11,464
1,752
6,220
_4
_4
_4
132
Run A
(1/9/90)

Reactor
Temp.-780°F

A_1
~i
41
202
178
<0.04
352
0.08
249
58.5
2.4
<0.04
NA
NA
1.3
--

A .2
mi
36
221
193
1.7
35
0.04
60
17.6
<0.1
<0.04
NA
NA
0.46
11.8

A-3

31
188
180
1.6
59
0.03
198
18.0
<0.1
<0.04
S5.7
5.1
1.5
~"
RunB
(1/10/90)

Reactor
Temp.«820°F

R 1
0*1.
29
87
83
0.9
2
<0.02
8J
6.1
<0.1
<0.04
NA
NA
<0.1
"*

R-2
D~m*
30
85
83
0.8
<2
<0.02
<20
5.0
<0.1
<0.04
NA
NA
<0.1
20.8

R3
mjtj
28
88
91
0.9
22
<0.02
46
10.1
<0.1
<0.04
31.6
2.5
<0.1
™"
RunC
(1/10/9Q)
Reactor
Temp.
-860°F

C-l
^* M.
22
35
34
0.3
<2

-------
          SLUDGE AND SOIL
    TREATABILITY STUDIES AT A
LARGE, COMPLEX SUPERFUND SITE
         BOFORS-NOBEL SUPERFUND SITE
             MUSKEGON, MICHIGAN


                    BY:
             Susan Roberts Shultz*
                Wendy Oresik
                Doug Graham
                 Larry Kiener
                 Sarah Levin
                John Trynoski
                 David Shultz
          DONOHUE & ASSOCIATES, INC.


                PRESENTED AT
        EPA SECOND FORUM ON INNOVATIVE
              HAZARDOUS WASTE
TREATMENT TECHNOLOGIES: DOMESTIC AND INTERNATIONAL
                     349

-------
                              1.0 INTRODUCTION

1.1  INTRODUCTION

1.1.1  Site Description

The Bofors-Nobel (Bofors) Superfund Site is located six miles east of downtown
Muskegon in  Egelston  Township,  Muskegon  County,  Michigan.    This site is
currently under the management of the Michigan Department of Natural Resources
(MDNR),  Grand Rapids  District,  in  cooperation  with  the U.S. Environmental
Protection Agency, Region V (EPA).  This 85-acre parcel of land is irregularly
shaped and includes  an  operating  specialty  chemical production facility, a
biological/carbon (PACT™) treatment  plant,  an  unused landfill, 10 abandoned
sludge lagoons used for disposal  of  wastes from the production facility, and
associated soils.  The southern portion  of  the  site is bounded by Big Black
Creek.   Figure 1  presents  a  map  showing  these  and  other important site
features.

1.1.2  Site History

Many investigations and studies of  this  site  have been performed during the
last 20 years.  A description of the site history is briefly summarized below.

Lakeway Chemicals began producing  industrial  chemicals  at the site in 1960.
Throughout the 1960s and  early  1970s,  the  State of Michigan placed various
restrictions on wastewater disposal from the site because of surface water and
potential groundwater contamination  from  wastewater  discharge  to Big Black
Creek.  In 1976, wastewater from the plant was accepted at the Muskegon County
Wastewater Treatment Plant, and  purge  wells  were  installed  at the site to
extract  contaminated  groundwater.    Bofors  Industries,  Inc.,  merged with
Lakeway in 1977 and  with  Nobel  Industries  in  1981.  In 1985, Bofors-Nobel
(Bofors) filed for bankruptcy  for  a  variety  of reasons, including reported
environmental expenditures in excess  of  $60 million.    As a result of legal
action in bankruptcy court, Bofors was  allowed to sell the operating chemical
plant to Lomac, Inc. (Lomac).   As  part  of the sale, agreements were reached
between Lomac, MDNR, and EPA  that  Lomac  could  not be liable for cleanup of
contamination existing prior to the  sale  of  the plant area property.  These
agreements allowed Lomac to operate  the  plant independently of previous site
activities.  The site  was  then  nominated  for  the National Priorities List
(NPL) and was placed on the NPL in March, 1989.

1.1.3  Site Operations

The  development  and  operation  of  the  industrial  chemical  facility  and
associated disposal activities at the  Bofors Site was investigated during the
RI by examining aerial  photographs  of  the  area  dating  from 1955 to 1987,
interviewing past workers  and  owners  of  the  facility,  and reviewing past
investigations and studies of the facility and adjoining property.
                                          350

-------
SOILS AROUND
LAGOONS
BOFORS
SITE BOUNDARY
                                 NUMBERS:SOURCE AREAS
                     Figure I

                  Site  Map
                            351

-------
Before 1970, the plant  produced  alcohol-base  detergents, saccharin, and dye
intermediates such as 3,3'-dichlorobenzidine (DCB),  benzidine, and azobenzene.
Raw materials included fatty  alcohols,  fatty  ether alcohol, sulfur dioxide,
aqua ammonia,  muriatic  acid,  sulfuric  acid,  nitrobenzene, o-nitrobenzene,
methanol, benzene, sodium chloride, and zinc.   During the 1970s, Lakeway pro-
duced a lauryl  alcohol  base  detergent,  dye  intermediates, pesticides, and
herbicides.  The lagoons were  used  for  plant wastewater and sludge disposal
until 1976.

Lagoon 1, the iron pond, was  used  for  the  disposal  of iron sludge or iron
scale from the production of DCB.    Lagoon 2,  the sodium formate pond, was a
synthetically lined lagoon and was abandoned after attempts to line the lagoon
were unsuccessful.  Lagoon 3 was used  as  a settling pond for DCB waste, zinc
oxide waste, and waste  from  cleanup  operations  and spillage.  Lagoon 4 was
used for the disposal  of  non-contact  cooling  water.    Lagoons 5, 6, 7, 8,
and 10 received calcium sulfate  sludge  pumped  to  the  lagoons as a slurry.
Lagoon 9 was used for discharge of calcium sulfate liquid and detergent waste.
As indicated by  plant  personnel,  Lagoon 9  berm  failures discharged sludge
southeast to the Big Black Creek floodplain in late 1974 and early 1975.  Berm
failure from Lagoon 10 also occurred into  the  floodplain in early 1975.  The
sludge from the berm failures was  removed from the floodplain and disposed of
in the southern portion of Lagoon 9.

1.1.4  Treatabilitv Studies

Identification and  screening  of  remedial  technologies  and process options
during the Feasibility Study for the  Bofors Site resulted in three innovative
process options determined  to  be  potentially  effective, implementable, and
cost-effective for the remediation  of  the  contaminated sludges and soils at
the Bofors Site:  soil washing, low temperature thermal desorption (LTTD), and
solidification/stabilization.  Because of  the  innovative nature of the tech-
nologies, particularly as applied to  the  unique, heterogeneous nature of the
wastes at  the  site,  treatability  studies  were  performed  to evaluate the
effectiveness of these process options and to obtain additional data necessary
to evaluate remedial alternatives  during  the  detailed analysis phase of the
Feasibility Study (FS).  The data  generated from the treatability studies was
used in the FS to identify materials handling issues and operating parameters,
reduce  cost  estimate  and  performance  uncertainties,  and  evaluate  post-
treatment requirements.   Treatability  and  analytical  testing was performed
during the bench scale treatability studies by vendors and laboratories desig-
nated by  Donohue &  Associates,  Inc.,  (as  part  of  a  joint  venture with
Goldberg, Zoino, & Associates) in accordance  with the Treatability Study Work
Plan  (May,  1989),  and  QAPP/SAP  Addendum  (GZA/Donohue,  May 1989)  to the
approved Work Plan for the  Bofors  Site (GZA/Donohue March 1988).  The treat-
ability studies were completed  between  May 1989  and  March 1990.  Delays in
receiving analytical laboratory data  and  vender  reports caused the extended
time frame.  Bench scale  treatability  studies were performed to evaluate the
general ability of the technologies to  treat  the  wastes at the site.  Pilot
scale treatability studies are beyond the scope of this RI/FS.
                                          352

-------
The lagoon sludges  and  underlying  soils  were  initially characterized from
samples collected in 1988 during Phase I RI activities.  Samples were analyzed
for volatile organic compounds  (VOCs),  metals, cyanide, semivolatile organic
compounds, polychlorinated  biphenyls  (PCBs),  pesticides,  and special semi-
volatile organic compound analytes.  No PCBs or pesticides were detected above
the analytical detection limit.    Compounds  of  concern for the contaminated
soil and sludge of the L.O.U. were  selected based on the results of the base-
line risk assessment and are listed below:

        ' Aniline
        0 Azobenzene
        0 Benzene
        * Benzidine
        0 3,3'-Dichlorobenzidine (DCB)

Samples from Lagoons 3, 8, 9, and 10  were selected for the bench scale treat-
ability studies.   Phase I  RI  data  indicate  Lagoon 3  contains the highest
detected concentrations of organic compounds  in the soil beneath the lagoons,
and Lagoon 9 contains the highest  concentrations  of organic compounds in the
sludge, possibly representing  the  worst  case  soil  and  sludge.  Lagoons 8
and 10 comprise the largest  volume  of  contaminated  sludge  and soil in the
L.O.U. and therefore may represent  average concentrations of contaminants for
both soil and sludge.  Approximately  62 percent of the lagoon sludge are con-
tained in Lagoons 8 and 10.

Sludge and soil samples were  collected  for chemical and geotechnical charac-
terization.  Samples  were  provided  to  vendors  performing the treatability
studies.  Four  samples  were  collected  from  the  four lagoons to represent
various conditions, as follows:

    1.   Lagoon 3  composite   soil   sample   to   represent  the  worst-case
         contaminated soil.

    2.   Lagoon 9  composite  sludge   sample   to  represent  the  worst-case
         contaminated sludge.

    3.   Lagoons 8 and 10 composite  sludge  sample  to  represent the average
         contaminated sludge.

    4.   Lagoons 8 and 10  composite  soil  sample  to  represent  the average
         contaminated soil.

Request for proposals were issued to nine vendors capable of performing treat-
ability studies  for  soil  washing,  LTTD,  and solidification/stabilization.
Three vendors were selected and contracted to perform the treatability studies
based  upon  their  qualifications.    BioTrol,   Inc.  (BioTrol),  of  Chaska,
Minnesota, performed the soil washing treatability study on the worst case and
average soil samples.  Chemical  Waste  Management, Inc.  (CWM),   of Riverdale,
Illinois,  performed the LTTD treatability study  on the worst case soil, worst
case sludge, and  the  average  sludge  samples.    Enreco,  Inc. (Enreco), of
Amarillo,  Texas, performed the solidification/stabilization treatability study
on the worst case soil, worst case sludge, and average sludge samples.
                                      353

-------
Results from the  treatability  studies  were  used  to  address the following
questions:

    1.   Are the treatment processes able to achieve adequate performances for
         the compounds of concern for the lagoon soil and sludge tested?

    2.   Do any residuals require further treatment prior to disposal?

Treatability experimental  procedures,  experimental  results, and conclusions
for each process option are discussed in the following sections.  Experimental
results from the treatability  studies  are evaluated according to performance
and the identification and characterization of residual waste streams.

Performance was evaluated either by  reviewing attainment of cleanup standards
(soil washing and LTTD),  or  by  reviewing potential reductions in contaminant
mobility (solidification/stabilization).   Attainment  of cleanup standards was
addressed by comparing the concentrations  of  the compounds of concern in the
treated material with 10"^ and  10"" risk-based cleanup standards developed in
the baseline risk assessment.  Because the treated material will be covered or
capped  after  replacement,  groundwater  ingestion  is  the  most significant
exposure route.  Therefore, cleanup standards  used in these studies are based
on remediation of sludges and  soils  to  prevent contamination of the ground-
water.  Potential reductions  in  contaminant  mobility  are evaluated by com-
paring TCLP  extract  concentrations  in  the  prestabilized    and stabilized
samples.

Data tables showing detected  compound  concentrations  are summarized in this
chapter.  Analytical holding  times  were  exceeded  by the laboratory on some
samples; these samples  are  noted  in  the  tables.    Some  samples were not
analyzed due to extreme exceedances on analytical holding times by the labora-
tory.  Because of these suspect or non-existent data, CWM-generated pre-treat-
ment data are sometimes used for the other treatability studies.  In addition,
because of the suspect data, the highest concentrations of each post-treatment
compound detected is used where multiple  analyses  of the same type of sample
have been performed.

Quantitation limits are often high  for  the  Bofors  sludges and soils due to
interferences in the samples.  Therefore, the risk-based cleanup standards are
sometimes below the quantitation limit on  the treated samples; thus, the per-
formance of the process options  cannot  always be confirmed for each compound
of concern.  In these cases,  the  results  of the TCLP for the treated sample
are evaluated to gain  additional  information on the potential concentrations
in the treated samples.  Since the  TCLP data are generated from an extraction
procedure, the concentration in the actual  treated sample is most likely sig-
nificantly greater than in the TCLP extract.  It is assumed in this study that
if the TCLP  extract  concentration  of  a  compound  of  concern is above the
quantitation limit,  then  the  actual  concentration  in  the  treated sample
exceeds the cleanup standard.    This  is  a conservative assumption; however,
since no information  on  the  relationship  between  TCLP  and actual concen-
trations is known, this assumption may  be  justified.  If the compound is not
present in the TCLP extract above the quantitation limit, then removal of that
                                       354

-------
compound is assumed to be adequate  to  achieve the cleanup standard.  In some
cases, the TCLP quantitation limit may  also  be higher than the cleanup stan-
dard.  In these cases, the limits of currently available analytical techniques
have been reached.  Since the  analytical techniques used in these studies are
the best available accepted  technology,  it  is  assumed that if TCLP concen-
trations are below  the  quantitation  limits,  the  technology has reached an
acceptable degree of cleanup.

1.2  SOIL WASHING

The soil washing process  extracts  contaminants  from  soil using water or an
aqueous solution  which  may  be  composed  of  chelating agents,  surfactants,
acids, or bases.  Chelating agents  are  organic compounds in which atoms form
coordinate bonds with metals.   Surfactants  are soluble compounds that reduce
the surface tension of liquids or  the interfacial tension between two liquids
or between a liquid and a soil.  These compounds help create phase changes and
disperse contaminants to facilitate contaminant removal.  The primary function
of this technology is to  reduce  the  volume of contaminated fine silt, clay,
and colloidal fractions from coarse  sand  and  gravel components.  Fine silts
and clays absorb organic contaminants  due  to  the surface chemistry and high
surface area of the natural  mineral  and organic material and, therefore, the
majority of contamination is associated  with  these size fractions.  Using an
intensive scrubbing  action,  soil  aggregates  are  broken  up and dispersed,
freeing highly contaminated fine particles  from  the coarser sand and gravel.
The surfaces of the coarser particles are cleaned by abrasive processes.

A typical soil washing removal treatment process is a multi-step process which
begins with excavating and screening the contaminated soil to remove oversized
material and debris.  The  contaminated  soil  is then slurried with water and
screened at  14 mesh,  or  approximately  0.05 inches  in  diameter.  Material
larger than 14 mesh can be washed further if necessary.  Material smaller than
14 mesh  enters  a  froth   flotation  circuit  where  hydrophobic  components
(compounds with a low water solubility)  are removed.  The flotation underflow
enters an  intensive,  multi-stage,  countercurrent  scrubbing  system.   Fine
particles leave the process and undergo  residuals treatment or disposal.  The
final washed product is dewatered and may  be  replaced on the site.  The con-
taminated water from the scrubbing system enters a treatment system and can be
recycled to the soil washing system.  Soil washing technology was evaluated in
the U.S. EPA Superfund Innovative Technology Evaluation (SITE) Program.

1.2.1  Treatability Experimental Procedures

As discussed previously,  chemical  analytical  results  from  soil and sludge
samples collected during the Phase I  RI  were  used to select samples for the
Phase II treatability study.   Lagoon 3 soil  (worst case), and Lagoon 8 and 10
soil  (average) samples were sent to  BioTrol.   Sludge samples were not tested
in the soil washing treatability study  because soil washing is not a suitable
process option for sludge remediation.
                                      355

-------
The soil washing treatability study involved the following processes:

    o  Sample preparation
    o  Soil scrubbing
    o  Soil leaching
    o  Gravity separation
    o  Flotation

Each of these  processes  can  be  used  as  individual processes for specific
applications or can be included as parts  of a treatment process train.  These
processes have all been used in  the hydrometallurgical industry and are being
combined here to form a soil washing remediation technology designed to remove
contamination from soil.  Each process is described below.

    1.2.1.1  Sample Preparation

Samples were provided by  in  one- to  five-gallon  containers.  These samples
were reduced in size using a riffle  splitter.  This device became coated with
a heavy grease-like substance  (tar)  when  used  once  to produce the smaller
samples.  This device was  not  used  again, and smaller samples were obtained
using a trowel.

    1.2.1.2  Soil Scrubbing

In this process each soil sample  was washed multiple times with the following
scrubber solutions:

         Water
    *    Orvus K liquid  (Proctor & Gamble Corporation)
    0    Citrikleen  (Penetone Corporation)
    *    Industrial Tide  (Proctor & Gamble  Corporation)
         Triton X-100  (Rohm and Haas Corporation)

The additives are emulsifiers used  to disperse hydrophobia compounds that are
not soluble in water.

Samples  of  Lagoon 3   soil   and   Lagoon 8   and 10  soil  were  preweighed
(1.0 kilogram) and combined with  water  and  additive to produce a 50 percent
solids slurry in the scrubbing  unit.    The slurries were thoroughly agitated
with an impeller for 5 minutes  and  then  diluted  to 2 liters and allowed to
settle.  The supernatant was decanted and stored for future chemical analysis.
A 50 percent slurry was made  with  the  remaining  solids.  The procedure was
then repeated.  After completing four  or five washing stages, the solids were
removed from  the  washing  unit,  dewatered,  and  chemically  analyzed.  The
supernatant from each washing stage was composited and submitted for analysis.

    1.2.1.3  Soil Leaching

To develop an understanding of how the Bofors soil samples react in an aqueous
environment, a series of  72-hour  leaching  tests were conducted.  Flasks were
prepared separately  for  Lagoon 3  soil  and  Lagoon 8  and 10  soil samples.
                                      356

-------
Flasks containing 20 to 50 percent solids for  each soil type were placed on a
shaker table which continuously mixed  the  samples for 72 hours.  The samples
were then filtered and air-dried.

    1.2.1.4  Gravity Separation

Gravity separation was used  to  separate  materials in an aqueous environment
using dense fluid.  Contaminants are  less dense than the soils, therefore the
contaminants should float as  the  "clean"  soil  sinks.  Calcium chloride was
used to separate contaminants from Lagoon  soil samples because of its availa-
bility and non-solvent properties.

    1.2.1.5  Flotation

In a flotation cell,  water  containing  the  fine  particles and chemicals is
brought in contact with very small  air  bubbles.  Chemical additives are used
to separate particulate solids by causing  one group to float.  Differences in
surface chemical properties of the  particles allow separation since some par-
ticles are entirely wetted by water,  and  others are not.  The bubbles attach
to the physically- and chemically-altered  surfaces  of the particles and rise
to the top of the flotation cell  where  they are skimmed.  This procedure was
tested with washed soil to  evaluate  the  potential of a flotation process to
act as a final polishing step in the remediation of the soil.

1.2.2  Experimental Results

    1.2.2.1  Laeoon 3 Soil

Lagoon 3 soil was sampled  to  represent  worst-case  contaminated soil in the
L.O.U.  Although soil characteristics indicated  that soil washing would be an
excellent technology, the soil  contained  a  heavy, grease-like substance (or
tar) which, when slurried with  water  and agitated, coated metallic equipment
surfaces in contact with the slurry.

Leaching tests on Lagoon 3 soil  indicated  that low intensity agitation could
be used to agglomerate the tar.  During leaching, the tar formed spherical tar
balls or tar granules that  were  generally  larger than the sand particles in
the soil.   The  spherical  tar  balls  appeared  when  the solids content was
greater than 30 percent in the testing.  Following leaching, this material was
removed by screening.  Analysis  showed  that  the leached soil was similar to
the soil subjected only  to  high-intensity  attrition scrubbing.  The concen-
tration of contaminants in the  tar  was extremely high, indicating that small
amounts of tar in clean soil  would  render the soil "contaminated."  Approxi-
mately 80 percent of the DCB  were  contained  in  less than 10 percent of the
soil.

Leached soil  was  then  subjected  to  high-intensity  agitation,  which sig-
nificantly reduced the concentration  of  contaminants.    By combining a tar-
removal step with soil washing, the Lagoon 3 soil was cleaned to approximately
the same level as Lagoons 8 and 10 soil.
                                      357

-------
Washed soil consisted of predominantly sand  and  small particles of tar.  The
tar was much  less  dense  than  sand.    Therefore,  gravity separation using
saturated solutions of calcium chloride  was used to separate the contaminated
tar particles and  improve  cleanup  of  the  coarse  material.  However, this
process did not achieve separation of  the  small  tar particles from soil.  A
flotation procedure was also implemented to remove the residual tar from soil.
The froth from this procedure appeared  to  carry the tar particles out of the
slurry.  However, no significant  differences resulted in the concentration of
the contaminants in leached, washed soil before and after flotation.

Performance Evaluation

The cleanup standards for the  Lagoon 3  soil  and the pre-treatment and post-
treatment concentrations for the compounds of concern are listed in Table 1.

Aniline was detected in the  pre-treatment  sample  below the 10"^ but greater
than the 10"^ risk-based cleanup  standard.    Aniline was not detected in the
post-treatment sample above the quantitation limit.  The quantitation limit is
greater than  the  10*6  cleanup  standard  but  less  than  the  10"^ cleanup
standard.  However, the TCLP  extract  concentration is below its quantitation
limit.  Therefore, as discussed  previously,  it is assumed that an acceptable
removal level has been  achieved  for  aniline.    Azobenzene and benzene were
detected in the pre-treatment sample at less  than the 10*^ and the 10"^ risk-
based cleanup standard; therefore, no  removal  is needed for these compounds.
Benzidine was detected in  the  pre-treatment  sample  above the 10"^ and 10"^
cleanup standards.  Benzidine was not detected above the quantitation limit in
the post-treatment sample.  The quantitation  limit is above the 10"^ and 10'^
standards.  However, the TCLP  extract concentration is below its quantitation
limit.  Therefore,  the  attainment  of  cleanup  standards  is  assumed to be
acceptable for benzidine.  DCB was  detected in the pre-treatment sample above
the 10"^ and 10"^  cleanup  standards.    The post-treatment DCB concentration
exceeds the 10*^ cleanup  standard;  therefore,  an acceptable removal has not
been achieved for a  10"^  remedy,  but  appears  to  be acceptable for a 10"^
remedy.

In summary, soil washing for  Lagoon 3  soil  appears to be successful in this
test for the removal of  aniline,  azobenzene,  benzene, and benzidine to 10"^
and 10"" cleanup standards.  However,  soil washing did not effectively remove
DCB to 10~6 levels, but may be appropriate for a 10"^ level remedy.

Process Residuals Evaluation

The treatment residual wastes for soil  washing  are the fines and the aqueous
phase  (wash water).   Chemical  analyses  of  the  fines and aqueous phase are
presented in Table 2.    Pre-treatment  compound concentrations are also shown
to indicate compounds originally present  in  the  sample.  Based on comparing
pre-treatment soil and  post-treatment  fines  concentrations, it appears that
benzine, azobenzene,  and  DCB  have  been  concentrated  in  the fines phase.
Aniline and benzidine are  not  present  in  the  fines above the quantitation
limit.  The five compounds of  concern  for  Lagoon 3 soil are also present in
the aqueous phase in significant  concentrations.  These results indicate that
soil washing of the sample has  caused  phase transfer and volume reduction of
contaminants of concern.

                                      358

-------
                                                       TABLE  1
                                                    SOIL WASHING
                                         ATTAINMENT OF CLEANUP STANDARDS
                                   FOR COMPOUNDS OF CONCERN FOR LAGOON 3 SOIL
                                                      Botors Site
                                                  Muskegon, Michigan
                              Cleanup Standards*
Compound
10"4 Risk
(ug/kg)
10* Risk
(ug/kg)
Pre-Treatment*
(Coarse)
(ug/kg)
Post-Treatment*
(Coarse)
(ug/kg)
Ouantitation
Limit
fug/kg)
Post-Treatment
TCLP
(ug/l)
Aniline
Azobenzene
Benzene
Benzidine
3,3'- Dichlorobenzidine
    97,000
42,000,000
   180,000
      110
    85,000
   970
420,000
  1,800
      1.1
   850
   9,200
  150,0000
     130 J
  590,000 D
1,100,0000
16,000 J
     3J

28,000 J
               20,000
96,000
LEGEND:
D :       Value reported from a diluted sample aliqout in order to stay within the linear calibration of GC/MS.
J :       Estimated value due to minor OC deviations or for tentatively identified compound (no standard available) or mass
         spectra indicates compound present below contract detection limit, but greater than zero.
— :      Compound was not detected.
Cleanup standards based on remediation for the groundwater ingestion exposure route.
TCLP:    Toxicity characteristic leaching procedure.
NA:      Not applicable to this sample.
*:        Soil concentrations and cleanup standards are on a dry weight basis.
                                                       359

-------
                                                   TABLE 2
                                                SOIL WASHING
                             CHEMICAL CHARACTERIZATION OF PROCESS RESIDUALS
                                              FOR LAGOON 3 SOIL
                                                  Bofors Site
                                              Muskegon, Michigan
                 Compound
Volatile Organic Compounds
Acetone
Benzene
2-Butanone
Ethylbenzene
Methylene Chloride
2-Methylpropane
Toluene
Total Unknown Hydrocarbons
Xylene (Total)
Semi-Volatile Organic Compounds
Aniline
Azobenzene
Azoxybenzene
Benzidine
2- Chloraniline
3,3' - Oichlorobenzidlne
Dichlorobenzidine Isomer
Methanone, (3-Chlorophenyl) (4-Chlorophenyl)
Nitrobenzene
Total Alkyl Benzenes
Total Unknowns
Metals
Aluminum
Arsenic
Barium
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Nickel
Potassium
Selenium
Sodium
Vanadium
Zinc
                                         Pro-Treatment*
                                             (ug/kg)
                                               130 J
                                                 15 BJ
                                                 3J
                                               490 BEJ
                                             18,343 J
                                                  4J
                                              9,200
                                           150,0000
                                            460,000 J
                                           590,000 D
                                             17,0000
                                          1,100,0000
                                              1,300 J
                                            450,000 J
                                             13,000 EJ
                                          1,080,000 J
                                            829,000 J
                                            185,000
                                               610 B
                                              4,800 B

                                            461,0008
                                              3,000
                                               790 B
                                              2,0006
                                            533,000
                                              6,400
                                             74,9006
                                              3,1006

                                             51,5008

                                            431,0006
                                               5308
                                            324,000
                                                                               Post-Treatment Residuals
    Fine*
   (ug/kg)
    4,000
   22,000
    3,800 B

       75 B
    4,600 BJ
   57,000
    3,300 BJ
3,200,000
13,000,000

   84,000 J
9,700,000

12,000,000
  785,000 J
     1,4008
     6,300 B
      7506
 1,580,000
    14,400

    16,600
 1,710,000
    96,800
  383,0006
    16,100
     2,4006
  200,0008
     1,400
  212,000 B
     4,3008
 1,920,000
Aqueous
  (ug/l)
    78 J
 11,000

    190 J
    85 BJ

 19,000

  2,000


    190 J
 13,0000
 26,000 J
 30,000 0
  4,300
 55,000 DJ
  3,900 J
  5,500 J
    250
  7,400 J
156,000 J
  6,300
      5.2 B
    378
      2.9 B
141,000
  1,130
      7.1 B
    228
  6,470
    360 J
 38,100
    132
    753
  8,630

 72,000
      8.3 B
  8,540
Legend:
B:
D:
E:
J:
Compound was also detected in the associated laboratory blank.
Value reported from a diluted sample aliquot in order to stay within the linear calibration of GC/MS.
Concentration exceeded the linear range of GC/MS calibration.
Estimated value due to minor QC deviations or for tentatively identified compound (no standard available) or mass
spectra indicates compound present below contract detection limit, but greater than zero.
Compound was below the quantifiable limit.
Concentrations reported on a dry weight basis.
                                                     360

-------
The  fines  and  aqueous  phase  contained  several  VOCs  (acetone, methylene
chloride, and  2-methylpropane)  and  metals  (cadmium,  nickel', and selenium)
which were not detected in the  original  sample.    This may be due to insuf-
ficient equipment decontamination, laboratory contamination, or differences in
compound quantitation limits.

