February 1993
                                         EPA/540/R-93/500
         TECHNICAL PAPERS

FOURTH FORUM ON INNOVATIVE HAZARDOUS WASTE
          TREATMENT TECHNOLOGIES:
         DOMESTIC AND INTERNATIONAL
                 San Francisco, CA
                November 17-19, 1992
           TECHNOLOGY INNOVATION OFFICE
    OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE
        U.S. ENVIRONMENTAL PROTECTION AGENCY
               WASHINGTON, DC 20460
                      AND
       RISK REDUCTION ENGINEERING LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
        U.S. ENVIRONMENTAL PROTECTION AGENCY
               CINCINNATI, OH 45268
                                     Printed on Recycled Paper,

-------
                                  ABSTRACT
On November 17-19,  1992, the U.S. Environmental Protection  Agency's Technology
Innovation Office and  Risk Reduction Engineering Laboratory, Department of Energy,
Corps  of Engineers,  and California Environmental Protection  Agency,  hosted  an
international conference in San Francisco, CA, to exchange solutions to hazardous waste
treatment problems. This conference, the Fourth Forum on Innovative Hazardous Waste
Treatment Technologies: Domestic and International, was attended by approximately
1,000 representatives from the U.S. and 25 foreign countries. During the conference,
scientists and engineers representing government agencies, industry, and academia
attended 42  technical  presentations  and  case studies describing  domestic and
international technologies for the treatment of waste, sludges, and contaminated soils at
uncontrolled hazardous waste disposal sites. Technologies included physical/chemical,
biological, thermal, and stabilization techniques. Presentations were made  by EPA, their
Superfund Innovative Technology Evaluation (SITE) program participants, other Federal
and State agencies and their contractors, international scientists and vendors. Over 70
posters were  on display.

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.

-------
                      TABLE OF CONTENTS

ABR-CIS bioremediation: full-scale experiences with an in-situ treatment system
for sediment volume-reduction and contaminant mineralisation in harbours and
waterways
      B. Malherbe, HAECON N.V., Harbour and  Engineering Consultants	  1

In Situ Bioventing of a Diesel Fuel Spill
      T.L. Bulman, Campbell Environmental  	  22

Remediation of Ground Water Contaminated with Organic Wood Preservatives
Using Physical and Biological Treatment Technologies
      J.G. Mueller, SBP Technologies, Inc	  35

The Design Criteria and Economics of Operating  a Full-Scale Above-Ground
Bioremediation Facility for the Treatment of Hydrocarbon Contaminated Soils
      Robert Mall, KJC Operating Company  	  40

Biopur®  An Innovative Bioreactor for the Simultaneous Treatment of
Groundwater and Soil Vapour Contaminated with Xenobiotic Compounds
      J.M.H. Vijgen, TAUW Infra Consult, B.V	  48

Bio-Rem Inc.'s Augmented In Situ Subsurface Bioremediation Process (TM)
Using Bio-Rem, Inc.'s Proprietary Biocultures (H-10) for the Remediation of
JP-4 Contaminated Soils at Williams AFB, AZ
      David Mann, Bio-Rem, Inc	  62

Bioremediation of Soils at a Waste Oil  Facility
      Gary Guerra, U.S. EPA, Robert  S. Kerr Environmental Laboratory	  64

Low Temperature Thermal Treatment LT3 Site Demonstration
      Michael G. Cosmos, Roy F. Weston, Inc	  71

Plasma Arc Vitrification
      Richard C. Eschenbach, Retech, Inc	  78

Thermal Desorption of PCB-Contaminated Waste at the Waukegan Harbor
Superfund Site (A Case Study)
      Robert Shanks, SoilTech ATP Systems, Inc	  90

The Babcock & Wilcox Cyclone Vitrification Technology for Contaminated Soil
      J.M. Czuczwa, Babcock & Wilcox R&DD  	  119
                                    HI

-------
The ECO LOGIC Process
      DJ. Hallett, ELI Eco Logic International, Inc	  129

BioGenesis"" Soil Washing Process Technology Summary
      Charles Wilde, BioGenesis Enterprises, Inc	  136

Steam Enhanced Recovery Process (SERP)
      William R. Van Sickle, Hughes Environmental Services, Inc	  142

Soil Venting to Remove DBCP from Subsurface Soil:  A Case History
      M.B. Bennedsen, Woodward-Clyde Consultants 	  146

5 Years Operational Experience with the Harbauer Soil Washing Plants
      Winfried Groschel, Harbauer GmbH & Co. KG  	  161

Field Demonstration of Soil Washing at the King of Prussia Superfund Site
      Michael J. Mann, Alternative Remedial Technologies, Inc	  175

The Lurgi-DECONTERRA® Process Soil Washing
      Ted J. Pollaert, Lurgi Corporation	  192

Efficient and Cost-Effective Programs for Soils and Groundwater Restoration
      J.J. Malot, Terra-Vac	  200

Radiolytic Remediation of a TCE Ground Spill Using an Electron Accelerator
      S.M. Matthews, Lawrence Livermore National Laboratory  	  206

Rethinking the hydraulic cage concept for hazardous waste disposal facilities
and mitigation of contaminated sites
      Charles Voss, Golder Associates, Inc	  212

Innovative Contracting Strategies for Equipment Procurement - Bofors Nobel
Superfund Site - Muskegon, Ml
      Ted H.  Streckfuss, U.S. Army Corps of Engineers	  226

Remediation of Groundwater Contaminated with Volatile Organic Compounds
in the Saturated Zone at a California Superfund Site in Mountain View,
California
      R.F. Battey, The Earth Technology Corporation	  254

Soil Washing of Lead-Contaminated Soil at a Former Gun Club Site
      Christine Parent, CAL/EPA Department of Toxic Substances Control  ....  261
                                     IV

-------
 Improvements in the Conventional High Temperature Rotary Kiln Incineration
 Technology
      Markku Aaltonen, EKOKEM Limited	  270

 Evaluation of Grouting Technology:  DOE Perspective
      W. Roberds, Golder Associates, Inc	  276

 Du Pont/Oberlin Microfiltration Technology (SITE)
      Ernest Mayer, E.I. du Pont de Nemours, Inc	  296

 Composting of Explosives-Contaminated Soil at the U.S. Army Depot Activity
      Captain Kevin R. Keehan, U.S. Army Toxic and Hazardous Materials
      Agency	  313

 United States/German Bilateral Agreement on Hazardous Waste Site Clean-Up
 Projects
      H. Stietzel, Federal Ministry for Research and Technology, Germany ....  324

 Groundwater Remediation: Extraction and Removal of TCE & Cr6+ at an NPL
 Site
      Subijoy Dutta, U.S. EPA, Office of Solid Waste	  340

 The Use  of Horizontal Wells in Remediating and Containing a Jet Fuel
 Plume - Preliminary Findings
      Mark Thacker, COM Federal Programs Corporation	  346

 Evaluation of Thermal Extraction Technologies for Treatment of Soils
 Contaminated with Coal Tars and Wood Preservatives at the Pacific Place
 Site, Vancouver, B.C.
      Sandra Whiting, SCS  Engineers	  358

CF Systems' Solvent Extraction Technology for Site Remediation:  A Post
 "Site Demonstration Update"
      Chris Shallice, CF Systems Corporation	  373

Commercialization of Innovative Technologies - Challenges and Solutions
 (A Technology Vendor's Perspective)
      Frederic A. Eidsness,  Jr., Canonie Environmental Services Corp	  385

Bioremediation of Chromium (VI) Contaminated Solid Residues Using Sulfate
Reducing Bacteria
      Louis J. DeFilippi, Allied-Signal Research and Technology	  417

-------

-------
 ABR-CIS bioremediation : full-scale experiences with an in-
     situ treatment system for sediment volume-reduction and
      contaminant mineralisation in harbours and waterways
Author :   ir. B. Malherbe
            HAECON N.V., Harbour and Engineering Consultants,
            Ghent, Belgium
ABSTRACT

The improvement of the in-situ natural microbiological degradation of organic matter
or organic compounds present a very attractive solution for  the management of
(contaminated) sediments in harbours and waterways.  To get efficient  biodegradation
as much aerobic bacterial  strains as possible have  to  be reactivated. This natural
biodegradation process is particularly  interesting for the achievement of an in-situ
volume  reduction of the deposit  and/or  in-situ  decontamination  of  organic
contaminants. ABR-CIS (Augmented Bio Reclamation -  Conditioning In-Situ) is such
an in-situ treatment technique which  has been applied on  full-scale. The projects
executed with the ABR-CIS system will be discussed  and the monitoring  results will
be presented. Moreover, the application of the ABR-CIS may represent major savings
with respect to the classical dredging and disposal costs.
The accumulation of mud  deposits in aquatic systems goes often parallel  with the
accumulation of organic micro-pollutants and nutrients and with eutrophication. This
causes problems for the management (disposal) of these sediments to be dredged for
navigation, calibration or even sanitation reasons.

-------
                                                                                 or
The "over-supply"  of organic wastes  in  river-systems  from  sewers, domestic
agricultural effluents has caused a saturation of the natural biodegradation processes in
the aquatic environment (figure 1).
    CONTAMINANT INPUT:

      - ORGANIC COMPOUNDS
      - NUTRIENTS
      - OTHERS
                                              ORGANIC MATTER INPUT:
                                                  -PLANCTON
                                                  - ALGAE
                                                  -LEAVES
                                                 - SEVWGES
IDECREASE OF OXYGEN-SUPPLY:
!    - EUTROPHICATION
j    - INCREASE OF VWTERDEPTH
'    - FLOATING ORG. PRODUCTS
                                               SEDIMENT ACCUMULATION:
                                                     - ANOXIC
                                                     - ENRICHED IN ORGANIC
                                                      MATTER
                CONSEQUENCE: SLOW, UNCOMPLETE ANAEROBIC
                              BIO-DEGRADATION
 Figure 1 : Mechanism of creation of saturated anoxic sediment deposits

 During  the  search  for  a  simple technique  to  re-activate aerobic  mineralisation
 processes, the ABR-CIS system was developed (ABR : Augmented Bio Reclamation -
 CIS : Conditioning In-Situ).

 This presentation goes through some elementary concepts in microbiological processes
 to illustrate the potential  of the ABR-CIS system which is an in-situ microbiological
 treatment system of sediments.

 The potentiality  of  ABR-CIS  lies in  a  mineralisation  of  organic  matter  ( =
 microbiological dredging) and  a  mineralisation  of organic  micro-pollutants  ( =
treatment in-situ).

-------
 BACTERIA AS  PART OF THE NATURAL  ECOSYSTEM OF
     SEDIMENT DEPOSITS

 Fine-grained sediments are deposited in a wide variety of aquatic environments. These
 deposits, often called  "muds", are a specific habitat for different organisms.  The
 specificity of mud habitats is related to  the  high accumulations in the deposits of
 decaying  biomass,  adsorbed  organic  compounds,   nutrients  or  man-induced
 contaminants. The sediment is making part of the global  aquatic ecosystem composed
 of atmosphere, water column and sediment deposit. The typical aquatic ecosystem is
 described by the trophic chains illustrated on figure 2.
Figure 2 :  Description of ecosystem elements in aquatic environment

Micro-organisms, such as bacteria, are occuring in the atmosphere, the water column
and the sediment deposit (micro-benthos). On figure 2 it is clear that bacteria play a
fundamental role in the mineralisation  of organic products in water and  sediment.
Within the ecosystem  bacterial activity is obviously related to the development and
activity of other organisms such as e.g. diatoms (Hamilton, 1987).
Types of bacteria occuring in mud deposits

The environment where mud is deposited is also favourable for the deposition of e.g.
dead phytoplancton (primary producers) or other fine-grained products. The presence

-------
 in  mud  of the  clay-humic complex  stimulates  particular  physico-chemical and
 biochemical reactions important to understand the environment of bacteria.

 Bacteria  are  micro-organisms  (diameter  ca.  0,0001   -  0,02  mm).  They  are
 systematically classified in 3 types (Vogel, 1970) :
 1.   Eubacteria (unicellular most common bacteria species) :
     - coccae
     - bacilli
     - vibrionic.b.
     - spirillic.b.
 2.   String bacteria (composed of micro-colonies interconnected with fibres).
 3.   Actinomycetes (sessile cells with radiating filaments).

 Bacteria can also be classified according to the feeding function,  i.e.  :
      autotrophic  :  assimilation  of  carbonic   acid  by  photosynthesis or,  more
      commonly,  by chemosynthesis of compounds present in the sediment  or the
      water ;
      heterotrophic  where the  assimilation  is  done  out  of living or dead  organic
      matter; as  such they are essential in the  C- and N-cyclus between lower and
      higher organisms.

 The above mentioned classification suggests consequently that bacteria are essentially
 food- and environment specific. In fact,  the bacterial flora of sediments show a wide
 range of adaptative variability relative to :
 -  the sediment  texture  related  to  the  penetration of water,  gases  or even  other
   organisms ;
 -  the type and  quantity  of organic matter  (governing the biodegradability or the
   specific food)  ;
 -  the bio-available oxygen content;
 -  the temperature;
 -  other organisms and their seasonal cycli (including respiration cycli) ;
 -  particular contaminants adsorbed to the mineral  or organic  mud particles and/or
   with a toxicity vs. bacteria.

Bacteria are also classified according to their respiratory mode (aerobic and anaerobic
bacteria). In fact, the classification of bacteria according to the respiratory mode is the
best adapted  to  describe  the   bacterial  ecosystem  in  aquatic  and  fine-grained
sedimentary environments. The  reason for this is related to the relative low vertical
diffusion of atmospheric oxygen through the water column and the sediment deposit ;
all ecosystem equilibria are governed by this oxygen input  which may  be strongly
affected by a lot of external reasons,  such as water circulation,  agitation,  water
depths, turbidity, dwelling organisms, ...

-------
Metabolism processes of bacteria

The above mentioned adaptative bacterial variability and heterogeneity related to the
cited factors is  a function of the availibility through the water of essential chemical
components for bacterial growth , such as oxygen , nitrates ,  phosphates, sulphates,
etc. ... The biodegradation products of bacterial activity (E^S, ammonium, N2, CC>2,
H2O) are diffused back to the water column and the atmosphere.

The carbonic acid  assimilation,  which forms  the  basis of bacterial  metabolism  (by
photosynthesis or chemosynthesis) is done by resorption through the cell-membrane
(diffusion, osmose). To activate  the assimilation,  bacteria are secreting selective and
species - specific exoenzymes (biocatalysators) able to decompose and  mineralize a
wide range of organic compounds.

Not only humic acids but also more complex organic chains such as phenols, PAH's ,
mineral  oils, ... can theoretically be decayed by bacterial metabolism processes (fig
3).
               CO, * H,0    —    CH,-COOH    ' —
                                               Kzm.-a>-oen>e: *
                                                                   OIO~A01PlNCZUUt
Figure 3 :  Biodegradation  process  of naphtalene,  a common PAH  contaminant in
           sediments

But not all organic matter is even bio-degradable and most of the mud deposits have a
lack of energy and nutrients to assure continuous bacterial activity and/or growth. The
important ecosystem cycli (e.g.  diatom cycli) are of primary  importance in bacterial
activity.  The illustration of typical  marine ecosystem cycli is illustrated on figure 4.
The spring and late-summer bloom of diatoms abundancy  is clearly visible on this
diagram.

-------
These time cycli are important because the supply of nutrients and organic matter to
the sediment bed, the temperature and the available oxygen (from respiration or from
diffusion)  will  know large fluctuations.  The  micro-organisms will  have to adapt
themselves to these fluctuations.

When  living   conditions  are   becoming   worse,   bacteria   have   interesting
defense/preservation modes, such as (Pommepuy et al., 1989) :

      spore formation (formation of a thick-walled resistant spore) .
      dwarfcell formation (formation of a small cell with limited metabolism) .
      somnicell stage (the bacteria are putting themselves in a kind of lethargic state).
                          SEASONAL VARIATIONS > PRIMARY PRODUCTION
                          PHYTOPLANCTON.NUTRIENTS AND ILLUMINATION
                  j  2 r 3  L, 4  »  5  „  6 j7j 8  ,9  8 10 o
                                                              12
                                    PARAMETERS

                     NUTRIENTS           	 DIATOMS ABUNDANCY

                     FLAGELLATE ABUND,     "«  DAILY ILLUMINATION
            'DIAGRAMS ON RELATIVE SCALE
                              (from e«tl from WiterkwililtiUDil"
             Noo'B«e 19EE i-c B "en 19891
Figure 4 :     Ecosystem cycli related to primary production in marine environment

When growing conditions are becoming better the bacteria are reactivated from these
spores, dwarfcells or somnicells. The survival  of the bacteria depends on their ability
to store energy reserves (e.g. in the form of glycogens).

Bacteria are able to improve the environmental conditions which they need to develop
their specific activities. They may adhere to sediment particles, to organisms or form
long chains of bacteria.  This adhesion is done with polysaccharic fibres  such as
glycocalix, which is secreted by the bacteria  themselves.  The  glycocalix (Moureau
and Richelle - Maurer, 1990) has specific characteristics :
-  it maintains bacteria on the substrate regardless current action  ;
-  it traps nutrients ;
-  it retains the digestive exoenzymes ;

-------
 -  it  forms a physico-chemical protection against toxins, virusses,  antibiotica and
   predators;
 -  it maintains appropriate physico-chemical conditions to allow bacterial growth and
   activity.

 Anyway, the survival time of bacterial population  in sediments can be very long and
 reach over several months to years (Rosack and Col well, 1987).

 As for all  living organisms and as mentioned before, the bioavailable oxygen is of
 primary importance for  bacterial metabolism  life  and growth.  Therefore, by many
 researchers, bacteria are rather classified according to their respiratory mode :
      aerobic bacteria  : bacteria which use free (dissolved)  oxygen; degradation of
      organic compounds is fast and complete (high burning efficiency) ;
      anaerobic  bacteria  :  bacteria  which use chemically bound  oxygen (no  free
      oxygen available):  biodegradation is slow and incomplete.
Aerobic and anaerobic bacteria in sediments

When mud is deposited in an aquatic environment the oxygen supply to the sediment
deposit  is  limited because of several reasons :
     mud is often  deposited in water with slow water circulation and consequently
     low  dissolved oxygen contents (< 5 mg C>2/1) ;
     the distance between the mud and the atmosphere increases (due to the presence
     of the water column and the mud for increased water depths) ;
     the  mud  has  low permeability due to  clay  minerals and  fine grains  and
     horizontal layering ; this decreases vertical oxygen diffusion ;
     the presence in mud of high proportions of Fe-hydroxydes increases the COD
     (chemical oxygen demand) ; in general the COD of mud deposits is high.

The main  electron terminal  acceptors used  during the decomposition of organic matter
are in decreasing order of efficiency (Claypool and Kaplan, 1974) :
1. free oxygen;
2. nitrates;
3. sulphates;
4. (bi)carbonates.

This biochemical succession is accompanied  by  chemical reduction processes in the
sediment and, consequently, by a decrease of the redox potential, Eh (redoxpotential
as low as -350 mV).

This vertical  layering in mud deposits has  been schematised by Claypool and Kaplan
(1974) and is also illustrated on figure 5.