Based on the results shown in the  table, the fines and the aqueous phase will
require further  treatment  prior  to  disposal  and  discharge. Metal concen-
trations in the aqueous phase are below literature values for inhibition of an
activated sludge process  and  should  not  present  a  toxicity  problem if a
biological process is employed  (Wastewater  Treatment Plant Design and Treat-
ment, Water Pollution Control Federation, 1977).  However, the relatively high
concentrations of  calcium  and  magnesium  may  result  in  equipment scaling
problems.

    1.2.2.2  Lagoon 8 and 10 Soil

A soil composite from Lagoons 8 and 10  was  used in the treatability study to
represent the average contaminated  soil  at  the  Bofors site. Lagoon 8 and 10
soil was washed with the following wash solutions:

         Water
     8    Orvus K liquid and  hot water
     0    Citrikleen and hot  water.

There did not appear to be a significant difference in performance between the
various wash solutions.  Leaching  tests  were also performed on Lagoons 8 and
10  soil.  The  results   of   these  tests  were   similar to those presented for
Lagoon 3 soil.

A second series of soil  washing  was  performed  with hot water and Citrikleen,
Industrial Tide, and Triton X-100 at  1 percent  doses.  Following this removal
step, the concentrations of  the  compounds  of concern were below quantitation
limits except for DCB.   Both the Triton X-100 and the Industrial Tide formula-
tions were able to  destabilize 'the  colloidal  dispersion  of fines and give
clear (but colored) aqueous  decants.  The Triton X-100 material generated much
less foam than the Tide  and was chosen as the additive for other iterations.

Performance Evaluation

The 10"^ and 10"^ risk-based  cleanup  standards for Lagoon 8  and  10 soil and
the pre-treatment and post-treatment concentrations  for  the compounds of con-
cern are listed in Table 3.    Aniline   and benzene were detected  in the pre-
treatment sample below  10'^  and   10"6 risk-based cleanup  standards;  therefore,
no  removal is needed for these compounds.   Azobenzene was detected  in the pre-
treatment sample below  the 10"^  cleanup  standard  but above the 10*° cleanup
standard.  Azobenzene was not  detected   in  the post-treatment sample above the
quantitation limit.  The quantitation  limit  is greater  than the 10*^ cleanup
standard.  As discussed previously,  the possible concentration in  the treated
sample was evaluated using   the  TCLP  extract   concentration.  Azobenzene was
present  in the post-treatment  TCLP  extract  (3 ppb).    Therefore,  the actual
                                     361

-------
                                                  TABLE  3
                                                SOIL WASHING
                                     ATTAINMENT OF CLEANUP STANDARDS
                           FOR COMPOUNDS OF CONCERN FOR LAGOON 8 AND 10 SOIL
                                                  Bofors Site
                                              Muskegon, Michigan
                           Cleanup Standards*
      Compound
                1CT Risk
                 (ug/kg)
10"6 Risk
 (ug/kg)
Pre-Treatment*  Post-Treatment*   Quantitation      Post-Treat-
   (Coarse)        (Coarse)          Limit         ment TCLP
    (ug/kg)          (ug/kg)	(ug/kg)          (ug/l)
Aniline                   360,000        3,600           900
Azobenzene                46,000          460          1,800
Benzene                   42,000          420            —
Benzidine                      9.2          0.092       7,500
3,3'- Dichlorobenzidine         140            1.4       120,0000
                                                                               660
                                                                               660
                                                                                  8
                                                                              3,200
                                                                 3J
                                                             6,500
LEGEND:
J:


TCLP:
Estimated value due to minor OC deviations or for tentatively identified compound (no standard available) or mass
spectra indicates compound present below contract detection limit, but greater than zero.
Compound was not detected.
Toxicity characteristic leaching procedure.
Soil concentrations and cleanup standards are on a dry weight basis.
                                                    362

-------
concentration in  the  treated  sample  is  assumed  to  be  above the cleanup
standard.  Benzidine was detected  in  the pre-treatment sample above the 10~4
and the 10"^ cleanup  standards,  but  was  not detected in the post-treatment
sample above the quantitation limit.  This  value is greater than the 10*4 and
10'° risk-based cleanup standards; however,  the TCLP extract concentration is
below its quantitation limit.  Therefore, removal is assumed to be acceptable.
DCB detected in the post-treatment sample exceeded the cleanup standards by at
least one order of magnitude for  10'4  risk and three orders of magnitude for
10'6 risk indicating that DCB was  not  effectively removed.  In summary, soil
washing does not appear  to  be  an  effective  process  option for this soil,
specifically for azobenzene and DCB.

Process Residuals Evaluation

The concentration of organic compounds and metals for the pre-treatment sample
and residual waste streams of Lagoon 8 and 10 soil are shown in Table 4.  Pre-
treatment compound concentrations are  shown  to indicate compounds originally
present  in the sample.    As  mentioned  previously, only semivolatile organic
compound analyses are available for the fines for Lagoon 8 and 10 soil.

Based on comparison of pre-treatment  soil and post-treatment fines concentra-
tions, only DCB appears to have been concentrated slightly in the fines phase.
Azobenzene and benzidine are not  present  in the fines above the quantitation
limit.  DCB is also the only compound of concern detected in the aqueous phase
at a concentration significantly above  the quantitation limit.  These results
indicate a phase transfer  and  volume  reduction of DCB contaminated material
but it is uncertain whether this has occurred for azobenzene and benzidine.

The results also  indicate  that  the  aqueous  waste  stream will most likely
require  treatment  for  organic  compounds  and  metals  prior  to discharge.
However, the concentration  of  calcium  and  magnesium  in  the aqueous waste
stream may cause equipment scaling problems and require preventive measures.

1.2.3  Conclusions

Results  of the soil washing treatability   study indicate that this process was
not effective in cleaning worst-case  or   average  soil in the L.O.U. to risk-
based cleanup standards, specifically  in  attaining cleanup standards for DCB
and azobenzene.   This  may  be  explained by  the  high  affinity of DCB and
azobenzene  for  soil  compared  to  the   other  compounds.    Soil  partition
coefficients (Kd) were developed during  the  RI  for the compounds of concern
using site soil information  (GZA/Donohue 1990), and are summarized as follows:

         aniline; Kd-0.027 cm3/g
         azobenzene;  Kd-14.9 cm-yg
         benzene; Kd-0.389 cm3/g
         benzidine; Kd-0.071 cm-yg
         DCB; Kd-7.634  cm3/g
                                      363

-------
                                                   TABLE 4
                                                SOIL WASHING
                             CHEMICAL CHARACTERIZATION OF PROCESS RESIDUALS
                                          FOR LAGOON 8 AND 10 SOIL
                                                   Bofors Site
                                              Muskegon, Michigan
                      Analyte
Volatile Organic Compounds
Acetone
Benzene
Chloroform
Methylene Chloride
Styrene
Toluene
1,1,1 - Trichloroethane
Xylene (Total)
Total Unknowns
Semi-Volatile Organic Compounds
Aniline
Azobenzene
Azoxybenzene
Benzidine
3,3'-Dichlorobenzldine
Dichlorobenzidine Isomer
Methanone, (3-Chlorphenyl)(4-Chlorophenyl)
Phenol, 4-(2,2,3,3-Tetranethylbutyl)
1-Propene, 2-Methyl, Tetramer
1-Tetradecanol
Total Unknown Alkene Derivatives
Total Alkyl Benzenes
Total Unknowns
Metals
Aluminum
Barium
Calcium
Chromium
Copper
Iron
Magnesium
Manganese
Nickel
Potassium
Sodium
Znc
                                               Pre-Treatment*
                                                   (ug/kg)
                                                     11J


                                                      7

                                                     49 B

                                                      1 BJ
                                                      gj
                                                    900
                                                  1,800

                                                  7,500
                                                120,0000

                                                  6,500 J
                                                  20,000 J

                                                  19,700 J
                                                 130,600 J
                                               1,150,000
                                                   7,8008
                                               7,140,000
                                                   6,100
                                                   1.400B
                                               1,990,000
                                                 892,000 B
                                                  24,400

                                                 177,0006
                                                 103,0008
                                                 311,000
                                                                                 Post-Treatment Residuals
   Fine*
  (ug/kg)
Legend:
B:
D:
J:
NA:
     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
380,000
390,000 J
  18,000 J
  15,000 J


     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
     NA
 Aqueous
   (ug/l)
      8B
      1 J
      4J
      SB
      2J
     10
     15B
      5
     30 J
    340 J

 23,000 DJ
 12,000 J
    570 J
    800 J
  2,000 J
    340 J
  2,290 J
  1.280J
 20.150J
    417
     19.3B
213,000
     27.8
     17.1 B
    406
 37,000
     45.8
     27.7 B
  6,710
 31,100
    192
Compound was also detected in the associated laboratory blank.
Value reported from a diluted sample aliquot in order to stay within the linear calibration of GC/MS.
Estimated value due to minor QC deviations or for tentatively identified compound (no standard available) or mass
spectra indicates compound present below contract detection limit, but greater than zero.
Compound was below the quantifiable limit.
Compound was not analyzed.
Concentrations reported on a dry weight basis.
                                                     364

-------
The K(j estimates ,the affinity  of  a  compound  to  soil; as K^ increases, the
affinity  increases.    Azobenzene  and  DCS  have  the  highest  K^  of these
compounds.  They are higher by one to three orders of magnitude than the other
compounds.

Questions that remain for  this  technology  are  effective tar removal before
soil washing and identification of the final polishing step to remove residual
tar from  the washed sand.  The agglomeration of tar and its subsequent removal
by screening are important  issues.    The  leaching  tests were performed for
72 hours  and have not been optimized.   Time constraints on the original study
precluded more than a cursory  assessment  of gravity separation and flotation
as final  process steps.  The effectiveness of gravity separation versus flota-
tion as a final polishing step could be examined in greater detail.

The aqueous phase and fines carry  a  relatively high level of chemical oxygen
demand (COD) and will require treatment  before being reused in the process or
before discharge.  Treatment of  these  residuals, which make up approximately
5 to 10 percent of the feed material, must be addressed.

Further bench scale testing  and  a  pilot  study  would  also be necessary to
further evaluate  attainment  of  cleanup  standards.    However,  this is not
recommended because  of  the  inability  of  this  soil  washing technology to
achieve cleanup standards for DCB at the bench scale.

1.3  LOW  TEMPERATURE THERMAL DESORPTION

Low temperature thermal desorption  (LTTD)  removes  contaminants from a solid
matrix by applying heat  to  the  material.    The  heat  allows the matrix to
release contaminants by several  mechanisms,   including direct volatilization
of the  organic  compounds  whose  boiling  points  have  been  reached, steam
stripping of organic compounds by water  within the solid matrix, and evapora-
tion of organic compounds due  to  their increased vapor pressures at elevated
temperatures.

1.3.1  Treatability Experimental Procedure

The following samples were  collected  and  sent  to CWM for LTTD treatability
testing:     Lagoon 3  soil  (worst-case  soil),  Lagoon 9  sludge  (worst-case
sludge),   and Lagoon 8 and 10 composite sludge (average sludge).

    1.3.1.1   Sample Preparation

Whenever visual  inspection  of  the  feed  material  indicated  that oversize
materials were present,  the  batch  of  feed  material  was  passed through a
1/4-inch  sieve.  The  screened  feed  was  then  thoroughly  mixed with a drum
roller or hand tools.

After the feed was screened and  mixed,  a  composite sample was taken by grab
sampling  at several locations  throughout  the  container.   A separate cleaned
volatile  organic analysis jar was  used  for  the VOC portion of the composite
sample.  This portion of the  sample  was refrigerated before analysis to pre-
                                      365

-------
serve its integrity.    The  feed  sample  was  then  taken from the remaining
material.  The resulting removal  efficiencies represent the material that was
actually processed, and no credit was  taken  by the LTTD system for losses of
volatile organic compounds due to feed preparation activities.

    1.3.1.2   Test Procedures

The bench scale system delivered feed  material  to a dryer using a volumetric
screw feeder.  Approximately 1,600 to  2,000 grams of material were fed into a
4-inch diameter cylindrical furnace  (kiln)  for  each  sample processed.  The
kiln was operated at maximum  temperature (800 degrees Fahrenheit) and maximum
residence time  (100 minutes).    These  operation  parameters were considered
economically viable by CWM.  The  dried  solids were collected in a hopper and
emptied through a valve.

Heated nitrogen gas entered and  exited  the  dryer  in a flow parallel to the
solids.   This nitrogen gas stream carried water vapor and the volatilized feed
constituents to  the  spray  tower  where  particulate  carry-over solids were
removed and the  gas  cooled  to  its  saturation  temperature (typically from
110 to 150 degrees Fahrenheit).   Some  contaminants  with high boiling points
were condensed in the  spray  tower.    The  spray  tower scrubbing fluid then
flowed to the phase  separator  where  the  solids and immiscible organic com-
pounds were separated from the scrubbing water.

The gas stream continued to the primary condenser where it was cooled to 80 to
120 degrees Fahrenheit.  The gas then entered the secondary condenser where it
was further cooled by a  mechanical  refrigeration  unit to a temperature less
than 40 degrees Fahrenheit.   Finally,  the  gas  stream  was passed through a
carbon adsorption unit before being purged to the atmosphere.

The treated solids were collected after the unit reached steady-state tempera-
ture.  After mixing, the test sample  was divided into five portions.  Four of
these subsamples were analyzed by laboratories selected by GZA/Donohue and the
fifth was analyzed by CWM.

The phase separator, which  supplies  the  spray  tower fluid, was filled with
17 liters of tap water at the start of each  test run.  At the end of the run,
it contained 12 to 16 liters.    A  representative  sample  of this liquid was
taken for analysis.

Liquid retained during the steady  state  portion  of the test period was com-
bined at the end of each  run.    The  combined  sample from each test run was
mixed and put in appropriate containers for analysis.

1.3.2  Experimental Results

    1.3.2.1  Lagoon 3 Soil

Lagoon 3 soil was bench scale tested to evaluate the upper limit of effective-
ness of LTTD for soil, and to identify potential limitations of this treatment
process at the Bofors Site.
                                      366

-------
Performance Evaluation

The 10"^ and 10"^ cleanup standards and the pre- and post-treatment concentra-
tions for the compounds  of  concern  are  listed  in Table 5.  Azobenzene and
benzene were detected in the pre-treatment  sample below 10*^ and 10~*> cleanup
standards.  Therefore, no removal is needed for these compounds.

Aniline was detected in the  pre-treatment  sample below 10*^ but greater than
10*6 cleanup standards.  Aniline was not detected in the post-treatment sample
above the quantitation limit.  The quantitation limit is below the 10"6 clean-
up standard; therefore,  an  acceptable  removal  level  has been achieved for
aniline, based on these results.   Benzidine was detected in the pre-treatment
sample above the 10"^ and 10*^ cleanup  standards, but was not detected in the
treated soil above the quantitation  limit.    The quantitation limit is above
the cleanup standards.  However,  the  benzidine  is  not detected in the TCLP
extract above the quantitation limit; therefore, it is assumed that an accept-
able removal level has been achieved for  benzidine.  DCB was also detected in
the pre-treatment sample above the  cleanup  standards and below the quantita-
tion limit in the treated sample.   Since the quantitation limit is below both
cleanup standards, an  acceptable  removal  level  has  been achieved for DCB,
based on these results for  Lagoon 3  soil.    In  summary, LTTD appears to be
effective in removing the contaminants of concern in worst case soil, Lagoon 3
soil.

Process Residuals Evaluation

Analytical results for post-treatment  residual phase separator and condensate
liquids are shown in Table 6.  Pre-treatment compound concentrations are shown
to indicate compounds present in the  pre-treatment soil.  The phase separator
is designed to contain solids  and  immiscible compounds removed from the con-
taminated soil.   The  condensate  is  designed  to contain volatile compounds
condensed out of the gas phase.  As shown in Table 6, benzene, a VOC, was only
detected in the  condensate,  not  in  the  phase  separator liquid.  However,
aniline, a semi-volatile compound, has  a  higher concentration in the conden-
sate than in the phase separator  liquid.  Azobenzene, benzidine, and DCB, all
semi-volatile compounds, have  greater  concentrations  in the phase separator
liquid, as expected.    Therefore,  the  compounds  of  concern generally par-
titioned into their expected phases.

The phase separator and condensate residuals contained several VOCs which were
not detected in the original  Lagoon 3  sample.   These compounds have several
possible sources.  They may be a result of insufficient equipment decontamina-
tion or laboratory  cross-contamination.    Differences  in compound detection
limits from one analysis to another  may  also have occurred (i.e., the detec-
tion limit for the pretreated  material  may  have  been an order of magnitude
higher than the post-treatment detection limit).

Because high  concentrations  of  contaminants  were  collected  in both waste
streams, it is likely  that  both  condensate  and phase separator liquid will
require further treatment before disposal or discharge.  Calcium and magnesium
may require treatment to prevent equipment scaling problems.
                                     367

-------
                                                   TABLE 5
                                   LOW TEMPERATURE THERMAL DESORPTION
                                     ATTAINMENT OF CLEANUP STANDARDS
                              FOR COMPOUNDS OF CONCERN FOR LAGOON 3 SOIL
                                                  Bofors Site
                                              Muskegon, Michigan
                          Cleanup Standards*
Compound
10"4 Risk
(ug/kg)
10"6 Risk
(ug/kg)
Pre-Treatment*
(ug/kg)
Post-Treatment*
(ug/kg)
Quantitation
Limit
(ug/kg)
Post-Treatment
TCLP
(us/l)
Aniline

Azobenzene

Benzene

Benzidine

3,3'- Dichlorobenzidine
   97,000

42,000,000

   180,000

      110

   85,000
   970

420,000

  1,800

      1.1

   850
   9,200

 150,0000

     130 J

 590,0000

1,100,0000
4,920

    7
                330
               1,600

                650
LEGEND:

D :       Value reported from a diluted sample aliquot in order to stay within the linear calibration of GC/MS.

J :       Estimated value due to minor QC deviations or for tentatively identified compound (no standard available) or mass
         spectra indicates compound present below contract detection limit, but greater than zero.

— :      Compound was not detected.

Cleanup standards are based on remediation for the groundwater ingestion exposure route.

         Cleanup standards and soil concentrations are on a dry weight basis.
                                                    368

-------
TABLE 6
LOW TEMPERATURE THERMAL DESORPTION
CHEMICAL CHARACTERIZATION OF PROCESS RESIDUALS
FOR LAGOON 3 SOIL
Bofors Site
Muskegon, Michigan
Post-Treatment Residuals



Pre-Treatment* Condensate
Analyte
Volatile Organic Compounds
Acetone
Benzene
2-Butanone
2-Hexanone
Toluene
1,1,1 - Trichloroethane
Semi-Volatile Organic Compounds
Aniline
Azobenzene
Azoxybenzene
Benzene, Isocyano
Benzidine
Benzole Acid
2-Chloroaniline
4-Chloroaniline
2-Chlorophenol
3,3'-Dichlorobenz!dine
Methanone, (3-Chlorphenyl)(4-Chlorophenyl)
2-Methylphenol
4-Methylphenol
Nitrobenzene
Phenol
1-Tetradecanol
Total Alkyl Benzenes
Total Unknown Trichlorobiphenyls
Total Unknowns
Metals
Aluminum
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Nickel
Potassium
Sodium
Vanadium
Zinc
Legend:
B : Compound was also detected in associated
(ug/kg)

_
130 J
15 BJ
—
490 EBJ
—

9,200
150,0000
460,000 J
—
590,000 D
—
170,000 D
820
—
1,100,0000
450,000 J
—
_
13,000 EJ
—
290,000 J
1 ,080,000 J
—
829,000 J

185,000
610 B
4,800 B
—
—
461, 000 B
3,000
2,000 B
533,000
6,400
74,9008
3,100 B
—
—
51,5006
431, 000 B
530 B
324,000

laboratory blank.
D : Value reported from a diluted sample aliquot in order to stay within
(ua/l)

89
30
72
46
65
10

60,000
6,000
3,600 J
760 J
6,700
—
28,000 EJ
390 J
3,100
4,800 J
—
480
—
1,300 J
17,400 E
—
760 J
2,570
—

34.8 B
R
3.6 B
1.8 B
—
2,640 B
—
17,400
437 J
52.6
685 B
53.2
3.9 J
23.4 B
—
1.170B
—
934


linear calibration of GC/MS.
Phase Separator
Liquid
(ua/l)

35
__
—
—
46
—

20,640
220,000 J
610,000 J
—
19,100
710 E
5,900 J
—
200
21 0,000 J
490,000
130
160
—
1 7,000 J
73,000 J
200,000 J
—
151.000J

1,870
2B
73.6 B
—
4B
70,000
135
430
3,650
131
19,400
53.2
1.1
126
2,860 B
14,900
7B
4,290



E: Concentration exceeded the linear range of GC/MS calibration.
J : Estimated value due to minor OC deviations or for tentatively identified compound (no standard
spectra indicates compound present below contract detection limit,
R: Data unusable due to major QC deviations.
— : Compound was below the quantifiable limit


but greater than zero.


available) or mass



*: Concentrations reported on a dry weight basis.
369

-------
    1.3.2.2  Lagoon 9 Sludge

Sludge from Lagoon 9 was  tested  to  determine LTTD performance on worst-case
sludge and to identify potential limitations  of the treatment process for the
L.O.U.

Performance Evaluation

Table 7 presents the cleanup standards  and pre- and post-treatment concentra-
tions for the compounds of concern  for Lagoon 9 sludge.  Aniline was detected
in  the  pre-treatment  and  post-treatment  samples  below  the  10*^ cleanup
standards but above the 10"*>  cleanup standard.  Therefore, sufficient removal
occurred only to attain a 10"^ risk  cleanup.  Azobenzene was also detected in
the pre-treatment samples below the 10"^ but above the 10**> cleanup standards.
The post-treatment concentration was less than both cleanup standards.  There-
fore, an acceptable level of  removal  has  been achieved.  Benzene concentra-
tions in the pre-treatment sample were greater than both cleanup standards and
were less than the cleanup  standards in the post-treatment sample, indicating
an adequate degree of  removal.    For  benzidine, the pre-treatment and post-
treatment concentrations  were  above  the  cleanup  standards;  therefore, an
acceptable removal level was not achieved for benzidine during this test.  DCB
pre-treatment concentrations exceeded both  cleanup standards; the post-treat-
ment concentration was below the standards.  Therefore, acceptable removal was
achieved for DCB.  In summary, an acceptable removal level was attained in the
worst case sludge  for  azobenzene,  benzene,  and  DCB.    However, LTTD only
achieved a 10'^ level removal  for  aniline  and did not achieve an acceptable
removal level for DCB.  Therefore, LTTD is not an effective process option for
remediating Lagoon 9 sludge.

Process Residuals Evaluation

The organic  compound  and  metal  analyses  for  post-treatment  residuals of
Lagoon 9 sludge are shown in  Table 8.    Pre-treatment compounds are shown to
indicate compounds originally present in  the  sample.  Benzene is distributed
evenly between the condensate and  the  phase  separator liquid.  Overall, the
semivolatile organic compounds appeared to be collected in the phase separator
liquid.  However, for  the  compounds  of  concern, aniline and azobenzene are
present in higher concentrations  in  the  condensate.   Benzidine and DCB are
more concentrated in the  phase  separator  liquid,  as  expected.  Both waste
streams contained very high levels  of VOCs and semi-volatile compounds, indi-
cating the need for further treatment.    Metals may also need treatment prior
to discharge.  Calcium  and  magnesium  treatment  may  be required to prevent
scaling of equipment.

    1.3.2.3  Lagoon 8 and 10 Sludge

A composite of  sludge  from  Lagoons 8  and  10  was  tested  to evaluate the
effectiveness of LTTD to remediate the average contaminated sludge that may be
encountered during remediation of the Bofors Site.
                                     370

-------
Units: ug/kg
                                                   TABLE  7
                                   LOW TEMPERATURE THERMAL DESORPTION
                                     ATTAINMENT OF CLEANUP STANDARDS
                             FOR COMPOUNDS OF CONCERN FOR LAGOON 9 SLUDGE
                                                  Bofors Site
                                              Muskegon, Michigan
                                        Cleanup Standards*
         Compound
1CT Risk
10"6 Risk
Pre-Treatment*
Post-Treatment*
Aniline
Azobenzene
Benzene
Benzidine
3,3'- Dichlorobenzidine
  1,100,000
920,000,000
    230,000
       220
  4,600,000
   11,000
9,200,000
   2,300
       2.2
   46,000
      52,000 J
  37,000,000 DJ
     980,000
   1,000,000 J
  11,000,000 DJ
      43,300
     326,700
        330
       3,700
      18,400
LEGEND:
D :      Value reported from a diluted sample aliquot in order to stay within the linear calibration of GC/MS.
J :      Estimated value due to minor QC deviations or for tentatively identified compound (no standard available) or mass
        spectra indicates compound present below contract detection limit, but greater than zero.
Cleanup standards are based on remediation for the groundwater ingestion exposure route.
*:       Cleanup standards and sludge concentrations are on a dry weight basis.
                                                   371

-------
                                                   TABLE 8
                                   LOW TEMPERATURE THERMAL DESORPTION
                             CHEMICAL CHARACTERIZATION OF PROCESS RESIDUALS
                                            FOR LAGOON 9 SLUDGE
                                                  Bofors Site
                                              Muskegon, Michigan
                                                                                Post-Treatment Residuals
                     Analyte
Volatile Organic Compounds
Acetaldehyde
Acetone
Acetonitrile
Benzene
2-Butanone
Chlorobenzene
1,4-Dioxane
Ethylbenzene
2-Hexanone
Methylene Chloride
Styrene
Xylene (Total)
Vinyl Chloride
Unknown
Semi-Volatile Organic Compounds
Aniline
Azobenzene
Benzidine
Benzole Acid
2-Chloroaniline
2-Chlorophenol
3,3'-Dichlorobenzldine
2-Methylphenol
4-Methylphenol
Phenol
Metals
Aluminum
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Manganese
Nickel
Pot8",sium
Silver
Sodium
Vanadium
Zinc
                                            Pre-Treatment*
                                               (ug/kg)
                                                  MA

                                                  NA
                                             980,000
                                                  NA
                                                6,000
                                               52,000
                                           37,000,000 DJ
                                            1,000,000 J

                                              600,000 J

                                           11,000,000 DJ
                                            2,720,000
                                                4,1006
                                               25,600 B

                                               12,400
                                           45,700,000 J
                                               28,700 J
                                              120,000 J
                                           11,100,OOOJ
                                              170,000 J
                                            4,910,000
                                              158,000
                                               16,900
                                              271,0008

                                            3,570,000
                                                9,7008
                                           83,300,000
 Condensate
    (ug/l)
    6,600 J
    9,300 E
      370 BJ
      738
    2,162
       50
      188
       14 BJ
       62
       50

       45 J
6,900,000
  320,000
 1,200,000

     460
    3,3106
       10.5
    7,350
      174
        6
      4166
        7.26
       27.6
      7766

      254
Phase Separator
    Liquid
     (ug/l)
     3,500 J
     3,890 E
     1.400BJ
      730
      710 J
        6
      160 J
       55

       59 BJ
       53
       20
       36 BJ
 1,070,000
  210,000
   64,000
   21,000
  444,000 J
     4,000
   20,700
     8,900
   25,300
   84,000 E
     2,230
        4.8 B
       56.2 B
        2.76
        8.4 B
   206,000
       17.4
      123
     7,170
      119
    35,900
      191
       29.6 B
     2,420 B

    42,400
       14.3B
    65,000
Legend:
B:
D:
E:
J:
NA:
Compound was also detected in associated laboratory blank.
Value reported from a diluted sample aliquot in order to stay within linear calibration of GC/MS.
Concentration exceeded the linea- range of GC/MS calibration.
Estimated value due to minor QC deviations or for tentatively identified compound (no standard available) or mass
spectra indicates compound present below contract detection limit, but greater than zero.
Compound was below the quantifiable limit.
Compound was not analyzed for.
Concentrations reported on a dry weight basis.
                                                     372

-------
Performance Evaluation

Cleanup standards and pre- and post-treatment concentrations for the compounds
of concern are presented in Table 9.  Aniline and azobenzene concentrations in
the pre-treatment sample are below  the  10*^  and the 10*6 cleanup standards;
therefore, no removal of  these  contaminants  is  necessary.  Benzene was not
detected above the quantitation  limit  in  the pre-treatment sample; however,
benzene was detected in  the  post-treatment  sample at a concentration higher
than the pre-treatment quantitation limit.    The  presence of benzene is most
likely a result of cross-contamination  during the treatability study testing.
Therefore, it is assumed that no removal  of benzene is needed.  Benzidine and
DCB are detected in  the  pre-treatment  sample  above  the  10'^ and the 10"^
cleanup standards, but  are  not  detected  in  the  treated  sample above the
quantitation limit.    The  quantitation  limit  is  greater  than the cleanup
standards.  However,  neither  benzidine  nor  DCB  are  detected  in the TCLP
extract above the quantitation limit; therefore, it is assumed that an accept-
able removal level has been  achieved  for  these compounds.  In summary, LTTD
appears to be effective in  removing  the  contaminants of concern in Lagoon 8
and 10 sludge.