-------
F
PHASE OXYGEN SUPPLY BACTERIA £ft MINERALISATION
PRODUCTS
AIR
WATER
^\
SED1MEN
//
y.

RESPIRATION
•DISSOLVED OXYGEN AEROBIC BAG. + CO2/H20
FERMENTATION ^AcS? +/~
T OEN.TR.HCAT.ON °™™™« -
SULPHATE SULPHATE
nEPUCT,ON -=A "
BICARBONATE METHANE
„„„,„.,,«*, PRODUCING 	 CH4/H20
REDUCTION BACTERIA
'igure 5 : Schematic presentation of bacterial layering in mud deposits

 The natural vertical layering explains why most of bacterial activity  in  mud is
 essentially anaerobic resulting thus in a slow and uncomplete degradation of organic
 matter.  The  vertical bacterial  layering  will  be analysed  more in  detail  below
 (Pommepuy et al.,  1989).
The aerobic surface layer

This  very  thin  layer is  colonized  by  an  abundant  microbial flora  (109-1010
bacteria/ml). Numbers decrease with depth  because of the reduction of available
substrates (Atlas & Bartha, 1981). The bacterial population may represent from 0,2 to
2  % of the total organic matter (Meyer-Reil,  1984).  The surface bacteria are mainly
Pseudomonas and  Vibrio-gram negative bacilli along with other bacterial-types such as
Nirrosomonas  and  Nitrobacter,  autotrophic  bacteria which  are responsible  for
nitrification (oxidation of ammonium salts to nitrites and  then to nitrates  (Martin
1979).

Other organisms like Diatoms (Navicula, Nitschia, etc. ...) and blue algae (Spirulina.
etc.  ...) are also  proliferating  in this aerobic surface layer especially if light  is
penetrating  as deep (Francis - Boeuf, 1949). Gen.  Thiorodo bacterium may oxydize
sulphur compounds and is part of the sulphur cyclus (similar to the N-cyclus).

-------
In the presence of molecular oxygen,  the  surface flora use aerobic  respiration  to
breakdown organic matter; this type of respiration is most energetically efficient.  In
this surface layer, aerobic bacterial production and decomposition of organic matter
are greatest and  total organic  carbon mineralisation (H2O/CO2  state) may occur.
There  is  intense  floral competition for  the utilization  of released intermediate
decomposition  products, and  bacteria with  the  highest  intrinsic growth rates are
favoured.

Fecal  flora can  settle at  the same time  as organic matter  from  anthropogenic
discharges.  The  main species found  are  fecal  Streptococci  and  Enterobacters
Escherichia coli, Klebsiella, Salmonella. These latter bacteria are pathogenic  and can
be dangerous for humans when contaminated shellfish are ingested. Hernandez (1985)
has recorded the  presence of very high concentrations  of fecal Streptococci in fine
harbour sediments.
Anaerobic levels

In the lower layers, there is an  observed reduction in the number of bacteria and a
floral change ,  which may be partially explained by the appearance of gram-positive
bacteria genus Clostridium ... (Bianchi et al., 1977; Bensoussan et al.  1981). Bacterial
flora  use either facultative or strict anaerobic  respiration, or fermentative processes.
At  all levels  Ferrobacteria are occuring reducing  the ferricompounds  to  ferro-
compounds and delivering the typical black color to reduced mud.
Fermentative bacteria

The genus Clostridium  (spore-forming) may account for a major part of the  strict
anaerobic fermentative bacterial population  (Atlas & Bartha,  1981).  These bacteria
play a major role in mineralisation of organic matter (cellulose, ...) since they are
capable of utilizing organic material which  other bacteria have difficulties breaking
down,  and   also  they  synthesize   exoenzymes  which  can  hydrolyze  many
macromolecules into small molecules that can be self-assimilated, or assimilated by
other micro-organisms (Marty et al.,  1989). Spore Clostridia which  are pathogenic
for  man  (Clostridium  perfringens  and possibly  Clostridium  tetani, Clostridium
botulinwri), are also found amongst these flora.
Denitrifying bacteria

Denitrifying bacterial numbers vary from 1C)4 to 10^ per gram of dry sediment (Jones,
1979). The Pseudomonas zndAeromonas strains are those most frequently isolated, as
well as some  vibrobacteria.  In  fact it seems that we know very little about the
bacterial species involved (Heitzer & Ottow, 1976). For denitrification processes, the

-------
 sediment must be anoxic, and there must also be nitrates in the underlying waters and
 sufficient  nitrate diffusion  within  the  sediment. Two  processes  are involved in
 denitrification :  the  first  process includes catabolic  reduction to transform nitrate
 (NOs) into nitrite and into nitrous oxide (N20) and molecular nitrogen (N2). More
 recently a  second  process has been discovered, which  involves formation  of an

 ammoniumion (NH^) from nitrite by catabolic fermentation. Current studies have not
 yet established whether or not this is a major phenomena. These two processes seem
 to operate simultaneously but their importance varies according to the concentration of
 organic matter. The percentage of nitrate transformed into ammonium was found to
 vary from 20 to 12 % in previously studied sediments (Koike & Hattori, 1978).

 Nitrobacteria and  Ferrobacteria  have wide  varieties  of  species able to  reduce
 anorganic compounds.

 Denitrification, which develops oxydation-reduction potentials that are higher than
 those accomplished by sulphate reduction,  takes place within the sediment above the
 sulphate-reduction zone.
Sulphate reducing bacteria
In seawater, oxygen associated with sulphur, in the form of sulphate, is 200-fold more
abundant than molecular oxygen.  Moreover, sulphate can be found deep within the
sediment (a few meters). In anoxic marine deposits, sulphate reduction is an important
phase in the mineralisation of organic matter (Marty et al.,  1989). The abundance of
these   sulphate   reducing  bacteria   (classified   in  7   genera  Desulfobacter,
Desulfobacterium, Desulfobulbus, Desulfococcus, Desulfomonas, Desulfonema and
Desulfosarcina)  is linked with sulphate availability and amounts of organic matter
(closeness  to  pollutant  discharges).  Their densities  are  estimated to  be  10^-10^
bacteria/ml.  These  bacteria are  involved  in  the final  phases of organic matter
mineralization within anoxic sediments  and could be responsible  for 50  % of the
oxidation of organic carbon (Jorgensen, 1983),  compared to 2-3 %  for denitrification
(Fenche & Blackburn, 1979; Jorgensen, 1983).

The high activity of  sulphate reducing bacteria in muds is also the cause of the typical
     smell problems  associated to waters with such sediment deposits.
Methane producing bacteria

At very  low redoxpotential values the bicarbonate molecule  is reduced by  specific
bacteria  to  methane.  Bicarbonate is  available from coprecipitation from  aquatic
plancton and necton deposits or from early diagenesis processes.
                               10

-------
Methane is produced e.g. in marsh areas and in the early petroleum-genesis processes
in mudstone.
Some interesting properties of bacteria

The  above mentioned  considerations  show  that bacteria, in good environmental
conditions of oxygen, nutrients, temperature, etc... can execute interesting actions to
serve man : mineralisation of organic mud or the clay-humic flocks (= microbiologic
dredging)  and   mineralisation of  organic   micro-pollutants  (=  decontamination
treatment in-situ).

It is important to notice some other particularities of bacteria :
     at different development stages some bacteria may have plasma-filaments which
     act as flagellae to execute displacements ;  however, most of bacteria are sessile
     organisms ;
     in good environmental conditions bacteria multiply by cell (mitosis)  ; each 20-
     40 minutes bacteria may divide (1 organism can deliver ca. 248 successors in 24
     hours) ;
     the small dimensions of bacteria facilitate the dispersion ;
     in contact  with  toxic  substances bacteria secrete organic acids or  produce
     metabolic products, causing a bio-leaching.

Bio-leaching of metals is a well-known phenomenon used in mining (Cu, Co, Au ore
extraction), in mineral valorisation or in corrosion protection.  All metals in the
aquatic environment are susceptible  to microfouling  by  bacteria, protozoa and algae.
Regarding corrosion, bacteria are thought to be responsible for ca. 50 % of the
failures of buried pipelines and cables.
CONTEXT OF MODERN  AQUATIC ENVIRONMENTS VS.
     THE ABR-CIS TREATMENT METHOD
It is now obvious that the early 60's idiom "The solution to pollution is dilution" is a
bad  and non-appropriate  management rule regarding the  protection  of aquatic
environments. The large supply of nutrients to surface water systems combined with
the input of organic-rich  and contaminated effluents  has caused  the well-known
problems of eutrophication contamination  of water (direct)  and contamination  of
sediments (Indirect).

The problem is  complex because  phenomena like eutrophication and desequilibria in
oxygen-balance  is also  linked to the  decrease of light penetration due to turbidity
increase. Turbidity increase may itself be caused by unappropriate agricultural sand-
                               11

-------
 use ( increased  run-off  of fine  sediments),  over-saturation  by organic compound
 effluents  (floating  at the  surface of the water)  and/or  other man-related  activities
 (dumping, dredging, ...) .

 Eutrophication is a situation where the aquatic  system  has an over-abundanccy of
 nutrients  such as  phosphates and  nitrates,  and where the  natural biodegradation
 process  by  micro-organisms has  become  over-saturated.  The  direct  effect  of
 eutrophication is oxygen  depletion and important deficits  in oxygen  supply.  This
 oxygen deficit is most important in the deeper parts of the  water column and in the
 sediment  deposits.  Indirect  effects  are e.g.  specific and intense  algae  blooms,
 reduction of diversity and abundancy of aquatic organism species, bad smell and other
 unattractive phenomena affecting the aquatic system.

 Phosphates can be treated in-situ by  Ca-rich  compounds  or FeCl3  (precipitation).
 Such treatment precipitates the phosphates to  poor-soluble salts. Nitrates are difficult
 to combine into  insoluble compounds. The problem with nitrates and phosphates is
 also the diffuse input into surface waters (from many small effluents or from run-off).

 The only  efficient treatment  method  of  organic and anorganic contamination  of
 aquatic systems is the tackling at the source, i.e. the strong reduction or prohibition of
 effluents.

 Muddy sediments  tend  to accumulate and  concentrate all  kinds  of organic and
 anorganic contaminants from the water column through  different mechanisms :
 - adsorption of contaminants to humic-clay complex or Fe-compounds
 - ion-exchange to clay particles (heavy metals) ;
 - co-precipitation ;
 - accumulation through plancton deposition.

 Consequently, highly industrialised countries face serious problems of environmental
 protection  of the aquatic  system due to the contamination of water  and sediments.
 Regarding the sediments the problem becomes accute because :
 1.    canals, waterways  and rivers  have to be maintained  at a given (artificial  ?)
      design profile to ascertain safe  navigation, to prevent  floodings or even  for
      sanitation reasons (smell, pollution,  ...)  ;  the dredged quantities to be  disposed
      (on land or in aquatic system) are quite important with respect to the available
      disposal  capacity : there is a lack of disposal facilities resulting- in drastic rise of
      disposal prices (upland disposal fee in  1990 up to 80 U.S. $/m3) ;
2.    industrial treatment of the sediments is generally considered as unfeasible due to
      (high) treatment costs, due to low treatment production rates and to too specific
     treatment methods with low efficiencies.

These facts have lead to the development of in-situ microbiological treatment  methods
where the  natural indigeneous bacterial flora  is  used  and  stimulated to obtain the
following :
                                12

-------
a)    mineralisation of organic matter resulting in microbiological dredging ;
b)    mineralisation of organic contaminants by treatment in-situ ;
c)    re-establishment of normal oxygen balances in the whole aquatic system.

This experimental research has led to the development of ABR-CIS.
Optimum conditions to re-activate normal bacterial degradation processes

The transformation of organic products into basic minerals such as CC>2 and H20 is
efficiently  done by aerobic  bacterial  degradation.  From  the  previous chapters it
appears how the normal biodegradation processes may be inhibited, i.e. essentially by
oxygen deficits.

To re-activate the desired aerobic biodegradation one should work to get optimum
conditions in the sediment deposit i.e. :
-  oxygen  supply   in very  dispersed  and  very  bio-available  way,  to allow  fast
   metabolism and population growth,
-  temperature : bacteria are active above ca. 8°C ;  the higher the temperature (up to
   30-40" C) the greater the bacterial activity ;
-  season : the season is important in relation to temperature and ecosystem cycli ;
-  toxicity  : no toxic elements to bacteria should be present in the environment;
-  organic  matter  :  type and quantity  of  organic  matter  is important for optimal
   aerobic biodegradation ;
-  specificity and mutant population :  for particular contaminants it is  important to
   assure complete biodegradation without formation of toxic intermediate products.
Augmented Bio Reclamation (ABR) - Conditioning In-Situ (CIS)

The ABR-CIS-system (Augmented Bio Reclamation - Conditioning In-Situ) is a joint
effort of the companies Ebiox A.G.  (Sursee, Switzerland), EMR b.v.b.a. (Ghent,
Belgium) and HAECON N.V. (Ghent, Belgium).
ABR micro-organisms are extracted and isolated from the sediment.  Thereafter they
are cultured on carbon substrates,  similar to the target contaminants to be treated. The
ABR  micro-organisms are then  grafted on  wood chips for  transport. ABR is  a
development and a trade mark of the company Sybron.

ABR micro-organisms are aerobic and the strains are selected in order to achieve an
as complete mineralisation of the contaminants as possible.

Conditioning In-Situ, CIS  is a technique in which carefully  selected  and blended
natural mineral products are injected into the sediment deposits. The CIS  products are
rich in free and bio-available oxygen (porous structure) and supply the ecosystem with
the required oligo-elements and  nutrients.  CIS  is  stimulating  and re-activating all
                                13

-------
 indigeneous natural aerobic bacterial activity causing an increased biodegradation of
 the  undesired sedimentation  products in  the  aquatic  system. CIS products are
 guaranteed to be harmless to components of the aquatic trophic chain, including man.

 ABR-CIS is  designed to be a microbiological in-situ treatment technique of organic
 muddy sediments contaminated with particular organic biodegradable micro-pollutants

      BTX and derivates ;
      mineral oil and compounds
      PAH's ;
      phenols, formaldehydes;
      pulp, cellulose, starch, proteins, fats ;
Process of ABR-CIS treatment

Each project is first analysed on the microbiological treatment of the mud sediments
and the organic contaminants. Therefore samples of the sediments are cultured in the
laboratory and microbiological proliferation  is monitored with respirometer-testing
(oxygen  uptake) and total  resp. selective  plate counts. When the treatability of the
mud  is confirmed  the  ABR-CIS treatment is designed (dosage,  type of bacteria,
injection scheme, ...).

The  treatment process is executed as follows :
1.    culture of ABR micro-organisms ;
2.    blending of CIS-conditioner ;
3.    mixing, drumming and transport of ABR-CIS powders ;
4.    dilution and aeration of ABR-CIS powder with water of the waterway ;
5.    injection  of ABR-CIS suspension  in the sediment by using adapted dredging
      equipment or high pressure jetting ;
6.    monitoring of the sediment-level and  sediment-quality.

Up till June 1992, 7 ABR-CIS full-scale projects have yet been executed on industrial
scale. Volumes ranging  between 100 m3 and 5.000 m3 have been treated in-site and
monitored during 12 to 20 months after injection.

Reasons for applying in-situ treatment are related to :
a)     volume-reduction of the sediment (and avoiding upland disposal) ;
b)    decontamination of mineral oils and/or PAH's.
                                14

-------
 RESULTS  OF MICROBIOLOGICAL TREATMENT OF
     SEDIMENTS
 To evaluate the efficiency of ABR-CIS treatments this last chapter will focus on some
 projects already executed and the treatment results obtained.
Re-establishment of oxygen balance

The best way to measure the physico-chemical and biochemical status of a sediment
deposit regarding oxygen balances in the determination of the redox potential, Eh. It
has been  stated previously that  anaerobic conditions are  associated to reducing
conditions, i.e. with negative Eh.
       REDOXPOTENTIAL Eh (mV)
    100
     60
    -60
    -100
    -200
                      ABR/CIS
                                                OXYDATION |
    -2SO
                       LJI
                                                REDUCTION
           34
                        0:00      04      41     1:53
                          DAYS (cfr. ABR-CIS-treatment)
                                                    2:24
                                                           4:10
Figure 6 :  Evolution of redox potential in mud deposits  after ABR-CIS treatment
          (project Zonh.)

Shortly after ABR-CIS treatment the black reduced mud deposit is changing to mud
with  neutral or even positive Eh-values. This is  illustrated on figure 6 where  Eh
values measured on the field show a quick evolution from ca. -200 m V to positive /
neutral values.
                            15

-------
Improvement of microbiological activity

The oxygen input by CIS allows the ABR and indigeneous bacteria to proliferate. This
is measured  by taking  samples at regular time intervals and by measuring the total
plate count.

This is illustrated on  figure 7 where the strong increase  in total  plate count  after
treatment can be seen.

Volume reduction of sediment deposit hy mineralisation

The biodegradation of organic matter in the mud deposit causes a volume reduction of
the sediment deposit (mineralisation of organic matter and/or disruption of the mud
flock-structure).  This   can be seen  on comparative  sounding-diagrams.  This  is
illustrated by cross-sectional soundings of waterways in which the sediment deposit
was treated with ABR-CIS. Figure 8 shows the volume reduction  of  the sediment
deposit  in   a   small   freshwater  river where  ABR-CIS  was   applied  (project
Krimpenerwaard, The Netherlands, project ordered and executed in collaboration with
Consulmij B.V.).
          TOTAL PLATE COUNT (CFU/g.dry »olid)
   1,0006* 11
   VOOOE-10
   1.000E-09
                            ABR/CIS,
   roooE-oe.	
   10000000 • — — •• -----
    1000000 >  '
      10000 '




"--^
*r ,
•:
c
&* V-
—



_^__




1
i i!
I 1 i
                        28/6/901-8)    S/9/»0(To)    8/10/90<«33)    14/11/901*70)
                            DATE OF SAMPLING (days w.r.t To)
Figure 7 :  Increase of total plate count (TPC) after ABR-CIS treatment
                               16

-------
         DEPTH (m)
      0.6 r^	
     -0.6
                            -••
     •1.6
                        23466
                           DISTANCE ALONG CROSS-SECTION (m)
                   DESIGN PROFILE
                                   DATE OF SURVEY
                                  •*-08/05/91   -8- 24/06/91
                                                          •22/07/91
Figure 8 :  Volume reduction of sediment deposit after ABR-CIS treatment (project
           Krimpenerwaard, the Netherlands)

A global volume reduction of ca. 50-60% of the original volume has been realized by
this microbiological dredging. Figure  9 shows a simular cross-secitonal  view with
sounding results at different time intervals showing the progressive volume reduction
of the  sediment  deposit  in  a brackish/fresh  water canal  (Bakhuistervaart, the
Netherlands, project ordered by J. Klip B.V.).
         DEPTH ( m  NAP)
     -0.8
     -1.3
     -1.8
        -2
                          4     6      8     10     12     14
                          DISTANCE ALONG CROSS-SECTION (m)
                                                            16
                                                                  18
                                                                        20
              DESIGN PROFILE
              8/08/90
SOUNDING DATE
24/04/90
21/09/90
                                                       14/06/90
Figure 9 :  Volume reduction of the sediment deposit after ABR-CIS  treatment over
           the whole deposit thickness (Bakhuistervaart, the Netherlands)
                                17

-------
 On figure 10 the reference cross-sectional profile of the Bakhuistervaart is shown to
 analyse the natural volume-evolution of the sediment deposit when this is not treated.
               DEPTH I m NAP)
                              | NON-TREATED REFERENCE SECTION
                             i     e    »    10    12    14    ie    IB   20
                              DISTANCE ALONG CROSS-SECTION (ml
                                    DATE OF SURVEY

                              -24/04/80  -H-11/07/BO
-B/00/80 I
Figure 10 :   Result  of  sounding  survey  of the  canal in  a  non-treated  section
              (reference section).

A global volume reduction of ca. 50% has been realised in ca. 18 months.
Biodegradation of organic compounds

For the Bakhuistervaart project  3.100 m3 of  sediments were heavily contaminated
with  mineral  oil  (2000-3000 ppm). After  ABR-CIS  treatment (18  months after
injection) the mineral oil content had significantly decreased down to acceptable legal
levels (fig.  11).
                  MINERAL OIL CONTENT (1000 no/kg 
-------
 A similar in-situ biodegradation of mineral oil has been obtained in the Park River
 (project Edegem ;  figure 12) and the Westeindse Waterweg (project Zoeterwoude,
 ordered by and executed with Consulmij B.V. ; figure 13).
          MINERAL OIL-CONTENT (1000 me/kg d.m.)
                3/28
                                 4/4              4/30
                             DATE OF SAMPLING (month/day) 1990
e/S
                                    LOCATION SAMPLES

                          (SAMPLE NR. 1  SOUTH  H3 SAMPLE NR. 2 CE NTER
Figure 12 : Biodegradation of mineral oil in Edegem Park River (Belgium)
         MINERAL OIL CONTENT (1000 mg/kg d.i.)
ABR/CIS
«




fc\S
fcr.
1 '-J
••.% ,
!* '\, -
&-
?;t
\..S',,
*••
hi
24/06/81
1

3IT-1

Hi
30/07/91 03/08/91 06/11/91 Ott/12/91
DATE OF SURVEY
SAMPLING LOCATION I
EDSAMPLE NR. 1 ^SAMPLE NR. II 1
Figure 13 :   Biodegradation of mineral oil in Westeindse Waterweg  (Zoeterwoude,
             the Netherlands).
                                19

-------
In the Zoeterwoude project a significant contamination with PAH's occured as well
(major PAH's : Fluoranthene, Pyrene).

On figure 14 the global evolution of the contaminant-content is shown illustrating the
progressive biodegradation of the considered contaminants although less spectacular
than with e.g. mineral oil content. Some fluctuations occur in the detected PAH levels
; these are supposed to be caused by analytical uncertainties, sample representativity /
reproducability and release due to microbiological action.
             TOTAL PAH CONTENT (mo/kgd.m.)
ABR-CIS
In
"
,11
i


•".


-
I . 1}

1
• i i , I

1 rn
i , ,
14/0t 30/07 lot 03/09 /10 04/11 /I! /01 /02 /03 14/04 /OC lOt IJ/07
1991 DATE OF SAMPLING 1992
SAMPLING LOCATION
tm BLANCO BCO L7J MONSTER SCO 1 D MONSTER BCO 2
•• BLANCO RIJNLANO 1 	 1 MONSTER RIJNLAND 1-2 dl MONSTER RIJNLANO 3'4
Figure 14 :   Biodegradation of PAH's in the mud of the Westeindse Waterway
             (Zoeterwoude, the Netherlands).

Further additional  scientific  research is  necessary  however to analyse the exact
biodegradation process and the variations in adsorption of contaminants  to the clay-
humic complex.
ABR-CIS process costs

The project preparation product and  injection costs for a typical ABR-CIS project
range between 20 and 30 U.S.S par m3 of sediment in-situ to be treated. When a 50%
volume reduction is achieved by microbiological ABR-CIS dredging, the cost of the
dredging of 1 m3 will consequently range between 40 and 60 U.S.$.

These  costs must  be compared with  the classical dredging and disposal costs  (for
contaminated sediments) :
                               20

-------
     dredging
     dewatering
     disposal

     total
 ca. 15 U.S.$/m3
 ca. 60 U.S.$/m3
 ca. 80 U.S.$/m3

ca. 155 U.S.$/m3
This has been  confirmed  by different  cost/benefit  calculations executed by the
managers of the waterways treated with ABR-CIS.

In the Bakhuistervaart  project the municipality  has officially declared  that the
application of ABR-CIS system represented a saving of ca. 70% with respect to the
classical approach of dredging and disposal.
LITERATURE

CLAYPOOL G.E. and I.R. KAPLAN, 1974. The origin and distribution of methane
     in marine sediments. Mar. Sci. 3 : pp. 99-139.
FENCHEL T.  and T.H.  BLACKBURN,  1979.  Bacteria and  mineral  cycling.
     Academic Press.
DE MEYER Chr.,  MALHERBE B.,  DE VOS K., 1992. ABR-CIS Bioremediation
     for  in-situ treatment of sediments.  Proc.  Forum for Applied Biotechnology,
     FAB '92, Brugge 24-25.09.92.
FRANCIS BOEUF Cl. and BOURCART J., La Vase,  Revue Scientifique, Paris,
     1942.
JONES J.G.,  1979. Microbial nitrate reduction in  freshwater sediments.  J.  Gen.
     Microbiol. 115 : pp. 27-35.
MARTIN  Y.  et BONNEFOND J.L.,  1986. Conditions  de decroissance en milieu
     marin des bacteries fecales des eaux usees urbaines, Oceanis  12, Fasc.  3  : pp.
     403-418.
POMMEPUY M., GUILLAUD J.F., LEGUYADER P., DUPRAY E., CORNIER
     M. The fate of the bacterial load of dredged sediemnts. Proc. of Int. Seminar on
     the Environmental Aspects of Dredging Activities, Nantes (Fr.) 27/11-01.12.89.
ROSZAK  D.B.  and COLWELL  R.R., 1987. Survival strategies of bacteria in the
     natural environment. Microbiol. Reviews 51  no. 3 : pp. 365-379.
                             21

-------
                      In Situ Bioventing of a Diesel Fuel Spill

        T.L.Bulman, Manager Consulting Services, Campbell Environmental
       M. Newland, Environmental Control Engineer, Campbell Environmental
                                   6 Gould St
                                  Osborne Park
                             Western Australia 6017

                      A.Wester, Industrial Chemist, Westrail
                                Montreal Rd East
                                    Midland
                             Western Australia 6056
       A diesel fuel spill  occurred at a state rail yard late in 1989.  The leak occurred
from a corroded buried pipe leading from an above ground tank.  Fuel accumulated at the
surface of a shallow (2.5 m) unconfmed aquifer comprising  sand  and silty  sands of the
Guildford formation.  An  area near the storage tank of 2000  m2 and over a  depth of 1.5
to 2.5  m was contaminated with fuel.  A total of 135,000 L of fuel was recovered using
oil skimmers in groundwater drawdown wells.  An estimated  50,000 L (approximately
1.5% w/w in soil) remained in situ, smeared in the areas of drawdown wells  to a depth of
3.5 m.
       The potential for active and passive in situ bioremediation was explored through
laboratory assessment of oil mobility  and biodegradation rates  and field assessment of
contaminant distribution and groundwater flow.   Oxygen  supply was  identified as  the
factor  most limiting biodegradation of the hydrocarbons, while the rate of degradation
was further increased by supply of the nutrients nitrogen and phosphorus.  Although  the
groundwater flow  velocity  achievable  under drawdown conditions was rapid  (2 x 10~3
cm/sec) it would  require  many  years  to  supply sufficient  oxygen  in situ  to  effect
bioremediation.  In contrast, supply of oxygen (as air) to unsaturated aquifer material was
predicted to achieve bioremediation in approximately 18 months.
       A  plan for  remediation  was  designed  in which the  contaminated area  was
dewatered, oxygen was supplied via air extraction from cased bores slotted throughout the
contaminated zone and  nutrients  were supplied  via a  micro-irrigation system.  A pilot
scale remediation on one quarter of the site is  currently underway to confirm rates of
degradation achievable in the field.  Progress to date has indicated hydrocarbon reductions
of 2000 to 4000 mg/kg  (10 to 30%) in the unsaturated soil to a depth of 3 m over a six
month  period  of venting only.  Dewatering  in the  3 to 3.5  m  zone, however,  was
incomplete and subject to fluctuating water levels.  A further reduction averaging 30%
took place over the subsequent six months  of venting  and nutrient addition and extended
to the  3.5 m  depth.  Further investigations to be carried out during  the pilot period
include 1) an evaluation  of the efficiency of oxygen supply via  vacuum or air injection, 2)
the fate of nutrients in unsaturated  and saturated zones,  3) extent of remediation  in the
unsaturated and saturated portion of the contaminated area.

Keywords: in situ bioremediation, bioventing, hydrocarbon contamination
                                            22

-------
                                  INTRODUCTION
       A  major diesel spill from  a  fuel line at the State Railway (Westrail) Marshalling
Yard occurred in 1989.  The spill  resulted in migration of fuel through a storm water
drain to the Perth Airport Southern Main Drain, necessitating immediate cleanup action
along a 3 km length of the  drain.   Further  investigations  on the Westrail site itself
revealed an area  near the storage  tank of approximately 2000 m2 and over a depth of 1.5
to 2.5  m was contaminated  with  fuel (Figure  i).   Fuel  recovery  operations were
immediately put  into effect and options were  explored for complete remediation of the
site.
                         ELECTRICAL
          STORM DRAIN
                                                          • monitoring bore

                                                          O bore (discontinued)

                                                            pumping bore


                                                                   10 m
                                        ROAD
      Figure 1 Site Plan of Spill Site at Westrail Forrestfield Marshalling Yard
                                            23

-------
       Due to the number of buried services and above ground tanks in the spill area, the
proximity of a carriage shed in full use and the low slope stability of site soil, excavation
was considered  undesirable for remediation  of the site.  The potential for active  and
passive in situ  bioremediation was  explored through  1)  site  investigations  of  diesel
concentrations, soil and  aquifer characteristics, 2) laboratory assessment of oil mobility
and  biodegradation rates and  3) pilot scale evaluation  of in  situ  bioventing.   Having
collected  information  in laboratory  and  field  trials,  it was  desired  to  determine if
commercially available bioremediation models would be useful in designing  full scale
remediation.
                                SITE DESCRIPTION
SPILL DELINEATION

       An initial survey  to determine the extent of contamination was performed by the
Geological Survey staff of the Western Australian Department of Mines.  A total of 29
holes were augered to a depth of 4 to 5 meters.  Soil was recovered for analysis  from
above the water  table at  1.8  to 2.5 m.   Soil samples could not be  recovered  from the
saturated zone as the sand could  not be held in the core barrel.   The soil samples
consisted of 2.25 m coarse to medium sand fill over Guildford Formation sediments of
silty sand.  Clayey sand  lenses were intersected in 5 boreholes at a depth of 4 to 4.5 m.
Each borehole was lined with slotted PVC pipe for future use in water sampling.
       A total of 280 soil samples were analysed by Westrail  staff and indicated a well
defined contaminated area extending in a radius of 20 to 25 m from the source of the spill
(approximately 2000 m2) from a depth of 1.5 to 2.25 m (average thickness 0.75 m).  Oil
concentrations ranged from 0.01 to 10% w/w.
       Free product  recovery  was  implemented  on the  site  through trenching  and
installation of soak wells, and finally through installation of four drawdown wells supplied
with skimmer pumps.
       Following  recovery  of  approximately  135,000  L  of free  product,  removal
efficiency was reduced to approximately 1 L/day.  At this point a second set of soil cores
were  collected to determine the content of  fuel remaining in the  soil and  to  collect
samples for laboratory treatabiUty testing.
                                             24

-------
 AQUIFER CHARACTERISATION

       The initial site survey performed by the State Geological Survey in March  1990
 described an unconfmed aquifer of coarse to medium sand  with clayey lenses at 4 to 4.5
 m depth.  The depth of the  aquifer, based on regional  information,  was estimated at 20
 m.  The natural hydraulic gradient  across the site was 0.003 m/m.  Subsequent intact
 cores collected in August 1990 following recovery of free product identified bulk densities
 ranging  from 1.5 to 1.8 g/cm3, porosities of 0.28 to 0.44 and oil contamination of 1.5%
 w/w in the 2 to 2.5 m depth and 2.5% w/w in the 2.5 to 4 m depth.
       Pumping tests were performed during installation of the drawdown wells.  These
 wells were fully slotted over a 12 m depth.  The depth averaged  hydraulic conductivity
 which was measured using these wells was 3 x  1Q4 to 2 x 10~3 cm/sec.
       To evaluate the practicality of an injection/withdrawal method of bioremediation, a
 field tracer test was conducted during continued water drawdown pumping.  The test was
 performed by  injecting  1500 litres  of  a solution  of 100  mg/kg  Li as LiCl.   The
 concentration of Li was subsequently monitored in all monitoring wells on the site over a
 21  day period.  Concentrations of Li in groundwater were used to estimate the rate and
 direction of groundwater movement under drawdown conditions.  Under these conditions
 the average gradient  from the point  of injection to the drawdown well was 0.1  m/m.
 Discrete flow directions, however,  were observed with differing flow velocities.   Flow
 circumvented  the  contaminated  zone  with  velocities  of 1.5  to 4  x  10"3 cm/sec,  in
 preference to the direction of the drawdown well in the contaminated zone which  had a
 velocity of 2 x  1CT* cm/sec.
                            TREATABILITY TESTING
SHAKE FLASK TESTS
       Preliminary screening experiments were conducted in shake flasks using treatments
of either water or nutrient solution under aerobic and  anoxic conditions.  These trials
demonstrated that aerobic conditions with added nutrients resulted in the greatest rate of
microbial  growth and  concurrent hydrocarbon degradation.   Very  limited degradation
occurred under anoxic conditions, even with nutrient supplementation. Results from trials
with nutrient  and oxygen  supplementation  and  the  control treatment (no oxygen or
nutrients supplied) are illustrated  in Figure  2.  These  trials confirmed that native soil
organisms were  capable of degrading  the diesel, however  oxygen  and nutrients were
limiting to degradation.
                                            25

-------
                    400
                •3  300
                 s
                 rt
                 4-1
                 o
a
o
,0
u<
rt
O
o
•o
>,
                    200
                    100
                                                     control
                                 oxygen and
                                    nutrients supplied
                                        10
                                15
20
25
                                        time (days)
        Figure 2  Results of Shake Flask Tests for Biodegradability of Diesel
        Product With and Without Supply of Oxygen and Nutrients
COLUMN TESTS

       Oxygen supply, in particular, was a problem in subsequent trials using saturated
soil columns  supplied with a  flow of well aerated  water.   Columns were set up to
simulate a control (water flow only) and supply of oxygen and nutrients in water at a rate
designed  to  simulate field  conditions.     Column  flow  rates  and effluent  nutrient
concentrations  were  monitored on  a  daily basis.   Flow was  initially restricted but
increased as  oxygen  and nutrient utilization increased.  Over  the   length of the 60 cm
column roughly 50%  of the added N, 80% of the added P and almost all supplied oxygen
was consumed. An encouraging increase in microorganism numbers was also noted. No
degradation of hydrocarbon was detected,  however, within the 8 week study. The initial
diesel contamination of the soil was very high (15000 mg/kg) and the amount degraded
according  to  stoichiometric calculations  (60 mg/kg)   was  well within  experimental
uncertainty.
       Trials were also conducted under unsaturated conditions in which a drip irrigation
system supplied nutrients and kept the soil moist.  Microorganism counts increased 1000
fold  within  a few weeks and  20%  of the residual  hydrocarbon  was degraded  after 4
weeks.
                                            26

-------
       Calculations based  on the stoichiometric amount of oxygen required  to  totally
mineralise the residual hydrocarbon   content in  site soil supported the results  of the
column  tests.   Based  on  an assumed  carbon  content  of 85%  in  the hydrocarbon,
approximately 3.5 mg of O2 would be required to mineralise 1 mg of hydrocarbon.  On
the basis of  measured  hydrocarbon  concentrations  of  1.5%,  supply of oxygen at a
solubility limit  of 8  mg/L in  water,  would  be insufficient  to mineralize  substantial
hydrocarbon in site soil within a  manageable time period.  In contrast, supply of oxygen
in the air phase, at a concentration of 20% by volume, would support degradation of the
hydrocarbon in site soil within 1 to 2 years.
       These  studies  confirmed  the  presence  of  indigenous  petroleum  degrading
organisms.  In view of its history as a fuel storage area this was not unexpected.  Oxygen
and nutrients are limiting degradation, however, under field conditions.  Residual diesel
concentration  on the  site was very high after recovery  of free product.   Consequently
bioremediation requires a significant supply of oxygen to meet the demand for total oil
removed. It is evident that groundwater flow through the  site would make saturated phase
supply of oxygen impractical.  Remediation under unsaturated conditions with supply of
oxygen through vented air was therefore viewed as the most practical approach.
                           PILOT TEST PERFORMANCE
OBJECTIVE
       A six month pilot scale remediation was designed involving dewatering one quarter
of the site  and encouraging in situ bioremediation through  bioventing.   The primary
objective of the pilot  trial was to confirm  the  effectiveness  of nutrient  supply  and
bioventing to meet proposed remediation criteria for diesel of 1500 mg/kg.  The extent of
remediation  was to  be determined by  periodic  monitoring of  soil  for  hydrocarbon
contamination and monitoring  of groundwater  for hydrocarbon and nutrients.  A second
objective was  to provide sufficient data  on rates  of oxygen uptake and  hydrocarbon
degradation  for calibration of the VIP and BIOPLUME II models.  These models would
then be used to predict  hydrocarbon loss for a full scale bioremediation design involving
the saturated and unsaturated  zones.   The design would subsequently  be  tested in  full
scale remediation.
DEWATERING

       The natural groundwater table resided at 2 to  2.5  m below grade, with seasonal
variation  within  that range.   This  was  depressed  by  the  weight  of diesel  in  the
contaminated area and  through  generation  of cones of  depression  to  facilitate  diesel
recovery.  As a consequence, residual contamination was smeared within the 1.5 to 3.5 m
zone.  Dewatering for bioventing of this zone was carried  out to a designed depth of 3.5
m via electric  submersible  pumps  installed in  three shallow wells spaced  around the
                                            27

-------
 northeast corner of the site  (Figure  3).   These wells are also supplied with skimmer
 pumps  for recovery of any additional oil  which becomes mobilised and  collects  in the
 wells as a result of dewatering.  The deep (12 m) recovery wells and skimmers have also
 continued in operation.
         06WATEF1ING WELL
                                           CARRIAGE  SHED
       PEWMNENT

       M3NITORING WELL
                                 BLOWER
                                                             DIRECTION OF   NATURAL