Process Residuals Evaluation

Analysis of post-treatment  residuals  is  shown  in  Table 10.  Pre-treatment
compounds are shown to indicate  compounds  originally in the sample.  Overall
the VOCs were predominantly  in  the  condensate, and the semivolatile organic
compounds were predominantly in the  phase separator liquid.  However, aniline
and azobenzene were detected at  greater concentrations in the condensate than
the phase separator liquid; while benzidine  and DCB were only detected in the
phase separator liquid.   Several  volatile  organic  compounds which were not
detected in the pre-treatment sample  were  found  in the waste streams.  This
may be indicative of different  detection  limits, or problems associated with
cross-contamination.  Both condensate  and  the phase separator residuals will
require further treatment for organic compounds prior to discharge.

Generally, the metals which were  detected  in the residual waste streams par-
titioned to the phase separator liquids, with the exception of copper.  Treat-
ment  may  be  required  for  calcium  and  magnesium  to  prevent  scaling of
equipment.

1.3.3  Conclusions

The bench-scale testing of LTTD for  these  samples indicates that LTTD may be
effective for treating Lagoon 3 soil and  Lagoon 8 and 10 sludge.  The testing
indicates that LTTD is not effective for Lagoon 9 sludge.

VOCs generally appeared to partition  into  the condensate phase, as expected.
However, the semi-volatile compounds,  including aniline and azobenzene, often
were concentrated in the condensate.   This  may  be due to the relative vola-
tility of  these  compounds  compared  to  the  other semi-volatile compounds.
Benzidine and DCB consistently  concentrated  into the phase separator liquid,
as expected.
                                       373

-------
                                                   TABLE  9
                                   LOW TEMPERATURE THERMAL DESORPTION
                                     ATTAINMENT OF CLEANUP STANDARDS
                          FOR COMPOUNDS OF CONCERN FOR LAGOON 8 AND 10 SLUDGE
                                                  Bofors Site
                                              Muskegon, Michigan
                          Cleanup Standards*
Compound
10"4 Risk
(ug/kg)
10"6 Risk
(ug/kg)
Pre-Treatment*
(ug/kg)
Post-Treatment*
(ug/kg)
Quantitation
Limit
(ug/kg)
Post-Treatment
TCLP
(ug/l)
Aniline
Azobenzene
Benzene
Benzidine
3,3'- Dichlorobenzidine
                         250,000       2,500         1,700             230 BJ
                       2,800,000      28,000         5,700             420 J
                          47,000         470            —          13,000 ESJ**
                              5.4         0.054     13,000             —
                         440,000       4,400      2,200,0000           —
3,300
1,300
LEGEND:
B:
D:
E:
J:

S:
        Compound was also detected in the associated laboratory blank.
        Value reported from a diluted sample aliquot in order to stay within the linear calibration of GC/MS.
        Concentration exceeded the linear range of GC/MS calibration.
        Estimated value due to minor QC  deviations or for tentatively identified compound (no standard available) or mass
        spectra indicates compound present below contract detection limit, but greater than zero.
         Instrument response was saturated, result represents an estimated minimum concentration present.
— :      Compound was not detected.
**:       Represents probable cross-contamination during treatability testing.
Cleanup standards are based on remediation for the groundwater exposure route.
TCLP:    Toxicity characteristic leaching procedure.
*:        Cleanup standards and sludge concentrations are on a dry weight basis.
                                                    374

-------
TABLE 10
LOW TEMPERATURE THERMAL DESORPTION
CHEMICAL CHARACTERIZATION OF PROCESS RESIDUALS
FOR LAGOON 8 AND 10 SLUDGE


Bofors Site
Muskegon, Michigan




Post-Treatment Residuals


Analyte
Volatile Organic Compounds
Acetaldehyde
Acetone
Acetonitrile
Benzene
2-Butanone
Hexane
Methylcyclopentane
Methylene Chloride
Toluene
Semi-Volatile Organic Compounds
Aniline
Azobenzene
Benzidine
2-Chloroaniiine
4-Chloroaniline
2-Chlorophenol
1 ,2-Dichlorobenzene
3,3'-Dichlorobenzidine
Dichlorobenzidine Isomer
Methanone, (3-Chlorophenyl) (4-Chlorophenyl)
2-Methylphenol
4-Methylphenol
Phenol
Total Unknowns
Metals
Aluminum
Arsenic
Barium
Beryllium
Cadmium
Calcium
Copper
Iron
Lead
Magnesium
Manganese
Nickel
Potassium
Sodium
Vanadium
Zinc
Legend:
B : Compound was also detected in associated

Pre-Treatment
(ug/kg)

_
—
—
—
30 BJ
—
—
210 BJ
4BJ

1,700
5,700
13,000
9,400
—
—
—
2,200,000 D
—
88,000 J
—
—
—
216,000 J

3,930,000
4,500
53,200 B
—
1,6006
241,000,000
18,200
1,330,000
11,200
2,470,000
32,900
—
79,600 B
284,000 B
4,0006
1,190,000

laboratory blank.
D : Value reported from a diluted sample aliquot in order to stay within

* Condensate
(Ufl/l)

1,500 J
6.557 E
1,300 J
20
249 E
—
—
22 BJ
19

62,000
28,100
—
85,600 J
2,600
20
—
—
—
—
—
30
210
—

_
—
—
—
—
2,0108
25,000
82.9 B
2.6 B
269 B
—
—
—
669 B
—
384


linear calibration of GC/MS.
Phase Separator
Liquid
(us/l)

20 J
71 J
98 J
11
—
20 J
8J
2BJ
66

8,000
15,800
32,000
3,700 J
670
—
30
21,000
7,000 J
16,000 J
30
—
240
22,900 J

4,650
3.2 B
80.86
1.8B
7.7
329,000
553
3,530
51
16,700
68.6
25 B
2,060 B
9,520
8.76
14,700



E: Concentration exceeded the linear range of GC/MS calibration.
J : Estimated value due to minor QC deviations or for tentatively identified compound (no standard
spectra indicates compound present below contract detection limit
— : Compound was below the quantifiable limit

, but greater than zero.

available) or mass


*: Concentrations reported on a dry weight basis.
375

-------
1.4  SOLIDIFICATION/STABILIZATION

On-site solidification/stabilization processes  involve  mixing excavated con-
taminated materials  with  proportional  amounts  of  treatment  reagents that
combine physically and/or chemically  with  contaminated materials to decrease
the mobility of waste constituents.    Depending  upon  the amount and type of
reagent added, the end product may be  a standing'monolithic solid or may have
a crumbly consistency.  Mixing  of  wastes  with treatment reagents is usually
performed in batch plants.  Mixtures may be conveyed to concrete mixing tanks,
where hydration water  is  added  if  needed,  and  thoroughly blended.  Then,
treated materials are generally placed in confining pits on-site for curing.

Portland cement and pozzolanic materials such as fly ash, ground blast furnace
slag, and cement kiln dust are  widely used as immobilization reagents because
of their availability and  effectiveness  in  binding contaminants to minimize
leaching.  These materials may be  used  singly or in combination. A number of
proprietary additives, including polymers  and absorbents, have been developed
for use with cement and  pozzolanic  materials to improve the physical charac-
teristics and  decrease  the  leaching  losses  from  the resulting solidified
material.

1.4.1  Treatability Experimental Procedure

The following samples  were  sent  to  Enreco  for  use in the solidification/
stabilization treatability study:   Lagoon 3  soil (worst-case soil); Lagoon 9
sludge (worst-case  sludge);  and  Lagoon 8  and 10  composite sludge (average
sludge).

Enreco performed a literature search to  identify reagents which could be used
to solidify/stabilize contaminated soil and/or  sludge from the Bofors Site at
this  site.    Based  on  findings  of  the  literature  search  and  previous
experience, three solidification/stabilization processes were identified:

    1.   Solidification and stabilization using Portland cement, fluidized bed
         material and Quickline;

    2.   Adsorption,  solidification,  and  stabilization  using organophyllic
         clays, high  carbon  coke  fines,  carbide  lime  and combinations of
         reagents;

    3.   Polymerization,  solidification,  and  stabilization  using  monomers
         polymerized by catalysts.

The third process was eliminated  during  the literature search because it was
determined to be uneconomical.

The testing program involved a  three-round iterative process to evaluate each
of the two remaining processes  to determine the optimal formulation to achieve
cleanup  standards.  Samples containing 200 grams of contaminated material were
mixed with varying ratios  of   the  fixation  reagents  and were then cured in
airtight containers.  Reagents  that  produced favorable results were selected
for further testing in succeeding  rounds.    Each additional round of testing
was used to refine the mix designs.
                                      376

-------
Round 1  mixing  and  testing  consisted  of  screening  techniques  with  the
reagents.    This  round  was  used  to  evaluate  the  physical  and chemical
properties of the  mixtures.    The  samples  were  tested for unconfined com-
pressive strength using  a  pocket  penetrometer,  and  for density and volume
expansion.

Round 2  mixing  and  testing  was   performed  with  the  preferred  reagents
determined in Round 1.    The  mixtures  were  prepared  to refine the reagent
formulations and  optimize  material  usage.    Some  additional mixtures were
prepared to narrow the field  of  potential  reagents.   After a review of the
results and an economic  analysis,  mix  designs  were  selected for the final
round.

Round 3 used the ratios determined in  Round 2 to prepare sufficient volume of
material which was then refrigerated and cured for seven days.  The stabilized
material was then distributed to laboratories to undergo TCLP extraction.

1.4.2  Experimental Results

As discussed  previously,  the  analytical  results  from  the solidification/
stabilization treatability  study  were  evaluated  differently  than the soil
washing and LTTD treatability  studies.    Soil  washing  and LTTD are removal
options that extract the contaminants from  the soil and/or sludge.  Solidifi-
cation/stabilization  does  not  extract  the  contaminants  from  soil and/or
sludge, but entrains the contaminants in the stabilized product so that leach-
ability is decreased.  Therefore,  concentrations  in the TCLP extracts of the
pre- and stabilized soil and/or  sludge  samples were compared to evaluate the
reduction in mobility.    If  the  compounds  of  concern  are detected in the
stabilized material TCLP  extracts,  this  may  indicate  that the stabilizing
reagents are not reducing the mobility of these compounds.

    1.4.2.1  Lagoon 3 Soil

The recommended stabilization reagent mix for final treatment of Lagoon 3 soil
to reduce mobility of  the  compounds  of  concern  and  be cost effective was
60 percent fluidized bed material and 10 percent carbide lime.

Performance Evaluation

Unstabilized and stabilized TCLP  extract concentrations and percent reduction
in concentration are presented in Table 11.     Aniline was  not detected in the
unstabilized extract but was detected in  the stabilized extract.  This result
may be due to laboratory  contamination or differences in quantitation limits.
There was  a  greater  than  90 percent  reduction  in  TCLP concentration for
azobenzene between the unstabilized  and  stabilized  samples.  However,  there
was only a 57 percent reduction  for  benzidine and a 79 percent reduction for
DCB.    In  summary,  azobenzene,  benzidine  and  DCS  were  detected  in the
stabilized TCLP extract.    Concentration  reductions  were achieved;  however,
significant  amounts  of  these  compounds  remained  in  the  stabilized TCLP
extract; indicating that the  mobility  of  azobenzene,  benzidine,  and DCB was
not sufficiently reduced by stabilization.
                                      377

-------
co
•sj
00
TABLE 11

SOLIDIFICATION/STABILIZATION
COMPARISON OF TCLP EXTRACT CONCENTRATIONS
Bofors Site
Muskegon, Michigan
Units: ug/l
Lagoon 3 Soil Lagoon
Pre- Post- Reduction in Pre- Po
Treatment Treatment Concentra- Treatment Treat
Compound TCLP8 TCLP8 tion(%) TCLP8 TC
Aniline — 110 * 2,200 €
Azobenzene 9.400 8800 90.64 1,500 1,1
Benzene 5J 24 * NA 9,C
Benzidine 21,000 9,000 D 57.14 16,000 5,7
3, 3'-Dichlorobenzidine 14,000 2,9000 79.29 2,300 1.E
LEGEND:
a : Concentration in TCLP extract from stabilized sample.



9 Sludge Lagoon 8 & 10 Sludge
st- Reduction in Pre- Post- Reduction in
ment Concentra- Treatment Treatment Concentra-
LP" tion(%) TCLPa TCLP8 tion(%)
>30 71.36 53 J — - *
00 26.67 160J 28 J 82.50
IOO * — 3J *
rOO 64 44 .... _.„ *
(00 J 34.78 3,500 3,000 14.29


D : Value reported from diluted sample aliquot in order to stay within the linear calibration of GC/MS.
J : Estimated value due to minor QC deviations or for tentatively identified compound (no standard available) or mass spectra indicates compound present below contract
detection limit, but greater than zero.
B : Compound was also detected in associated laboratory blank.
TCLP : Toxicity characteristic leaching procedure.
— : Compound was below quantifiable limit or no associated risk.
*: Reduction in concentration cannot be calculated or is not applicable.
NA: Compound not analyzed.







-------
    1.4.2.2  Lagoon 9 Sludge

The recommended stabilization reagent  mix  for Lagoon 9 sludge was 60 percent
fluidized bed material and 2 percent organophillic clay.

Performance Evaluation

As shown in Table 11, aniline, azobenzene, benzidine, and DCB were detected in
both the unstabilized and stabilized  TCLP  samples.  Benzene was not analyzed
in the unstabilized TCLP sample due to  loss of sample.  Benzene is present in
the stabilized TCLP sample.    Percent  reductions in concentration range from
34 percent to 71 percent; however, significant amounts of all six compounds of
concern were still detected in  the  stabilized  TCLP extract.  This indicates
that  the  mobility  of  the   compounds   was  not  sufficiently  reduced  by
stabilization.

    1.4.2.3  Lagoon 8 and 10 Sludge

The recommended formulation  of  stabilization  reagents  for  Lagoon 8 and 10
sludge was 32 percent fluidized bed material and 5 percent carbide lime.

Performance Evaluation

As shown in Table 11, aniline  was  detected  in the unstabilized TCLP extract
but was not detected  in  the  stabilized  TCLP extract, indicating sufficient
reduction in mobility.  Benzene  was  not  detected in the unstabilized sludge
but was detected in  the  stabilized  sludge  at a low concentration, possibly
indicating laboratory contamination.    Benzidine  was  not detected in either
sample.  Azobenzene and DCB achieved some reduction in concentration; however,
significant  amounts  of  these  compounds  remained  in  the  stabilized TCLP
extract, indicating that their mobility was not sufficiently reduced.

1.4.3  Conclusions

Different formulations were developed  to  combine  with sample materials from
Lagoon 3 soil, Lagoon 9  sludge,  and  Lagoon 8  and  10  sludge.  The reagent
common to all three formulations was fluidized bed material.

A comparison of unstabilized and stabilized TCLP extract concentrations showed
that, while some compounds  achieved sufficient concentration reductions, most
compounds were still present  in  the  stabilized  TCLP extract in significant
concentrations.  This indicates that  stabilization  is not an effective tech-
nology for use on any of the samples tested.

1.5  UNCERTAINTY OF TREATABILITY STUDIES

Treatability studies were  conducted  to  obtain  additional data necessary to
adequately evaluate remediation alternatives  during  the  RI/FS.  The results
will be used for costing and determining performance uncertainties, as well as
to eliminate alternatives which are not technically feasible for site cleanup.
Bench scale tests may simulate  the  full  scale processes, but the results of
these studies should be  used  with  discretion  to develop cost estimates and
design parameters.
                                      379

-------
Soil washing  is  an  innovative  technology  with  respect  to Superfund site
remediation.  The soil washing  process  utilizes existing technology which is
common to the metallurgical and  mining  industries.    The soil washing  unit
operations are standard, yet the process itself  is unique to the site and for
the treatment objectives.    Unit  operations  are  typically  much larger for
industrial uses than those recommended for the L.O.U.  Additional problems and
costs may be incurred in designing a  full scale process because of the use of
smaller, specialty equipment in the bench scale studies.  Another concern with
the soil washing process involves the volume  of water used to clean the soil.
The accuracy of this estimate  from  bench  scale  tests is dependent upon the
mixing intensity, chemicals  used,  and  contaminant  concentrations.   If the
mixing intensity is greater for the  bench  scale process than for full scale,
the cleaning action will be greater and  less wash water will be needed.  This
uncertainty also relates to the type of  detergents used to scrub the soil and
their effectiveness under lower agitation.

For LTTD, the  initial  concentration  of  semivolatile organic compounds will
drastically affect the other process streams  associated with the system.  The
volumes and  contents  of  the  condensate  and  phase  separator  liquids are
directly related to the  initial  moisture  content  and concentrations of the
volatiles and semi-volatiles.  The  bench  scale  test uses only a kilogram of
material; therefore, these waste streams  may  not be accurately simulated due
to their low volumes.    The  bench  scale  results may not accurately predict
problems which may  occur  in  the  condensate  and  phase separator treatment
process.  The treated soil fraction  should represent a reasonable estimate of
the final product  in  the  full  scale  process  since removals are primarily
dependent upon kiln temperature and retention time for a given initial concen-
tration of contaminants.

Variations in contaminant concentrations will  also  affect the results of the
solidification/stabilization  bench  scale  tests,   but  the  major  area  of
uncertainty is the ability of the  bench  scale tests to simulate the addition
of the reagents during the full scale  process and whether TCLP is an accurate
predictor of long-term performance and reduction in mobility.


RP/EPASSTS/AA1
                                     380

-------
                                BIBLIOGRAPHY


Water Pollution Control Federation,  Wastewater Treatment Plant Design, Manual
of Practice No. 8, 1977.

Remedial Investigation Report for the Bofors Site, MDNR, February 1990.

Draft Phase  II  Remedial  Investigation  Report  for  the  Bofors Site, MDNR,
March 1990.

Draft Feasibility Study Report  for  the  Lagoon  Operable  Unit at the Bofors
Site, MDNR, May 1990.


RP/EPASSTS/AA1
                                      381

-------
                CONCEPTUAL COST EVALUATION
               OF VOLATILE ORGANIC COMPOUND
             TREATMENT BY ADVANCED OXIDATION
                       GLENN J. MAYER, P.E.
                          CH2M HILL, Inc.
                     2510 Red Hill Avenue, Suite A
                      Santa Ana, California 92705
                       WILLIAM D. BELLAMY
                          CH2M HILL, Inc.
                           P.O. Box 22508
                       Denver, Colorado 80222
                     NEIL ZIEMBA/LEE A. OTIS
                         U.S. EPA Region DC
                          1235 Mission Street
                     San Francisco, California 94103
Presented at:      Second Forum on Innovative Hazardous Waste Treatment
                Technologies: Domestic and International, May 15-17, 1990,
                Philadelphia, Pennsylvania
LAO22033U31 007.51
                              382

-------
                   SUMMARY AND CONCLUSION
Cost estimates based  on conceptual designs of groundwater treatment facilities for
volatile organic compounds (VOCs)  using three different treatment technologies have
been developed for a local area of  the San Gabriel Basin. These technologies are:
1) packed tower air stripping  with  vapor phase carbon adsorption, 2) liquid phase
carbon adsorption, and 3) ozone/peroxide  advanced oxidation. Traditional design data
and costs are used in evaluating the first two technologies.  Bench-scale testing on site-
specific groundwater is used for the development of conceptual treatment plant design
and cost estimates for the advanced  oxidation process.

Using the same plant sizing, influent  water quality, and target treatment levels for each
treatment technology, a cost comparison  has been prepared.  Costs for two  influent
conditions are evaluated:  the probable design case, representing the conditions most
likely to be encountered at the  treatment plant given data currently available; and, the
maximum credible deviation case representing the worst conditions that may  credibly
be encountered during the next 30 years.  The results  of the cost comparison are as
follows:
Treatment Technology
Air Stripping
Carbon Adsorption
Advanced Oxidation
Cost in $/l,000
gallons
Probable
Design
Case
0.19
0.43
0.25
Maximum
Credible
Deviation
Case
0.35
__a
0.25
*Liquid phase carbon adsorption is not eco-
nomically feasible in the maximum credible
deviation case due to high carbon loading.
Considering the costs of treatment alone,  advanced oxidation appears to be more
economical in the maximum credible deviation case.  However,  a number of other
criteria, including reliability, ease of operation and maintenance, expected lifetime, and
presence of treatment process hazards, must also be considered in evaluating treatment
options.
                                    383
LAO22033\231 008.51

-------
                              BACKGROUND

The  170-square-mile San Gabriel Basin is located in the northeastern portion of Los
Angeles County. The water-bearing formations of the basin consist of gravel, boulders,
and coarse sediments ranging to a thickness of over 4,000 feet or more. Groundwater
movement is generally from the perimeter of the basin toward the pumping centers in
the basin and then toward the Whittier Narrows, which is the only identified subsurface
discharge from the basin. Figure 1 shows the location of the San Gabriel Basin.

In recent years, over 200,000 acre-feet of water have been extracted annually from the
basin by more than 45 water purveyors.  Groundwater  provides water  supply  for
approximately 90 percent of the basin population of over 1  million people.

In December 1979,  VOC contamination  of the groundwater was discovered.  Further
sampling by the Los Angeles Regional Water Quality Control Board and the California
Department of Health Services confirmed contamination over a wide area. EPA pro-
posed in September 1983, and subsequently listed in May 1984, that four general areas
of groundwater contamination be included on the National Priority List  (NPL).   In
February 1986, EPA initiated remedial investigation/feasibility study (RI/FS)  activities in
a combined effort for the entire San Gabriel Basin.

A primary objective  of the RI activities was to develop appropriate treatment technolo-
gies  for contaminated groundwater.  During  the  review and development of techno-
logies for treatment of  synthetic organics, especially  VOCs, it was  determined that
activated carbon, air stripping, and oxidation were all possible treatment  techniques.
Carbon adsorption and air stripping have been accepted and are  proven technologies
for many of the organic constituents  present in  the basin. Oxidation, on the other
hand, has only recently been proposed  as a cost-effective treatment technology for
drinking water.

Since this technology is relatively new and unproven and it has not yet been extensively
used for large municipal water treatment systems, it was decided that additional inves-
tigation into the technology was appropriate.  A literature review was conducted, and
recent work in the field, most notably by Dr. William Glaze of the University of North
Carolina, suggests that some VOCs (e.g., trichloroethene [TCE] and tetrachloroethene
[PCE])  can  be oxidized at  an  equivalent cost  to air stripping  with off-gas  carbon
adsorption.  Oxidation can result in destruction of the VOCs with minimal  or no
recognized by-products in the water and no significant VOCs released to the air.

Reaction kinetics for traditional ozonation have not been considered favorable for the
oxidation of most chlorinated organics in potable water systems. The inefficient oxida-
tion of most VOCs by ozone (O3), relatively low contaminant concentrations, and large
water volumes  associated with VOCs and potable water treatment create excessive
capital and operating costs for a conventional ozone system.
                                   384
LAO22033N231 008.51

-------
CO
CO
en
                                                   RAYMOND
                                                    BASIN
         SAN  GABRIEL  MOUNTAINS
                                                            SAN  GABRIEL  BASIN
                                                      GROUNDWATEH
                                                          FLOW
                                                        DIRECTION
WHITTIER
NARROWS
                              CENTRAL
                                                                    ORANGE COUNTY
                                                                     GROUNDWATER
                                                                         BASIN
                                                                                                 NOT TO SCALE
                                                                                               FIGURE 1

                                                                                            LOCATION  MAP

-------
Developments in oxidation technology for aqueous  organics, known as "advanced
oxidation  processes" (AOPs), appear  more  feasible  for  the  treatment of  VOC
contaminated potable water sources.  AOPs involve the generation of highly reactive
free  radical intermediates from the decomposition of ozone or hydrogen peroxide
(H2O2). Organics are oxidized much faster by ozone decomposition by-products (e.g.,
free radicals) than by O3.  Consequently, AOPs are designed to promote the formation
of free radicals, thereby enhancing oxidation  of organics  to levels  practicable  for
drinking water treatment.

Formation of free  radicals by decomposition of ozone can be initiated by hydroxide
ions  (OH  "), hydrogen peroxide, ultraviolet (UV) light, and some transition metal ions.
The  AOP systems  currently receiving attention include Oj/high pH, O3/H2O2, O^V
light, and  H2O2/UV light.  On the basis of the literature review, the O3/H2O2 system
appears to be the most practical for potable water treatment when the organics
oxidized do not react with UV light.  Hence, it was decided to conduct a bench-scale
evaluation of the O3/H2O2 process for oxidizing VOCs in groundwater from the San
Gabriel Basin.

AOPs  can effectively oxidize VOCs.  This is a  potential advantage  over treatment
methods currently  designated  as best available technologies (BATs) for VOC removal
(packed tower aeration or granular activated carbon  [GAC] adsorption; EPA, 1979).
The  BATs transfer the contaminants from water to another phase or medium which, in
turn, has to be treated or disposed of.  Air stripping followed by gas  phase oxidation
has the potential to destroy the VOCs, resulting in the same benefits as the AOPs. Gas
phase oxidation was not evaluated in this study because, at the time of initiation of the
oxidation  study, gas phase oxidation had not been demonstrated adequately.

Ozone/peroxide bench-scale testing was undertaken as part  of the remedial investiga-
tion process of the San Gabriel Basin RI/FS.  The purpose of the bench-scale tests was
to evaluate  the applicability of  the  ozone/peroxide  process to  the  treatment  of
contaminated groundwater from the basin.  Results of the tests were used to predict
effectiveness of the technology and treatment costs.  This paper briefly highlights the
advanced oxidation process, describes the bench-scale testing methods and results of
the testing, and presents a conceptual cost comparison of the ozone/peroxide technol-
ogy with packed tower air stripping/vapor phase granular activated carbon adsorption
(PTAS/VGAC) and liquid phase granular activated carbon adsorption (LGAC).  Cost
estimates are expected to be accurate within plus 50 percent to minus 30 percent. Cost
is only one of nine factors considered in the selection of a remedial action at an NPL
site.
                                    386

 LAO22033\231 008.51

-------
                   ADVANCED OXIDATION PROCESS

Aqueous ozone can oxidize dissolved organics by two mechanisms:  1) direct reaction
or 2) decomposition of the ozone to intermediates (e.g., radicals), which in turn reacts
directly or  indirectly with  the  organics.  The direct  ozone  reactions are  highly
substrate-specific and relatively slow, occurring on the order of minutes. Hoigne and
Bader (1983) reported second order rate constants for the direct oxidation of organics
in the range or 1 to 1,000 per molar per second.  In contrast, reactions between ozone
decomposition intermediates and solutes are nonselective and fast, occurring in micro-
seconds.  For example, rate constants for oxidation of organics by the hydroxyl radical
(OH •) are typically 10* to 1010 per molar per second (Farhataziz and Ross, 1977).

Figure 2 illustrates the complex cyclic reactions involved in the  aqueous chemistry of
ozone. The primary free radicals formed by ozone decomposition may: 1) enter into
chain reactions  with other ozone  molecules,  perpetuating  (autocatalyzing)  ozone
decomposition; 2) react with organic solutes to form oxidized carbon constituents and
secondary radicals that also participate in chain reactions with ozone; or 3) react with
radical-scavenging solutes (carbonate, bicarbonate, certain organics), forming inefficient
oxidation species and effectively inhibiting radical chain reactions.

The observed oxidation of solutes in any ozonated system is a combination of direct
and radical reactions. The relative concentrations of the two types of reactants and the
importance of the two types of reactions are highly dependent on the  chemistry of the
water being treated.  This is due to the presence of naturally occurring  constituents that
promote or inhibit ozone decomposition and radical chain reactions.  Since the overall
rate at which organics  are oxidized is a function of both direct  and radical  reactions,
bench- or pilot-scale testing is currently required to evaluate the  effect of source water
chemistry on the efficiency of an advanced oxidation system.
                     BENCH-SCALE TEST PROGRAM

Bench-scale testing of groundwater collected from the San Gabriel Basin was under-
taken to determine the rate of destruction of select VOCs. The design of the bench-
scale testing system was based on the apparatus described by Glaze and Kang (1988).
A 75-liter, stainless steel, continuously mixed batch reactor (CMBR) was used for the
study.  The reactor was equipped with a dual-impeller mixing system; a 20-cm-diameter
ceramic ozone sparger; and sampling and viewing ports. A laboratory ozone generator
was  used to generate ozone from bottled air. Hydrogen peroxide solution was fed to
the  reactor through teflon  tubing  by a positive-displacement  chemical feed pump.
Figure 3 is a schematic of the bench-scale system.
                                    387

-------
OH*
                             Direct Oxidation of Substrate
                                   Slow, Selective

                                Cyclic Decomposition
                                                                               Products
                              ofOjby Secondary Radicals
                   Cyclic Decomposition
               /"     of O, by OH"
                      Radical Formation  ~.i.»   Radical Oxidation
                                     • Utl •
             COjandHCOj
                                              Fast, Nonselective
Secondary
 Radicals
                                                                           Products
                                                   Radical Consumption
       CO* and HCO*
                                        Figure 2
                             Cyclical Reactions of Ozone
                                 (from Ai eta etal., 1988)
                                            388

-------
                               Ozone
                               Monitor
              Vent -*•
CO
00
IO
Oa Purge

-«	'
          *
     Flow
   Controller
                                                    Vent
                                                     i
 Ozone
Generator
                                            CD
                                                                                                    Vent
                             Mixer
                             Drive
                                                      #      :
                                                                    -Vent
                                                                          a
                                 Flow
                               Controller
4
4
4
                                                                                                 k—t**—-O
                                                                                                           Peroxide
                                                                                                            Feed
                                                                                                            Pump
                                                                            Peroxide
                                                                            Reservoir
                                                                     Vent
                                                                                               ft\  Needle Valve

                                                                                               if I  3 - Way Valve
                                                              Figure 3
                                                 Schematic of Bench-Scale System

-------
Groundwaters for the AOP bench-scale experiments were collected from wells near the
Whittier Narrows (Southwest Suburban Water System) and near the town of Azusa
(Azusa Valley Water Company) in the San Gabriel Basin and shipped by air freight in
55-gallon polyethylene barrels to the laboratory.

Four target VOCs were used to evaluate the oxidative efficacy of the ozone/peroxide
process:

      •     Trichloroethene
      •     Tetrachloroethene
      •     Trans- 1,2-dichloroethene (t-l,2-DCE)
      •     Carbon tetrachloride (CTC)

Because groundwater concentrations of these VOCs in the wells sampled were well
below the desired testing concentration range,  the groundwaters were spiked using a
semisaturated stock solution of the four test organics.  Initial concentrations of the four
target compounds were either 50 or 500 micrograms per liter (ng/1).

A total of 15 laboratory tests were carried out under varying conditions of initial con-
centration, ozone mass feed rate, and ratio of peroxide to ozone molar dose. Table 1
summarizes the bench-scale test program.
                 RESULTS OF BENCH-SCALE TESTING

The results of bench-scale testing of the ozone/peroxide AOP for San Gabriel Basin
groundwaters are detailed in a summary technical memorandum prepared for EPA
(CH2M HILL, 1989).

Reaction rate constants for removal of target VOCs in this study are expressed in terms
of time as k, (sec ~1). In addition, empirical stoichiometric constants dependent on dose
have been calculated and are expressed as kD (liter/milligram [1/mg] ozone). Results
support the following conclusions:

      •      kt values for the removal of TCE and PCE in test runs utilizing an
             H2O2/O3 molar ratio of 0.50 equalled or exceeded rate constants from
             runs using other ratios (i.e., 0.25 or 1.00).  In addition, lower residual O3
             and H2O2 concentrations resulting from use of the 0.5 molar ratio suggest
             that it provides the most efficient use of oridants.
                                  390

LAO22Q33\231_008.51

-------
                               Table 1
                 Conditions for VOC Oxidation Runs
Test Number
2
3,14
4
5
6
11
8
7,15
9
10
13
12
Water Source
Suburban
Suburban
Suburban
Suburban
Suburban
Suburban
Azusa
Azusa
Azusa
Azusa
Azusa
Azusa
Target VOC
Concentrations
(ng/i)
500
500
500
500
50
50
500
500
500
500
50
50
Ozone Mass
Feed Rate
(mg/l-min)
0.205
0.205
0.205
0.051
0.205
0.051
0.205
0.205
0.205
0.051
0.205
0.051
HjOj/Oj
Dose Ratio
(Molar)
0.25
0.50
1.00
0.50
0.50
0.50
0.25
0.50
1.00
0.50
0.50
0.50
Note: Test 1 was a dry run using tap water.
LAO22033U31 001.51
391

-------
            As expected from process theory, a higher ozone mass feed rate yielded
            higher k, constants for PCE and TCE than a lower ozone mass feed rate.

            Groundwater from the two wells studied were very similar with respect to
            pH, alkalinity, hardness, and total dissolved solids (TDS).  As a result, k,
            values were not noticeably different for the two waters.

            k, values were essentially independent of initial VOC concentration.

            For conditions yielding the highest reaction rates  (O3 mass feed rate =
            0.205 mg/L-min; H2O^/O3 molar ratio = 0.50), k, values were:

                  PCE:  15-30 x 10" sec'1
                  TCE:  32-46 x 10-4 sec"1

            kj values for  t-l,2-DCE generally were too high to be quantified by the
            sampling protocol employed (i.e., concentration was below detection limits
            in the first sample taken after initiation of oxidant feed).  Based on semi-
            quantitative data, Iq values for t-l,2-DCE appeared to be two to  three
            times higher than those for TCE.

            k, and kD values for TCE and PCE are consistent with those reported in
            similar studies.

            Control experiments conducted to compare kt values for the H2O2/O3
            process with rate constants for gas stripping and oxidation by ozone alone
            suggest the following:

                  PCE:  Less than 10 percent of observed ozone/peroxide removal is
                  attributable to gas stripping.

                  TCE:  Approximately 5 percent of observed ozone/peroxide remo-
                  val is attributable to gas stripping.

                  t-l,2-DCE:  Gas stripping is negligible; the rate of oxidation  by O3
                  alone  is on the  same order of magnitude as that of oxidation by
                   ozone/peroxide.

                  CTC:  Virtually all observed ozone/peroxide removal is attributable
                  to gas stripping, not oxidation.
                                    392
LAO22033U31 008.51

-------
                  Note:   The relative contribution of gas stripping to  observed
                  removals of VOCs in this study is applicable only to a batch system;
                  stripping would be  expected  to  have less effect in a properly
                  designed full-scale, flow-through system.

            The test results indicated that a stoichiometric constant, kD, of 0.61/mg O3
            would be appropriate for estimates of ozone dose requirements.  The fol-
            lowing exponential equation would apply:

                  ln(C/C0) = kD [03]

            where:

                  C0 is initial contaminant  concentration, C is the final contaminant
                  concentration, kD is as defined above, and [O3] is the required
                  ozone dose in mg/1.
                  CONCEPTUAL COST EVALUATION

The above observations are used in the evaluation of the ozone/peroxide AOP for treat-
ment of groundwater in the Whittier Narrows area  of the San Gabriel Basin.  The
Whittier Narrows Operable Unit Feasibility  Study  (OUFS) has been  prepared to
evaluate the appropriate action to be taken to mitigate contaminant migration out of the
San Gabriel Basin and into the Central Basin,  an adjacent aquifer.

Preparation of a feasibility study for an EPA NPL (Superfund) site requires consideration
of the following nine criteria for each remedial alternative:

      •     Overall protection of human health and the environment

      •     Compliance with applicable or  relevant and appropriate regulations
            (ARARs)

      •     Long-term effectiveness and permanence

      •     Reduction in mobility,  toxicity, and volume

      •     Short-term protectiveness

      •     Implementability

      •     Cost
                                  393

LAO22033\231 008.51

-------
            State acceptance
      •      Community acceptance

A number of alternatives are presented in the Whittier Narrows OUFS, encompassing
not only treatment technologies, but also plume containment measures (i.e., where and
how much to pump) and treated-water distribution measures. The combined alternatives
are evaluated with respect to the above nine criteria. The remainder of this paper
presents only the  evaluation of the cost effectiveness of the following three treatment
technologies considered in the Whittier Narrows OUFS:

             PTAC/VGAC
             LGAC
      •      Ozone/peroxide advanced oxidation

BASIS OF THE COST EVALUATION

A common  set of input conditions  are  required to accurately  compare the costs
associated with each of the treatment technologies. Conditions that are anticipated for
remedial action at the Whittier Narrows area of the San Gabriel Basin are used in this
cost evaluation.

The Whittier Narrows OUFS has been prepared using a  new approach to feasibility
studies at EPA Superfund sites.  This approach, referred to as  the "observational"
approach, requires the evaluation of two conditions for each remedial alternative: 1)  the
probable design case, and 2) the maximum credible deviation case. The probable design
case represents the conditions  most likely to be encountered during remedial action,
given information  available at the time  of FS preparation.  The maximum credible devia-
tion case represents conditions that may credibly occur within the next 30 years.  An
evaluation of the  two conditions  allows the remedial alternatives  flexibility to handle
changing conditions. In the preparation of the Whittier Narrows OUFS, treatment costs
are evaluated for  both of these cases.

The following groundwater flow rates are used in the development of costs for the treat-
ment portion of the remedial alternatives.


Flow Rate
Peak (gpm)
Average (gpm)

Probable
Design
Case
14,600
7,200
Maximum
Credible
Deviation
Case
17,700
11,800
                                   394
LA022033\231 008.51

-------
The  concentration of contaminants in groundwater  requiring treatment as part of a
Whittier  Narrows remedial action is based on observed concentrations in both the
Whittier  Narrows and  upgradient  in  the San  Gabriel  Basin.  Probable design
concentrations and maximum credible deviations from the probable concentrations were
estimated using regional hydrogeological  modeling. Table 2 presents the influent
concentrations  used  in developing the  costs  for  treatment of Whittier  Narrows
groundwater, as well as the target treatment levels  (state and federal maximum
contaminant levels [MCLs]).

Using the estimated influent concentrations and groundwater flow rates, costs are esti-
mated  for the treatment of the groundwater extracted  in the Whittier Narrows  area.
Additional assumptions and input data needed to develop costs for each specific
technology are presented below.  Capital and operation and maintenance (O&M) cost
estimates are expressed  in terms of dollars per 1,000 gallons of treated water.

Capital cost estimates are based on the equipment required to meet the peak  flow.
Average  flow rates are used to calculate annual O&M costs. The annualized cost of
capital is calculated assuming a  30-year facility life  and 5 percent interest rate.  The
annual cost of the capital investment is added to the annual O&M cost to determine the
total annual treatment cost for the facility.  This value was divided by the total volume
of water  treated by the facility, assuming average flow conditions of 24 hours per day,
365 days per year.

COST OF AIR STRIPPING

A conceptual design layout is required prior to estimating costs of treating groundwater
by air stripping.  Because of the location of the project in the South Coast Air Basin of
Southern California, off-gas treatment is a required process for both probable design and
maximum credible deviation conditions.

Due to the large flow rate on which the sizing of a groundwater treatment plant is based,
a number of air strippers operated in parallel is required.  Using computer modeling of
the  air stripper, single units capable  of  meeting the target treatment levels  were
conceptually designed. The specifications for these units are outlined below.
                                    395

LAO22033\231 008.51

-------
                               Table 2
            Influent Water Quality and Treatment Criteria
                          (All Values in ppb)
Treatment Plant Influent


Contaminants
PCE
TCE
Carbon Tetrachloride
1,1,1-TCA
1,1-DCA
1,1-DCE
cjs-U-DCE
trans-l,2-DCE
U-DCA
UA2-TCA
Acetone
Benzene
Ethylbenzene
Methylene Chloride
Toluene
Total Xylenes
Vinyl Chloride
MEK
Bromoform
Dibromochloromethane
Chloroform
Freon 113
Probable
Design
Condition*
75
15
1.1
3 ntn
.08 ntn
3 ntn
2.7 ntn
20 ntn
0.6
0.25 ntn
3.3 ntn
1 ntn
1 ntn
3 ntn
1 ntn
1 ntn
0.4 ntn
5.5 ntn
2 ntn
2 ntn
6 ntn
3 ntn
Credible
Deviation
Condition1*
240
170
2.5
135 ntn
16 ntn
10
22 ntn
65 ntn
0.9
1
250
3.4
5.6 ntn
2,800
40 ntn
55 ntn
1
10 ntn
8 ntn
2.6 ntn
6.5 ntn
4 ntn


Treatment Level
5
5
0.5
200
.20
6


0.5
1
na
1
680
40
1,750
0.5
200




18,000

Source of
Regulation
Cal MCL
Cal MCL
Cal MCL
Cal MCL
CalAL
Cal MCL


Cal MCL
Cal MCL

Cal MCL
Cal MCL
CalAL
Cal MCL
Cal MCL
HEA ADV




CalAL
'Probable design conditions are based on migration of contaminants found in production wells
within the 30-year groundwater travel zone. Vertical and areal dilution have also been accounted
for.
''Maximum credible deviation conditions are the same as (a) with the exception that concentra-
tions used were from source investigation sampling within the 30-year travel zone.
ntn - no treatment necessary (the concentration is below the target treatment level)
na - no standard was available
Cal AL - California Action Levels
MCL - Maximum Contaminant Level
HEA ADV - Federal Health Advisories
                              396
LAO22033\231 002J1

-------

Diameter (ft)
Packed Depth (ft)
Water Flow Rate (gpm)
Air: Water Ratio
Air Flow Rate (cfm)
GAC Off-Gas
Treatment
Probable
Design
Case
12
20
3,000
20:1
8,000
Nonregenerable
Maximum
Credible
Deviation
Case
12
20
3,000
75:1
30,000
Regenerable
Table 3 summarizes the cost estimate for packed tower air stripping with VGAC, as
estimated for the Whittier Narrows  OUFS.  These  costs are  estimates based  on
preliminary design data and neglect such factors a facility location, delivery system
requirements, etc. The cost estimates are intended to represent order-of-magnitude
(+50 and -30 percent) estimates.  These costs are expected to change during detailed
remedial design.  Development of these conceptual costs is based on the  following
additional assumptions:

      •      Regenerable off-gas carbon units in the maximum credible deviation case
             are  Hastelloy-constructed units.

      •      Condensate will be shipped to a local treatment facility at a total cost of
             $1.00 per gallon in the maximum credible deviation case.

      •      Each   3,000-gpm  stripper   operating   at  full   capacity   requires
             1,090,000 pounds of steam per year to regenerate the vapor-phase carbon
             in the maximum credible deviation case.

      •      The regenerable carbon  will require replacement every 3 years, and the
             annual replacement cost  is based on replacing one-third of the full charge
             each year.

      •      Carbon replacement for  the probable design case is based on an average
             150  ng/1 VOC loading and  a 0.044 Ib VOC/lb GAC available capacity.
                                    397
LAO22033\231 008.51

-------
                                             Table 3
                       Treatment Costs Using Packed Tower Air Stripping*




CAPITAL COST ESTIMATES
Air Stripper (3,000 gpm/unit)
Air Supply fans

Tower Recirculation Pumps
Off-Gas Prebeaters
Off-Gas Carbon Units

Off-Gas Fan

Water Feed Pump
Equipment Cost Subtotal
Bid Contingencies @ 15%
Scope Contingencies @ 25%
Construction Total
Service During Construction @
10%
Land Acquisition
Total Implementation Cost
Engineering, l^gai, and
Administrative Costs @ 22%
TOTAL CAPITAL COST


Unit Cost (VUnit)
105,750
4,200 (Design)
11,800 (CredDev.)
4,100
1,600
93,750 (Design)
152,000 (Cred.Dev)
7,100 (Design)
11300 (Cred.Dev.)
13,800
_
_
_
_
__

300,000
M
_

-

Units Required


Probable Design
Sea
6ea

Sea
6ea
6ea

6ea

6ea
_
_
-
_
„

2 acres
_
_

-
Maximum
Credible
Deviation
6ea
7ea

Sea
18 ea
24 ea

7ea

7ea
—
_
-
—
_

2 acres
_
_

-
Treatment Facility
Cost(s)

Probable
Design
S 528,800
25,200

12300
9,600
562^00

42,600

82,800
$1,2453,800
189,600
316,000
$1,769,400
176,900

600,000
$2^46300
560,200

$3,106Ł00
Maximum
Credible
Deviation
$ 634400
82,600

12300
28300
3,648,000

82,600

96,600
$4,585,400
687,800
1,146,400
$6,419,600
642,000

600,000
$7,661,600
1,685,600

$9347,100
LA022033\231_003.51-16
                                             398

-------
                                                Table 3
                                             (Continued)




O&M COST ESTIMATES
Electrical Power
Carbon Replacement
Natural Gas
Boiler Fuel for Steam in Carbon
Regeneration
Water Sample Analysis
Air Sample Analysis
Condensate Disposal
Operating Labor
Maintenance @ 2% of Equipment
Costs
TOTAL O&M COSTS
Annualized Cost of Capital
Total Annul Costs
UNIT COST FOR TREATMENT
(V1.000 gal)a


Unit Cost (VUnit)
0.07
Z50
3.00
0.45
110
250
1.00
35
-
_
-
_
-

Unite Required


Probable Design
1,913,700 kW-hr
110,100 Ib
8,400 MM Btu
Ogal
72 ea
72 ea
Ogal
730 hr
-
_
_
_
-
jLffjMrliiingn
Credible
Deviation
7,633^00 kW-hr
82,400 Ib
37,700 MM Btu
4,400
84 ea
216 ea
523,400 gal
730 hr
-
_
_
_
-
Treatment Facility
Cost(s)

Probable
Design
$134,000
275,200
25,200
0
7,900
18,000
0
25,600
25300
$511,200
202300
$713,400
0.19
Mttdmnm
Credible
Deviation
$534300
206,000
113,100
2,000
9,200
54,000
523,400
25,600
91,700
$1^59300
608,500
$2,167,800
OJ5
aUnit treatment costs are estimates based on preliminary design data and neglects factors such as facility location, distribution
system requirements, etc. The cost estimates are intended to represent order-of-magnitude (+50 and -30 percent) estimates.
These costs will change during detailed design.
                                                399
LAO22033\231 003.51-17

-------
As shown in Table 3, the unit treatment cost is estimated at $0.19 per 1,000 gallons for
the probable design case and $0.35 per 1,000 gallons for the maximum credible deviation
case.  Treatment costs for the maximum credible deviation case are higher because of
higher capital and operating costs. Capital costs are higher because corrosion-resistant
Hastelloy carbon vessels are needed for the regenerable GAC system.  Operating costs
are higher primarily because of higher power costs and condensate disposal costs. Power
costs are higher because of the increase in airrwater ratio from 20:1  to 75:1.

COST OF LGAC

A conceptual design layout for LGAC  was  developed prior to estimating  costs for
treatment.

Based on the probable design and maximum credible deviation influent concentrations
and the required removal efficiencies to achieve target treatment levels, the  following
design criteria were developed.

Diameter (ft)
GAC depth (ft)
Empty bed contact time (minutes)
Liquid Loading (gpm/ft2)
Probable
Design
Case
12
8
6
10
Maximum
Credible
Deviation
Case
—
~
—
—
The use of LGAC in the maximum credible deviation case is unrealistic because of the
very large carbon replacement requirements.  In this case, air stripping was assumed.
The design criteria and costs for air stripping were discussed above.

A liquid loading rate of 10 gpm per square foot of GAC and a bed diameter of 12 feet
results in a maximum hydraulic capacity of 1,130 gpm per carbon bed. To achieve the
required treatment capacity of 14,600 gpm, a total of 14 carbon beds are required. The
cost estimate for carbon treatment assumes a system of two carbon beds in series to
avoid  breakthrough.  Additionally, one extra set of beds is included to accommodate
shutdowns. Therefore, treatment of Whittier Narrows groundwater with LGAC would
require a total of 30 carbon beds.
                                   400
LA022033\231 008.51

-------
Table 4 summarizes the cost estimate for LGAC, as estimated for the Whittier Narrows
OUFS.  These costs are estimates based on preliminary design  data and neglect such
factors as facility location, delivery system requirements, etc.  The cost estimates are
intended to represent order-of-magnitude (+50 and -30 percent) estimates. These costs
are expected to change during detailed remedial design.  As shown in Table 4, the cost
of treating Whittier Narrows groundwater by LGAC in the  probable design case is
estimated at $0.43 per 1,000 gallons treated.

COST OF O3/H202 ADVANCED OXIDATION

Development of costs for ozone/peroxide advanced oxidation treatment of the Whittier
Narrows  groundwater is more  difficult than the traditional treatment technologies
described above because of the lack of data on implementation costs.  Significant
assumptions were necessary to develop a cost estimate for the OUFS.  The following
conditions were used to develop the costs:

      •      The  reactor configuration  is  a flow-through  system  constructed  of
             concrete.  Dimensions of the  reactor were based on a residence time
             within the reactor of 5.5  minutes. This residence time is an assumed value
             which must be verified during pilot testing.  The reactor was designed to
             consist  of four  individual cells with  ozone  and peroxide feed in the first
             three cells.

      •      Three reactors operating at 50 percent capacity were assumed to allow for
             operating redundancy and flexibility in operating with the new technology.

      •      Ozone  and peroxide chemical  usage were determined using conditions
             determined during the bench-scale testing described earlier. The following
             conditions were used:

                   peroxide/ozone molar ratio = 0.5

                   empirical stoichiometric constant = 0.61/mg O3

      •      Electrical usage for generation of O3 from air =  12 kW-hr/pound Oj, cost
             of electricity = $0.07/kW-hr, and cost of peroxide  = $1.10/pound.
                                    401

LAO22033V231 008.51

-------
                                            Table 4
                    Treatment Costs Using Liquid-Phase Carbon Adsorption"'1'

CAPITAL COST ESTIMATES
Carton Adsorption Vessels
Chlorine Treatment System
Inlet Water Pumps
Air Strippers
Air Supply Fans
Retirculation Pumps
OS-Gas Reheaters
Off-Gas Carbon Vessels
Off-Gas Fan
Water Feed Pump
Equipment Cost Subtotal
Bid Contingencies @ 15%
Scope Contingencies @ 25%
Construction Total
Service During Construct 10;
@10%
Land Acquisition
Total Implementation Cost
Engineering, Legal, and
Administrative Costs @ 22%
TOTAL CAPITAL COST
Unit Cost (VUnit)
167,000
15,000
11,000
105,760
11300
4,100
1,600
152,000
11,800
13,800
—
_
_
—
-
300,000
_
-
-
Units Required
Probabk Design
30 ea
lea
Sea
0
0
0
0
0
0
0
H
_
_
H
-
2 acres
H
-
-
Maximum
Credible
Deviation
0
0
0
6ea
7ea
3ea
18 ea
24 ea
7ea
7ea
_
—
—
_
-
2 acres
_
-
-
Treatment Facility Cost(s)
Probabk
Design
$5,010,000
15,000
55,000
0
0
0
0
0
0
0
$5,0*0,000
762,000
1,270,000
$7,112,000
711,200
600,000
$8,423400
1353,100
$10,276^300
Maximum
Credible
Deviation
0
0
0
634,500
82,600
12300
28,800
3,648,000
82,600
96,600
$4,585,400
687,800
1,146,400
$ 6,419,600
642,000
600,000
$7,661,600
1,685,600
$ 9^47,100
LA022033\231 004.51-20
                                            402

-------
                                                Table 4
                                             (Continued)

O&M COST ESTIMATES
Electrical Power
Carbon Replacement
Condensate Disposal
Natural Gas
Boiler Fuel
Water Sample Analysis
Air Sample Analysis
Operating Labor
Maintenance @ 2% of Capital
Costs
TOTAL O&M COSTS
Annualized Cost of Capital
Total Annual Costs
UNIT COST FOR TREATMENT
(Vl.000 gal)«
Unit Cost (S/Unit)
0.07
130 (Design)
Z50 (Cred.Dev.)
1.00
3
0.45
no
250
35
-
_
—
_
-
Units Required
Probable Design
693,800 kW-hr
600,000 Ib
0
0
Ogal
31 ea
0
730 hr
-
w
_
H
-
p^yicfni^m^
Credible
Deviation
7,633300 kW-hr
81,400 Ib
523,400 gal
37,700 MM Btu
4,400 gal
84 ea
216 ea
730 hr
-
M
_
H
-
Treatment Facility Coct(s)
Probable
Design
$48,600
780,000
0
0
0
3,400
0
25,600
101,600
$959,200
669,000
$1,628,200
0.43
Maximum
Credible
Deviation
$534300
206,000
523,400
113,100
2,000
9,200
54,000
25,600
91,700
$1,559300
608,500
$2,167,800
035"
'Unit treatment costs are estimates based on preliminary design data and neglects factors such as facility location, delivery system, etc.
The cost estimates are intended to represent order-of-magnitude (+50 and -30 percent) estimates. These costs will change during
detailed design.
Tjquid phase carbon adsorption is not economically feasible in the marimum credible deviation case due to high carbon loading.
Costs shown are for air stripping.
                                                403
LAO22033\231 004.51-21

-------
Using the above conditions and additional data for the capital cost of ozone generation
equipment, a preliminary design for ozone/peroxide  treatment  of Whittier Narrows
groundwater was developed. The following conditions were used to develop treatment
costs:

Number of reaction chambers
Volume of reaction chambers (gal)
Outer dimension (ft, including concrete thickness)
Ozone demand (Ib/day)
Total installed ozone generating capacity (Ib/day)
Peroxide demand (Ib/day)
Probable
Design
Case
3
40,000
15 x 43 x 15
534
1,300
190
Maximum
Credible
Deviation
Case
3
49,000
15 x 49 x 15
1,150
2,300
407
Note that the ozone and peroxide usages per day are based on an average flow rate.
Usage would be higher during peak demand.

Table 5  summarizes  the  cost estimate for ozone/peroxide  advanced  oxidation as
estimated for the Whittier Narrows OUFS. These  costs are estimates based on pre-
liminary design data and reglect such factors as facility location, delivery system require-
ments, etc. The cost estimates are intended to represent order-of-magnitude (+50 and -
30 percent) estimates. These costs are expected to change during detailed remedial
design.

As shown in Table 5, the  cost of treating Whittier Narrows groundwater by advanced
oxidation in the probable design case is estimated at  $0.25 per 1,000 gallons treated and
$0.25  per  1,000 gallons for  the maximum credible deviation case.  As mentioned
previously, these costs are based on  bench-scale  data  collected  using  a batch-type
reactor.  Actual pilot evaluation of a flow-through system is required to more accurately
quantify the costs of treating  groundwater.

The advanced oxidation process has not yet been applied  at the large scale required for
the Whittier Narrows OUFS (e.g., 21 to 25.5 million  gallons per day [MOD]).  Although
the  cost estimates  are  based  on  sound  engineering  principles,  the  actual costs
experienced during startup of such a large system are likely  to be higher than predicted
using accepted cost estimating techniques. Such costs were not accounted for in this
study because they are as yet unquantifiable.
                                     404
 LAO22033\231 008.51

-------
                                             Table 5
                            Treatment Costs Using Advanced Oxidation


CAPITAL COST ESTIMATES
Reactors
Ozone Generation System
Peroxide System
Building
Electrical and Instrumentation
Equipment Cost Subtotal
Bid Contingencies @ 15%
Scope Contingencies @ 25%
Construction Total
Service During Construction
@10%
Land Acquisition
Total Implementation Cost
Engineering, Legal, and
Administrative Costs @ 22%
TOTAL CAPITAL COST

Unit Cost (VTJnlt)
339,000 (Design)
401,500 (CredXtev.)
2,193,500 (Design)
3,193,000 (Cred.Dev.)
30,000 (Design)
42,000 (Cred.Dev.)
120
579,800 (Design)
859,500 (Cred.Dev.)
_
-
_
_
-
300,000
_
-
-
Units Required

Probable Design
lea
1 ea
lea
5300ft2
lea
H
_
„
_
-
2 acres
_
-
-
Maximum
Credible
Deviation
lea
lea
lea
S^OOft2
lea
..
_
_
»»
-
2 acres
H
-
-
Treatment Facility Coct(s)

Probable
Design
$339,000
2,193,500
30,000
636,000
579,800
$3,778300
566,700
944,600
$5,289,600
529,000
600,000
$6,418,600
1,412,100
$7,830,700
MBtam
Credible
Deviation
$ 401,500
3,193,000
42,000
996,000
859,500
$5,490,000
823,500
1372400
$7,686,000
768,600
600,000
$ 9,054,600
1,992,000
$ 11,046,600
LAO22033\231 00541-23
                                            405

-------
                                               Table 5
                                            (Continued)

O&M COST ESTIMATES
Electrical
Peroxide
Water Sample Analysis
Air Sample Analysis
Operating Labor
Maintenance @ 2% of Capital
Costs
TOTAL O&M COSTS
Annualized Cost of Capital
Total Annul Costs
UNIT COST FOR TREATMENT
(S/1,000 gal)"
Unit Cost ($/Unlt)
0.07
1.10
110
250
35
-
H
_
H
-
Units Required
Probable Design
1,543,400 kW-hr
69,000 Ib
48 ea
36 ea
4,400
-
H
_
_
-
Maximum
Credible
Deviation
5,261,900 kW-hr
148,700 Ib
48 ea
36 ea
4,400
-
_
_
—
-
Treatment Facility Coat(s)
Probable
Design
$108,000
75,900
5300
9,000
154,000
75,600
$427,800
509,800
$937,600
0.25
Maximum
Credible
Deviation
$368400
163,600
5,300
9,000
154,000
109,800
$ 810,000
719,100
$1,529,100
0.25
aUnit treatment costs are estimates based on preliminary design data and neglects factors such as facility location, distribution system
requirements, etc. The cost estimates are intended to represent order-of-magnitude (+50 and -30 percent) estimates. These costs will
change during detailed design.
                                               406
LAO22033\231_OOS.Sl-24

-------
 In addition, it is assumed that additional treatment (e.g., filtration) is not necessary to
 control by-products such  as  assimilable  organic carbons  (AOC)  or other ozone
 byproducts.