                                                            GROUM)WAT6R FLO*
                                                           AIR MANIFOLD
                                                            AIR EXTRACTION WELL
                   PILOT WORK AREA
                                                         SCALE:
    Figure 3 Schematic Diagram of Bioventing Pilot Trial - Plan View
       Intermittent failure of  one or  more  of the groundwater extraction pumps  has
resulted in a fluctuating water  table between 3 and 3.5 m depth for the first four months
of the pilot study.  This problem was compounded by water entering the venting  system
and tripping out the air extraction exhauster.
AIR SUPPLY

       Soil aeration has been performed via air extraction from 9 bores  installed to 4 m
with 50 mm PVC casing slotted between 2.5 and 4.5 m bgl  (Figures 3 and 4). A coarse
sand filter pack was installed around the slotted interval with a bentonite seal to the soil
surface.   The  wells  are  joined to  a valved and instrumented  PVC  header system
connected, through a moisture  "knock-out pot", to the suction side of  a small  blower.
Exhaust is direct to atmosphere.
                                             28

-------
               GPOUNDWATER
        NUTRIENTS
            DIESEL CONTAMINATION
                                                                 NUTRIENT FLOW
    Figure 4 Schematic Diagram of Bioventing Pilot Trial - Section View
       Influx of air is facilitated by the existing fully slotted monitoring and  drawdown
wells, as well as  exposed soil  surfaces.   Trenches  which  were used  for free product
recovery have been back filled and provide zones of high permeability  for air flow.  A
vacuum of less than 2 kPa has been measured at all air extraction wells during continuous
operation.   All wells are  in an  area  of the site which is not sealed  (i.e. no  asphalt or
concrete cover).
       Dewatering problems have resulted in intermittent function of the aeration system
over the first four month period.  A strong hydrocarbon odour was initially evident in the
exhausted air, which has decreased with time.
       The area of influence of the air extraction bores has been assessed by applying a 3
to 5 kPa  vacuum to  each well separately and measuring vacuum pressures in the other
wells.   Vacuum  pressures of approximately 20 Pa were measured at radii of 3 to 4 m
from the  source  well.  Well spacing and configuration at final scale will be optimized
based on these results.
                                             29

-------
NUTRIENT MOBILITY

       The supply of nutrients to  the study area was delayed  and began six months after
initiation  of  site dewatering.   Nitrogen and  phosphorus are  supplied  through a  drip
irrigation  system installed  15 cm below  the  soil surface.   Stock  nutrient solution  is
prepared using potable water and is added  to recovered groundwater via a dosing pump
for irrigation.   Addition  of nutrient solution is adjusted  to  balance the  uptake rate  of
nutrients and maintain a soil moisture content of approximately 60 % of the water holding
capacity.
       To date the mobility of nutrients has been monitored in groundwater only.  Three
new monitoring wells fitted with dedicated bladder pumps were installed down-gradient of
the site ..(Figure 3).   Groundwater  concentrations  of nitrogen  or phosphorus  measured
down-gradient of the site have demonstrated no response to nutrient addition levels.  It is
planned  to  install soil  pore  water  monitoring  devices  to collected  additional data on
nutrient  availability  in  order  to better estimate uptake rates  occurring in the  field and
allow better optimization of nutrient addition.

OXYGEN UPTAKE

       The  rates  of  oxygen uptake in the unsaturated zone were monitored towards the
end of the six month pilot trial (July 1992).  A total of 15 holes were hand augered to the
capillary zone  (3 to 3.5 m) along transects through the pilot  study  area, with varying
distance  from the air  extraction wells,  and into  the uncontaminated  and unremediated
portions  of  the site.  Soil samples  were collected from  all  locations.  In 10  locations,
samples were taken at several  depths.  PVC probes  fitted with  soil oxygen sensors (Jensen
Instruments) were installed in all locations.   To  date,  oxygen  consumption  has been
assessed  at  4 locations, including in the  uncontaminated  and  contaminated unremediated
areas of the site.  Initial oxygen readings were recorded (during venting).  The blower
was  then turned off and oxygen concentrations were recorded  over a period of  2.5 hours
and  3  days.  The rate  of  oxygen consumption  was  assessed  as the decrease  in oxygen
content over time.   Initial concentrations and rates  of oxygen uptake are provided in
Table  1  with  the measured  hydrocarbon  content.   High  hydrocarbon  concentrations
correspond  to  high  initial oxygen  uptake  rates to  a  maximum  of approximately 2.5
%/hour.  Uptake rates did not correspond  to initial oxygen concentration.
                                            30

-------
Location
unremediated
#11 (2.8m)
uncontaminated
#2 (2.8 m)
pilot area
#4 (2.8m)
(2.5 m)
#5 (2.8 m)
(2m)
(2.5 m)
DOC0
(%)

1

11

22
21
22
21
21
DO Rate0
(%/hr)

-0.4

-0.429

-2.43
-2.4
-2.49
-0.686
-1.37
Diesel
(mg/kg)

3800

0.2

3600
na
7700
0.2
860
Microbial Number
(CFU/gram)

9 x 102

4x 102

8x 103
8x 103
8x 103
5 - 70 x 103
8x 103
Table 1. Diesel concentrations, microbial numbers and initial oxygen concentrations and
uptake rates measured during bioventing pilot trial
DIESEL CONCENTRATIONS

       Diesel  concentrations in soil are currently available for five monitoring periods.
The first three periods were prior to the pilot study, as follows; 1) prior to free product
recovery (March 1990), 2) following recovery of most free product (September 1990), 3)
immediately prior to the pilot test (June 1991).  The last two monitoring periods included
4)  following six  months venting but  prior to  nutrient  supply  (January,  1992)  and 5)
following six  months of both nutrient  supply and bioventing (July, 1992).  Cores taken
following free product recovery  were located according to distance from drawdown and
oil  recovery wells.  Those taken  following the initiation of venting were located according
to distance from the air extraction wells.
                                            31

-------
                    35000
                                          MONTHS
         Figure 5 Average Trend  of Diesel Concentrations in Pilot Trial Area
         During Product Recovery  and Bioventing Operations
       The trend of overall concentration (averaged over depth and location) with time is
 illustrated  in Figure 5.  The line illustrated has been fitted to  all data using  a distance
 weighted least squares smoothing routine  (SYSTAT).  The majority of the decrease in
 concentration (55%) occurred over the period of free product recovery of 15 months.  A
 further 13%  decrease occurred over  the following  6 months,  during the  period  of
 dewatering and venting.  During the subsequent 6 month pilot  trial period (venting and
 nutrient  supply)  hydrocarbon   concentrations  decreased   by   18%   of  the  original
 concentration, or 50%  of the concentration  remaining after free product recovery.
       This decrease in concentration  is not consistent over  the  site,  however.   The
concentrations  measured  at  specific depths are  illustrated  in Figure  6 for  the  four
sampling  periods  during  remediation  of the site.  A distance  weighted least squares
routine was used  to illustrate average  trends in hydrocarbon  concentration  with depth.
The initial spill was located  in the 2 to 2.5 m  depth.  The  first six months of product
recovery  involved  cones  of depression which  reduced the average  concentration  but
extended  it over  a greater  depth  (2  to 3.5 m) in some locations.    More extensive
dewatering lowered the water  table to  3.5  m  and extended the contamination over a
greater depth (2 to 4.5 m) prior to the pilot test  (t=15).  Following 6 months of venting,
the average  concentrations decreased  in  the  unsaturated zone above  a 3 m  depth.
                                           32

-------
Concentrations remained  high, however,  in  the  3  to  3.5  m  depth  interval  which
constituted a fluctuating water  table depth  (t=21).   Following 6 months of venting  and
nutrient supply (t=27), average concentrations had decreased over all depths.
   35000

   30000

   25000

J  20000
w
VI
s
S  15000

   10000

    5000
                                                     (4>t=21
                                               (2>t=
                                                            (3)1=15
                                          METRES
        Figure 6 Average Trends of Diesel Concentrations  in Pilot Trial Area
        at Discrete Depths at 6, 15, 21 and 27 month Sampling Periods
       During  the  course of the pilot  trial  a sequence of visible  changes was also
observed  in  free hydrocarbon collecting in the monitoring  wells.   Initially increased
darkening (oxidization) was observed in  the free product  followed by  formation of a gel
(emulsification) and a grey slime.  Finally only clear water was evident in the wells.  A
small amount of free product continues to collect in  certain monitoring wells within the
pilot study area, while  FP2, outside the  study area,  shows evidence  of oxidization and
emulsification.
       Although average  concentrations  in both  Figures 5 and  6 show decreases  with
time, the range of concentrations showed little change over the last 12 months of study.
Work is continuing on analysis of the data according to distance from air extraction wells
and oxygen supply.
                                            33

-------
CONCLUSIONS

       Both venting  alone  and bioventing  with  the addition  of a nutrient supply have
proven a valuable adjunct to free product recovery for attainment of cleanup criteria. The
average rate of hydrocarbon removal was greater with the supply of nutrients and venting
than with venting  alone.   Much work remains  to be done in  relating  the decrease in
hydrocarbon concentration to localised oxygen and  nutrient  supply within the pilot area.
Data collected to date suggest that the cleanup criteria of 1500 mg/kg will be met  within
the next 12 month period.   The data will also provide a good  basis for design of the full
scale remediation.
                                          34

-------
        REMEDIATION OF GROUND WATER CONTAMINATED WITH ORGANIC WOOD
    PRESERVATIVES USING PHYSICAL AND BIOLOGICAL TREATMENT TECHNOLOGIES

                          J.G. Mueller; S.E. Lantz; R.L. Thomas:
            SBP Technologies, Inc., Gulf Breeze, Fla. 32561, Tel.:1 904 934 9282
                              D.P. Middaugh; P.H. Pritchard:
                U.S. EPA,  ERL, Gulf Breeze, Fla. 32561, Tel.:1  904 934 9200
Abstract

        Pilot-scale field studies at the American Creosote Works Superfund Site, Pensacola,
Florida, evaluated two technologies for their ability to treat ground water contaminated with
creosote and pentachlorophenol (PCP): 1) Hyperfiltration (volume reduction), and 2)
Bioremediation using specially-selected microorganisms (terminal destruction).  The
hyperfiitration unit was operated in a cross-flow mode yielding "concentrate" (containing excluded
chemicals) and "permeate" (clean, aqueous material passing through the membrane.) Operating
over a 6-day period on site, a total of 6,300 gallons of creosote- and PCP-contaminated ground
water (average total semi-volatile concentration was 88.5 mg/L) was processed thereby reducing
the volume of contaminated material >80% while removing >95% of the PAHs.  Simultaneously,
the concentration of chlorinated dtoxins and furans were reduced from 22.5 ppb in the feed to
0.047 ppb (cumulative) in the permeate. Based on chemical analyses and biological toxicity and
teratogenicity assays, the permeate stream was acceptable for direct discharge.
        A  two-stage, continuous-flow, sequential inoculation bioreactor strategy for the
bioremediation of ground water contaminated with creosote and pentachlorophenol (PCP) was
also evaluated.  Performance of continually stirred tank reactors using specially-selected
microorganisms was assessed according to chemical analyses of system influent, effluent and
bioreactor residues, a chemical mass balance evaluation, and comparative biological toxicity  and
teratogenicity measurements. When specially-selected bacteria capable of utilizing high-
molecular-weight polycyclic aromatic hydrocarbons (HMW PAHs) as primary growth substrates
were used in pilot-scale bioreactors (454 L), the concentration of creosote constituents was
reduced from ca. 1,000 ppm in the ground water feed (flow rate = 114 L/day) to <9 ppm in the
system effluent (removal efficiency of >99%). Notably, the cumulative concentration of 8 HMW
PAHs (containing 4 or more fused rings) was reduced from 368 ppm in the ground water feed to
5.2 ppm in the system  effluent.  Moreover, the toxicity and teratogenicity of the bioreactor effluent
was significantly reduced.  Biodegradation of PCP was limited (ca. 24%) due in large part to poor
inoculation and a high degree of abiotic loss (bioaccumulation and adsorption).

Introduction

        Since 1988, SBP Technologies, Inc. (SBP) has been developing the concept of a "multi-
phasic bioremediation strategy" in cooperation with the U.S. EPA Environmental Research
Laboratory at Gulf Breeze, Florida (1-3). This strategy integrates physical and biological treatment
processes for effective and cost-efficient remediation of a variety of substrates (soil, sediment,
aqueous) contaminated by myriad hazardous wastes; organic and inorganic wastes and mixtures
thereof. Depending on the nature of the contaminated substrate (i.e., soil or aqueous material),
the treatment strategy may consist of several discrete phases.  For example, treatment of soil or
sediment contaminated with organic and inorganic wood preservatives could entail three phases:
Phase 1, soil washing;  Phase 2, dewatering of soil slurry, volume reduction of wash waters, and/or
fractionation of pollutant chemicals by hyperfiitration; and Phase 3, terminal destruction of
concentrated pollutants or recovery/recycling of valuable resources (eg. heavy metals).
                                           35

-------
         In the event that the concentrated pollutants are biodegradable, terminal destruction is
 accomplished most effectively via some form of bioremediation. Since the material has already
 been concentrated, bioreactor systems are usually more appropriate than other approaches (eg.,
 land farming). For pollutants not amenable to biological treatment (eg., heavy metals), then other
 technologies, such as incineration, chemical oxidation or solidification, become
 more cost effective due to the reduced volume of material requiring treatment.
         Other advantages of the multi-phasic strategy using integrated technologies include: 1)
 Holistic approach to site remediation: simultaneous treatment of surface water, ground water, soil
 and sediment, 2) Ability to treat complex mixtures of organic, inorganic and radioactive wastes, 3)
 Pollutants are immediately removed from contaminated matrices and separated into discrete
 fractions which greatly facilitates biological degradation, 4) Toxic compounds (i.e., metals,
 radionuclides) are removed or recycled prior to biological treatment, 5) Contained systems are
 amenable to the future integration of intergeneric microorganisms capable of degrading
 recalcitrant compounds, 6) Sole source utilization of high-molecular-weight PAHs by patented
 microorganisms facilitates the efficient removal of recalcitrant, carcinogenic and teratogenic
 creosote constituents which commonly persist during conventional biological treatment, 7) Ability
 to recover/recycle valuable products from hazardous waste sites/industrial effluent: heavy metal
 (lead, chromium), petroleum products, 8) Entirely mobile treatment systems foron-site
 remediation, and 9) Flexible, versatile systems with broad applicability that can be specially
 designed and readily modified for particular needs.
        In this paper, we describe pilot-scale field demonstrations of two technologies for the
 treatment  of ground water contaminated with organic wood preservatives (creosote,
 pentachlorophenol) and associated chemicals  (chlorinated dioxins/furans): 1) hyperfiltration
 (volume reduction), and 2) bioremediation using specially-selected microorganisms (terminal
 destruction).  These studies were performed at the American Creosote Works Superfund site
 (ACW), Pensacola, Florida, in cooperation with the U.S. EPA Superfund Innovative Technology
 Evaluation (SITE) Program. The other component of the multi-phasic treatment strategy,  namely
 soil washing, was not demonstrated since this is basically a "proven" technology, with experience
 and expertise being provided by Dredging International, Belgium, and their subsidiaries.

 Hyperfiltration System

        Technology description:  The hyperfiltration system consists of porous, sintered,
 stainless steel tubes coated with multi-layered inorganic and polymeric "formed-in-place"
 membranes.  Membranes are coated at microscopic thickness on the inside diameter of the
 stainless steel tubing by the recirculation of an aqueous slurry of membrane formation chemicals
 exclusive to SBP. Membranes are specially designed and formulated on a site-specific basis
 according to the characteristics of the chemical pollutants and the contaminated substrate. As
 shown in Figure 1, the hyperfiltration unit operates in a cross-flow mode yielding  "concentrate"
 (containing excluded chemicals) and "permeate" (clean, aqueous material passing through the
 membrane.) The device can perform filtration ranging from microfiltration to hyperfiltration.  In-
 house and independent analyses have demonstrated the ability of SBP membrane systems to
 extract (retain) heavy metals, caustics and various organic pollutants including petroleum
 products, PCP, chlorinated dioxins/furans and creosote (4).  Most recent efforts are geared
 toward the design and construction of hyperfiltration systems for the recovery of heavy metals
 (eg., arsenic, mercury), PCBs, and chlorinated solvents such as TCE.

        Pilot-scale field demonstration; Ground  water was removed from the shallow aquifer at the
 ACW site and transferred to an equilibration tank where oil-phases were passively separated
 (upper and lower oil phases and aqueous phase). The resultant aqueous phase was used as the
feed water containing an average total semi-volatile concentration of 88.5 ppm with
phenanthrene (17.1 ppm) and naphthalene (12.9 ppm) as the major contaminants plus 22.5 ppb
cumulative chlorinated dioxins/furans. Operating over a 6-day period on site, a total of 23,850 L of
creosote- and PCP-contaminated ground water was processed (ca. 4000 L7day) with the
objectives of 80% volume reduction and >90%  removal efficiency of monitored contaminants.
Process operating data collected included flow rates, temperatures, pressures and electrical and
potable water usage.  The feed water, permeate, concentrate and wash water were analyzed for
volatile organic compounds, semi-volatiles, chlorinated dioxins/furans,  total dissolved and
                                       36

-------
suspended solids, oil and grease, total organic carbon and chemical oxygen
demand.
               Figure 1: Cross section illustrates stream flow within tubing.
                         Stainless Steel Tubular Housing
                         Porous Stainless Steel Tubing
                  oncentrate
ermeate
Biodegradation System

       Technology description: In conjunction with the U.S. EPA Environmental Research
Laboratory at Gulf Breeze, Florida (GBERL), SBP isolated and characterized the first axenic
bacterial cultures capable of utilizing high-molecular-weight PAHs as sole sources of carbon and
energy for growth (U.S. Patent No. 5,132,224). The action of these and other organisms toward
43 priority creosote constituents has been reported previously (5-8). These findings represent a
major advancement in bioremediation technology since the targeted pollutants (i.e.,
fluoranthene, pyrene, benzo[a]pyrene) typically persist during conventional biotreatment
operations (eg.,land farming). Unfortunately, this  same group of compounds represents the most
serious threat to human and environmental health. Thus, with the use of these proprietary
microorganisms, the efficiency and applicability of bioremediation technologies has been greatly
expanded to include creosote, coal tar and related wastes.

       Pilot-scale field demonstration: A two-stage, sequential inoculation, bioremediation
process was field-tested with ground water, highly contaminated by creosote and PCP, recovered
from the ACW Superfund Site. The first stage of the pilot-scale system used a 454 L completely
stirred tank reactor (EIMCO, Denver, CO) designated FBR1  (field bioreactor 1). The second stage
of the process made use of three, 227 L stirred tank batch reactors (bioreactors FBR2a, 2b and
2c).   The bioreactor was operated at a flow rate of 114 L/day with raw ground water containing ca.
1000 ppm semi-volatile creosote constituents and ca. 15 ppm PCP.  Additional information on
bioreactor design and operation, including inoculation, sampling and analyses, can be found
elsewhere (7). Efficiency and performance of the  bioremediation system using bioreactors and
specially-selected microorganisms was assessed  according to chemical analyses  of system
influent, effluent and bioreactor residues, a chemical mass balance evaluation, and comparative
biological toxicity and teratogenicity measurements.  Liquid/liquid extraction of PCP and creosote
constituents from bioreactor feeds, aqueous effluents, settled biomass and reactor residues was
performed as previously described (3,8); creosote and PCP were extracted from  activated carbon
samples as described previously (3,8). Toxicrty and teratogenicity analyses using embryonic fish,
Menidia beryllina, Microtox™ assays, Mysidopsis bahia and  Ceriodaphnia dubia tests were
performed as previously described (4,7,8).