 Finally, during the San Gabriel bench-scale testing, the advanced oxidation process
 proved infeasible for the treatment of carbon tetrachloride due to the observed stripping
 effect prior to chemical oxidation.  This effect may only be a function of the con-
 figuration of the bench-scale equipment.  However, if pilot- or full-scale equipment
 exhibit a similar phenomena, pre- or postoxidation treatment may be necessary to control
 the CTC. This will result in increased treatment costs.

 COMPARISON OF TREATMENT COSTS

 Table 6 summarizes the capital, O&M, and unit costs for treatment of groundwater using
 each of the three technologies discussed earlier.  Based on review of this summary table,
 air stripping with vapor phase GAC may be most economical for the probable design
 case.  However, advanced oxidation may be most economical of each of the treatment
 technology options under maximum credible deviation conditions.

                                   Table 6
                 Summary  Comparison of Treatment Costs8
Cost in $1,000 Gallons
Air Stripping
Carbon Adsorption
Advanced Oxidation
Probable Design
Case
0.19
0.43
0.25
Maximum Credible
Deviation Case
0.35
__b
0.25
"These costs are estimates based on preliminary design data and neglect such
factors as facility location, delivery system requirements, etc. The cost
estimates are intended to represent order-of-magnitude (+50 and -30 percent)
estimates. These costs are expected to change during detailed remedial design.
bLiquid phase carbon adsorption is not economically feasible in the maximum
credible deviation case due to high carbon loading.
As mentioned above, eight additional criteria must be considered in choosing a remedial
alternative for the preparation of a feasibility study for an EPA NPL site. Some of these
criteria include:

      •     Reliability
      •     Ease of operation
LAO22033\231 008.51
                                   407

-------
      •      Expected lifetime
      •      Ease of maintenance
      •      Presence of process hazards

When these criteria are evaluated alongside costs, the selection of a remedial treatment
technology becomes more subjective. For example, although advanced oxidation appears
less costly (for the maximum credible deviation case), each of the above criteria are
much more easily met by air stripping and carbon adsorption. Additionally, the presence
of gaseous ozone and liquid hydrogen peroxide onsite creates additional safety concerns.
All of these concerns must be carefully weighed prior to selection of the appropriate
technology for groundwater treatment.
                                    408
LA022033\231 008.51

-------
                             REFERENCES
U.S. Environmental  Protection Agency.   National Interim  Primary Drinking Water
Regulations; Control  of  Trihalomethanes  in Drinking Water.  Federal Register
44(231):68624 November 29, 1979.

Hoigne, J. and H. Bader.  Rates Constants of Reactions of Ozone with Organic and
Inorganic Compounds in Water. I.  Non-Disassociating Organic Compounds. Water
Resources 17 (1983): 173.

Hoigne, J. and H. Bader.  Rates Constants of Reactions of Ozone with Organic and
Inorganic  Compounds  in Water.  II.  Disassociating Organic  Compounds.  Water
Resources 17 (1983): 185.

Farhataziz, P. C. and A. B. Ross.  Selective Specific Rates of Reactions of Transients in
Water  and Aqueous  Solutions.   III. Hydroxyl Radical and  Perhydroxyl Radical and
Their Radical  Ions.   National Standard Reference Data Service  59, U.S. National
Bureau of Standards.  1977.

Glaze,  W. H. and J. W. Kang.  Advanced Oxidation Processes for Treating Ground-
water Contaminated with TCE and PCE: Laboratory Studies. JAWWA 80 (1988):51.

U.S. Environmental Protection Agency. Agency Review Draft Technical Memorandum,
San Gabriel Basin, Results of Ozone/Peroxide Bench-Scale Treatability Test. Prepared
by CH2M HILL January 10, 1989.
                                  409
LAO22033\231 009.51

-------
RECOVERY OF METALS FROM WATER
       USING ION EXCHANGE
           A CASE STUDY
              PREPARED FOR:

           SECOND FORUM ON
       INNOVATIVE HAZARDOUS WASTE
        TREATMENT TECHNOLOGIES
               PREPARED BY:

  B&V WASTE SCIENCE AND TECHNOLOGY CORP.


                MAY 1990

                410

-------
                       RECOVERY OF METALS FROM WATER
                            USING ION EXCHANGE

                               A CASE STUDY

                                    By

         Thomas A. Hickey, B&V Waste Science and Technology Corp.
                  David K. Stevens, Pritchard Corporation

Since 1984, B&V Waste Science and Technology Corp. has been assisting an
industrial client in the remediation of a wood treating facility in
northern California.  The wood treating facility consists of a 15-acre
paved wood treating and storage yard, the wood treating facilities
themselves, and a storm water retention pond which was recently closed.
Because of past operating practices, wood treating chemicals have entered
into the soil and ground water at the facility.  Wood treating chemicals
have also entered storm water by rinsing from freshly treated wood that is
stored on the yard.


The wood treating chemicals used by the client are inorganic metals,
chromium, copper, and arsenic.  When wood treatment operations began at the
site in 1966, a chromated copper arsenate, or CCA, solution was used, which
was a blend of sodium dichromate, copper sulfate, and arsenic acid.  The
ratio of the metals in the solution was about 60 percent chromium, 25
percent copper, and 15 percent arsenic.  In 1982, the solution was modified
to use only sodium dichromate and copper sulfate in a solution called acid

                                  411

-------
copper chromate, or ACC.  The metal ratio in this solution is about 60



percent chromium and 40 percent copper.








The wood treating chemicals each have different characteristics.    The



Safe Drinking Water Act criteria limit the total chromium in potable water



to 50 parts per billion (ppb), although recent recommendations call for



revising the maximum to 100 ppb.  At the wood treating facility, chromium



is found



in two valence states:  as trivalent chromium and hexavalent chromium.



Trivalent chromium generally acts as a cation;  hexavalent chromium is



present as chromate8 or dichromates and acts as an anion. The solubility of



trivalent chromium species is very low at a pH greater than 5.  Hexavalent



chromium species are quite soluble in the normal pH range of ground water.








Arsenic commonly occurs as the anion arsenate.  The maximum contaminant



level  (MCL) established by the the U.S. Environmental Protection Agency for



arsenic in potable water is 50 ppb.








Copper is only found as a cation.  The oxidized state, cupric, is copper's



more  stable form and has a valence of +2.  A  secondary standard in the Safe



Drinking Water Act limits copper to 1.0 ppm.
 Site  investigations  revealed  that  the  soil and storm water runoff contained



 elevated  concentrations  of  all  three metals, but only hexavalent chromium



 had migrated  in detectable  quantities  to  the ground water.  These findings
                                    412

-------
are consistent with the chemical characteristics of the metals and the site



soil.  A number of previous studies have documented that common clays, such




as montmorillonite and bentonite, act as strong adsorbants for arsenic,




copper, and trivalent chromium compounds.  The highly soluble hexavalent




chromium compounds are very mobile in soil, however, and are readily




transported through soils with normal water penetration.








Hexavalent chromium concentrations in the ground water vary across the site




from not detectable in the outer regions of the affected area to greater




than 50 ppm directly beneath the wood treating facilities.  Ground water




extracted from a variety of locations in the affected area and blended




exhibits relatively consistent characteristics.  The ground water does not




contain any of the other wood treating chemicals, and it has a low




concentration of suspended solids.  We have estimated that an extraction




rate in excess of 300 gallons per minute (gpm) or 160 million gallons per




year is needed to control contaminant migration and accomplish ground water




remediation in a reasonable time frame.








Approximately six million gallons of precipitation fall annually on the



wood treating facility's paved yard.  All storm water runoff is contained



on the yard and treated.  Storm water runoff accounts for approximately




five percent of the potential water volume treated through the treatment




plant.  The characteristics of storm water are much less consistent than




those of ground water.  The concentrations of all three wood treating




chemicals vary with the runoff patterns  of each storm as well as the timing




of the storm during the winter, or rainy, season.  The first rains of the
                                    413

-------
season carry much higher loads of the contaminants than later storms.



Suspended solids and oil and grease loads are also present in storm water



runoff and vary significantly.  A treatment rate of approximately 200 gpm



is needed to limit containment of storm water on the yard after major



storms to a period of 2 or 3 days.  During that period, the storm water is



treated and discharged.








In 1987, the client began operating an ion exchange water treatment plant



that our firm designed.  We selected ion exchange technology for the water



treatment plant because it allows removing of the wood treating chemicals



from storm water and extracted ground water in a manner that enables



recovery and reuse of the chemicals in the wood treating operations.  The



ion exchange water treatment plant allowed the client to fulfill the basic



intentions of RCRA:  to conserve and recover resources.








Other major design objectives included:







     o    Process and equipment  flexibility to treat both storm water and



          ground water  influents.







     o    Process performance to meet  stringent effluent  requirements.








     o    Cost  optimization and  operator time minimization.
                                     414

-------
Several factors allowed implementing the ion exchange technology for this



application, including:








     o    The client's philosophy encouraged waste minimization and



          resource recovery.








     o    The client understood the technology and had experience with



          similar programmable controller-driven equipment.








     o    Ongoing wood treating operations at the facility presented the



          opportunity.








     o    The characteristics of wood treating chemicals in the waters were



          suitable for the technology.








Our remedial design work began with an evaluation of the technologies



available for removing metals from contaminated water.   The technology



evaluation focused on hexavalent chromium removal processes for ground



water treatment with secondary emphasis on removing the other metals from



storm water.








The chromium removal processes that were considered included chemical



reduction and precipitation, proprietary processes using electrochemical



co-precipitation, and co-precipitation with iron.  These processes produce



a sludge that is unsuitable for reuse in the wood treating process and



requires landfill disposal.
                                     415

-------
Ion exchange was selected as the most desirable technology to meet our




client's objectives.  We looked for the following characteristics when




selecting resins for the application:








     o    High selectivity for the chromium species present.








     o    High capacity for the chromate radical at normal pH.








     o    Efficient regeneration to achieve the highest possible




          concentration of metals for reuse in the wood treating facility.








     o    Regeneration in a medium suitable for reuse in wood treatment.








     o    Physical  stability for long bed life.








A  considerable amount of research was conducted in the 1970's on the




removal  of  chromates from cooling tower discharges, which may contain up to




20 ppm of chromium.  This work  showed that styrene resins are highly




selective for the chromate  radical compared to acrylic resins.








Weak base resins exhibited  the  other desirable properties listed.   To




establish the effectiveness and economics of  the process, bench-scale and




pilot tests were conducted.  Various weak base styrene resins were  tested




for the  following parameters:








     o    Chromate  capacity measured  in bed volumes  treated against
                                     416

-------
          hexavalent chromium leakage concentration in the effluent.








     o    Concentration of chromium in the regeneration solution by




          determination of the elution curve.








     o    Effect of inlet pH on the capacity of the resin.








These studies confirmed that satisfactory performance was achieved with a




weak base anion resin in the sulfate form.  The optimum inlet pH to




maintain the sulfate resin form was found to be 4.5, and the most efficient




regeneration rate for chromium concentration was achieved with 2 to 2-1/2




bed volumes of 3 percent caustic soda solution.








After elution with the caustic solution, the resin is returned to the




sulfate form with a sulfuric acid conditioning step.  In this step, the




hydroxyl ions supplied by the caustic are in turn replaced by sulfate ions.




The sulfate radical is readily displaced by anionic chromate.








A small amount of hexavalent chromium was found to reduce to trivalent



state by oxidation of the resin.  For this reason, as well as to remove the




lower concentrations of trivalent chromium and copper present in storm




water, a cation bed was installed after the anion beds.  Through a  series




of steps similar to those described in the selection of the anion resin, a




weak acid cation resin was selected.  Figure 1 illustrates the system




selected.  Regeneration of the cation bed is accomplished with one bed




volume of 5 percent sulfuric acid.
                                     417

-------
                        WATER TREATMENT PLANT

                         ION EXCHANGE SERVICE
Water from _
Pre-lreatment
System
_H2SO4
                     Pri
                    Anion
                                 	NaOH
                Sec
               Anion
Cation
                                                               To Discharge
                                                    NaOH

-------
The final effluent water from the cation exchanger is adjusted to near



neutral pH with caustic soda for discharge.  The regeneration streams from




the anion and cation exchange columns are combined in the recovery tank.




The resulting mixed metal concentrate is returned to the wood treatment




process by incorporating it into the treating solution, as illustrated in




Figure 2.








The ion exchange system resins selected are more suitable for treating




waters with low contaminant load, and cannot tolerate significant organic




or solids loadings.  To protect the resins, we installed a multifunctional




pre-treatment system, as shown in Figure 3.  The pre-treatment system




consisted of chemical feed, coagulation and flocculation chambers, settling




basin, flow equalization basin, strainer, and banks of filters.  Because




storm water has the greatest variety of extraneous materials, it is treated




through all of the pre-treatment components, which allows for removing




suspended solids, some organics, and arsenic.  We found that the arsenate




radical was not adsorbed by the anion resin, so separate treatment of the




surface water containing arsenic was required.  Ferric chloride addition is



a successful flocculant for arsenic removal.  Ground water is processed



only through the strainer and filters before ion exchange.  Solids from the



settler, strainer, and backwashable filters is held in sludge drying beds




until they are disposed of offsite.








The completed ion exchange water treatment plant has been in operation for




three years and has produced water in compliance with the NPDES permit




conditions even though effluent limits were significantly reduced after
                                    419

-------
                        WATER TREATMENT PLANT

                      ION EXCHANGE REGENERATION
ro
o
                     Pri
                     Anion
                       Cr'
                 Recovery Tank
                                          NaOH
 Sec
Anion
                     H2SO4
Cation
                                 Cu + Cr+3
      To Wood Treating
      Chemical Storage Tanks
                                                                 FIGURE 2

-------
                             WATER TREATMENT PLANT

                             PRE-TREATMENT SYSTEM
     Groundwater
     Stormwater
r\>
     Pre-Treated Water
     to Ion Exchange
     Train
                                                             1
Flow Equalization
                                          Lamella Separator
                                            Sludge to Drying Beds
                             Cartridge Filters
 Bag Filters
Automatic
 Strainer
                                                                            FIGURE 3

-------
plant design and construction were completed.  For example, the original



hexavalent chromium effluent limit was set at SO ppb.  As part of NPDES



permit revisions, the limit was reduced to 11 ppb.  The effects of this



revision are discussed in detail later in this paper.  Other major findings



during the past three years of operation are summarized as follows.  Each



of these problems is described in detail in following sections.








     o    The resins have been fouled  by iron and organics despite



          pretreatment and other precautions.








     o    The breakthrough of hexavalent chromium from the secondary anion



          bed has limited treatment plant run lengths between regeneration



          cycles.








     o    Normally, anion and cation exchange resins have been fully



          regenerated simultaneously, and the use of regenerant streams in



          the wood  treating process has been successful.








     o    The balance between the  supply of  recovered chemicals and the



          demand  for the chemicals and makeup water  in the wood



          treating  facility is critical to  successful operation.








 Figure  4  illustrates the behavior  of  the water  treatment  plant in terms of



 effluent  hexavalent chromium as  a  function  of gallons of  water treated.



 The  pH  of the  influent water has been found to  affect the chromium capacity



 of  the  anion resins and  the duration  of the runs  before  breakthrough.
                                    422

-------
                           WATER TREATMENT PLANT

                              PERFORMANCE DATA
ro
ca
 Effluent
Concentration
    (ppb)
                                                               /• urigmai urn
                                       Pri Anion Effluent
                                                           Sec Anion Effluent
                                                             Reduced Limit
                                        Gallons Ireated (x106)
     Primary Anion Inlet pH @ 4.5
     Primary Anion Inlet pH @ 7.5
                                                                          FIGURE 4

-------
Under the design condition of influent pH control to 4.5, chromium




breakthrough occurs with a gradually increasing effluent concentration, or



a "slow leak".  This operation method maximized throughput before the




original effluent limit of 50 ppb was achieved.  Under  normal pH or




uncontrolled pH, chromium breakthrough concentrations are very low




initially and rise sharply at the end of the run. This behavior allows for




maximizing throughput against the revised limit of 11 ppb.  Sengupta [5]




has reported similar results and has shown that the effect of pH is a




result of the hexavalent chromium equilibrium characteristics, not the



result of mass transfer effects.








We have been able to combine the effects of pH control to optimize run




lengths by operating the beds under "hybrid pH control".  In this operating




mode, the primary anion bed operates under acidic conditions and secondary




anion bed operates under near neutral conditions.  The result allows the




primary anion to generate a "slow leak" of hexavalent chromium and the




secondary bed to remove the "slow leak" under favorable equilibrium




conditions.  The total hexavalent chromium retention capacity is greater in




this mode of operation than it would be for the current effluent limit, as



indicated in Figure 5.








Resin fouling has primarily resulted from iron carry-over from the




pre-treatment facility and organic fouling from materials in storm waters.




Resin fouling has been evidenced by excessive pressure losses through resin




beds that have not been reduced by backwashing and by low capacity for




wood treating chemicals on the resin.  The following steps have been taken
                                  424

-------
                      WATER TREATMENT PLANT

                         PERFORMANCE DATA
                                                             Original Limit
           +6
  Effluent Cr
 Concentration
    (ppb)
             40-
              3d-i
              20-
              10-
                               r
Pri Anion Effluent
                    Sec Anion Effluent
                              w
                                                t  / Reduced Limit

                                0.5
               I
              1.0
 i
1.5
                                                    6
                                    Gallons Treated (xlO )
Primary Anion Inlet pH @ 4.5

Primary Anion Inlet pH @ 7.5
                                                                     FIGURE 5

-------
to alleviate the problems associated with resin fouling:








     o    Periodic resin samples have been taken from the resin beds and




          submitted to the resin manufacturer for analysis.








     o    The pretreatment systems have been operated properly to minimize




          the potential of iron carry-over and to maximize removal of other




          deleterious materials in the waters.








     o    Periodic surfactant and brine-caustic cleanings have restored




          resin activity and performance.








The balance between the "supply" and the "demand* for recovered chemicals




has, at times, limited the effectiveness of resource recovery.  When the




demand is low, during periods of limited wood treatment, water treatment




has been  curtailed.  Obviously, this seriously affects the effectiveness




of ground water extraction for site remediation.  To counter this effect,




we have optimized the recovery process by increasing water treatment



plant runs between regeneration cycles and by minimizing the amount of




regenerant produced during each regeneration.








The major methods employed to maximize treatment plant runs are those




described previously:  pH control for hybrid operation and resin




maintenance for maximum capacity.  We have also explored utilizing detailed




aquatic toxicity testing to modify the effluent limits currently imposed on




the plant's performance.  Preliminary investigations indicate that effluent
                                   426

-------
limits could be raised without increasing aquatic toxicity to the most




sensitive organisms.








Several methods have been used to reduce the amount of regenerant produced




during each regeneration.  Careful timing  and control of the regenerant




steps have produced marginal reductions in regenerant volumes.  Recycling




of certain chemical rinse steps has been implemented to reuse lightly




contaminated caustic solutions as the first elution step in the subsequent




regeneration cycle.








In summary, hexavalent chromium and other soluble metals can be




successfully recovered in a condition suitable for reuse.  We have seen




that  this recovery  is effective and successful when the wastewater




characteristics are suitable for the resins selected, the operator commits




to operation and maintenance practices to keep the resin and system in




optimum  condition and balance, and the "demand" equals or exceeds the




"supply" of the recovered chemicals.  The ion exchange process was applied




to produce a usable regenerant concentrate, and it was economical when




concentrating  to 50 times the extracted ground water contamination level.
                                   427

-------
                                REFERENCES
1.  Bartlett, R. J. and Kimble, J.  M.,  "Behavior of Chromium in Soils",
     Journal of Environmental Quality:   Volume 5, Ho.  4,  1976 and Volume 8,
     No. 1, 1979.

2.  Lang's Handbook of Chemistry, Ed. VII.

3.  U.S. Environmental Protection Agency:  "Hazardous  Waste Land
     Treatment", Office of Solid Waste, SW-274, April  1983.

4.  Sorg, Thomas J. and Logsdon, Gary S., "Treatment Technology to Meet  the
     Interim Primary Drinking Water Regulations for Inorganics", Journal of
     the American Water Works Association:  July 1978.

5.  Sengupta, A. and Clifford, D., "Some Unique Characteristics of Chromate
     Ion Exchange", Reactive Polymers, 4  (1986) 113-130,  Elsevier
     Scientific Publishers;  Amsterdam.

6.  Kumin, Robert,  "New Technology for the Recovery of Chromates from
     Cooling Tower  Slowdown", Amber Hi-Lites, No. 151, Rohm & Haas Co.
     publication, May 1976.
                                     428

-------
        SECOND FORUM ON
 INNOVATIVE HAZARDOUS WASTE
   TREATMENT TECHNOLOGIES:
 DOMESTIC AND INTERNATIONAL
   CRITICAL FLUID SOLVENT EXTRACTION
        Presented May 17,1990
               Authors:


   Cynthia Kaleri, EPA, Region VI, Dallas, Texas
Louis Rogers, Texas Water Commission, Austin, Texas
  Calvin Spencer, Roy F. Weston, Houston, Texas

                429

-------
                        TABLE OF CONTENTS
SECTION        DESCRIPTION                                  PAGE


1.0            INTRODUCTION                                 1-1

2.0            SITE DESCRIPTION AND HISTORY                 2-1

3.0            TECHNOLOGY DESCRIPTION
               3.1  Technology Theory                       3-1
               3.2  Simplified Critical Fluid
                     Solvent Extraction                     3-1
               3.3  Propane Solvent Extraction Process      3-4
               3.4  Treatment                               3-7
               3.5  Characterization                        3-7
               3.6  Sample Preparation                      3-8

4.0            RESULTS
               4.1  Treatment                               4-1
               4.2  Material Handling Problems              4-1

5.0            TREATABILITY CONSIDERATIONS                  5-1
                               430

-------
                           SECTION 1.0

                           INTRODUCTION
As mandated by the Superfund Amendments and Reauthorization Act of
1986 (SARA) and in accordance  with the  National Contingency Plan
(NCP),  EPA must consider  several  factors in  selecting remedial
action for a  superfund site.   Overall protection of human health
and the environment and consistency with other environmental laws
are primary criteria  for selection of a  cost  effective response
action.    However, the  statutory  preference   for  treatment  in
permanently and  significantly reducing  toxicity,  mobility,  or
volume must also  be  considered  and balanced  accordingly.   In
addition,  state and community  acceptance  influence  the selection
of remediation at  a Superfund site.

EPA  realizes   the  need   for   less costly,  effective  treatment
technologies to manage hazardous waste sites.  In response to this
need, EPA established the Superfund Innovative Technology Evalua-
tion (SITE) Program in 1986 to promote the development and use of
new  or  innovative  treatment technologies to clean up Superfund
sites  across  the  country.   The  objective of  the  demonstration
portion of the SITE program is  to develop  reliable performance and
cost information of the technologies selected so that they can be
adequately considered in the Superfund decision making process.

Although  the   SITE program identifies   the  feasibility  of  an
innovative technology for particular types of  waste,  the need to
evaluate and refine a technology for site specific application is
a practical necessity  for selection of the most appropriate remedy
for a  site.   EPA has  encouraged the use  of  treatability studies
during the early phases of the Superfund investigative process in
order to effectively  evaluate  the  use of  innovative technologies
at individual  Superfund sites.

This paper discusses the  results of a treatability study performed
for  the  United Creosoting  Superfund site and the  evaluation of
those  results  in  relation to the  remedy  selection  process.
Although Critical  Fluid  Extraction is  an innovative technology
under the SITE program, the demonstration of the pilot scale unit
at this site  was  carried  out through a  separate contract between
the vendor and Roy F. Weston,  Incorporated (WESTON).   WESTON was
selected by  the   Texas  Water   Commission [TWC]  to  conduct  the
treatability studies  and  incorporate the  results into an Amended
Feasibility Study  Report.1
     1     An  abbreviated  site history  from the  1989 Record  of
          Decision  [ROD]  is  presented  prior  to the  technology
          description and results discussion.  The 1989 ROD should
          be consulted for in-depth discussions  in all areas to be
          presented in this paper.

admin:crflpapr.425            431

-------
                           SECTION 2.0

                   BITE  DESCRIPTION AND HISTORY
The United Creosoting site is  located 40 miles north of Houston in
the City of Conroe, Montgomery County,  Texas [Figure 2-1].  Bound
on the west and south by Alligator Creek,  on the north by Dolores
Street, and on  the east by the Missouri-Pacific  rail  lines,  the
property is approximately one hundred acres in size.

The  United Creosoting  Company  operated  as a  wood  preserving
facility from 1946 through the summer of 1972.  With the exception
of the process building where  timber was  debarked and  cut to the
desired product, the process areas became scarred by an accumula-
tion of the  black oily chemicals  used for treating the lumber.
Historical aerial photographs  and analytical data obtained to date
have been utilized to describe the process areas  as they existed
during active operations.

Formed lumber,  such as telephone  poles  and railroad  ties,  were
treated in a  two-step  process by the  pressurized addition  of
pentachlorophenol [PCP]  and creosote.  The pressure cylinders were
rinsed and the wastewater  routed to one of the  two process waste
ponds located onsite.  Segregation of the two waste streams allowed
possible reclamation and reuse.  The larger  pond  held  mainly the
creosote waste and the smaller pond the PCP process waste.

No evidence exists that  PCP was produced onsite.  However, PCP was
stored in one or more of  the  storage tanks onsite.   Creosote was
produced via  a coal tar  distillation  unit onsite  and  stored  in
lined pits just east of the  process  waste ponds.   Creosote and
other distillate fractions  of coal  tar included polycyclic aromatic
hydrocarbons [PAHs] of varying molecular weights.  Coal tar pitch,
a dark brown  to black amorphous residue, was an unusable byproduct
which was  apparently disposed  of in the larger process waste pond.

The  physical  characteristics  of  the  site  have been  altered  by
redevelopment of the property, which has  resulted in residential
and  light  industrial  structures   typical  of suburban  settings.
Other residential areas surround the site to the immediate north,
west and  south, while  industrial  and commercial land  uses  are
evident to the  east.  Approximately 13,000 people currently live
within a two-mile radius of the site.

The Texas Water  Commission [TWC]  submitted the  United Creosoting
site as a  candidate for  cleanup  under the Superfund  program  in
August 1982.   The immediate concern at that time was contaminated
surface water runoff flowing from the former waste ponds area into
Tanglewood East Subdivision, the residential portion of the site.
TWC collected additional soil,  water, and air samples from the site
and in September 1983, the United Creosoting site was included on
the  proposed  National  Priorities  List  by EPA  and thus  became
eligible for remedial funding.

                               432

-------
433

-------
In early December 1983,  EPA  initiated an immediate response action
at  United Creosoting.    Surficial soils  samples  taken in  the
vicinity of the former waste ponds and within the Tanglewood East
subdivision revealed PCP contamination  and chlorinated dioxins and
dibenzofurans, trace byproducts of commercial  grade PCP.   It was
suspected that the source of the contamination might  be storm water
runoff  from  former waste  pond areas  located  on the  commercial
portion of the site.  Therefore, exposed sections of contaminated
soils in the  former waste ponds area were regraded and covered with
a  synthetic  membrane and  at least 6  inches of compacted  clay.
Access  to the cap  area was restricted by  fencing the  area and
constructing drainage ditches to channel cap  area runoff away from
the subdivision.