Results and Discussion

       Hyperfittration system: Over the entire course of the  6-day period, the goal of reducing
the volume of contaminated material >80% was consistently achieved.  For every 1000 gallons of
feed water processed, 200 gallons of concentrate were generated and 800 gallons of permeate
was produced.  Chemical analyses showed that >95% of the PAHs were removed, and the
concentration of chlorinated dioxins and furans were reduced from 22.5 ppb in the feed to 0.047
                                           37

-------
 ppb (cumulative) in the permeate. The system was less effective in removing the lower-molecular-
 weight contaminants, specifically phenol (30% removal efficiency). However, physical analyses
 and biological toxicity and teratogenicity assays showed that the permeate stream was acceptable
 for direct discharge to the public water treatment plant.

        Biodegradation system: Mass balance chemical analyses showed that ca. 64, 77, and
 70% of the Group 1 (2 ring PAHs), Group 2 (3-ring PAHs) and Group 3 PAHs (4 or more fused
 rings) were biodegraded, respectively (Table 1). Relatively poor performance toward PCP (24%
 bfodegraded) was attributed to the use of inocuia of insufficient quality (incomplete induction and
 tow cell viability) which was not discovered until after the time of application. During continuous-
 flow operations, an average 113 L of ground water feed containing approximately 31, 539,368,
 106, <1 and 25 mg/L Group 1 PAHs, Group 2 PAHs, Group 3 PAHs, heterocycles, phenolics, and
 PCP, respectively, was added to FBR1 on a daily basis. Based on these removal efficiencies, the
 system effluent contained, on average, 0.1,1.6,5.2,1.8 and 5.6 mg/L Group 1,2 and 3 PAHs,
 heterocycles, and PCP, respectively.  A secondary treatment process using the same
 hyperfittration unit to separate of residual contaminants has been developed in the event that
 reactor effluent of this quality requires subsequent treatment prior to discharge.

 Table 1.  Mass balance evaluation3 of pilot-scale studies on the biodegradation of 30 monitored
 creosote constituents and PCP by specially-selected microbial strains.
Chemical
Group
Group 1
PAHs
Group 2
PAHs
Group 3
PAHs
Hetero-
cyclics
PCP
Total
System
Input


28256.2

484936.8

331599.9

95506.3
12061.9
System
Discharge
(Output)


47.2

1298.8

5548.0

1603.0
2674.0
Removal0


99.8

99.7

98.3

98.3
77.8
Residue
Settled
Bio mass
fry*

186.0

1320.0

5160.0

1090.0
2376.1
n
Bioreactor
and lines




10020.6


107811.1


88115.3


28437.7
Amount
Volatilized0


44.3

32.9

1.5

47.4
4101.8 <1.0
Amount
Biodegraded (%)d


17958.1

374474.0

232775.1

64328.2
9151.8

(63.6)

(77.2)

(70.2)

(67.4)
(24.1)
Cumulative values following 14 days of field operation (8 days continuous-flow operation);
b% removal - (total system input - system discharge)/total system input x 100;
Calculated value based on intermittent measurements; d% biodegradation =
   100 x total system input - (system discharge + biomass residue + line residue + volatilization!
                                    total system input

        Whereas analytical chemistry data indicate toss of parent compounds, it is often wrongly
assumed that this loss correlates with a removal of the hazards associated with the original wastes.
This is especially true in the case of bforemediation where slight modification of contaminants can
result in their apparent "degradation". Therefore, in these studies, a series of toxicity and
teratogenicity assays were performed on all starting materials, bioreactor effluent and membrane
permeate. For both the hyperfittration and bioreactor technologies,  toss of parent compound
correlated with a significant reduction in the toxicity/teratogenicity of treated material, potential
hazard associated with the treated wastes along with a decrease in the concentration of parent
compound.  Based on Microtox™ assays, the reduction in toxicity of ground water by specially-
selected microorganisms was approximately 230- and 130-fold  for FBR1 and FBR2, respectively;
tests with mysids indicated a 34- to 75-fold reduction in toxicity for FBR1 and FBR2,  respectively,
when compared to ground water.  In tests with daphnids, a 75- to 100-fold reduction  in toxicity was
noted.  Exposure of Menidia embryos to a 1% concentration of  ground water resulted in rapid
embryotoxicity (100%).  In contrast, the FBR1 effluent was teratogenic (an indication of reduced
                                           38

-------
toxicity), and 86% of the embryos exposed to effluent from FBR2 developed normally and
hatched without discernible terata; these larva were essentially identical to controls.
Conclusions

       The effectiveness of two key components of SBP's multi-phasic remediation strategy was
rigorously evaluated under field operating conditions and ultimately proven at the pilot-scale:
volume reduction via hyperfiltration, and biodegradation of persistent organic contaminants in
bioreactors using specially-selected microorganisms. The integrated use of physical and
biological treatment technologies expands the applicability of bioremediation technologies, and
provides a cost-effective, viable remedial approach for creosote- and similarly-contaminated sites.

Acknowledgments

       Technical assistance was provided by Myke'He Hertsgaard, Barbara Artelt, Beat Blattmann,
Maureen  Downey, Sol Resnick (Technical Resources, Inc), Cherie Heard, Ellis Kline, Edward
Kouba and Arjan Van Buul (SBP Technologies, Inc). Assistance from Richard Colvin, Derek Ross
(The ERM Group, Exton, PA), Scott Beckman, K.C. Mahesh (SAIC), Kim Kreiton (U.S. EPA,
RREL, SITE Program), and Madolyn Streng (U.S. EPA, Region IV) is also gratefully
acknowledged.  Financial support for these studies was provided by the U.S. EPA SITE Program,
Cincinnati, OH.  This work was performed as part of a Cooperative Research and Development
Agreement between the Gulf Breeze Environmental Research Laboratory and SBP
Technologies, Inc. (Stone Mountain, GA) as defined under the Federal Technology Transfer Act,
1986 (contract no. FTTA-003).

References

[1]. Mueller, J.G., P.J. Chapman, R. Thomas, E.L. Kline, S.E. Lantz, S.E., Pritchard, P.H.:
Development of a sequential treatment system for creosote-contaminated soil and water: bench
studies. Proceedings U.S. Environmental Protection Agency's Symposium on Bioremediation of
Hazardous Wastes: U.S. EPA's Biosystems Technology Development Program. EPA/600/9-
90/041, p. 42-45 (1990).
[2]. Mueller, J.G., Lantz, S.E., Blattmann, B.O., Chapman, P.J.:  Alternative biological treatment
processes for remediation of creosote-contaminated materials: bench-scale treatability studies.
EPA/600/9-90/049. 89 p. (1990).
[3]. Mueller, J.G., Lantz, S.E., Blattmann, B.O., Chapman. P.J.: Bench-scale evaluation of
alternative biological treatment processes for the remediation of pentachlorpphenol- and
creosote-contaminated materials: slurry-phase bioremediation.  Environ. Sci. Technol. 25:1055-
1061 (1991).
[4]. Middaugh, D.P., Mueller, J.G., Thomas, R.L., Lantz, S.E., Hemmer, M.J.,  Brooks, G.T.,
Chapman, P.J.:  Detoxification of creosote- and PCP-contaminated groundwater by physical
extraction: chemical and biological  assessment. Arch. Environ. Contam. Toxicol. 21:233-244
(1991).
[5]. Mueller, J.G., Chapman, P.J., Blattmann, B.O., Pritchard, P.H.: Isolation and characterization
of a fluoranthene-utilizing strain of Pseudomonas paucimobilis,  Appl. Environ. Microbiol.
56:1079-1086 (1990).
[6]. Mueller, J.G., Chapman, P.J., Pritchard, P.H.:  Action of a fluoranthene-utilizing bacterial
community on polycyclic aromatic hydrocarbon components of creosote. Appl.
Environ. Microbiol. 55:3085-3090 (1989).
[7]. Mueller, J.G., Lantz, S.E., Middaugh, D.P., Colvin, R.J.,  Ross, D., Prftchard,  P.H.: Strategy
using bioreactors and specially-selected microorganisms for bioremediation of ground water
contaminated with creosote and pentachlorophenol.  Environ. Sci. Technol.  (submitted).
[8]. Mueller, J.G., Middaugh, D.P., Lantz, S.E., Chapman, P.J.: Biodegradation of creosote  and
pentachlorophenol in groundwater: chemical and biological assessment. Appl. Environ. Microbiol.
57:1277-1285 (1991).
                                            39

-------
   THE  DESIGN CRITERIA AND ECONOMICS OP OPERATING
 A PULL-SCALE ABOVE-GROUND BIOREMEDIATION FACILITY
FOR THE TREATMENT OF HYDROCARBON CONTAMINATED SOILS
                         by

                    Robert Mall
             Environmental Specialist

               KJC Operating Company
                         40

-------
SUMMARY

The bioremediation of soil contaminated with certain hydrocarbons,
has been demonstrated as an  effective and economical alternative
waste treatment technology.   This paper focuses on a project which
was  specifically   designed   and  implemented   to   treat  soils
contaminated with a synthetic heat transfer fluid.

INTRODUCTION

Due to increasing hazardous waste disposal costs, rapid regulatory
changes, and a corporate policy of sound environmental practices,
the operators  of the world's  largest solar  electric generating
systems (SEGS)  embarked  on a program to eliminate the need  for off-
site disposal of hydrocarbon contaminated soils.

The SEGS  technology  employ  line-focus  parabolic trough mirror
collectors to focus and concentrate sunlight onto heat collection
elements (HCE).  The HCE's are filled with a unique heat transfer
fluid (HTF)  which is circulated through the solar fields and heated
to between  600 and  750 degrees Fahrenheit.   The HTF  is pumped
through a series of  heat exchangers, generating steam which in turn
passes  through  a  turbine generating electricity.    HTF is  an
eutectic mixture of 26.5% Biphenyl and 73.5% Diphenyl Oxide.

Accidental releases of HTF from the SEGS systems sometimes result
in contamination  of soil.   Historically,  this  meant excavation,
transport, and disposal  in a  class  I hazardous waste landfill.  In
an effort to reduce  the cost of disposal,  conserve limited landfill
space, and implement a soils  recycling program, bioremediation was
studied as an on-site alternative treatment technology.

TREATMENT PROCESS DESCRIPTION

The treatment technology employs solid phase bioremediation which
is environmentally controlled to optimize degradation processes.
By  utilizing   naturally  occurring  microorganisms,   HTF   soil
contaminants are metabolized  into harmless by-products.  Treatment
is conducted within a fixed containment facility.

MEDIA AND POLLUTANTS TREATED

The waste stream consists of well-defined coarse and medium grained
sands and a small percent of clayey silts and gravel contaminated
with HTF (biphenyl and diphenyl oxide).

PROJECT DEVELOPMENT STATUS

Two pilot scale projects, conducted over an 18-month period,  were
successful  in  reducing  HTF  concentrations to  below  the target
"clean-up"  level of  1000  ppm.    The positive  results  achieved
prompted the implementation of a full-scale project.  A full-scale
treatment demonstration project has been underway for the last 24
months.
                                41

-------
 INITIAL AND FINAL POLLUTANT CONCENTRATION

 Initial contaminant concentrations vary with the quantity  of  HTF
 released  and  surface  area  of  contamination.    Concentrations
 typically average  about  7,500  parts per  million  (ppm); with  an
 observed range from less  than 2,000  ppm to as high  as 20,000 ppm.
 Treatment requires  approximately 90 days to accomplish contaminant
 removal to below 1000 ppm.   Treatment to levels less  than 100  ppm
 are possible, however this lower treatment  level is contingent upon
 initial concentrations, favorable  environmental conditions  and is
 generally time dependent.

 PROCESS LIMITATIONS

 Treatability studies have shown that the most critical  limiting
 factors are:
 1.   Soil pH
 2.   Moisture
 3.   Temperature
 4.   Nutrients
 5.   Oxygen supply
(6 to 8)
(40 to 60 % of saturation)
(65 to 85  degrees F)
(Variable)
(Variable)
The  ability  of the  operator  to manage  these  conditions  will
determine whether an ideal environment can be created to stimulate
bacterial population growth and enhance the  degradation process.

To enhance aerobic bacterial growth and reduce the potential odor
problems associated with anaerobic conditions, the soils are tilled
on a^ 72-hours  schedule.   A good oxygen supply,  in concert with a
sufficient  moisture  contentr  stimulates bacterial   growth  and
promotes mobility,

PROCESS WASTE STREAMS

If properly  managed,  the process produces no residual hazardous
waste or other conditions of concern.  Treatment  is viewed  as a
form of "geo-reclamation"; the recycled material is to be used as
"clean" backfill on site.

AGENCY INVOLVEMENT

As with any project of this nature, it is necessary to consult with
a variety of regulatory agencies.  It is important to gain a full
understanding of the regulatory conditions before spending too much
time and money  on development.   In  order  to  avoid duplication of
agency requirements,  a workshop  should  be  set  up to bring  all
concerned parties together to discuss  the project in full.  In the
case of this  facility,  a meeting was held to  discuss and establish
the treatment process and goals.   The proposed facility design and
construction parameters were compared  to  regulatory  statutes  for
adequacy and compliance.
                                42

-------
 Sampling  protocol  and  monitoring  requirements,   as   well   as
 regulatory concerns  and comments,  were  addressed.   The  "lead"
 agency was established and  the necessary  approvals obtained.

 The California State agencies  involved  in the project are:

 CAL\EPA - Region 4 (Facility Permitting Unit)      (CAL\EPA)
 CAL\EPA - Alternative Technology Division         (ATD)
 Regional Water Quality Control Board              (RWQCB)
 Air Pollution Control District                    (APCD)
 County Environmental Health Services              (EHS)
 California Energy Commission                      (CEC)*

      *  Conditions of Certification  issued by the  CEC required
         their involvement in this project.  The  CEC would not
         normally be involved in waste treatment  projects.

 DESIGN AND CONSTRUCTION CRITERIA

 There are several factors which must be considered when determining
 a facility's design.  While  economic considerations may dictate  the
 options selected, regulatory requirements will ultimately direct or
 limit those options.

 The primary consideration should be environmental protection. This
 usually means   containment features   to  ensure   prevention   of
 groundwater and  vadose zone  degradation,  but  may also include
 contaminant emissions  to the atmosphere.

 Regulatory  requirements will dictate minimum protection standards
 and tolerances.   The selection of materials used  for  containment
 should  also   consider  durability,   exposure   to   environmental
 elements,  load factors, and the  contaminant  being contained  and
 treated.

 High-density  polyethylene  (HOPE)  or  similar materials are widely
 used  for this type of  application,  and are  easily installed  and
 relatively  inexpensive.   Natural  clay  barriers  can also satisfy
 containment requirements, however, they can  be very expensive  to
 install  because  of  intensive labor requirements  and  import costs.

 To prevent run-on and/or run-off from precipitation, the facility's
 containment   area   properly   configured  and    the   perimeter
 appropriately graded.   In some cases, the entire facility can  be
 covered  or,  at  additional  cost, a roof  constructed over  the
 treatment  area.   This  may  be  necessary  in areas where average
 ambient temperatures are below  65 degrees Fahrenheit for prolonged
periods of time,  or regions  where high precipitation occurs.  Cost
 figures for roofing applications are  not considered in  this paper.

 Because water is applied to the facility,  leachate collection and
recovery  systems  (LCRS)  must   be installed  directly below  the
treatment zone.  The LCRS acts  as a first line of defense against
a possible release from the facility.
                                43

-------
The LCRS  allows  collected  fluids to be returned to the treatment
zone and warns of a possible facility breach.  Should fluids enter
the  LCRS,  inspection  and  repair of  the  facility  should occur
immediately following confirmation that the fluid is leachate.

Installation of vadose zone and groundwater monitoring devices are
often required to detect possible releases beyond the LCRS  and into
the unsaturated  zone or water table below the facility.

Once  appropriate containment  and monitoring  systems  have  been
determined,  facility  sizing  should  be   calculated.     Facility
dimensions should be sufficient to handle only the anticipated rate
and  volume  of  waste  material  generated.   A minimal  storage
structure  should  be appropriately  sized  and  considered as  a
contingency, should unusual events require  temporary  storage of
material prior to treatment.

It may  be of value  to  partition the treatment  zone into two or
three individual treatment areas  to allow a staggering of start and
end dates of smaller batch treatment processes.

FULL-SCALE FACILITY AS DESIGNED

This  full-scale  facility's   design  incorporates   primary  and
secondary  containment  features,  a  triple  LCRS  installation,
equipment access requirements  and environmental control  devices.
The facility is  225' long,  60'  wide  and 2.51  deep with an annual
throughput capacity of approximately 1,500 cubic yards or roughly
2,000 tons.

The facility was constructed with 7 inches of reinforced  concrete
as the primary  containment  structure.  The concrete was coated with
a chemically resistant sealant.  An elaborate triple-liner system
of HOPE was installed below the concrete.  This required secondary
containment is made  up of  two  40 mil  and  one 80 mil HOPE liners.
There is  a "geo-fabric" material between  the two  40  mil liners
which acts as a wick to  assist  in the  recovery of fluids which may
collect between  these  two liners.    A LCRS  exists between  the
concrete and the upper 40 mil liner;  between the  two  40 mil liners;
and between the  lower 40 and bottom 80 mil liners.

The area  between the  40  and  80 mil  liners was backfilled  and
compacted with 2' of native soil, and designated the lower vadose
(unsaturated) zone; the third LCRS acts as the primary vadose zone
monitoring system,  therefore no  additional subsurface  monitoring
devices   (i.e.  lysimeters,  neutron probe  casings, etc.)   were
installed below the 80 mil liner.

A broadcast watering system was  installed  and  set up on a timing
device to maintain  proper soil moisture content.   Soil moisture
readings are taken every three days.
                                 44

-------
 MONITORING REQUIREMENTS

 Because  of the relatively  small  size  of  this  facility,  it was not
 practical  to install  groundwater monitoring wells  for purposes of
 detecting  a leak from the facility.  The depth to groundwater is
 greater  than  150 feet,  and contaminant  pathways are not  well
 defined.    As  a result,  it would  be  virtually  impossible  to
 construct  a well monitoring system that could adequately detect any
 degradation of the  groundwater  due to a leak from this facility.
 Hence the  triple liner and LCRS system.   The  LCRS are monitored on
 a quarterly basis.

 ECONOMICS

 An economic  comparison of various alternatives to the operation of
 a  bioremediation  facility  show that  it  is  cost-effective  to
 construct  and  operate a bio-facility.  This assumes, however,  that
 either a substantial volume  of waste  exists, or is expected to be
 generated  over the operating life of  the project.

 INITIAL CAPITAL COST

 Pilot-scale  testing conducted at this site cost about  $15,000.00
 over  18  months.  The  full-scale facility has  thus  far  required
 agency fees totalling  $85,000.00, including financial assurance for
 eventual facility closure.

 Full-scale facility construction requirements  and associated costs
 made up the bulk of first year capital expenses.  Facility design,
 plan  check fees, material  and  labor, as  well  as  administrative
 oversight, has resulted  in a capital  expenditure of approximatelv
 $220,000.00.                                                    *

 Total project  costs amounted to about $320,000.00.    Actual costs
 may be higher if the purchase of earth moving and tilling equipment
 is required.

 TREATMENT  COSTS

 Annual treatment costs  for bioremediation are  relatively low by
 comparison  to other methods of treatment.   Costs are directly
 related  to manpower and equipment needs, as well as  materials
 consumed (i.e.  nutrients,  water, etc.).   Sampling  and analytical
 costs will vary depending  upon the  frequency  of sampling and the
 type and quantity of tests performed on each sample.

 Overall operating costs  will also be  contingent upon  the size of
 the facility and number of treatment batches  completed.   For this
 facility (which processes three 500 cubic yard batches per year),
the total annual cost of operation runs  approximately $35,000.00.
This figure includes all of those factors indicated above.
                                 45

-------
ECONOMICS OF OPERATIONS

In deciding whether on-site bioremediation is practical, the focus
is primarily on  economic  payback.   By comparison  to  some of the
alternatives, such  as  landfilling,  incineration,  solidification,
thermal desorption,  and vapor extraction, bioremediation can be
very attractive.

The key questions as to whether  or  not on-site bioremediation is
attractive or not are:

How much waste do I currently have to treat?
How much am I generating over what period of time?
Are there regulatory limitations in dealing with this waste stream?
Do I have land availability to conduct on-site treatment?
Am I under some time constraint to solve the problem?
Do I have the means in-house to attend to on-site treatment?
Is the waste stream hazardous or not?  And finally,
What are my long term goals?

If  the answers  to  the above  questions are   favorable  towards
implementing an  on-site bioremediation  facility, the operating
costs are minimal and payback can be relatively  rapid.   The factors
that play a role  in facility operating costs are: frequency of soil
aeration  (labor),   purchase   of  supplemental  nutrients,  water,
sampling and analytical expenses, and administrative oversight.

For this particular facility,  and a total first-year investment of
$220,000.00  (construction), and  $35,000.00  (operations),  on-site
bioremediation produced a  per treated  cubic  yard cost  of about
$170.00, or $30.00  per  yard less  than the cost for landfilling the
waste.    In the second year,  given the same volume  treated,  the
cost  drops to  about  $26.00  per  yard without amortization  of
facility  construction.   An additional  cost reduction,  to about
$21.00 per cubic yard,  was realized  once  sampling and analytical
parameters were reduced to minimum requirements.  Experience gained
over time will  allow the operator to reduce sampling to  the minimal
requirements.

A comparison of the first year costs of selected alternatives, are
generally  competitive.   In  the  second and  subsequent  years  of
operation,  the  cost  of   conducting  on-site  bioremediation  is
dramatically less on a per volume basis.  For example:  Treating
1,500  cubic yards  of  contaminated  soil would   result  in  the
following first-year expenditure, using the treatment  or disposal
method shown in the following  chart.  In the second year, there is
a significant difference in the cost of bioremediation  compared to
the other methods.
                                 46

-------
METHOD

Incineration
Solidification
Landfill
Thermal Desorption
Bioremediation
 1st year

$787,500.00
$165,000.00
$322,500.00
$300,000.00
$255,000.00
2nd year

NO CHANGE
    it
    ii
    n
$39,000.00
3rd year

NO CHANGE
    n
    n
$31,500.00
These  comparison figures are  averages of  actual  quotations and
information  obtained from  articles  printed  in  various  industry
publications for the method indicated.  The  bioremediation figures
are actual costs based on this project.

It  should be  remembered that  the economic  variables affecting
individual on-site  bioremediation  projects  will  be influenced by
regulatory requirements,  the cost of  construction materials and
facility scale, the type and concentration of the  contaminant being
treated,  the climatic conditions of the area, and the amount of
waste  to  be  treated annually.    Therefore,  each  of these factors
must be considered individually and specifically  for each project.

CONCLUSION

This project and the effectiveness of bioremediation has proven
itself, both  environmentally  and  economically.    In a world of
tighter  and tighter  regulatory  requirements,  rising   cost  of
hazardous waste  treatment and  disposal,  and  public  awareness of
environmental   problems,   there   is   no   doubt  that   on-site
bioremediation is a sound,  cost-effective solution  to  a variety of
hydrocarbon contaminated soil problems.
                                47

-------
                                          Biopur*
                  An Innovative Bioreactor for the Simultaneous Treatment
           of  Groundwater and foil Vapour Contaminated with Xenobiotic Compounds

         TAUU Infra Consult  B.V.. P.O. Box 479,  7400 AL Oeventer, The Netherlands,
Introduction

At  many  places  the soil,  groundwater  and soil  vapour  are contaminated  with  organic
compounds. A bioreactor was developed to treat  groundwater and soil  vapour simultaneously.
This  system   is  called  Biopur*.   Contrary  to  conventional  physico/chemical  methods,
biological treatment does not cause waste problems and  is  therefore an attractive method.
The Biopur* system is patented.

Background

Biological  groundwater  treatment  in  fixed film  reactors, has  been  used  by  TAUW  Infra
Consult  B.V.   since  1986.  Trickling  filters  were used initially,  followed by  Rotating
Biological Contactors (RBC).  Groundwater heavily contaminated with  benzene,  mono-chloro-
benzene and HCH could be treated. Removal efficiencies of up to 98X were obtained.
Furthermore, at in situ  remediation sites,  there was a need to develop a system to treat
soil vapour in combination with groundwater. For  these  reasons the  Biopur* was developed.
The Biopur*  is  a fixed  film  bioreactor with  polyurethane as  a carrier  material  for  the
bioroass.  The  flow of air and  water  is co-current to prevent  the  volatile compounds being
stripped  (see  Figure 1).
    Cenpartaent filled with
    Polyurethane iponeei therein
    air and water *  '•'
    are being treated
                                                                        Treated grounduater
 Figure 1; An Overview of an  In Situ Application of the Biopur* Reactor  for  the
           Simultaneous  Treatment of Soil Vapour and Groundwater
                                          48

-------
 Results - Pilot Plant Scale

 The Biopur* was tested for groundwater treatment at pilot plant scale, the results of  one
 experiment are given in Table 1.

 Table 1;  Results of  Pilot  Plant Research
Site
Former Gasworks
Configuration 1
BTEX
PAH (16 EPA)
Mineral Oil
Configuration 2
BTEX
PAH (16 EPA)
Mineral Oil
HRT Influent
(hours) 
-------
 Table 2: Results of Full Scale Experiments
Site
Raalte I
BTEX
Mineral Oil
Overslag
BTEX
Mineral Oil
Utrecht
BTEX
Mineral Oil
Raalte II
BTEX
Mineral Oil
HRT Influent
(hours) «i9/O
0.25
650
< 100
1
39
350
0.25
1300
323
1.2
261
1050
Effluent
Otg/l)

< 0.5.
< 100

< 1
55

3
36

< 1
50
Removal
Efficiency X

> 99.9
not applicable

> 97.4
84.2

99.7
88.8

< 99.6
95.2
 Results •  Simultaneous Treatment of Groundwater and Soil Vapour

 The Biopur* was used at a site to simultaneously treat contaminated grounduater and soil
 vapour. A  mass  balance of the system is shown belou.
soil
vapour  •-"•
(50 m3/h)
210 g/h
                       Biopur*
                                                •>  < 3  g/h exhaust fumes
groundwater — 10 g/h
(15 ra3/h)
                                               -»• < 2 g/h effluent
The  mass  balance shows  that  5 kg of  petrol was  removed daily with  hydraulic  retention
times of less than 15 minutes.

The  exhaust  fumes did not  contain detectable organic compounds  (<0.1 ppm) and could  be
disposed of  without  being treated.  The treated  groundwater from the  site  was  discharged
into surface water.
Treatment Costs

Related to  the  entire remedial  action the treatment costs of soil vapour  and groundwater
were less than DM 0.40 per m  of groundwater.  For calculation purposes only,  the treatment
of the soil vapour was free of charge.

In Table  3 a comparison  of the treatment  costs of four  types of groundwater  purifying
plants is given.
                                      50

-------
 Table 3:  Comparison  of  the Treatment Costs of Four Types of Groundwater Purifying Plants
 Type
             Costs (DM) per m  of grounduater
             10 m /hour  20 m /hour  40 m3/hour
1
2
3
4
Biopur*
Rotating biological contactor
Stripper/activated carbon
Stripper/compost filter
0.53
0.78
0.85
1.16
0.44
0.65
0.54
0.87
0.35
0.59
0.36
0.72
Conclusions

(1)    Groundwater contaminated with organic compounds, such as BTEX, PAH, mineral oil end
       chlorinated hydrocarbons can be treated by using the Biopur* process.
(2)    The Biopur* can be used for simultaneously treating biological groundwater and soil
       vapour.
(3)    Using the Biopur* process good  removal  results  were obtained with relatively short
       hydraulic retention times.
(4)    Application of  the Biopur*  process  in groundwater/soiI  vapour treatment  is  cost
       effective.
We are prepared  to  discuss  specific  problems  and to find a way of  cooperating  in complex
environmental fields.

Should you require further information please  contact:-
B.A. Butt (H.Sc.)
Head, Water Technology Sector
Telephone number +31-5700-99862
H.B.R.J. van Vree (H.Sc.)
Project Leader, Research & Development
Telephone number +31-5700-99561
                                      51

-------
to
 Optimizing Bioventing in Shallow Vadose

      Zones  and Cold Climates: Eielson

     AFB Bioremediation of a JP-4 Spill


                Gregory D. Sayles, Ph.D.1

                          and

Robert E. Hlnchee2, Richard C. Brenner1, Catherine M. Vogel3 and Ross N. Miller4

               'U.S. Environmental Protection Agency
              Risk Reduction Engineering Laboratory
                       Cincinnati, OH

                 2Battelle Memorial Laboratory
                       Columbus, OH

           3 U.S.  Air Force Civil Engineering Services Agency
                      Tyndall AFB, FL

          4 U.S. Air Force Center for Environmental Excellence
                      Brooks AFB, TX

-------
          Eielson AFB:  Background
           JP-4 jet fuel contaminated
           unsaturated soil
CA
         • Contamination from leaks in
           underground fuel distribution system
           Soil contaminated from 2 to 6 ft
           (groundwater at 6 ft)

-------
Eielson AFB:  Background (continued)
   U.S. EPA, U.S. Air Force, and Battelle
   began work here in July 1991


   3-year study


   Funded by

    - U.S. EPA Bioremediation Field Initiative

    - U.S. Air Force

-------
Eielson AFB:  Project Objectives

 Document bioventing in a cold climate

       -  average annual air temp = 0*C
 Evaluate various soil warming methods to
 enhance bioventing rates

       - active warming with warm water

       - passive solar warming

-------
    \
o\
        N
              Schematic of Eielson AFB Three Test Plots
                 Passive
                      Site Trailer
            Active
           Warming
            System
                                        Control
                                                  Groundwater monitoring well
                                                  Air injection/withdrawal well
                                                  Three-level soil gas probe
                                                  Three-level thermocouple probe
 25'
Scaie
50'

-------
              Cross-Section of EPA Active Plot Showing General
          Site Lithology and Bioventing System Construction Details
 Ground Surface

- Plastic Sheeting

   I— Styrofoam Insulation
                                            Groundwater
                                             Monitoring
                                            Well (MW-3)
  Three-Level
Thermocouples
                                                   Air Injection/Withdrawal Well
                                                          /         Three-Level
                                                        r         Soil Gas Probe
                                                                          B
                                                                           C(S6)
         — Plywood
             - Blue Vinyl
               Cover
                                               rial  B
                                               Valve A
                     Brown Sand
                     and Gravel
 Heat
 Tape
                   Backfill
                                              fi h > > > > > > >
                                              ;. A-Vxy^ V A-
   %%%%%%%*&          Sand and Gravel x;\
   8S\\\N\VN
  - i**>W.%W.%V
                     Gray Sandy Gravel
                                                                              Sand
                                                                              Bentonite
10-

-------
 Oxygen Concentration in Three Test Plots and Background
                   August 22, 1991 to July 11, 1992
25-
Initiation of Air Injection
  Initiation of-  r-Pressure Pulse
  Soil Heating
                                  Reduction of Water Circulation
                   on Soaker Hose
                  1	1	1	1	1	1	1	1	T
      Sept.   Oct.   Nov.    Dec. t, Jan.   Feb.   Mar.    Apr.    May   June   July
               1991                               1992
                Active Warming
                Passive Warming
                                  Contaminated Control
                                  Uncontaminated Background

-------
  Soil Temperature in  Three Test Plots and Background
VO
28

24

20

16
      fi
       12
      0)
      Q.
      £  8
        -4
                 Active Warming
                 Uncontaminated Background
                 Contaminated Control
                 Passive Warming
           Sept.  Oct.  Nov.  Dec.  Jan.   Feb.  Mar.  April  May  June  July  Aug.
                                Time (Days)

-------
          Biodegradation Rates at Eielson AFB
8
     10
   I 8

   ^

   O)
     6-
O)





5
DC



I  4


I

2
O)
0)
     0
              • Active Warming


              • Passive Warming


              A Contaminated Control
         October
                    January
April
August

-------
            Eielson AFB:  Conclusions
o\
Active soil warming is successful


Optimization of active heating
and aeration necessary


Passive solar heating looks
very promising


Biodegradation via bioventing is
occurring year-round in all plots

-------
            BIO-REM,  INC.'S AUGMENTED IN SITU SUBSURFACE
         BIOREMEDIATION PROCESS (TM)  USING BIO-REM, INC.'S
         PROPRIETARY BIOCULTURES (H-10)  FOR  THE REMEDIATION
           OF JP-4 CONTAMINATED SOILS AT WILLIAMS AFB f AZ

 In 1987, ten 25,000 gallon JP-4 aviation jet fuel tanks were
 removed from the Primary Fuels Storage Area at  Williams AFB, AZ.
 Extensive contamination was discovered; laboratory analysis
 revealed contaminant levels in excess of 88,000 ppm of JP-4, and
 the site was closed and subsequently placed on  the National
 Priority List (Superfund List) for future clean-up.   In August of
 1991, after the base had been listed for closure  by the U.S.
 Government, an award .was made for the initial clean-up of the
 contaminated site, designated as Operable Unit-2  of this
 Superfund site.

      In September of 1991, the clean-up contract  was  awarded to a
 construction firm from Stockton, California to  excavate and pile-
 remediate 16,000 cubic yards of JP-4 contaminated soil.  However,
 due to State of Arizona and Maricopa County,  Arizona  Air Quality
 standards and regulations concerning volatile organic compound
 (VOCs) emissions, this methodology was rejected.   The Bio-Remfs
 Augmented In Situ Subsurface Bioremediation Process  (TM) and Bio-
 Rem's H-10 microaerophilic bacterial cultures were selected to
 remediate the site.

 REMEDIATION METHOD:   The  initial stages of  the bioremediation of
 the Fuels Storage Area at Williams AFB,  AZ  began  in April  of
 1992.   Over 320 borings and  monitor wells were placed on the
 site.   Bio-Rem H-10 Product  was propagated  and subsequently
 placed into the contaminant  plume,  through  these borings and
 wells, over a period of ten  (10)  days.   Then  the site was  covered
 with plastic sheeting to  reduce evaporation.

 I.  Technical Information.

     A.   Description of Technology.

     Bio-Rem,  Inc.'s Process and BiocuJture are commercially
 available and have been used in over 75  sites in 14 states.   The
Augmented In Situ Subsurface Bioremediation Process (TM)
 incorporates four (4)  steps  in  the subsurface bioremediation  of
hydrocarbon  contamination  in soil and water.  This process  occurs
 totally underground.   It  utilizes Bio-Rem, Inc.'s Product  "H-10",
 a proprietary blend  of microaerophilic bacteria and
micronutrients.   After placement, Bio-Rem H-10 does not require
additional injections  of nutrients,  oxygen, or oxygen-producing
 compounds (i.e. hydrogen peroxide).   Furthermore,  the by-products
of  the degradation process are  naturally occurring elements;
therefore, this process complies fully with Title I of the  Clean
Air Act.
                                62

-------
10,000
 5,000
 1,000
  500
  400
  300
  200
  100
    0
BIO-REM, INC.
WILLIAMS AFB, ARIZONA
                                  10/8  10/22

-------
BIOREMEDIATION OF SOILS AT A WASTE OIL FACILITY
                         Gary Guerra
               On-Scene Coordinator, USEPA Region VI
                          Dallas, TX

                        John Matthews
                         Bert Bledsoe
                     Environmental Scientists
           Robert S. Kerr Environmental Research Laboratory
                           Ada, OK

                        Daniel F. Pope
                      Senior Staff Scientist
                      Dynamac Corporation
                           Ada, OK
   ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
           OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                       ADA, OK 74820
                            64

-------
                         Bioremediation of Soils at a Waste Oil Facility
        The Baldwin Waste Oil site is located approximately 1.5 miles north of Robstown, a small town near Corpus
Christi, Texas. Site weather is typical of Gulf Coast conditions: the summers are hot and humid, and winters are mild
with frost conditions averaging about seven days each year.  The site, an abandoned waste oil facility occupying
slightly more than one acre, is surrounded by agricultural lands.  The north half of the facility is occupied by two
tank farms surrounded by  earthen containment dikes and a caliche drive.   The south half is  a  low area with
vegetation and debris, bordered on three sides by a drainage ditch located just inside the property line. A small pond
filled with sludge and water is located beside one group of tanks.  Oil, water, and sludges were located in fifteen
large storage tanks, five 55 gallon drums, numerous small containers and two containment berms surrounding the
two groups of storage tanks. Areas between the tanks are covered with a sludge layer.

        During operation,  the facility received a large portion of its waste  oil from the local navigation district
ballast pits, used to store waste oil collected from slop tanks of ocean-going oil tankers. Waste oil was transported
to the site and pumped into one of several mixing tanks where the oil was blended with solvents (benzene, toluene
or xylene) for resale.

        The native soil at the site is the Victoria clay. Characteristics of this clay material include:  low permeability,
high water-holding capacity, high compressibility, high to very high  shrink-swell potential, poor drainage, low to
depressed relief, low shear strength, and high plasticity. A particle size analysis of the Victoria clay soil indicated
a sand content of 43%, a silt content of 10%, and a clay content of 46%. A caliche fill material, classified as a silty
clay,  is also present  Particle size analysis of  the caliche material indicated a sand content of 53%, a silt content
of 10%, and a clay content of 36%.  Analysis of soil samples indicated levels of benzene, ethylbenzene, toluene,
xylene (BTEX)  of 50 ppm or less; elevated levels of lead,  chromium, barium, arsenic, and total petroleum
hydrocarbons.

        The remedial plan for the site provided for removal of tanks, barrels, buried piping, debris and  sludges.
However, approximately 1500 to 2200 cubic yards of contaminated soils  that had been under an oily sludge layer
on the tank farm area remained to be disposed. This soil contained  levels of total petroleum hydrocarbons up to
50,000 ppm.  Treatment processes considered for the soil included incineration or in-situ bioremediation based on
the land treatment concept. Bioremediation technology was chosen for pilot testing to determine the feasibility of
the innovative use of this technology  in-situ during the removal action. The treatment goals to be achieved during
the removal program are concentrations of less than 1% oil and  grease (O&G) and 10,000 mg/kg total petroleum
hydrocarbons (TPH).