The Remedial  Investigation/Feasibility  Study  [RI/FS] was completed
in  May 1986.   Although incineration  was  considered a  feasible
alternative,  no commercial facilities were available at that time
to accept dioxin contaminated soils;  the residents opposed incin-
eration so close to their homes.  In August 1986, EPA formally
proposed  natural  attenuation of the groundwater and a  temporary
consolidation  and capping remedy  for  contaminated soils at the
site.  EPA further specified that  innovative  technologies would be
evaluated in the  next 5 years  for permanent  soils remediation in
order  to  mitigate further groundwater  contamination.   Following
formal public comment, EPA selected the proposed remedial action.

In March 1987, two treatability studies were  initiated to evaluate
innovative technologies  as possible remedies for soils at the site.
These  treatability  studies  involved  biological   treatment  and
critical  fluid extraction.  This paper will  focus on the critical
fluid extraction technology treatability results and  the evaluation
of  these  results  in relation to the  remedy selection criteria in
place  at  the signing of  the September  1989 Record of Decision.
admin:erfIpapr.425

-------
                           SECTION 3.0

                      TECHNOLOGY DESCRIPTION
3.1  Technology Theory

Solvent  extraction,  at  its simplest,  is  the separation  of the
constituents of a liquid  solution by contact with  another insoluble
liquid.   If  the  constituents in the original solution distribute
differently  between  the two  liquid phases, a certain  degree of
separation  will  result.   In  such an  operation,  the  original
solution  is called the feed, and the liquid with  which the  feed is
contacted is called the  solvent.  The solvent-rich product of the
operation is called the extract and the  residual  liquid from which
the solute is removed is called the raffinate.

In the case  of  Critical  Fluid Solvent Extraction, the solvent is
a pure liquid whose temperature and pressure are controlled at or
near the materials' critical point or state.  At the critical point
(Point C on Figure 3-1),  distinctions between the  materials' liquid
or vapor  phase disappears  and the  properties of  the liquid, such
as density,  viscosity,  refractive index,  etc.,  are  identical to
those of  the vapors.

Figure 3-1 is a simplified vapor-pressure curve which shows phase
differences  (solid,  liquid,  gas),  the  critical point  (C)  with
corresponding critical pressure and critical temperature, boiling
point temperature, and the triple point  (T).  The triple point is
that physical state at which all three phases coexist.

For the United Creosoting site, a technology using propane as the
solvent was selected.  Propane is an excellent solvent because it
exhibits  an excellent solubility of hydrocarbons, is inexpensive,
is readily compressed, and  if released, does not pose  a significant
environmental hazard.

Although a relatively new technology for hazardous waste treatment,
propane  solvent  extraction  was used by the  petroleum  refining
industry  in the late  1940's and early 1950's for the de-asphalting
of lube   oil  feed stocks.   Since  the  early  1970's fundamental
research  in  critical  fluid extraction  has been  ongoing.   The
technology presented below is a product of this ongoing research.

3.2  Simplified Critical Fluid Solvent Extraction

As shown  in Figure  3-2, a  liquid  or slurried  feed such  as  an
organic-containing hazardous  waste is  admitted  to  an  extractor
along with the solvent at  or  near  the  solvent critical  point and
the organics in  the  waste dissolve into the  solvent.   Extracted
organics are  removed with solvent from the top of  the vessel, while
clean water and solids exit through the bottom.  The extract then
                                435

-------
Pcrlt
                                                           GAS
                                                 tnbp     Wit
                             TEMPERATURE
                                                     FIGURE 3-1
                                                 VAPOR  PRESSURE OF
                                                    A PURE UQUID
                            436

-------
FEED



^^—
EXTRACT T
SE
CC
EXTRACTOR
4 i
SOLVENT
:PARATOR
)LUMN
r
mt^m
*\
^.
                                                                      SOLVENT
                                                                      MAKEUP
                                                                 COMPRESSOR
                                         EXTRACT
               RAFFINATE
                                                               FIGURE  3-2
                                                         SIMPLIFIED FLOW CHART
                                                               OENERMjGeNERU7.TV i-J-SO 1:1
                                    437

-------
goes to  a  second vessel, where the  temperature  and pressure are
decreased, causing the organic to separate from the solvent.  Clean
solvent is recycled to the extractor, and the concentrated organics
are recovered from the bottom of the separator.

3.3  Propane Solvent Extraction Process

The treatment process for the extraction of organics using propane
from  soils  and  sludges  is outlined   in  Figures  3-3 and  3-4.
Although this represents a commercial process,  the pilot unit used
at United Creosoting  was operated to simulate this process.  Figure
3-3  illustrates  the extraction  of hydrocarbons  from soil  and
sludges and Figure 3-4 outlines the solvent recovery system.

Extraction

The Extraction Unit operates  on a continuous basis in a countercur-
rent manner.  Feed is processed in a series of extractors/separa-
tors.  Maximum extraction efficiency is achieved by  feeding the
fresh  solvent  to  the last  extraction  stage  and  operating with
solvent  flowing  in  a  countercurrent mode.   The  following  is  a
description  of a  three  stage  extraction  unit.    The number  of
extraction stages  is decreased  or increased to meet the require-
ments of the specific application.

Slurry from the  feed surge drum (D-101) and the  solvent/organics
steam from the second-stage  decanter (D-103)  are combined in the
first-state  extractor  (XT-101).     After  thorough  mixing,  the
resulting  stream is  sent to the first-stage decanter  (D-102).
Substantial separation of  the solvent/organics solution  from the
water/solids residue  occurs  in  this decanter,  and a concentrated
solvent/organics solution exits  overhead to the decanter/coalescer
(D-105) in the solvent recovery  section.  The water/solids residue
from  the first  decanter  (D-102),  which contains  some  residual
organics, and the  solvent/organics  solution  from the third-stage
decanter  (D-104)  are combined and mixed in  the  second-stage
extractor (XT-102).

Separation of  the water/solids  residue  and the  solvent/organics
solution from  the second-stage  extractor  (XT-102)  occurs  in the
second-stage decanter (D-103);  the  solvent/organics  solution  is
pumped to the first-stage extractor  (XT-101) and the water/solids
residue  is  sent to  the  third-stage  extractor  (XT-103).    The
water/solids  residue from  the  second-stage  contains water  and
solids  with  some residual  organics.    Fresh  solvent  and  the
water/solids residue from the  second-stage decanter  (D-103)  are
combined in the third-stage extractor (XT-103).  Final separation
of solvent/organics solution and water/solids residue occurs  in the
third-stage separator  (D-104).

Water/solids  residue  from  the  third-stage  decanter  (D-104),
consisting  of water and  solids,  together with water from  the
decanter/coalescer (D-105), are  sent to  the dewatering  system.  If
needed, a portion of the water recovered in the dewatering system
                                438

-------
CO
                      XT-101
                   FIRST STAGE
                    EXTRACTOR
              XT-102
           SECOND  STAGE
             EXTRACTOR
              XT-103
            THIRD STAGE
            EXTRACTOR
            D-101
      FEED SURGE DRUM
    D-102
  FIRST STAGE
   DECANTER
    D-103
SECOND STAGE
  DECANTER
   D-104
THIRD STAGE
 DECANTER
                                                                                     EXTRACT TO D-105
                                                                                  «   CLEAN SOLVENT FROM P-104
     SLURRIED
     FEED
                  P-101
                P-101
           MAIN FEED PUMP
     P-102
 FIRST INTERSTAGE
PRESSURE BOOSTER
      PUMP
                                                                P-103
       E-103
 SECOND  INTERSTAGE
  PRESSURE BOOSTER
       PUMP
                                                                                                 « WATER FROM
                                                                                                   D-105
                                                                                                   WATER
                                                                                                 #• SOLIDS
                       FIGURE 3-3
                                                                                          PROCESS FLOW DIAGRAM
                                                                                           EXTRACTION SECTION

                                                                                               CENBMUCENEIU38.JP1V B-2-W t:t

-------
D-105 T-101 C-101 C-102 T-102
DEACANTER/ PRIMARY SOLVENT PRIMARY SECONDARY SECONDARY SOLVENT
COALESCER RECOVERY STILL COMPRESSOR COMPRESSOR RECOVERY STILL
E-101
PRIMARY STILL
REBOILER HEAT * "~ " ' " > '
EXCHANGER
D-105 TT-IOI I
. . . . s^~-\ r-*. ^
f X N 	 1
TYTPAPT ... . , M , l~ ....- .-*

/r-ooki ri •i/T'A v - - -• 	 ' ^ ? \^«^ •***.
^rKUM U— IVZ) ^<^ "^y ^^
C-101 1 C-102
WATER TO n^>E-101
nrwATrmisir < \\^ j 	 *>
SYSTEM
^
<• 	 _. 	 k


T-102
^

D-106 ^^


SOLVENT ^^
Em^
IUj
i
^TO XT-103) ( > 	 *
r~\
P-104- P-105
J " E-102
., k .., 	 	 ^ nit
-y
P-105
E-102
SECONDARY STILL SECONDARY STILL
CIRCULATING PUMP HEAT EXCHANGER
— IU4 U— lUu L— 1UO
SOLVENT RECYCLE SOLVENT SURGE SOLVENT
PUMP VESSEL SUBCOOLER/
CONDENSER
RGURE 3-4
MATERIAL BALANCE
SAMPLE 1

-------
is recycled as slurrying liquid.  Effluent water is sewered or sent
to an on-site wastewater treatment facility.

Solvent Recovery

The decanter/coalescer  (D-105)  removes entrained water  from the
solvent/organics  solution  stream prior  to the  primary  solvent-
recovery  still  (T-101) .   The  recovered  water is  sent to  the
dewatering system.

The water-free extract stream  is sent  to the  primary  solvent-
recovery  still (T-101) where  the solvent  is stripped  from the
organics.   Vaporized solvent goes  overhead while a  highly con-
centrated organics-solvent mixture is withdrawn from the bottom of
the still  and sent  to  the secondary  solvent-recovery still (T-
102).  The remaining solvent is stripped from the organics in the
secondary still,  and the  recovered organics  are pumped  from the
still bottom to  the Client's  storage facilities  (outside CFS1
battery limits).

Vaporized  solvent  from  the  secondary  still  is  compressed  to
primary-still pressure, combined  with  the  vapor from the primary
still, and  further  compressed to extractor pressure.  The  high-
pressure vapor  is then used as the heat source for  the  primary-
still reboiler (E-101), where it is partially  condensed.  Complete
condensation is achieved in  the condenser/subcooler  (E-103), and
the condensed high pressure solvent  is  stored  in the solvent surge
drum  (D-106)  prior  to recycle  to the  extraction system.   Thus,
thermal  energy requirements of  the  solvent  recovery  system are
minimized by using the hot compressed solvent vapor from the main
compressor,  to  supply the  heat required to vaporize  most of the
solvent from the organic in the primary still.

3.4  Treatment

Treatment may be simply stated as the process by which compound x
is  removed  from compound  y.   The  theory  of  the critical  fluid
solvent extraction  system  has been described  in detail,  but the
theory is only a portion  of the  removal process.   The treatment
process  must pay close  attention to  the  physical  and  chemical
characteristics of the material to be treated.   Removal of the PAHs
and dioxins from the parent material to be treated.  Removal of the
PAHs and  dioxins  from the parent material  studied at the United
Creosoting site  requires as close a scrutiny  of the physics and
chemistry as is cost effective.

3.5  Sample Characterization

For treatment,  the  contaminated material was  mixed  with potable
water from  the City of Conroe;  the chemical makeup  of  which is
known.   However, budgetary  constraints did  not allow  chemical
characterization of the parent material.  The physical mixture of
the contaminated  material  consisted of clayey  sand,  clay,  sand,
rocks and woody  material  in the form of  bark,  sticks and old
boards.    This wide  range  of compositions and  sizes  presented a


                               441

-------
materials handling problem  which may best be solved  by the many
vendors who produce machinery such as  that  used in the mining or
road building industry.

3.6  Sample Preparation

This treatment demonstration consisted of treating two batches of
contaminated soil from separate locations at the site, the former
waste pond  and  the former  operations  area.   The  sample  for the
first batch was  collected in the  former waste pond and weighed 238
pounds.  Contaminated soil material was prepared for treatment by
sifting  the material  through  a  1/2"  sieve  and  discarding all
material which did not pass the sieve.  The  solids which passed
through the 1/2" screen were combined with potable water from the
City of Conroe and mixed with an  electric mixer  for  20 hours.  The
resulting slurry was  passed  through a 1/8" screen and all material
not passing the sieve was also discarded.   The slurry passing the
1/8" screen was  very homogeneous and showed marked tendencies for
the particulate  to stay in suspension during the time the mixture
was being processed.   Figure 3-5 shows  a schematic  of the feed
preparation/treatment  process  with  the  weight of the  various
material streams added or lost.

Additional water was added during the actual treatment process to
rinse soil solids from the walls of the feed preparation tank and
the treated slurry collection tank.  Losses of both water and soil
occurred by accumulating in  the pilot units pipes, filters, pumps,
extractor vessels, and overflows  into a second raffinate tank from
the first  tank.   Some  loss occurred  due to  a clogged  pump and
circulation piping which  required dismantling  in  order to return
the system to operation.  The additions and losses identified in
Figure 1 for both water and soil  resulted from conditions which are
more characteristic of a  pilot scale operation  than a full scale
unit.  Proper  design and selection of mechanical  equipment for full
scale operation will eliminate such inefficiencies.

The sample for the second batch  to be  treated was  collected from
the former operations area and weighed 280 pounds.   The sample to
be treated in the pilot  unit was prepared in  a manner similar to
that of the  first batch.  However, this  soil exhibited considerably
different working characteristics than the material from the pond
area.  The second batch contained heavier sand particles and what
appeared to be black flakes of tar.   When mixing ceased, the sand
would immediately drop out of suspension in the mixing tank.  The
tar flakes, even  with thorough  mixing, would not  dissolve; pre-
senting potential problems with treatment.  A decision was made to
treat only a  portion  of  the material which  had been prepared and
dilute it with more water in an  attempt to  reduce  the heavy sand
load and provide for  better  suspension  of the remaining particles.
Figure  3-6  shows a  schematic  of the  feed  preparation/treatment
process  with  the  actual weight of  the material  treated.   The
material wasted may have  contributed to the organic loading, but
the PAH concentration of the slurry treated was attributed entirely
to the soil mass within the slurry.


admin:erfIpapr.425

-------
oo
        SAMPLE
                                          4-1/2  EXTRACTIONS  *
                                               TREATMENT
 1/2"    <1/2"                   1/8"

•^-»  205#  +   183# H20    -§SVE_  346#  ^55L
                 20  HRS. MIXING
         238#
I
                        >1/2"33#
                             >1/8"42#
                                                                          TREATED  &
                                                                          RECOVERED
                                                                               I
                                                                           93# SOIL
                                                                           177# H20
                                                                          1# EXTRACT
                                                                              i
                                                                          TO  ANAYLSIS
       * AVERAGE SOLVENT TO FEED RATIO = 2.2
                                                                               ADDITION TO PROCESS
                                                                               75# H20
                                                                                            -*• LOST IN  SYSTEM
                                                                                              69# H20
                                                                                              81 # SOIL
                                                                                                   RGURE 3-5
                                                                                               MATERIAL BALANCE
                                                                                                    SAMPLE 1
                                                                                             F: \MAWMGS\GENERM.\GENERMO.JP S-2-90 4:1

-------
 SAMPLE
1/2"
 280#
                  267#  +   281 # H20
                         18 HRS. MIXING
                >1/2"13#
* AVERAGE SOLVENT  TO  FEED  RATIO =  3.2
                                                    3-1/2  EXTRACTIONS *
                                                         TREATMENT
                                         1/8"
448#    SLURRY  >
                                         >1/8"15#
                                            HEAVY
                                             SAND
                                             85#
                  TREATED &
                  RECOVERED
                                                                    59# SOIL
                                                                    341# H20
                                                                  2# EXTRACT
                                                                  TO ANAYLSIS
                                       ADDITION TO PROCESS
                                       95# H20
                                                                             LOST IN SYSTEM
                                                                             34# H20
                                                                             107# SOIL
                                                                                           RGURE 3-6
                                                                              MATERIAL BALANCE
                                                                                  SAMPLE 2
                                                                                     F:\fMA«HWS\GEMERM.\aENERMI.JP S-2-W 1:1

-------
                           SECTION 4.0

                             RESULTS

4.1  Treatment Results

Removal efficiency  may be calculated by dividing  the difference
between the  initial reference and  the  final reference  with the
initial reference point.  However, the reference point  (mass versus
concentration versus  toxicity)  influences the  interpretation of
removal efficiency.   For example,  a mass basis  may provide high
removal efficiencies  for PAHs; from  the  two tests  completed at
United  Creosoting,  removal  efficiencies  of 94%   and 98%  were
achieved for total mass PAHs.  In comparison, using concentration
as the indicator for removal efficiency - - expressed in terms of
toxicity equivalents  or  Benzo(a)Pyrene  Equivalents  [BAPE]   -  -
removal efficiencies are slightly lower for each  test,  86% and 91%
for total BAPE concentrations.

Table  4-1  presents the  laboratory results  for  the two  tests
conducted  at the  United Creosoting  site  in terms  of mass  of
individual and total  PAHs, PCP,  and individual  and total dioxins
and furans.  An interesting consideration in the variance among in-
dividual PAHs  is the different  structures  of  individual  PAHs.
Figure 4-1 demonstrates  that those  similar structures correspond
in the percent of mass removal.   More polar compounds have simi-
larly high removal efficiencies, as anticipated since the solvent
propane is polar.

Table 4-2 presents the concentrations of the chemicals of concern
in terms of individual carcinogenic PAHs,  naphthalene, total PAHs
versus  total BAPE, PCP,  and  dioxins/furans versus  total  TCDDE
(toxicity  equivalent  similar to  BAPE).   Again, those chemicals
similar in structure have similar removal efficiencies in terms of
toxicity and/or concentrations (noncarcinogens).

4.2  Material Handling Problems

The problems encountered  in  handling  and  treating  the waste pond
soil were  very different than those encountered during treatment
of the residential surface soil, demonstrating that variability in
soil  composition across the  site  will  present  dynamic  soils
handling problems.  The variability in soil will require that the
full scale unit be equipped and operated to respond effectively to
meet the changing influent conditions.  Problems  encountered  in the
pilot  study  should  provide valuable  insight   into   design  and
operating  parameters  that must  be  incorporated into full  scale
treatment.
     The 1989 ROD has  an  in-depth discussion on the use of these
     equivalents and  the  determination of action  levels for the
     United Creosoting site.

                               445

-------
                                                            TABLE 4-1


                                                       tENOVAL EFFICIENCY
                                            TEST ONE
                                                                                              TEST TWO
COMPOUND
ACENAPTHENE
ACENAPHTHYLENE
ANTHRACENE
BENZO(A)ANTHRACENE
BENZO(A)PYRENE
BENZO(B)FLUORANTHENE
BENZO(GHI)PERYLENE
BENZO(K)FLUC«ANTHENE
CHRYSENE
DI BENZOCA, H)ANTHRACENE
FLUORANTHENE
FLUORENE
INDENO(1,2,3-CD)PYRENE
NAPHTHALENE
PKENANTHRENE
PYRENE
TOTAL MASS OF PAH (HG)
MASS IN
UNTREATED
SOIL (HG)
26,640
1.110 J
24,420
7,400
3,552
3,774
1,480 J
3,700
8,140
ND
26,640
28,120
1,406 J
10.360
43,660
26,640
213,046
MASS IN
TREATED
SOIL (HG)
143 J
126
373.8
331.8
504
407.4
504
714
382.2
181 J
462
160 J
462
63 J
546
462
5,149
REMOVAL
EFFICIENCY (X)
NA
NA
98X
96X
86X
89X
NA
81X
95X
NA
98X
NA
NA
NA
99X
98X
98X
MASS IN
UNTREATED
SOIL (MG)
ND
675 J
2,957 J
1,441 J
24,413
20,319
120,020
136,245
44,733
54,892
34,573
40,714
28,508
12,813
2,047 J
12,207
529,437
MASS IN
TREATED
SOIL (HG)
ND
246 J
ND
ND
670 J
616 J
5,089
5,357
2,679
3,482
4,018
2,170
2,679
2,009
964 J
2,089
29,572
REMOVAL
EFFICIENCY (X)
NA
NA
NA
NA
NA
NA
96X
96X
94X
94X
88X
95X
91X
84X
NA
83X
94X





COMPOUND
PENTACHLOROPHENOL
TOTAL TCOD
TOTAL PeCDD
TOTAL HxCDD
TOTAL HpCDD
TOTAL OCDD
TOTAL TCOF
TOTAL PeCDF
TOTAL HxCDF
'OTAL HpCDF
•OTAL OCOF



MASS IN
UNTREATED,
SOIL (UG)
28,120 *
ND
ND
1,184
26,640
96,200
ND
74
2.220
11,840
11,840


TEST ONE
MASS IN
TREATED .
SOIL (UG)
2,436*
ND
ND
202
7,560
28,980
6
109
756
3,150
3,654
PCP AND COO/CDF
RESULTS
TEST TWO
MASS IN MASS IN
REMOVAL UNTREATED TREATED REMOVAL
EFFICIENCY (X) SOIL (MG) SOIL (MG) EFFICIENCY (X)
91X 3,563 J 455 J NA
NA
NA
83X
72X
70X
NA
NA
66X
73X
69X
XA « Not Applicable (because  results are below detection limits or not detected)

•
   • (UG) « micrograms, pentachlorophenol results in mi IKgrans


lote:  Dioxin analysis was  not performed on soil from drum 2.
                                                           446

-------
  NAPHTHALENE
  ACENAPHTHALENE
        CHRYSENE


BENZ(A)ANTHRACENE
  ACENAPHTHENE
BENZO(B)FLUORANTHENE
      FLUORANTHENE
                                    BENZO(K)FLUORANTHENE
    PHENANTHRENE
                                     BENZO(A)PYRENE
      ANTHRACENE
                      1NDENO(1,2,3-CD)PYRENE
FLUORENE
                             DIBENZO(A,H)ANTHRACENE
   PYRENE
                                BENZO(G,H,F)PERYLŁNE
                                                 FIGURE 4-1
                                            STRUCTURE OF MAJOR
                                               COMPONENTS OF
                                                 CREOSOTE
                                                  OENOWjflOeM4S.1V 5-7-90 1:1
                       447

-------
                            TABLE  4-2
           CRITICAL FLUID EXTRACTION TREATABILITY STUDY
Chemicals of Concern

PAHs:
  Benzo(a)Anthracene
  Benzo(a)Pyrene
  Benzo(b)Fluoranthene
  Benzo(k)Fluoranthene
  Benzo(g,h,i)Perylene
  Chrysene
  Dibenzo(a,h)Anthracene
  Indeno(1,2,3-cd)Pyrene

  Naphthalene
                                        Test  1
         78 /
         36 /
         38 /
         39 /
         16 /
         85 /
        194 /
         15 /

        105 /
 8
12
11
14
12
 9
 4
11
                        Test 2
  19 /
 322 /
 268 /
1797 /
1583 /
 590 /
 724 /
 376 /
 nd
 25
 23
200
190
100
130
100
        169 /   75
Total PAHS
Total BAPE*
       2502  / 123
        175  /  16
       6983 / 1104
        875 /  121
Pentachlorophenol
Total TCDDE
        295  /  55
        5.3  /  2.7
         47 /
         17
Chlorinated Dioxins:
Tetra-, Total
Penta-, Total
Hexa -, Total
Hepta-, Total
Octa -, Total
Chlorinated Furans:
Tetra-, Total
Penta- , Total
Hexa -, Total
Hepta-, Total
Octa -, Total

nd
nd
16
360
1300

nd
1
30
160
160

/ nd
/ nd
/ 5
/ 180
/ 690

/ nd
/ 3
/ 18
/ 75
/ 87
C0::Concentration Before
Cx::Concentration After Extraction
nd::Not Detected; 1/2 detection  limit used for those compounds
    known to be present at the site.

 *::Concentrations for PAHs in Test  1  are  averages based upon
    duplicate samples (both C0 and Cx)  available  for analysis.

**::Benzo(a)Pyrene equivalency, only carcinogenic compounds are
    included in  equivalencies;  i.e., those PAHs  listed,  less
    Naphthalene; the  2,3,7,8-disubstituted isomers of dioxin and
    furan.
                               448

-------
Soils handling problems  that  were encountered due to limitations
in the design of the pilot unit should not be considered along with
legitimate concerns for  full  scale design.   In other words, some
of the  problems  encountered  in  the pilot study were due  to the
generic design of the pilot unit.  The pumps, pipe sizes, cooling
capacity, etc., of the pilot unit were designed  for the wide range
of operating conditions  to which  a pilot unit  is  subjected.
Proper  selection  of  pumps and pipe sizes for  a  full  scale unit
would eliminate several  pilot scale materials handling problems.
Also, if accumulator  tanks  were  designed to be emptied through a
hopper-type mechanism instead  of  through  valves, plugging problems
encountered during blowdown procedures could be  eliminated in full
scale treatment.

Another  intrinsic  pilot  scale  problem  involved  temperature/
pressure capabilities. The pilot unit is designed  to operate under
limited pressure  conditions.   Cooling  of the  unit  and complete
condensation of the  propane is provided  by tap water circulated
through  a  condenser.   When  ambient  temperature  at  the  site
increased, there  was  some trouble  maintaining  a  complete  liquid
phase of propane.   This  problem  is  eliminated  on the full scale
unit by two design modifications:  1) higher  operating pressure and
2) inclusion of a chiller to provide colder cooling water.

The proper choice of  pumps in  full scale  design  is very important.
Pumps that are designed to handle slurries with varying  solids
composition should be chosen.  Variable  speed pumps are required
since control  of  the  flow rate through  the unit  can effectively
eliminate  several handling  problems associated  with lighter soil
particles that are pumped too fast.

Careful  consideration should  be given  to  design  of  dewatering
equipment.  The differences in settling properties  of the two types
of soil treated  in  the pilot  study are  anticipated to  cause
variable dewatering conditions in  full scale treatment.  The light,
black tar  mat particles in the  residential soil tended to stay
suspended  in water, even after centrifugation.

Proper feed preparation equipment will also be crucial to effici-
ent pretreatment.  There was a significant difference between the
two soils  regarding the  material  that was dry sieved through the
1/2 inch screen.   The >  1/2 inch material of  the  waste pond soil
consisted mainly  of large gravel and pieces of wood.  The >l/2 inch
material from  the residential surface  soil consisted  mainly of
organic material  (leaves, grass,  small  gravel).  Proper crushers
or shredders to handle this material must be included in full scale
design. The >l/8  inch material that was sieved after the soil was
slurred also differed  between  soil types.  The  >l/8 inch waste pond
material was mainly composed of uniformly sized gravel.   The >l/8
inch material  from the  residential soil contained more  organic
matter than  rock.   Therefore, it  is  important  that pretreatment
equipment  be selected  that  is  capable  of providing  effective
handling of a variety of soil  types.  If the screened material is
to be crushed and treated, these  factors  should  also be considered
in pump selection.


                               449

-------
Another important design consideration deals with the configuration
of the vessels, mixers, and piping.   Although complete mixing is
not possible in any vessel, some  of  the handling problems in the
pilot unit could be rectified  by changing the process flow in full
scale treatment.  The loss of material during treatment is mainly
attributed to the fact
that the  slurry flows out of  the top of the  extractor into the
decanter.   Heavier solids that  do  not stay suspended  will not
overflow  into  the decanter,  but will  settle in  the  extractor.
Process flow should be designed so  that maximum  mixing and easy
flow-through of material can be achieved to minimize losses due to
settling and sticking.  Efficient  mixers should be  included in the
extractor design.

General maintenance and mechanical problems on the  full scale unit
will be similar to those of the pilot unit.  However, some of the
mechanical  problems with  the pilot unit were  due  to  lack  of
preventive maintenance  and wear and tear  from transportation to
many sites.

From  a solids  handling standpoint,  if the  problems  that  were
encountered in the pilot study are given serious consideration in
the  design of  a  full scale  unit,  full scale remediation using
solvent  extraction  technology  appears  feasible.    Particular
attention  should be given to  feed preparation equipment, pumps,
dewatering equipment,  pipe  size, efficient process flow,  and vessel
blowdown procedures.
admin:crflpapr.425              459

-------
                           SECTION 5.0

                   TREATABILITY CONSIDERATIONS
Some of the questions which were  asked  in the case of evaluating
the feasibility of using the Critical Fluid Extraction Technology
at United Creosoting included:

     •    Can  contaminants  be  extracted  from the soil  down  to
          health based levels? Levels associated with ARARs (i.e.,
          LDRs)?