        Bioremediation based on the land treatment concept involves use of natural biological, chemical and physical
processes in the soil to transform organic contaminants. Biological  activity apparently accounts for most of the
transformations of organic contaminants in soil.  Bioremediation is carried out by utilizing techniques for enhancing
the development of the microorganisms and bringing them in contact with the contaminants.  Conventional land
treatment has been shown in a number of instances to be a viable treatment alternative for soils contaminated with
oils and hydrocarbons.  However, there was some concern that compounds or elements might be present  at the
Baldwin site which could inhibit microbial activity in a land treatment unit. Laboratory tests conducted by Dr. Alan
Marker of the  University of Oklahoma indicated the presence of a  population of active petroleum hydrocarbon
degrading microorganisms, so it was  decided to conduct a pilot scale land treatment study in the field.

        The pilot scale land treatment system was designed to evaluate the feasibility of utilizing in-situ remediation
of the contaminated soils.  Objectives of the pilot scale demonstration were to:

        1)      Demonstrate the efficacy of in-situ bioremediation for the oil contaminated soils at Baldwin Waste
                Oil site under field  conditions;

        2)      Collect statistically defensible data on which to base conclusions of bioremediation efficacy;

        3)      Identify potential problems in implementing a full-scale bioremediation project;
                                                  65

-------
          4)      Collect data which will be necessary for the full-scale implementation of in-situ bioremediation at
                 the site.

          The pilot system was constructed outside the bermed area so the system could be left in place during
 removal operations associated with the site tank waste disposal and demolition.  A treatment plot approximately 33
 feet long and 12 feet wide was constructed with the bottom sloped (about 2-4% slope) toward the middle of the area
 and about a 1-2% slope toward one end. A two foot berm was provided to prevent any run-on or run-off from rain
 or irrigation. A high density polyethylene liner was anchored within the bermed area and covered with five inches
 of pea gravel and eight inches of sand to allow for drainage and help protect the liner during tilling of the unit.  A
 slotted PVC pipe was installed down the medial axis of the plot for leachate collection.

         The two contaminated soil types (the native clay and caliche fill material)  were  kept separate in the
 treatment plot in order to determine the effect of soil type on treatment management and efficacy.  For sampling,
 the plot was divided into six cells (6 ft by  10 ft.), with three cells representing each of the two soil types. The north
 side of the treatment unit (cells NW, N, ME) contained the caliche material, and the south side of the treatment unit
 (cells SW, S, SE) contained the native clay material. An eight inch thick layer of each contaminated soil was placed
 on the plot, and two inches of composted chicken manure was tilled into the soil for conditioning and as a source
 of nutrients. The manure contained about 11,000 ppm EDTA extractable phosphate phosphorus, 30,000 ppm EDTA
 extractable potassium, 2.4 %  total kjeldahl nitrogen (TKN), and 14% total organic carbon.  Nutrient content of the
 amended soil is  shown in Table 1. Nitrogen and phosphorus were monitored throughout the study, and fertilizer was
 added as needed to maintain  the appropriate levels of these nutrients.  A movable overhead irrigation system  was
 used to add water as needed.  Leachate from the pilot unit was collected for analysis.

         Samples for the starting concentration (Week 0) were collected after the compost had been tilled into the
 soil. Samples taken at later sampling times were taken immediately after tilling to aid in sample homogenization.
 Five evenly spaced subsamples were taken within each cell. These subsamples were mixed to get one composite
 sample for each cell.  Volatile organic compounds were measured on grab samples to minimize loss of the volatile
 components. The grab samples were taken at three locations in each cell using  a soil coring sampler with acetate
 sleeves.

         Week 0 soil samples were analyzed for nutrients, pH, oil and grease (O&G), total petroleum hydrocarbons
 (TPH), volatile organics (VOC), semi-volatile organics (S VO), polynuclear aromatic hydrocarbons (PAHs) and total
 organic carbon (TOC) (See Tables 1 - 3 and Figure 1). O&G, TPH, pH, and nutrients also were measured at Weeks
 3, 6, 10, 21, 28,  36.40 and 43 as operational parameters to determine the course of treatment. Once the operational
 parameters reached a steady state at Week 43, the full suite  of analyses was performed to determine the treatment
 cndpoints.

        The project was started in November of 1991 since November is normally the beginning of the annual  dry
 season for the Robstown area.  However, the rainfall for the  next ten weeks was far above normal  (Figure 3). The
 project was proposed to end by the 12th  week, but the unusually high and continued rainfall precluded tilling and
 other operations at the site during this period, so the project was extended. Significant degradation of O&G and TPH
 occurred during  this period even though the LTU was saturated at times.  Analyses of leachate from the pilot unit
 indicated TPH and O&G levels were below detection limits. Treatment levels at the 43rd  week were considered
 acceptable since TPH and O&G concentrations  approached or were  below  the removal program cleanup goals.
 Volatile organics and semivolatile organics,  present at low parts per billion levels at the beginning of the study, were
 at nondetect levels by the end of the  study. Based on reduction of TPH and O&G to cleanup goals and reduction
 of volatile and semivolatile organics to nondetect levels,  in-situ bioremediation based on the land treatment concept
has been selected as the remediation  alternative for full  scale clean up under  the removal program.

        The pilot scale demonstration study was conducted through the cooperative effort  of USEPA Region VI
Superfund Program and their Technical  Assistance Team contractor (Ecology  and  Environment, Inc.), and  the
USEPA Robert S. Kerr Environmental Research Laboratory Subsurface Fate and Transport Technology Support
Center and their  technical support contractors (Dynamac Corporation).
                                                  66

-------
Tabk 1. Status Of LTU Soil At Week 0
Electrical Phosphate
Sample Location .. Conductivity Total Total Kjeldahl Phosphorus
(Cell) mmhos/cm Organic Carbon % Nitrogen % ppm
SW 4.5
S 5.4
SE 5.2
NW 5.4
N 5.1
NE 5.2
1.1
2.4
1.1
1.0
1.3
1.2

0.06
0.20
0.05
0.16
0.13
0.13
42.8
76.1
68.5
42.4
38.0
40.3
Potassium
ppm
1250
1176
1283
1144
1048
1175

Table 2. Total Petroleum Hydrocarbons In LTU Soil
Sample
Location Type
. to
NW
NW D
NW R
NE
N Mean
S
SW
SW D
SW R
SE
SMean
Week 0 Week 3
31900
38100


48700
39567
14400
14300


14500
14400
Table 3. Oil And Grease In
Sample
Location Type
N
NW
NW D
NW R
NE
N Mean
S
SW
SW D
SW R
SE
5 Mean
22300
21200


25400
22967
7800
25200


19800
17600
LTU Soil
WeekO Week 3
5.8
8.6


8.4
7.6
2.1
3.8


2.4
2.7
2.5
2.6


4.1
3.0
0.9
2.9


2.3
2.0
Total Petroleum Hydrocarbons (mg/kg)
Week 6 Week 10 Week 21 Week 28 Week 36
16100
17900


19500
17833
5800
7300


9100
7400

Week 6
3.5
25


2.7
2.9
1.6
1.4


1.5
7.5
16000
13400


15600
15000
4200
4700


95OO
6133

Oil and
Week 10
5.2
2.4


3.0
5.5
0.9
2.6


1.9
1.8
40590
12770


25750
26370
6876
4292


11390
7519

Grease (%
Week 21
5.1
2.0


3.5
3.5
1.5
1.0


1.9
1.4
17570
16345


16090
16668
6675
9728


7075
7826

Dry Weight)
Week 28
2.1
1.9


2.0
2.0
0.8
1.2


1.3
1.1
10590
9479


13120
11063
4248
4845


6450
5181

Week 36
1.9
1.9


2.2
2.0
1.4
1.2


1.5
1.3
Week 40
8929
7415
6759
7007
10240
8070
5024
4586
5761
4713
9454
5908

Week 40
1.2
1.0
0.8
0.9
1.4
1.0
0.5
0.6
0.7
0.5
1.3
0.7
Week 43
6854
8015
6917
8426
9264
7895
3771
2233
2707
4375
5361
3689

Week 43
1.2
1.3
1.2
1.6
1.6
1.4
0.7
0.6
0.7
0.9
1.0
0.8
                                                  67

-------
                                         TOTAL PETROLEUM HYDROCARBONS IN LTU SGML
                                                       NORTH CELLS
       soooo
       44000
       4OOOO <
     UJ MOOO'
     I30000
f       2 5000
       20000
     E 18000
       10000
        (000
          0
                                                          W**k21
                                                                                                      WMk43
                                        TOTAL PETROLEUM HYDROCARBONS IN LTU SOIL
                                                       SOUTH CELLS
       80000 *
       49000
       40000<
     u ttOOOi
     i 30000'
     « 25000
     ^20000
     fe 1SOOO
       10000
        5000
          0
               WM*O
                                                          WMk21
                                                        Jomptng CXrt*
                                                                     WMk2«
                                                                                WMkW
                                                                                           W««k40
WMk43
Figure 1. TPH in LTU Soil.
                                                          68

-------
      10 • <
       • ' >
       »< >
    S  7..
                                             OH. 4 GREASE WLTU SOU-
                                                 NCR TH CELLS
                                                  Sonping Dot*
                                             OB. & GREASE IN LTU SOU.
                                                  SOUTH CELLS
    •i. 7< '
    E ••
    5 s
    a «•
    =!
    o
                      WMk3
                                          WMK 10     WMK n
                                                   Samplng Da)*
Figure 2.  % O&G in LTU Soil.
                                             MONTHLY PRECIPITATION
Figure 3. Rainfall During the Pilot Study.
                                                     69

-------
                       WeekO
                       Week 6
                       Week 43
Figure 4. Gas Chromatography Chromatogram of a Hexane Extract of Treated Soil.
                                              70

-------
             Low Temperature Thermal Treatment LT3 Site Demonstration
                            By Michael G. Cosmos, P.E.
                                Roy F. Weston, Inc.


 The LT3. was used as part of a full-scale demonstration and to perform the clean up of
 wastewater lagoon at the Anderson Development Company Superfund Site located in
 Adrian, Michigan. The contaminate of concern in the lagoon was 4,4' methylene bis 2-
 cMoroaniline commonly known as MBOCA.  Roy F. Weston, Inc. (WESTON.) of West
 Chester, PA, was contracted by Anderson Development  Company to perform  the
 remediation of the lagoon and underlying soil. WESTON employed its patented (Patent No.
 4,738,206) Low Temperature Thermal Treatment System (LT3) to treat the material.  As
 part of the consent order Anderson Development Company and WESTON prepared a
 Remedial Action Plan, Health and Safety Plan and Quality Assurance Project Plan.

 Following approval of these plans by United States Environmental Protection Agency (US
 EPA) Region V and Michigan Department of Natural Resources, a demonstration test was
 conducted as  part of the US EPA Superfund Innovative Technology (SITE) program. A
 rigorous testing program was conducted on eight process streams. The process streams were
 sampled and analyzed for volatile organics, semivolatile organics, metals, dioxin and furans.
 In addition, the stack emissions were sampled for particulate and hydrochloric acid to
 determine if regulatory criteria were met.

 Contaminate  of Concern

 MBOCA is a chemical  intermediate used in the production of high-tech plastics and
urethane foams.  MBOCA which is insoluble in water is a semivolatile with an extremely
low vapor pressure. MBOCA was present  as a result of previous disposal operations. The
lagoon covered approximately 2 acres.   The MBOCA was present  in the sludge and
underlying naturally occurring clay in concentrations up to 1,600 ppm.  A cleanup criteria
                                      71

-------
of 1.6 ppm of MBOCA was established as the result of site specific risk analysis.

In addition to the contaminant of concern, the lagoon sludge also contained trace amounts
of other organic  compounds including  dichlorobenzene and  toluene.   The sludge also
contained a high concentration of manganese potassium permanganate which was added
during previous operations to oxidize the organics present in the lagoon.

LT3 System

The LT3 process is based upon indirect thermal heat exchanger. The screw type processor
is used to dry  and heat the soil causing volatilization of the contaminates and moisture. The
process is capable of heating of the soils up to 580°F as was demonstrated during the
remedial activities at Anderson Development Company. By maintaining a constant exhaust
from the processor the evaporated water and organics are removed. The vented gases are
subsequently  treated in a series of condensers.  The condensers remove the bulk of the
evaporated moisture. The vapor stream is then treated  by carbon adsorption packs and
exhausted to  the atmosphere. The liquid phase generated in the condensers consists of
water and chemicals from the soil.  The condensate is treated and either recycled as quench
water for the  processed soil or shipped to an off-site treatment facility.

The LT3 is divided into three main processing areas: solids processing, emissions control and
wastewater treatment.  A schematic diagram of the  system is shown in Figure 1.  The
general arrangement of equipment is shown in Figure 2.  The LT3 equipment used on the
Anderson Development Company  was mounted on three tractor trailers. The equipment
remained mounted on the trailers throughout operations.

Solid Handling System

The excavated soil and dewatered sludge from the lagoon was transported by track loader
to a double deck vibrating screen to remove debris greater than 2 inches.   The screened
                                        72

-------
material was then stockpiled.  Awheel loader removed the screened soil from the stockpile
and fed the surge hopper.  The wheel loader passed  over a registered weigh scale to
determine total quantity of material treated.  The surge hopper is maintained full to form
a seal at the inlet to the thermal processor

The thermal processor is  constructed of two jacketed screw conveyors.  The jacket and
screws are constructed with double walls that allow the circulation of a high temperature
heat transfer fluid. Each screw conveyor consists of four parallel intermeshed screws.  The
two units are constructed  in piggyback fashion.  Intermeshing screws allow for  improved
heater  transfer  and breaking up of  soil.  The high  clay  content soil at Anderson
Development Company did not cause balls to be formed. The screw conveyors are driven
by  a variable  speed drive which allows adjustment  in  soil discharge temperature  and
residence time.   During  operations at Anderson Development  Company the soil  was
maintained at temperatures above 540 °F and residence times of 90 minutes.

The treated  soil is discharged into a mixing screw conveyor. The mixing screw conveyor is
a ribbon flight conveyor where water is injected to quench and wet the soil. The material
is then stacked with a belt conveyor  into daily storage piles.

Emissions Control

The volatilized organics and  water  from the LT3 are removed and passed through the
pollution control system.  The exit gas which is  typically maintained at 300°F. The gas is
drawn through  a pulse-jet type  baghouse dust collector.   The  dust collector removes
particulate matter. On a daily basis two to four drums of "fly ash" or dust are generated.
The gases are then drawn into an air  cooled condenser. The air cooled condenser is a tube
type heat exchanger that  uses force ambient  air to indirectly cool the exhaust gases.
Temperatures exiting the air cooled  condenser ranged from 125°F in summer to 40°F in
winter.  The cooled exhaust gases then are drafted into a refrigerated condenser. In the
refrigerated  condenser the exhaust gases are indirectly cooled to 60°F by a recirculating
                                         73

-------
glycol and water solution.  The glycol and water solution is maintain at 30 °F by a
refrigeration system. The exhaust gases are subsequently heated to 70°F by three inline
electric resistance heaters.  By heating the exhaust gases to 70 °F the relative humidity of
gas is reduced to 70% preventing condensation in process piping and enhancing carbon
adsorption efficiency.  The final emission control device is the carbon adsorption column.
Two standard carbon rental packs are used to treat the emissions prior to release.  One
system  is on line  at all tunes.   The second carbon pack is activated  automatically if
breakthrough occurs.

Monitoring of the pollution control system effectiveness is provided by a Continuous
Emission Monitoring system (CEM). The CEM is used to monitor the exhaust stack for
concentrations of oxygen, carbon dioxide, carbon monoxide and total hydrocarbons.  The
system also monitors the exhaust gases from the baghouse dust collector for oxygen and total
hydrocarbons.

Wastewater Treatment

The two condensers generate  a liquid stream.  The liquid consists primarily of water,
however, it also contains heavier organics.  On the Anderson Development Project a simple
overflow decanter was used to separate out heavy condensate that contained high levels of
the insoluble MBOCA.  The overflow was then filtered to remove entrained heavy.  The
water was then treated by two carbon packs in series and stored in a 6,000 gallon tank. The
water at the Anderson Development project was then shipped off-site to an industrial waste
waster treatment plant for disposal.

Utilities

The operation of the LT3 requires the utilities which included 480 Volt electrical power,
natural gas and potable water.  At previous operations the LT3 has used propane as a fuel.
                                         74

-------
 SITE Demonstration

 The SITE demonstration was conducted in December 1991 on dewatered sludge from the
 lagoon. The Application Analysis Report and results of the testing are currently undergoing
 internal EPA review and are expected to be released shortly. The emission data from this
 test have shown considerable control of organics and other pollutants.

 The paniculate emission was on average less than 1.1 x KF* grains per dry standard cubic
 feet.   Similarly  control of acid emissions were less  than 6.0 X 10"5 pounds  per hour.
 Exceptional control of acid gases was expected since the LT3 is a desorption process and not
 expected to generated destruction or combustion of chlorinated compounds.

 The emissions of volatile organics were measured by a VOST sampling train. The total non
 methane organic emissions were only 8.9 x 10"3 pounds per hour. Measurements of volatile
 organics entering and exiting the carbon column demonstrated 95% removal of the volatile
 organics by the carbon unit. Similar measurements on semivolatile organic compounds using
 a Modified Method  5 sampling train demonstrated only 3.86 x 10~5 pounds per hour of
 detectable semivolatile organics. The inlet and discharge streams of semivolatile organics
were used to measure a 98.4% removal of semivolatiles in the carbon bed.

Stack sampling using a  Method 23 sampling train was also completed to determine the
emission of dioxins and furans. The emission of 2,3,7,8 TDCF and 2,3,7,8 TCDD were both
less than detection at <2.1 x 10'3 and <3.4  x 10'3 ng/dscm, respectively.  The total emission
of TCDF was < 3.0 x 10'3 ng/dscm and total TCDD was determined to be 1.1 x -1 ng/dscm.