     •    Is the process realistically controlled through various
          parameters in order to increase removal efficiency?

     •    Once treated,  can the soils be replaced without concern
          of residuals from the treatment process?

     •    Does  implementation of  the process  present short-term
          risks to human health or the environment? Air emissions,
          etc.?

     •    How can the extract  be handled (i.e., further treatment,
          off-site disposal)?

Health based remediation levels for soils at United Creosoting were
determined  from an EPA in-house  risk assessment  [documentation,
1989  ROD].    In addition, the  soils at  United  Creosoting  were
determined to be contaminated with a listed hazardous waste under
the  Resource  Conservation  and Recovery  Act  [RCRA],  K001  Wood
Preserving Waste  [1989 ROD, 40  CFR 261.32].   Treatment standards
for  nonwastewaters  of  this  category,   invoked   under  the  Land
Disposal Restrictions [LDRs]  of RCRA, included chemicals separate
of  the risk-based contaminants  assessed for human health  and
environmental impacts.

Results of the treatability tests were evaluated as favorable since
the toxicity of PAHs, dioxins,  and PCP  had been  decreased if not
to, then close to,  the desired concentration levels in soils; those
contaminants of concern under LDR were similarly reduced.  A full
scale unit, with  the process  designed specifically for the soils
at United Creosoting, could very likely meet,  if not better, those
levels necessary for allowing soils to be placed back onsite.
          Under Superfund, keeping  treated  soils onsite would be
          preferred to transporting such a large quantity off-site
          to a commercial facility.
                               451

-------
Most of the  remaining questions were easily answered  due  to the
properties inherent of the process, as previously discussed.  For
example, the use of propane as the  solvent near the critical state
should not leave significant residue  in the treated soil.  Several
factors influence the efficiency of the process and these factors
can be modified to  increase the efficiency  over the pilot unit's
performance.  Although  removal  efficiency may be based upon the
mass transfer of contaminants from  the soil, the concentrations of
contaminants remaining  in the  soil  is  of most  consequence;  the
efficiency of the process to decrease toxicity of the soils can be
improved.   Finally, the process is conducted in an enclosed unit,
and space  is  available  on the commercial portion of the site to
manage the treatment of soils with  minimal risk to workers and the
surrounding community.

Treatment  alternatives  for  a  Superfund site  must satisfy such
fundamental requirements in order to contend for remedy selection
under the  Superfund program.   In  many cases,  various treatment
technologies may be combined to  form  alternatives which adequately
address individual  site  circumstances.   In  the case  of  United
Creosoting, application of the Critical Fluid Extraction Technology
would mean that once contaminants were removed from the soils, the
resulting  concentrated  organics  must also be  dealt  with in order
for the remedy to be complete.

In consideration  of the small  volume of extract  and  the charac-
teristics  of the  concentrated contaminants, and due to the close
proximity  of the  residential  area,  off-site  management  of  the
concentrate  was selected.   Incineration  (the  Best Demonstrated
Available  Technology  [BDAT]  for dioxins -  total destruction) was
selected as the most effective means of permanent remediation for
site contaminants.4
          As  a  separate  alternative,  incineration of  the soils
          presents  an  interesting  point of  comparison  for  the
          effective use of more than one treatment technology.

          For  example,   off-site  incineration  of  soils  would
          necessitate  off-site transport  of a  large  volume of
          contaminated  material and  ultimately  a large volume of
          ash  for disposal  (off-site).   Onsite  incineration of
          soils would necessitate a temporary incinerator being set
          up fairly close to the residential area; a similar plan
          opposed in past discussions with the residents at United
          Creosoting.

          In comparison,  onsite critical fluid extraction coupled
          with off-site incineration of the  extract offers certain
          advantages:  a  tremendous  reduction  in  the  volume of
          contaminated  material   to  be  shipped  off-site  for
          destruction, with minimal ash anticipated from the liquid
          concentrate.  In addition,  the critical fluid  extraction
          process  is  entirely enclosed and  is  perceived  more
          favorably by the public than an onsite incinerator — so
          close to the  residential area.
admin:erfIpapr.425
                                452

-------
                           SECTION 6.0

                            REFERENCES
Nelson, W. L., 1969.  Petroleum Refinery Engineering, McGraw-Hill
Book Company, Fourth Edition, 1969.

Texas Water Commission,  1989.   United Creosoting Superfund Site,
Feasibility Study Amendment - Preferred Alternatives Analysis, July
1989.

Treybal, R. E., 1980.  Mass-Transfer Operations, McGraw-Hill Book
Company, Third Edition, 1980.

U.S. EPA,  1989.    Record of Decision for the  United Creosoting
Superfund Site, Conroe, Montgomery County, Texas, September 1989.
admin:erfIpapr.425
                                453

-------
       	C«« Study	

       Bench-Scale Solvent Extraction
                    May 1990
CKMHILL                   Joseph A. Sandrin
Milwaukee, Wisconsin ^___^_i^^ John Flelssner
                   454

-------
                                       Case Study:
                           Bench-Scale Solvent Extraction
            Treatability Testing of Contaminated Soils and Sludges
           from the  Arrowhead Refinery  Superfund Site, Minnesota

                                      Joseph A. Sandrin
                                        John Fleissner
                              CH2M HILL, Milwaukee, Wisconsin
ABSTRACT

        Solvent extraction is a separation process that has emerged as an  effective hazardous waste
treatment technology.  It  has been  applied successfully to industrial wastewaters, soils, and sludges
contaminated with hydrocarbons, petroleum products, and heavy organic compounds.
        Solvent extraction was evaluated as an alternative  treatment technology at the Arrowhead
Refinery Superfund Site in Hermantown, Minnesota, the location of a former waste oil recycling facility.
Highly acidic, metal laden sludge bottoms and oil saturated clay filter cake were disposed of in a 2-acre
lagoon.  The peat layer underlying the lagoon and the surrounding soils are contaminated with oil,
metals, and numerous organic compounds.   Under contract to CH2M HILL, Resources Conservation
Company (RCC) conducted bench-scale tests of its Basic Extractive Solvent Technology (B.E.S.T.™).
The results of bench-scale testing and a discussion of the applicability of the process to the wastes at this
site are presented.
INTRODUCTION

Site History
        The Arrowhead Refinery Site occupies about  10 acres in northeast Minnesota  near  Duluth
(Figure 1). According to the Minnesota Pollution Control Agency (MPCA), milk cans were retinned at
the site before 1945.  From  1945 to February 1977 the site was used as a waste oil recycling  facility.
During oil processing, waste oil was treated with sulfuric acid to deemulsify the oil/water mixture. The
wastewater recovered from this process was discharged to the wastewater ditch, and  the waste  oil was
filtered through a clay/sand filter. The sludge from the deemulsification process and the filter cake were
disposed of in an unlined 2-acre  lagoon in a wetland on the site. The filter cake was also used as fill in
the process area adjacent to the sludge lagoon (Figure 2).

Nature and Extent of Contamination
        The U.S. Environmental Protection Agency and MPCA investigated the environmental effects of
onsite waste disposal from 1979  through  1984. The results of their investigations indicate that  various
organic and inorganic hazardous substances are present at the site in the subsurface soil, sediment,
surface water, groundwater, and sludge lagoon. The two major contaminant sources defined during the
remedial investigation (RI) were the sludge and filter cake disposed of in the  sludge lagoon and the
contaminated soils in the process area.
        The surface soils consist of gravelly sand, silt, and fill that was deposited during site operations.
Much of the soil is visibly stained and saturated with waste oil.  The lagoon contains a viscous, black oily
liquid sludge, and a black filter cake that consists of an oily clay and a silty sand and gravel fill layer.
The entire lagoon  is underlain  by  peat that appears to  persist  across the  site and  to be highly
contaminated.  Hazardous substances  detected at the site included polycyclic aromatic hydrocarbons
(PAHs), volatile organic  compounds, lead, zinc, and  small  quantities of polychlorinated biphenyls
(PCBs). Estimated contaminated waste volumes as known at the time of the treatability study are:


                                         455

-------
ARROWHEAD
REFINERY
                                             FIGURE 1  Kw.
                                          VICINITY MAP
                      456

-------
                 LEGEND

                *-—•••  EPA DITCH

                ........ SITE BOUNDARY


                 NOTE: Arrows Indicate direction of (low.
rr
                     FIGURE 2 r*:ri:nii
                     SITE MAP •••

-------
        Soil              19,000 yd3
        Peat              13,000 yd3
        Sludge             4,600 yd3

        As documented in its Record of Decision (ROD) for the site, the  EPA's selected remedial
action  was thermal treatment.  The  EPA and  MPCA were  both interested in the application of
alternative treatment technologies that  might achieve similar levels of treatment more economically than
thermal treatment.  As a result, the EPA agreed  to fund a treatability study using a solvent extraction
treatment process.
        CH2M HILL had already performed a remedial investigation and feasibility study of the site for
the EPA.  Under contract to  the EPA, CH2M HILL  subcontracted the  treatability study to Resources
Conservation Company (RCC) for testing of its Basic Extractive Solvent Technology (B.E.S.T.™).
SOLVENT EXTRACTION

Background
        Solvent extraction technology has been used  for many years as solvent leaching to recovery
valuable minerals from ores, to remove unwanted materials from coal processing operations, and to de-
oil quench waters in refinery processing operations.  More recently it has been  used to treat sediments
and soils contaminated with PCBs, chemical manufacturing wastes, and oily, hazardous and toxic wastes.
        Organic solvent extraction is particularly suited for treatment of oily wastes because the process
separates wastes into product oil, solids, and water fractions.  Solvent extraction can effectively extract
the oil fraction of a waste including such contaminants as PCBs. The remaining solids can sometimes be
disposed of as non-hazardous wastes and the water discharged to a wastewater treatment plant.
        The  successful application  of solvent  extraction  depends  on  solvent  selection,  process
configuration, the nature of  the waste and  the contaminants, the economic value of the potentially
recoverable compounds, and the treatment or remedial goals.  Solvent extraction can be an especially
attractive treatment  alternative where (1) valuable products  can  be recovered from the waste; (2) the
process can yield nonhazardous  or uncontaminated residual products;  and (3) inordinate wastewater
disposal or air emissions problems are not encountered.


B.E.S.T.™ Solvent Extraction Technology
        RCC is the owner of the B.E.S.T. solvent extraction technology, a patented process that  takes
advantage of the peculiar solubility behavior of certain aliphatic amines.  The process  was originally
developed for NASA as a waste treatment process for  space travel, where resource recycling is critical.
        Triethylamine (TEA) has chemical and physical properties that make it a good candidate for use
in solvent extraction. At temperatures below 65°F, TEA is completely miscible with water arid is also a
good solvent for many organic compounds such as PCBs, PAHs,  and petroleum products. At 65°F the
soluble organic and water components of a waste  can be separated from the solid components.
        When TEA is heated, the solubility of water in TEA drops to less than 2%, separating the  water
fraction from the soluble organic fraction. The TEA is then removed to yield an organic fraction. The
B.E.S.T. process separates the waste into three waste product streams or fractions: a solid with soluble
organic contaminants removed; a water that may require treatment before discharge; and an oil product
that can be recycled for energy recovery or incinerated (Figure 3).
         TEA is  a chemically basic compound that reacts with acids in the  waste to yield ammonium
salts.  Excessive  reaction will result  in the loss  of expensive solvent.   To  minimize solvent loss the
B.E.S.T. process includes the addition of caustic to increase the pH of the waste to greater than 11. The
high pH minimizes solvent loss  and has the side  benefit of precipitating low concentrations of metals
into  the product solids and possibly decreasing the teachability of metals in the toxicity characteristic
leaching procedure (TCLP).
                                            458

-------
                                                                                  0.065569 TS Rl SM* FIG 1 5-10 «0
               Contaminated Sludge,
               Peat, or Soil
-Pa
cn
                                           B.E.S.T.^ Process
                                        I Oversized
                                        I Contaminated
                                        I Material

                                      ^
                                                                     Product Oil
Product Solids
                                                                     Product Water
                                                            ®             FIGURE 3
                                                      B.E.S.T. BENCH-SCALE SOLVENT
                                                      EXTRACTION PROCESS STREAMS

-------
        RCC conducts bench-scale treatability studies at its laboratory to evaluate  the ability of the
process to treat a given waste.  In the treatability tests, 1-kilogram batches of waste are subjected to the
same unit processes as a full-scale facility to simulate full-scale operation.  Performance is evaluated at
each step, and samples of the products and process intermediates are analyzed to determine if the wastes
are being treated effectively.


TREATABILITY TESTING
       CH2M HILL developed a sampling and analysis plan to evaluate the treatability study.  Besides
the wastes processed to meet RCC's treatability protocol, additional samples were processed  to generate
additional quantities of treated solids and treated oil product.  Samples of the three raw  wastes and
treated solids and treated oil from each raw waste were analyzed through the EPA's Contract Laboratory
Program (CLP) to provide an independent assessment of process performance.  Samples were analyzed
for metals, PAHs, and PCBs.  Because of cost limits, it was not  possible to process sufficient waste to
provide treated water  for CLP analysis.
       Lead and PAHs are the primary contaminants at the site. Low levels of PCBs are also present.
The overall purpose of the treatability study was to assess whether a solvent extraction process could
effectively treat the  Arrowhead Refinery site wastes and to assess whether it could be a cost-effective
alternative to incineration for site  remediation.
       Samples of contaminated soil, sludge, and peat wastes were sent to RCC in May 1989 for bench-
scale treatability testing. An observer from CH2M HILL was present at RCC's laboratory  throughout
the bench-scale test.  RCC conducted its tests according to the protocol presented in Figure 4.

Sample Preparation
       Before bench-scale testing began, the raw waste feed composition (Table 1) and various feed
preparation steps were performed. Samples of the soil, peat, and sludge were screened through a 1A*
screen. Samples were then mixed  with caustic to determine the amount of caustic required  to raise the
pH to  11  as  required before the addition of TEA.  The sludge and peat had low pH and substantial
caustic addition was required.
Table 1
RAW WASTE FEED COMPOSITION
RCC Analysis
Waste
Type
Sludge
Peat
Soil
Oil
(wt%)
42
22
5
Water
(wt%)
43
53
12
Solids
(wt%)
15
25
83
Ash (550'F)
(wt%)
6.3
9.1
79
        After determination of caustic requirements, samples were mixed with required amounts of 50%
sodium hydroxide and the TEA was added. Each waste type was extracted three times with TEA and the
extracts were combined.  The first extraction was performed with cold TEA (40°F) to maximize water
solubility.  Subsequent extractions were performed with heated TEA (140°F) to  maximize extraction  of
organic constituents.  Solids were separated from the TEA/water solution by centrifugation.
        The extracts were combined and heated  to  140°F to decrease  water solubility.  Under ideal
conditions  the water should have separated cleanly from the TEA, and the two layers would be separated
                                              460

-------
                Raw Waste Feed
              (sludge / peat / soil)
      Contaminated
         Rajact
      ^ Material
Screened
 Waste
  Feed
                             NaOH
                               TEA
H20
                          TEA
                                  40°F
                                 1st Extraction
                                     i
                                 Centrifugation I
                                 TEA

                                iJ
                                       Solids
                                2nd Extraction
                                «^^^^J^»»~^
                                Centrifugation I
                                 TEA

                                11
                                  140°F
                                       Solids
                                 3rd Extraction
                                      I
                                I Centrifugation

                                 Solids I

                                   JL
                          TEA
                   NaOH  (recycle)


                    I     t
                                                       Decantatlon

                                                       7  140°F
                                                      Water /
                                                       TEA
                                               Liquid
                                             (water / oil /
                                               TEA)   ^
                                                        Evaporation
                  Liquid
                 (oil / TEA)
                                                     Liquid
                                                    (oil/TEA)
 Drying
        PRODUCT
         SOUPS
                               PRODUCT
                                WATER
                                            TEA
                                          (recycle)
                                oil /TEA
t
                                          Distillation
                                                                            PRODUCT
                                                                               OIL
•a
e
                          a    FIGURE 4
                  B.E.S.T.® GLASSWARE
          BENCH-SCALE PROCESS FLOW
                                           461

-------
 in a separatory funnel.  Separation was not good for any of the wastes, and it was necessary to recover
 the water from the mixture by evaporation.
        The TEA was then removed by distillation and steam stripping, leaving the treated oil fraction.
 The treated water fraction was recovered from the evaporation, and the treated solids were recovered
 from the centrifuge. Samples of each treated product and the raw wastes were analyzed for PCBs, lead,
 and PAHs.  Mass balances were performed to compare actual recoveries of treated products to predicted
 values from the initial waste composition analysis (Table 2).  RCC also performed a mass balance for
 lead comparing the lead content of the raw wastes to the total lead content of the treated products.
Table 2
PHASE FRACTION MASS BALANCE
RCC Analysis
Waste Type
Sludge
Oil
Water
Solids
Peat
Oil
Water
Solids
Soil
Oil
Water
Solids
Compositional Assay
(wt%)
42
43
15
22
53
25
5
12
83
Bench-Scale Test
% Recovery (wt %)
43
14
13
19
16
22
5
7
82
Analytical Results
        Mass balance recoveries for solids and oil were acceptable compared to the initial compositional
analysis.  However, recovery of water was poor and RCC was unable to account for the losses (Table 2).
Lead recoveries were 80% from the soil, 180% from the peat, and 200% from the sludge. The high lead
recoveries may be due to the oily matrix of the raw wastes that interfered with lead analysis. The solvent
extraction process separated and concentrated much of the lead in the treated solids product. Based on
the  high lead recoveries, it  is possible that the actual  lead concentrations in the raw wastes are
substantially higher than indicated by the raw waste analyses.
        The treated product water had no detectable contaminants except for elevated pH due to the
TEA distilled with the water in evaporation.  The low level of contaminants should be expected since
the water was recovered by evaporation.
        Solvent  extraction concentrated  the inorganic  contaminants in  the treated  solids  while
concentrating the organic compounds in the oil fraction separated from the TEA. The treated solids
were also analyzed using the EP toxicity procedure.  The solids from the peat and the sludge were at or
exceeded the allowable lead concentration of 5 mg/1 in the extract and  would be considered hazardous
wastes (Table 3).
                                             462

-------

Waste
Type
Sludge
Peat
Soil

Waste
(mg/kg)
12,700
13,000
2,600
Table 3
LEAD Analysis
CLP Analysis
Product Oil
(mg/kg)
6,500
6,000
1,700
Product
Solids
(mg/kg)
32,000
38,000
2,900
Product Solids
EP Toxicity
(mg/l)
7.9
5
1.9
        The organic contaminants were concentrated in the treated oil fraction.  However, substantial
concentrations of lead were  also detected in  the  treated oil fraction.  The concentrations were high
enough that it might be necessary to dispose of the oil as a hazardous waste by incineration.


PERFORMANCE EVALUATION

        The B.E.S.T. solvent extraction  process was able to successfully separate  the refinery wastes
(sludge, peat,  soil) into three treated product fractions (water, solids, oil).
        The treated product  water could possibly be discharged to a publicly owned treatment works
(POTW) with little or  no additional  treatment  following pH adjustment, since it was recovered by
distillation.  However, water  mass balance showed poor recovery for  water, and RCC was not able to
explain this satisfactorily.
        The treated product  oil was high in total lead content and might have to be disposed of as a
hazardous  waste.   The  changing  nature  of  the waste oil market  would not allow  final disposal
requirements to be determined until the oil was available.
        Treated product solids  would probably have to be disposed of as a hazardous waste because of
the high total  lead content in  all three treated solids products. The solids from the peat and sludge were
at or  above the allowable limit  for  the  EP toxicity  extraction  procedure, which  designated  them
hazardous wastes at the time of the study.  The  evolving RCRA  land ban  regulations and the recent
adoption of the TCLP test to replace the EP toxicity test increase the likelihood that the solids would
have to be disposed of as hazardous wastes.
        The estimated  hopper-to-hopper treatment  cost  for  solvent  extraction  in  this  case  was
comparable to  incineration.    The estimated cost for solvent  extraction  provided by RCC  was
approximately $289/ton as compared  to  $300/cubic yard for incineration.  The requirement to use
distillation  to  recover  the water  because of poor decantation  in  the extraction  step contributes
significantly to the treatment cost for solvent  extraction, as does the high caustic requirement for pH
adjustment of the acidic waste materials.
        The cost  estimate for solvent  extraction  is  from the  vendor and is  based  on  minimal
demonstrated  field experience, whereas  the estimate  for incineration  is based on actual costs from
remedial actions at other sites. The uncertainty of the solvent extraction cost estimate is,  therefore,
likely to be greater than that  developed for incineration.
        Materials handling considerations would also affect the implementability of a solvent extraction
remedy at this site.  Solvent extraction processes typically require that  the waste be delivered to the
treatment system at sizes  less  than 1 inch, as opposed to a range  of 4 to 6 inches for rotary kiln
incinerators.  At a site like Arrowhead, with the large  debris, tree stumps, and other oversize wastes,
considerable waste preparation would be required  for incineration, and even more for solvent extraction.
                                               463

-------
        The poor water recovery demonstrated in the bench-scale testing, and the materials handling
questions would require that a pilot-scale  study  be performed  to  better evaluate performance  and
estimate costs for solvent extraction.
        The  bench-scale tests did not monitor or evaluate  the  fate of volatile contaminants in the
treatment system. The site is known to contain substantial volatile contamination. The fate and control
of volatile emissions would have to be addressed in any selected remedy at the site.
        It is difficult for new and innovative technologies to demonstrate a substantial savings based on
the uncertainties inherent  in  the fact  that  they are new and innovative.   If  new or  innovative
technologies  are to be selected and used they must  therefore offer other advantages to overcome the
uncertainties. The recovery of a potentially recyclable product and a reduction of the volume of waste
that must be  handled as a  hazardous waste are examples of such advantages. Unfortunately, at this site,
the difficult nature of the wastes did not provide demonstration of such process advantages.
CONCLUSION

        The complex nature of the wastes at this site, the low pH, and the high lead content make the
wastes at the Arrowhead Refinery site difficult to treat using any technology.  Solvent extraction did not
present an advantage over incineration in performance or cost.  It was, however, able to separate the
water, solids, and oil fractions from the waste.  At  properly selected sites with a less difficult matrix,
solvent extraction may offer more of an advantage as  a remediation technology.  It should continue to be
examined and evaluated at sites where it is appropriate.
GLT977/022.51
                                                 464

-------
FIELD DEMONSTRATION OF A CIRCULATING BED COMBUSTOR (CBC)
       OPERATED BY OGDEN ENVIRONMENTAL SERVICES OF
                    SAN DIEGO, CALIFORNIA
                    Prepared for Presentation at:

                     EPA's Second Forum on
          Innovative Hazardous Waste Treatment Technologies:
                     Domestic and International
                        May 15-17, 1990
                          Prepared by

                        Nicholas Pangaro
                        Charles W. Young
                        Douglas R. Roeck
                                                            ALLIANCE
                                                      JWt TK,-,:-.; .:

                              465

-------
    FIELD DEMONSTRATION OF A CIRCULATING BED COMBUSTOR (CBC)
            OPERATED BY OGDEN ENVIRONMENTAL SERVICES OF
                            SAN DIEGO, CALIFORNIA
This paper describes several support aspects of a TSCA Demonstration Test Burn and  site
remediation program conducted by Alliance Technologies Corporation and Ogden Environmental
Services (OES) at the Swanson River Oil  Field on the Kenai Peninsula, along the south Coast
of Alaska.  The intent is to summarize the planning requirements associated with conducting
testing of this nature at a remote site, to describe problems encountered and problem resolution,
and to summarize the results of the Demonstration  Test Burn.  Because of the site location,
Alliance focused strongly on project planning to anticipate and minimize technical and logistical
problems that otherwise could have caused  significant time delays and additional costs. Problems
that did arise were successfully addressed. The merits of careful, conservative planning should
become clear through reference to this case study.

HISTORY OF THE PROBLEM1

PCBs were used  at the Swanson River Field between 1962  and  1972 as a coolant material in
electrical transformers and in heat transfer oils.  A compressor  explosion  occurred in 1972,
resulting in the release of an unknown quantity of PCB laden oil.  Later, in 1983 and 1984, PCB
contaminated sand and gravel from the explosion area were used for dust suppression along  two
miles of roads within the Swanson River site.  Soil sampling conducted during and after the
summer of 1984 demonstrated PCB contamination throughout the area. Based on extensive site
characterization studies, most of the identified contaminated soil areas were shown to have PCB
levels of less than 50 ppm.  A  cleanup target level of 12  ppm PCB was  established with a
relaxed target of 24 ppm applicable to areas with difficult access. About 75,000 tons of soil are
estimated  to be contaminated, requiring  treatment.   PCB  mixtures  of concern are primarily
Aroclor 1242/1248.

PROBLEM RESOLUTION

A feasibility study was conducted to  identify alternative cleanup methodologies.  This study
resulted in a list of five possible thermal technologies and six possible non-thermal technologies.
Based on cost, feasibility, permitting and  policy issues, cleanup capability, future liability, and
other related considerations, the list of candidate technologies was narrowed to the five  thermal
techniques or landfilling.  Landfilling was ruled out due to concerns about future liability.

Based on  the results of  the feasibility  study, the  facility owner  prepared a  remediation
specification and released it for open competition.   The solicitation process resulted in the
selection  of Ogden  Environmental  Services  (OES) of LaJolla, California to conduct the
remediation program. OES owns and operates  transportable incineration units which are based
on the OES circulating fluidized bed combustion technology.
                                                                            ALLIANCE
                                          466

-------
THE OGDEN CBC2

The Circulating Bed Combustor (CBC) is an advanced fluidized bed system that employs high
velocity combustion air (14-20 ft/sec) to circulate solids through a highly turbulent loop. The
system's high  combustion efficiency is  well suited to burning materials  with relatively low
heating value, such as contaminated soils.  Residence times range from 2 seconds for gases to
30 minutes for circulating solids.  Waste solids and limestone are combined and added to the bed
at an  inlet located  between  the  cyclone  and  the combustion  chamber.   Upon calcination,
limestone reacts with acid gases such  as HC1 and SO2 to form  calcium chloride and calcium
sulfate.   The feed materials are rapidly heated after introduction to the unit.  A uniform
temperature (over the operating range of 1450-1800°F, ± 50°F) is maintained in the system  by
virtue of the highly turbulent well-mixed combustion zone, the circulating soil and limestone, and
use of auxiliary fuels.  Ash is periodically removed and cooled. Hot off gases pass through a
flue gas cooler/feed water heater and then to a baghouse for removal of fine particles,  prior to
exiting the stack. A schematic diagram of the CBC is shown in Figure 1.

OPERATING REQUIREMENTS

In order to obtain the requisite permits for conducting on-site destruction of PCB (>500 ppm) or
PCB contaminated  (50-500 ppm) material, EPA's Office of Toxic  Substances (EPA-OTS)
imposed a number of  requirements for data  gathering and monitoring.   To  comply with
regulations promulgated under the Toxic Substances Control Act (TSCA), OES was required to
demonstrate:

             99.9999% destruction and removal efficiencies for PCBs;

       •     99% removal of hydrochloric acid; and

             paniculate emission rates of less than 0.08 gr/dscf.

Due to the relatively low levels of potential HC1 precursors present in the soils being incinerated
(600-700 ppm total PCB), the less stringent RCRA standard for HC1 removal (total emissions of
less than 4 pounds per  hour) was applied to this case in  lieu of demonstrating 99% removal
efficiency.  However, in addition to the regulatory requirements, OES  was required to:

       •     monitor for a  number of potential products of incomplete combustion (PICs)
             during the Demonstration Test Burn, including a demonstration that the stack
             exhaust contained less than 10 ng/m3 of polychlorinated dibenzo dioxins and
             furans; and

       •     continuously monitor levels of PCBs in the solid effluents from the incinerator
             (bottom and baghouse ash) to assure that materials contaminated with PCBs above
             2 ppm were not redeposited at the site.
                                                                           ALLIANCE
                                         467

-------
                                                                  STACK
COMBUSTOFK.
                                 COOLING
                                 WATEn
        Figure 1. Schematic of the Ogden Circulating Bed Combustor
                                                                  ALLIANCE
                                 46R

-------
ALLIANCE'S ROLE

OES secured the services of Alliance Technologies Corporation of Bedford, Massachusetts to
assist in securing proper operating permits. Alliance's role was to address test requirements to
be used during the test burn and to implement the test burn. In addition, Alliance assisted in the
set up  and initial operation of an  on-site laboratory where routine analysis  and monitoring of
incinerator effluents and feeds would be conducted.

ON-SITE LABORATORY OPERATION

During routine operations, it  was necessary to verify decontamination as batches of soils were
treated by the CBC.  Since the storage capacity for post treatment soils at the site was limited,
the need for rapid turn-around analysis for PCBs was obvious.  In addition, analysis of the feed
materials was required to provide  information on  the amount of limestone needed to maintain
CBC operation at approximately 10% above stoichiometric conditions, to record the amounts of
PCB material treated, and to  verify compliance with permit requirements on concentration and
homogeneity limitations. Considering the remoteness of the site, the ability to obtain rapid turn-
around analysis at an off-site location was not possible, leading to the establishment of an on-site
laboratory. This on-site laboratory, housed in a 40'xl2' skid mounted transportable building, was
equipped with preparatory equipment and  a Varian Model 3400 Gas Chromatograph with an
Electron Capture Detector  (GC/ECD).

Although the methods used for these analyses were based on standard EPA procedures, certain
adaptations were made to allow for the limitations of working in a field laboratory and to provide
data  in a rapid (overnight) manner. Standard Operating Procedures (SOPs) were prepared to:

       •      Supply personnel at the on-site laboratory with guidance on analyzing both pre-
             and post-treated soils for PCB contamination levels;

       •      Define the application of the standard methods to this project and modifications
             made to allow for the limitations mentioned;

       •      Define acceptance/rejection criteria for evaluation of the analytical data.

Since the PCB material found in the contaminated soils were a mixture of several individual PCB
isomers, it was likely that residual contamination would not retain  its identity as an Aroclor
mixture in the post-treated soils. The analytical scheme made allowances for this by using two
sets of parameters for analysis.  Samples from the pre-treatment streams (feeds) were analyzed
for the predominant PCB mixtures  on the site, Aroclor 1242/1248.  Samples from post-treatment
solid streams (treated soils and ash) were analyzed for PCB homologue groups (mono- through
deca-chlorinated biphenyls) to verify treatment to a  level below the limit applied for landfill
disposal, i.e.,total PCB material <2.0 ppm.  In order to make the analysis as routine as possible,
Alliance  developed  several   computerized calculation  schemes, based  in  a  LOTUS-123
spreadsheet, to allow fast reduction of raw analytical data.
                                                                            ALLIANCE

-------
Since the period required to obtain and ship replacement parts to the South Coast of Alaska made
potential down time unacceptable, the laboratory was equipped with an overabundance of backup
materials, including a  complete set  of  expendable  and breakable items for the  GC/ECD.
Additional problems were encountered in supplying power to the laboratory as the analytical
instrumentation and computers proved to  be sensitive  to the industrial grade power available at
the site.  This problem was finally overcome by  supplying power to the laboratory from a
dedicated generator.

DEMONSTRATION TEST BURN PROGRAM

Logistical Considerations

Due to the location of the site, planning for the Demonstration Test Burn required much more
attention  than is  usually associated  with such  efforts.  Although Alliance had previously
conducted Trial Burns at locations which  are quite distant from its Bedford, Massachusetts base
(for instance, Houston,  Texas, McDowell, Missouri, and San Diego, California), the difficulty
involved  in procuring additional equipment or spare  parts within a short  time frame required
packing redundant materials, supplies, and instrumentation to be used on the program.  Secondly,
transportation of equipment and personnel to a remote  site in Alaska provided a challenge in
logistics and scheduling.

Two Alliance employees  provided overland transportation of equipment to the Port of Seattle,
Washington in a company truck, a five-day trip.  The truck was then loaded on a barge for the
three-day journey to the Port of Anchorage, Alaska.  Since the barges leave Seattle only once
every three days, scheduling was critical to allow cost-effective use of the traveling employee's
time.  After a flight  to Anchorage, the drivers continued the overland  trip  to  the site,
approximately 160 driving  miles  south  of Anchorage.  In  all, the equipment was hauled
approximately 4,700  one-way miles to the site,  taking nine days at a cost of approximately
$10,000.

Redundancy in equipment and material was considered a necessity. Commercial shipping of
spare parts between Bedford, Massachusetts and the site would have required at least 3 days in
the event of breakage or a mishap. All equipment and material, from metering boxes to duct tape
to aluminum foil, was backed up. Materials lists, formulated based on Alliance's past experience,
were checked, rechecked, and  all  amounts were increased.   Other  materials were brought as
contingencies (for instance, several  additional solvents were packed  despite the low probability
of use).

Permitting Issues Associated with the Demonstration Test Burn

In order to satisfy the regulatory agency's needs  for adequate data from the test bum, Alliance
had to meet OTS' strict Quality Assurance specifications.  Alliance prepared a Demonstration
Test Burn Quality Assurance Project Plan (QAPP), which underwent two revisions prior to final
approval  by OTS. Alliance also had to accommodate OTS laboratory audits, and coordinate with
OTS on scheduling the burn to allow EPA observation of the field testing.
                                             470                       /**  ALLIANCE
                                                                      /iftft V.-.:•:; -:-•-•

-------
EPA-OTS guidelines on performing laboratory audits resulted in the expenditure of considerable
time and effort to meet the requirements.  Previous exper'ence  with laboratory audits  had
consisted of one-day  site visits with review of laboratory procedures and capabilities through
review of records and interviews and personnel.  In many cases, receipt of audit check samples,
with results due some time after the visit, is included in the audit OTS, however, required the
analysis of check samples during the time of the audit, with the auditor observing all aspects of
the analysis.  To maintain  confidentiality of other client's data during the auditor's visit, the
laboratory did not accept other work during this time frame.  As a result, OES compensated the
laboratory for time required to conduct the visit.

The audit lasted for approximately three and one-half days.  All results from the audit were
determined to be well within acceptable ranges and laboratory data submitted in conjunction with
the Demonstration Test Burn were accepted by the agency.

RESULTS OF THE DEMONSTRATION TEST BURN PROGRAM

During the September 24-27, 1988 time period, Alliance completed three replicate runs at each
of two process operating conditions.  Results of the program are summarized in Table  1.  The
major change in the process for the second operating  condition  as compared to the first was a
7 percent increase in the soil feed rate and an increase in the CBC exit temperature from  1615°F
to 1700°F.  Along with these changes, the CBC exit velocity increased from 20 to 22.5 ft/sec.

As shown in Table 1, all applicable emission limits as imposed by OTS were met.  DREs were
essentially the same for both test conditions despite the low average PCB concentrations in the
soil feed.  HC1 emissions were well below the 4.0 Ib/hr limit and dioxin/furan concentrations in
the post-treated ash were well below the imposed limit of 2 ppb.  Paniculate emissions doubled
from Condition 1 to Condition 2  but remained  well below the standard of 0.08  gr/dscf @
7 percent O2 standard.

During the test burn, few delays were encountered due  to operation of the CBC unit itself. Some
process delays were experienced, due mainly to jamming of the rotary valve in the soil feed line
and/or the feed auger.

The final report for this project was submitted in December 1988.  On June 23,  1989, Ogden
announced that they had been issued a nationwide federal permit for use of the CBC system in
remediation of PCB contaminated soils.  At the time of this announcement, the Swanson River
project represented the world's  first major remediation program  using CBC technology  and the
largest PCB/soil cleanup to date.

In conclusion, the successful completion of any field  sampling program can be traced back to
adequate planning prior to initiation of the program. Without adequate planning, factors involved
in the Swanson River Demonstration Test Burn, such as the distance and remoteness of the  site,
may have combined to cause major problems, delays, and cost increases in the program. Because
of the approach followed, the project team was able to readily address most problems ultimately
encountered.
                                                                             ALLIANCE

                                           471

-------
TABLE 1. SWANSON RIVER TEST RESULTS SUMMARY
Parameter
Soil Feed Rate (Ib/hr)
Sorbent Feed Rate (Ib/hr)
CBC Exit Temp (°F)
CBC Exit Velocity (ft./sec.)
Ash Rate (Ib/hr)
Ash (% of feed + sorbent)
Soil PCB Concentration (ppm)
Soil RCl Concentration (ppm)
Ash PCB Concentration (ppb)
Ash PCDD/PCDF Concentration (ppt)
PM (gr/dscf @ 7% 02)
HCl Emissions (Ib/hr)
Avg PCB ORE (%)
PCDDs/PCDFs Concentration (ng/dscm)
Total Organic Chloride (Ib/hr)
Condition 1
8,500
170
1,615
20
8,233
95
683
162
< 9
< 175
0.0077
1.3
> 99.99994
<1.5
< 7.3E-5
Condition 2
9,100
170
1,700
22.5
8,790
95
507
229
< 16
< 195
0.016
1.4
> 99.99994
<2.0
< 2.4E-4
                      472
                                               ALLIANCE

-------
REFERENCES

1.     Ives, J.A. (ARCO Alaska, Inc.) and D.T. Young (OES). PCB Remediation in Alaska.
      Prepared for Presentation at the PCB Forum in Houston, Texas. August, 1989.

2.     Anderson, B.M.,  and  R.G. Wilbourn (OES).   Contaminated  Soil Remediation by
      Circulating Bed Combustion - Demonstration Test Results.  Prepared for Presentation at
      Superfund '89 in Washington, D.C. November 1989.
                                                                         ALLIANCE
                                                                      14 >:v,;.:,;••;:. :••;: "
                                        473

-------
      LOW TEMPERATURE THERMAL TREATMENT (LT3) OF SOILS
  CONTAMINATED WITH AVIATION FUEL AND CHLORINATED SOLVENTS

                      Roger K. Nielson
                    Weston Services, Inc.
                   West Chester, PA  19380

                             and

                       Craig A. Myler
       U.S. Army Toxic and Hazardous Materials Agency
             Aberdeen Proving Ground, MD  21010
                          ABSTRACT

Successful pilot study of Low Temperature Thermal Treatment
(LTJ) (Patent No. 4,738,206) of soils contaminated with
volatile organic compounds led to a full scale demonstration
to evaluate the use of this technology at DOD installations.
The LT  process developed by Weston Services, Inc. under
contract with the U.S. Army Toxic and Hazardous Materials
Agency (USATHAMA) was used in a test of remediating soils
contaminated with aviation fuel (JP4) and trichloroethylene
(TCE).  Tinker Air Force Base, Oklahoma was selected as the
demonstration site.  An abandoned landfill area had high
concentrations of contamination in a clay soil which was
determined to be ideal for this test.  The LT  process
performed better than expected and achieved cleanup levels
at lower temperatures, higher processing rates and shorter
residence times than previously thought possible.  This
directly translates into cost savings for future remediation
sites.  Test results indicitive of the performance and
potential application of this technology will be presented.
INTRODUCTION

    Thermal processes for decontaminating hazardous waste
laden soils have focused on incineration.  Raising the soil
to temperatures at which the contaminants present undergo
decomposition is an effective means of cleaning the soil.
Effective that is in terms of contaminant removal, not cost.
In raising the soil to the elevated temperature, energy is
expended needlessly, causing increased operating costs.  In
addition, contact heating between the combustion gases and
the contaminated soils produces a large volume of
contaminant laden vapor which must be processed through
pollution abatement equipment.  The larger the volume of
this vapor phase, the larger (and costlier) the down stream
equipment.
    To decrease energy requirements and simultaneously
reduce downstream pollution abatement equipment and
operations costs, a low temperature thermal treatment system
                           474

-------
has been developed.  The rationale for this system centers
around those drawbacks to incineration listed above.  First,
by heating the contaminated soil only enough to volatilize
the organic contaminants, energy is saved.  Second, using
indirect heating of the soil, downstream pollution abatement
equipment cost is reduced, both in capital and operating
costs.                   -
    Development of the LT  process began in August, 1985
with the operation of a pilot scale thermal processor.   Use
was made of an indirect heat exchanger which operated by
contacting the soils with heated augers which both
transported and heated the soil.  Volatile contaminants were
released from the soil and processed through an afterburner.
The pilot system schematic is shown in Figure 1.  The
success of this pilot scale system prompted the performance
of an economic evaluation  and bench scale tests  that
resulted in a full scale LT  system.
    The bench scale system was operated to determine if
results from the pilot scale could be effectively reproduced
on a smaller scale, thereby providing a low cost means to
pre-screen potential soils and contaminants.  The bench
scale system was also used to determine the effects of
temperature and residence times on different soils.  Of
particular note during this study was the processing of
soils containing semi-volatile contaminants from diesel fuel
and JP4.  Results indicated the potential for bench scale
data to be used in scale up.  This would greatly simplify
pre-screening of soils by reducing the costs of sampling as
well as the amount of soil required.  A central laboratory
was established at Weston to perform pre-screening tests
rather than constructing a pilot unit at each site.
    On April 19, 1988 U.S. Patent Number 4,738,206 was
issued for a Low Temperature Thermal Treatment Process for
removing volatile and semi-volatile organic contaminants
from soil.   A full scale LT  system was designed and
constructed by Roy F. Weston, Inc. under this patent based
on the results of the pilot and bench scale systems.
Initially, the full scale system was used to treat petroleum
contaminated soils at a site in Illinois. The first test of
this equipment in the remediation of hazardous wastes was
performed at Tinker Air Force Base, Oklahoma.  The results
of the test program at Tinker will be described.

PROCESS DESCRIPTION

    The full scale low temperature thermal treatment system
used two, thermal processors operated in series.  The first
processor is mounted on top the second.  Four intermeshed
screws in the first processor convey the soil through a
jacketed trough where it falls by gravity into the second
processor.  The second is identical to the first except the
direction of travel of material is reversed. In this way,
soil enters the unit on one side and exits the unit on the
opposite side thereby reducing the space requirements
                             475

-------
                                                               Air to
                                                            Atmosphere
  Hot Oil
 Reservoir
                                               Air Containing
                                              Stripped VOC's
                                                       After-
                                                       burner
       Oil Heating
        System
                                                 Combustion Air
                                                    Blower
Air In
                            Air
                         Preheater
Figure  1:   Schematic Diagram of  the Pilot Scale LT  System
                               476

-------
necessary for operations as well as reducing the flow of
contaminated soil outside of the processors.
    A 7.6 million KJ/hr hot oil system provides a heat
transfer fluid, Dowtherm HT, to the hollow screws and
jackets of the thermal processors.  Temperature controls
allow the oil to be maintained at temperatures up to 343 C.
A portion of the combustion gases from the oil heater are
passed through the thermal processor at about 375 C.  This
is done for two reasons.  First, waste heat recovered from
this stream maintains the volatile stream exiting the
thermal processor at a temperature high enough to avoid
condensation of the contaminants on  the walls of the
exhaust ductwork.  Second, the exhaust acts as an inert
atmosphere to maintain the gases in the thermal processor
below their lower explosive limit.
    The vapor stream from the thermal processor consists of
the contaminants being removed, water vapor from the soil
and exhaust gases from the hot oil heater.  This stream
exits at approximately 150  C (maximum) and flows through a
fabric filter, condenser, afterburner and caustic scrubber
system.  The fabric filter removes particulate carried over
from the processor.  The vapor stream then passes through an
air cooled condenser which reduces the temperature to
approximately 52 C.  Water and organics condensed reduce the
load on the afterburner.  The afterburner is a 3.7 million
KJ/hr gas fired, vertical, fume incinerator operating at 982
 C.  The afterburner is operated at a minimum 3% excess
oxygen and the exhaust is continuously monitored for 0~/ CO
and total hydrocarbons.  The 982  C exhaust from the
afterburner is quenched to approximately 82  C.  It then
passes through a packed bed absorber where acid gases
produced in the afterburner are neutralized with a caustic
solution.
    A liquid stream is produced by the condenser which is
water rich but does contain some hydrocarbons.  The aqueous
phase is separated from the organic phase in an oil-water
separator.  The aqueous phase is processed through a water
treatment system consisting of fabric filters followed by
granular activated carbon.  This water is then used as
makeup water for the scrubber and for dust control on
processed soil.  The organic phase from the separator is
either drummed for off site disposal or injected into the
afterburner.
    A block diagram of the system is shown in Figure 2 and a
schematic layout of the system is shown in Figure 3.  The
system described here is mobile and can be transported on 5
trailers.  Utilities required for operation are propane or
natural gas, electricity and process water.
discharges from the system include the scrubber stack
exhaust, the processed soil, the granular activated carbon
and filter cake, and, if not injected into the afterburner,
                         477

-------
00
                   Contaminated
                       soil
                      storage
                     Classifier
                     Oversize
                                                                        Sweep gas
                                                                                   To
Drag (light
conveyor
Surge
hopper
                                                                Feed soil
                                                               Dust
                                                        Fuel/
                                                     combustion
                                                         air
                                      Hot
                                      on
                                               atinuauhsie
                                                  f  Hoti
                                                                                      Hot on
                                                                                     system
                                   Hotel burner off-gases


                                          — Fuel/combustion air
                        Cool
                         oil
                                                                                 Treated soil
 Thermal
Processor
                                            rested«
Discharge
conveyor
Truck feed
 conveyor
Processed
soil storage
   bin
                                                                        Fabric fror
                                                                                  off-gases
                                                                       Condonsor
                                                                                         ONAMtor
                                                                                                   Organtes
                                                                                Condensate    1 Water
                                                                       55-gtlon
                                                                         drum
                                                                         LD.fan
                                                     Carbon
                                                   adsorption
                                                      unit

                                 Attorbumar
                                   Water
                                   tank
                                                                        Scrubber
                                                     Carbon
                                                     UMMfctL
                                                     sorpo
                                                      unit
                                                                        To stack
                                        Figure 2:   Block  Diagram  of  the Full  Scale LT   System Used
                                                       at  Tinker Air  Force  Base,  Oklahoma

-------
                                                                                          	1
10
                                  w
                                                                                    O
                                                                                          _ J JL
                                                                                               70"
A  Thermal Processors
B  Baghousa
C.  Motor Control Center
D  Thermal Processor Drive Units
E  Hot CM System
F  Condenser
G  Induced Draft Fan
H  Afterburner
I   Scrubber ID Fan
J  Scrubber System
K  Exhaust Slack
L  Caustic Storage Tank
M.  Frvsn Watsr  Systtni
N  Slowdown Carbon Adsorptnn Units
   Oa*«—.	*niii« Tii*a»m
   oiwoown oynvin
P  Recycle Water Tank & Pump
0  Oil/Water Saparator
R  Walar Carbon Adsorption Units
S.  day Shredder
T  Drag Conveyor
U.  Discharge Conveyor
V.  Dump Truck
W  Continuous Emissions Monitoring Trailer (CEM)
                                                      -W-
                                Figure 3:   Schematic  Diagram Showing  Equipment  Layout  of  the
                                                 Full  Scale LT    System Used at  Tinker  Air  Force
                                                 Base,  Oklahoma

-------
the organic phase from the oil-water separator.  The maximum
utility load for the unit is as follows:

          Liquid Propane         2.27  m^/day
          Process Water         40.88  m /day
          Electricity          600 amp/460 V (3 phase)


Operation requires 8 personnel for continuous operations
including a site manager and an instrumentation technician.

SITE DESCRIPTION

    To test the capabilities of the full scale LT  system a
site contaminated with compounds of specific Army concern
was desired.  Objectives in site selection were to find a
location which had a clay soil matrix contaminated with JP4.
JP4 was selected as the contaminant of choice as spills of
this fuel on Army bases may require remediation and diesel
fuel contaminated soil had already been tested in this
unit.D
    Coordination with the U.S. Air Force resulted in Tinker
Air Force Base, Oklahoma being selected as the test site.
Remedial investigations on this active Air Force support
base turned up an abandoned sludge dump in one of the bases
landfill areas.  Survey and analysis of the area estimated
600 m  of sludge contaminated clay soil.  Analysis of soil
and water from this site indicated JP4 and trichloroethylene
in high concentrations along with other potential
contaminants in lesser concentrations, probably constituents
of diesel, kerosene and other solvents.  The land fill was
no longer in use and outside of everyday operational
traffic.  The Air Force personnel at Tinker Air Force Base
as well as the state and federal regulators were in favor of
the test.  The mixed wastes which included TCE, a listed
hazardous waste, provided an opportune site for
demonstration of the LT  technology on a full scale.  The
site layout is shown in Figure 4.
    Due to the presence of listed hazardous wastes, a permit
was required for the test.  An application for a Research,
Development and Demonstration (RD&D) permit was filed in
accordance with the requirements of RCRA.  An approved RD&D
permit was issued by Region VI, EPA 120 days after
application.  This permit authorized the conduct of the test
plan for the LT  process.

TEST PLAN AND RESULTS

    The test plan for the full scale LT  system was designed
to meet the following objectives:

     - determination of effectiveness of the full scale LT
       system at removing JP4 and chlorinated solvents from
       soil.
                                480

-------

















00










i SCALE
4 «TACKC* 6tir CotvMoR
« o*-
t. SCtuaBEft 5VSTEM
7. C*u5TiC STelTASt SVSTSrt
6, PCW«« T*A«SFcCMe*





H MCON T«AlLM







1 I





















































*?


























.?.


























1
i
M*


|
*Mta1
«»r
















,
/
/
f

/

\ b ~

i M«r
••••!g«3"-
m







I
r





/
X
/
n
GD n;
i — i ^

1 ^* 1 **
if







iwr
VZc



1







l—ii-l
» 1
1
1

r >
Sl^


nn

L««nLL «OA.


iTScBT^l


















!



"2



















ft 	


••
SDK



•*
















m
n
Wt





«TN




















*TI
!•»
tk»*





'




















U


















'S'J'" ^"W* ASP-AIT



.MMn* M.
"•-«"•«" ""

o«!<*m»

Ml >U
ft rD
55 ar
r •*





Ct»t 
-------
             Table 2:  Results of First Test Run

Contaminant            Feed Soil     Processed Soil     GCL
                    Concentration    Concentration      (Ppb)
                        (ug/kg)        (ug/kg)

TCE                     37,250           5.4            70
1f2-Dichlorobenzene     35,000           BDL            125
1,3-Dichlorobenzene      3,500            ND            NS
1,4-Dichlorobenzene      8,700            ND        10,800
Toluene                  8,700            ND            330
Napthalene               4,300           BDL            NS
Total Xylene             5,900            ND            150

     ND - not detected
     BDL - below detection limit
     NS - not specified
            Table 3:  Results of Second Test Run

Contaminant            Feed Soil     Processed Soil     GCL
                    Concentration    Concentration
                        (ug/kg)        (ug/kg)

TCE                    111,000            5             70
1,2-Dichlorobenzene     15,000           BDL           125
1,3-Dichlorobenzene      BDL             BDL            NS
1,4-Dichlorobenzene       ND              ND        10,800
Toluene                  8,300            ND           330
Napthalene               5,000           BDL            NS
Total Xylene            11,400            ND           150

     ND - not detected
     BDL - below detection limit
     NS - not specified

    Testing at higher production rates (less residence time)
and lower temperatures resulted in mechanical failure of the
process equipment. The system was overloaded. The maximum
soil processing rate achieved was 60% above the design rate
at 11,364 kg/hr.  Although this rate was not demonstrated
for a sustained period, a sustained rate of 9091 kg/hr was
established. A final test was conducted at a higher
production rate but at a previously tested intermediate
temperature (204  C) to confirm the data of the 3 previous
tests.  The results are presented in Table 4.
    At this time in the test program, analysis of the soil
indicated the presence of Arachlor 1260, a polychlorinated
Biphenyl (PCB).  The system was never intended to process
wastes of this nature nor were they expected.  This
discovery halted all operations and investigation eventually
led to the cancellation of the remainder of the test.  A
                             482

-------
     - determination of impact of system parameters on
       effectiveness.

     - evaluation of the use of a stripping agent to enhance
       contaminant removal.

     - determination of optimum operating conditions.

     - determination of stack emissions.

Goal cleanup levels (6CL) were established for the test
based on previous requirements or established standards.
These goals were accepted by the EPA regional office in the
RD&D permit.  Table 1 lists the GCL for contaminants found
above the detection level in soil processed.


        Table 1:  Partial List of Goal Cleanup Levels

	Contaminant	GCL (ppb)

Trichloroethylene (TCE)                       70
1,2-Dichlorobenzene                          125
1,4-Dichlorobenzene                       10,800
Toluene                                      330
Total Xylenes                                150


    The first test run was conducted at the high end of the
hot oil temperature to be tested, 315  C.  This run
established that the process is capable of remediating soils
contaminated with JP4 and TCE.  Results are presented in
Table 2.  Following this run, the system was operated at a
reduced hot oil temperature (204  C) in an attempt to
establish a lower operating limit for hot oil temperature
from which an optimum would be determined.  The results of
this run are presented in Table 3.  Comparison of the
results shown in table 2 and 3 indicate no significant
difference between the two operating conditions in terms of
contaminant removal.
     At this point, an attempt was made to reduce the
residence time to establish a maximum processing rate.  The
temperature was also reduced to 150  C to determine the
lower effective operating temperature.  The results of this
test are not conclusive as the test was aborted due to
mechanical failure caused by overloading the processors.
                             483

-------
complete description of the events following the PCB
discovery and the actions taken will be reported at a later
time.
       Table 4:  Results of Fourth Test Run

Contaminant            Feed Soil     Processed Soil     6CL
                    Concentration    Concentration     (ppb)
                        (ug/kg)        (ug/kg)

TCE                     10,575           23.4            70
1,2-Dichlorobenzene     53,000           BDL            125
1,3-Dichlorobenzene      BDL              ND             NS
1,4-Dichlorobenzene     14,750           BDL         10,800
Tolune                   BDL             BDL            330
Napthalene               BDL             BDL             NS
Total Xylene              ND              ND            150

     ND - not detected
     BDL - below detection limit
     NS - not specified
CONCLUSIONS

    Although the test plan was cancelled prior to
completion, the results of the test at Tinker Air Force Base
are positive.  Three of the five objectives of the test were
met.  It was determined that low temperature thermal
treatment technology is effective at removing JP4 and TCE
from soil.  Optimum operating conditions for the full scale
system were established as the mechanical limits of the
processing equipment.  At the levels of contamination in the
soils tested, the range of operating conditions had no
discernible effect on removal efficiency.
    Two objectives were not met in the test plan due to the
premature cancellation of the test.  First, the use of
stripping agents to improve contaminant removal was not
investigated.  The results of tests performed showed removal
of contamination to near or below detectable limits in all
cases.  The use of a stripping agent would not have improved
this result at a detectable level.  The second objective,
determination of emissions, was not met as stack testing was
not performed for the test runs made.  Stack sampling was to
be conducted only at the optimum operating conditions.  As
the PCB discovery halted testing during the determination of
the optimum conditions, no stack sampling was performed.
This is not felt to be a significant impediment to fielding
the LT  technology.  The afterburner and scrubber system are
standard unit operations used in similar treatment systems
for contaminant laden streams.  Although no field data was
collected on this system, its standard configuration is
expected to meet or exceed the required discharge criteria.
                            484

-------
    The LT  full scale system tested was capable  of
remediation of soils contaminated with volatile and
semi-volatile compounds.  Processing rates obtained
indicated a cost for remediation between $90 and  $100  per
metric ton of soil fed to the system.  The unit was
determined capable of processing at a rate of  9090 kg/hr.
Removal of the contaminants present was consistantly below
the criteria established for cleanup.
                      References Cited

1.  Pilot Investigation of Low Tennperature Thermal
Stripping of Volatile Organic  Compounds  (VOC's)  From Soil,
U.S. Army Toxic and Hazardous  Materials  Agency Report No.
AMXTH-TE-CR-86074, APG, MD, June 1986.

2.  Economic Evaluation of Low Temperature  Thermal  Stripping
of Volatile Organic Compounds  from Soil, U.S.  Army  Toxic and
Hazardous Materials Agency Report No. AMXTH-TE-CR-86085,
APG, MD, August 1986.

3.  Bench-Scale Investigation  of Low Temperature Thermal
Stripping of Volatile Organic  Compounds  (VOC's)  From Various
Soil Types, U.S. Army Toxic and Hazardous Materials Agency
Report No. AMXTH-TE-CR-87124,  APG, MD, November  1987.

4.  United States Patent Number 4,738,206,  Apparatus and
Method for Low Temperature Thermal Stripping of  Volatile
Organic Compounds from Soil, Inventor: John W. Noland,
Issued April 19, 1988.

5.  Nielson, 5-K. and Cosmos,  M.G., "Low Temperature Thermal
Treatment  (LT  ) of Volatile Organic Compounds  from  Soil:  A
Technology Demonstrated", Presented at the  1988  Summer
National Meeting of the AICHE, Denver, Colorado, 21-24
August, 1988.
                           485
                                  tt US GOVERNMENT PHNTNG OFFCE 1990-748-159/20442

-------