For further information,  please contact Michael G. Cosmos, P.E., Project Director at (215)
430-7423 or Adolfo G. Murphy, Principal Project Manager at (215) 344-3724.
                                         75

-------
    Contaminated
     soil storage
    Clay Shredder
      Oversize
      material
                                                          Sweep gas
                                                    To atmosphere

                                                           Hot oil burner off-gases
                                                                                            Fuel/combustion air
Feed conveyor
                                                                                          Spray water
                                                                                           Makeup
                                                                                            water
                                                                                       Processed soil
                                                                                        dump truck
                                                                           3-Phase
                                                                           oil/water
                                                                           separator
                                                                                        To backfill
                                                                                      excavation area
424-6340
                                  To atmosphere
                            FIGURE 1  SCHEMATIC FLOW DIAGRAM OF THE LTl PROCESS

-------
Equipment
A. Thermal Processors
B. Fabric Filler
C. Process Control Trailer (Oflice)
D. Thermal Processor Drive Units
E. Hot Oil System
F. Air-cooled Condenser
G. Induced Draft Fan
H. Refrigerated Condenser
I.  Vent Condenser
J. Glycol/Water Pumps
K. Glycol/Water Reservoir
L. Heater
M. Vapor Fan
N. Vapor Phase Carbon Columns
O. Oil/Water Separator
P. Liquid Phase Carbon Pumps
Q. Organic Collection Drums
R. Liquid Phase Carbon Columns
S. Clay Shredder
T. Drag Conveyor
U. Discharge Conveyor
V. Dump Truck
              FIGURE 2  GENERALARRANGEMENT
                         OF LTi SYSTEM

-------
                   PLASMA ARC VITRIFICATION
 Richard C. Eschenbach, Retech, Inc.
 Abstract:  The paper updates results obtained with Retech's Plasma
 Centrifugal Furnace (PCF) process for treating hazardous waste. Data from
 three SITE Demonstration tests with a PCF-6 in Butte, Montana, are
 reported, as well as activities in Europe.

 Process Concept:  The Plasma Centrifugal Furnace (Figure 1) is a thermal
 technology that uses heat generated from a plasma torch to treat
 hazardous waste containing metals and/or organics. Metal-bearing solids
 and soil are melted by the process, and organic contaminants are thermally
 destroyed. The molten material forms a hard, glass-like leach-resistant
 mass after cooling in a mold. The major elements of the process are the
 feeder, plasma torch, rotating reactor well, afterburner, secondary
 combustion chamber, and off-gas treatment system.

 The process operates as follows: contaminated soil is placed in a bulk
 screw feeder and gradually fed into the rotating reactor well. Solid
 material is retained in the well by centrifugal force while a plasma arc
 heats the  material in an enriched oxygen atmosphere to temperatures
 sufficient to melt soil (typically on the order of 3000°F). At this
 temperature organic contamination  is volatilized  and burned. Any
 incompletely burned gases or products of incomplete combustion formed
 are incinerated by an afterburner located downstream of the reactor well.
 Once a sufficient amount of feed material has been  treated, the rotating
well is slowed, and the molten mass flows through a secondary chamber
 into a slag collection chamber.

 Initial tests with a lab-size Plasma Centrifugal Furnace having an 18"
diameter (=1.5 ft.) centrifuge (i.e. PCF-1.5) were conducted in the fall of
 1987. Leach tests showed that  the vitrified material which solidified in
the tub  upon run termination was extremely leach-resistant (1).

The encouraging results from the lab-size unit led to acceptance of the
Retech process into EPA's second Superfund Innovative Technology
Evaluation (SITE) solicitation in 1988.
                                   78

-------
 In order to meet EPA's desire for a demonstration, rather than
 development project, Retech designed and built a larger unit with a six foot
 diameter centrifuge (PCF-6).  In all PCF's, the centrifuge is located inside a
 double-walled, water-cooled shell which can withstand substantial positive
 or negative pressures without leakage. The PCF process is covered by two
 U.S. patents (2).

 Test Setup at Butte. Montana

 Construction of the PCF-6 was completed in March of 1989, and a
 preliminary test, supported by the EPA, was conducted at Ukiah April 25,
 1989 with monitoring by Radian.  The test disclosed problems with the
 feeder and SCC temperature.  Modifications were made, and ten additional
 shakedown runs were made in Ukiah in May before the furnace was
 disassembled and shipped to Butte.

 The Component Development  and  Integration Facility (CDIF) of the U.S.
 Department of Energy, in Butte, MT was picked as the test site for three
 main reasons:  1) Butte had a Superfund location with pentachlorophenoi
 contaminated soil leaving a high content of heavy metals - this
 combination is readily treated with the PCF,  2)  there was vacant lab space
 at the CDIF with good support facilities, and 3)  at that time the CDIF was
 under the administrative supervision of the Idaho National Engineering
 Lab (INEL) and INEL was interested in evaluating the utility of the PCF for
 treating several different types of problem wastes at INEL.

 About thirty shakedown tests were performed  by MSE personnel between
 October 1989 and June 1991 (MSE, Inc. is the M&O contractor for DOE at
 the CDIF). Results of some of the shakedown tests have been reported (3).

 During these shakedown tests, modifications were made to the original
 system design.  These modifications included the installation of an
 afterburner in the secondary combustion chamber, the installation of a
 chiller on the gas treatment system to compensate for this additional heat
 input, and the elimination of the surge tank.  In June the  Environmental
 Assessment was approved by both DOE and the EPA, and  the SITE tests
were  scheduled for the week of July 22, 1991.

The waste selected for treatment consisted of heavy metal-bearing soil
from  the Silver Bow Creek Superfund site mixed with 10% by weight No. 2
 diesel fuel. The mixture was spiked  to provide 28,000 ppm of zinc oxide
 and 1,000 ppm of hexachlorobenzene.
                                   79

-------
 Zinc was added as a tracer to determine the leachabiiity of the slag and the
 hexachlorobenzene was the Principal Organic Hazardous Constituent (POHC)
 used to determine organic Destruction and Removal Efficiency (DRE).

 The entire process was controlled on the first level at a central control
 panel. Here the torch operator, aided by T.V. monitors, controlled the
 screw feeder and the position of the torch. Cameras were installed in
 viewports attached to the primary chamber. All process parameters  were
 monitored by an automated  Data Acquisition System (DAS) which collected
 information every 30 seconds. Auxiliary equipment was monitored by
 MSB personnel performing visual inspections and recording data every 30
 minutes.

 On Monday, 22 July 1991, furnace preheat started at 9:05 a.m. and feeding
 of the first load commenced  at 1:10 p.m. Four loads were fed, a pour was
 made (277 pounds) and operations were shut down.  On Wednesday,  24.
 July 1991, furnace preheat started at 8:20  a.m. and feeding of the first load
 commenced at 11:20 a.m. Exhaust fan problems limited the number of soil
 loads to three (265 pounds of slag poured). On Friday, 26 July 1991,
 furnace preheat started at 8:20 a.m.  A water leak from  the torch
 protective sleeve interrupted the preheat; repair required three hours.
 Preheat was again initiated at 12:40  p.m. and feeding of the first load
 started at 3:50 p.m. Five loads were fed; feeding of the fifth load was
 completed at 9:06 p.m. The melt was poured (595 pounds) starting at 9:23
 p.m.

 As part of the SITE Program, a sampling strategy was  designed and
 employed to evaluate the performance of the Plasma Centrifugal Furnace
 technology developed by Retech, Inc. Samples were collected in
 accordance with the "Demonstration  Plan for Plasma Centrifugal Furnace
Technology" (4). Minor changes to the original sampling plan were made.
Figure 2 presents a schematic of the  sampling locations for the
Demonstration Tests.

 A somewhat more  complete summary of the test procedure and the result
was given at the 1992 Incineration Conference (5). The  detailed final
report appeared in two volumes (6, 7). An Applications Analysis Report
has also been issued by the EPA (8).

After the three SITE tests were completed, the PCF-6 at Butte was used to
treat surrogates of wastes now at INEL. These tests have been
summarized by MSE (9).
                                  80

-------
Test Results at Butte

The Destruction and Removal Efficiency (DRE), used to determine organic
destruction, was determined by analyzing the feed soil and the stack gas
for the Principle Organic Hazardous Compound (POHC). For these tests, the
POHC was hexachlorobenzene, which is harder to destroy than PCB's. The
estimated mean level of hexachlorobenzene, based on ail the feed soil
samples for the three tests was 972 ppm. The 95% confidence interval for
the estimated mean was 864  to 1,080 ppm.  No hexachlorobenzene was
detected in the stack gas; therefore, all DREs determined are based on the
detection limit from the appropriate tests. Table 1 gives the 95%
confidence interval for the DREs based on the confidence interval of the
feed soil and the detection limit for the particular test.

                               Table 1

                   DRE Results for Demonstration Tests
Hexach.loroben.zene
Test 1 I Test 1
1 Duplicate
Test 2
Test3
Lower 95% Confidence Interval >99.9964 >99.9982
Mean >99.9968 >99.9984
Upper 95% Confidence Interval >99.9971 >99.9986
>99.9990 >99.99989
>99.9991 >99.99990
99.9992 99.99991
As can be seen from Table 1, the estimated average DRE values for these
tests ranged from >99.9968% to >99.9999% for a highly chlorinated
compound, hexachlorobenzene.  It can be reasonably assumed that this
level of DRE (if measurable) can be achieved for most chlorinated or
halogenated compounds.

Throughout each of the three Demonstration Tests, CO, C02, 02, NOX, and
total hydrocarbons (THC) were monitored continuously to present a real
time image of the combustion process and to determine if regulatory
standards were being exceeded. MSE, Inc. also monitored the same
combustion products for DOE during shakedown tests and other testing of
the PCF.

Since the installation of the afterburner in the secondary combustion
chamber, the level of total hydrocarbons exiting the system has been low
(<4  ppm) even with at least 10% organics in the feed soil.
                                  81

-------
 This gives a good indication that effective thermal destruction of the
 organic compounds is occurring. Another indication of the ability of the
 process to treat organic contaminated media is the low level of CO in the
 exhaust (approximately 1.4 ppm) and the level of C02 (approximately 8%).
 02 monitors show significant variation throughout the treatment process as
 pure 02 is fed to the primary chamber at approximately 18 scfm while
 waste is being fed to the furnace.

 The PCF is designed to encapsulate inorganic compounds in the vitrified
 slag and render the treated soil nonleachable.  Testing activities have
 demonstrated that the process can effectively bind inorganic compounds
 into the treated soil. The Toxicity Characteristic Leaching Procedure (TCLP)
 was performed on both the feed soil and the treated soil (i.e. slag). The
 feed soil was tested to ensure that the material was leachable for organic
 and inorganic compounds.  The vitrified slag underwent TCLP to meet the
 testing objectives.                   t

 TCLP analysis of the feed soil for metals showed that the only elements
 which exhibited significant leachabiiity characteristics were calcium and
 the spiked zinc. The results of the TCLP metals analysis of the feed soil are
 summarized in Table 2.  The presence of sodium in the leachate is not
 unexpected because of its high concentration in the soil and the fact that,
 unlike other metals, it will have an affinity to go into solution.  If the
 solution is even slightly acidic (as in the TCLP)  this phenomenon is
 enhanced. None of the eight RCRA characteristic metals found in the feed
 soil leachate were above the regulatory limit, therefore, the evaluation of
 the leachabiiity of the vitrified slag was based on calcium and zinc.

 Calcium was chosen, in addition to zinc, because of its tendency to  leach
 from the feed soil. During the Demonstration Tests, sodium was not used
 in evaluating the soil leachabiiity because it is not typical of regulated
 TCLP metals.

 The treated soil TCLP metals analysis is also shown in Table 2.  None of the
 metals, with the exception of sodium, show any strong characteristic for
 leaching.  Sodium is  probably present in the leachate for the reasons stated
 above and is not considered in this evaluation.  Both tracer metals, calcium
 and zinc, showed significant reductions in leaching properties in the
 treated soil as compared to the feed.  In fact, all of the metals with the
 exception  of aluminum and  iron showed reduced leaching characteristics.
The leachabiiity for the aluminum in both feed  and treated soil is low and
the values reported for the treated soil are only estimates (less than the
 quantity limit).
                                   82

-------
Therefore, it is quite probable that the leachability of aluminum from the
feed soil to the treated soil did not change.  The apparent increase in
leachability of iron in the treated soil occurred is because approximately
100  pounds of mild steel was placed in the  furnace to aid in initiating the
arc.  This considerably increased the iron content of the slag in comparison
to the feed soil.

The only organic constituents that were found to be leachable from the
feed soil were 2-methylnaphthalene and naphthalene.  Although the feed
soil was spiked with a high level of hexachlorobenzene  (1000 ppm), it did
not leach from the soil.  No organic compounds were found to leach from
the treated slag.

The Toxicity Characteristic Leaching Procedure requires samples to be
ground into small particles.  In this manner, a large amount of surface area
is available for leaching. Since the PCF produces a monolithic slag after
treatment, the surface area per pound of treated soil is much smaller than
that for the  TCLP test. The TCLP results, therefore, present a conservative
assessment of the actual leachability of the  monolithic slag.
                                   83

-------
                                 Table 2

                    TCLP Results for Demonstration Tests
Average
Compound Soil Lea
Concent
(mg/L)
Metals:
Treated Soil
Feed
chate
ration Test 1 Test 2
(mg/L) (mg/L)

Aluminum 0.23 J 0.45 J 0.41 J
Barium 0.14 0.08 0.08
Cadmium 0.07 ND ND
Calcium 175
Copper 4.6
2.1 J 2.5 J
0.15 0.35
Iron 0.06 2.5 2.95
Magnesium 8.12 ND ND
Manganese 4.82
Nickel 0.02
0.06 0.06
ND 0.01 J
Potassium 4.57 ND ND
Sodium 1475
Vanadium 0.1
Zinc 982
Semivolatiles:
1500 1400
ND ND
0.45 0.36

Mapthalene 0.397 ND ND
2-Methyinaph- 0.282 ND ND
thalene
JJeiachloro- ND
senzene

ND ND

Leachate
Test 3
(mg/L)

0.32
0.07
ND
2.05
0.3
31.1
ND
0.24
0.1
ND
1400
ND
0.3

ND
ND

ND

Concentration
Regulatory
Limit*
(mg/L)

J NR
100.0
1.0
J NR
NR
NR
NR
NR
NR
NR
NR
NR
NR

NR
NR

0.13





















ND - Not Detected
NR - Not Regulated
' - 40 CFR (07/01/90 Edition) §261.24, Table 2
J - Estimated result (less than five times the detection limit)

Note that slightly less than half the zinc was captured in the slag. If the
scrubber had been effective, the  zinc not captured when charged could
have been subsequently re-charged.

The post-test scrubber liquor did not contain any significant quantities of
organic compounds.  Nitrated compounds and phthalates were the only
compounds present.  The nitrated compounds were most likely produced
from the high levels of NOX in the exhaust gas reacting with the water from
the scrubber and any organic compounds present.
                                    84

-------
The phthalates and volatile organic compounds probably entered the
scrubber sump from the scrubber make-up liquor.

The lack of organic compounds in the scrubber liquor and, as stated earlier,
the absence of volatile or semivolatile organic compounds in the exhaust
gas, indicates that combustion of the organic compounds was complete.

The scrubbing unit was very inefficient in the capture of the inorganic
compounds. The scrubber did capture some of the volatile metal elements
but not at the levels that would typically be expected from a well-designed
system. As stated previously, the exhaust gas contained a variety of
metals that should have been captured by the scrubbing unit. The types of
metals found in the scrubber liquor were similar to those found in the
stack gas;  that is, arsenic, iron, and zinc were the elements in abundance.
High sodium levels found in the liquor were probably a consequence of the
scrubber make-up (sodium hydroxide).

Results of the tests run by MSB for DOE after the SITE tests again showed
excellent TCLP results.  Cerium was added as a surrogate for radioactive
compounds found in some INEL wastes; the mass balance for cerium was
inconclusive since less than half the feed cerium was accounted for (9).
Additional tests are planned.

Other Applications

In 1989 a Swiss firm, MGC Plasma ordered a PCF-8 furnace which was
installed in the summer of 1990  in Muttenz, a suburb of Basel (10).  A gas
cleanup system built by Ceiicote  was  installed to permit the effluent to
meet the very strict Swiss standards. The furnace has a very unique
charging system: whole drums are inserted, five at a time, into a drum-
feeder chamber. A drum manipulator clamps the lid of the lead drum and
holds the drum in the furnace while an auxiliary cutting device opens up
the drum to dump the contents gradually into the furnace. There have
been some problems with this feeder and a different feed system is under
construction.

Retech is currently designing three more systems for three European
customers. One is a PCF-8 for treating soil contaminated with military
wastes. Most of the PCF feed in this application will arrive from a soil-
washing process which will enable at least 90% of the contaminated soil to
be returned as "clean" soil.
                                   85

-------
 The second is a PCF-8 for treating low/intermediate level radioactive
 wastes.  In this application, three categories of feed will be charged to the
 furnace - organic liquids, combustible solids and metal waste.  All three
 categories will come to the furnace in 200 liter drums.

 The third project is a smaller furnace, a PCF-2, which will be used to test
 first radioactive surrogates, then radioactive mixed wastes. Figure 3 is a
 schematic of this lab-size furnace.

 Acknowlede ments
 Much of the detail in this report came from the Butte SITE tests, and
 special thanks are due to the co-authors of the paper summarizing this
 work:  Rob Haun of Retech, Corry Alsberg and Dan Battleson of MSB and
 Trevor Jackson of SAIC.  I am also grateful for the support of the EPA SITE
 Demonstration Program which is conducted by the  Risk Reduction
 Engineering Laboratory.  Ms.  Laurel Staley was the Technical Project
 Manager for the SITE tests.

 References
 1.   J. W. Sears, R. A. Hill and R. C. Eschenbach, "Stabilization and
     Decomposition  of Toxic and Radioactive Wastes by Transferred-Arc
     Plasma", 1989  Incineration Conference, Knoxville, TN
 2.   U.S. Patents 4,770,109 and 5,136,137
 3.   A. J. Viall, J. W. Sears, R. C. Eschenbach, "Test Results with the Plasma
     Centrifugal Furnace", 1990 Incineration Conference, San Diego, CA
 4.   SAIC, Rev. 1, July 11, 1991
 5.   Rob Haun, R. C. Eschenbach, Dan Battleson, Corry  Alsberg, Trevor
     Jackson, "SITE Test Results with the PCF-6", 1992 Incineration
     Conference, Albuquerque, New Mexico
 6.   "Technology Evaluation  Report of Retech's Plasma Centrifugal
     Furnace", Vol. I PB92-216 035-V1
7.   "Technology Evaluation  Report of Retech's Plasma Centrifugal
     Furnace", Vol. 2 PB92-216 043-V2
8.   Retech, Inc., Plasma Centrifugal Furnace, EPA  Applications Analysis
     Report, EPA/540/A5-9 1/007, June 1992
9.   C. G. Whitworth, L. G. Twidwell, T. W. Jenkins  and G. F. Wyss, "Slag
     Chemistry and  Metals Volatilization in the Plasma Arc Furnace
     Experiment", Proceedings  of Spectrum 92 - Nuclear and Hazardous
     Waste Management International, August 1992, Boise ID
 10.  M. R. FUnfschilling, R. C.  Eschenbach, "A Plasma Centrifugal Furnace
     for Treating Hazardous Waste,  Muttenz, Switzerland", Electrotech 92,
     Montreal, Canada
                                   86

-------
00
-0
                  FEEDER
             SECONDARY
             COMBUSTION
             CHAMBER—
             SLAG
             CHAMBER
                  n
           EXHAUST
            STACK
                                            PLASMA TORCH
                                              ROTATING
                                              REACTOR WELL
SURGE
TANK
GAS TREATMENT
                \/^^^^^
                        FIGURE 1: COMPONENTS OF PCF SYSTEM

-------
                         SCREW
                         FEEDER
                  PR 1 MARY
                  CHAMBER
AFTER '
BURNER
SECONDARY
 CHAMBER
00
00
              VENT
               TO
            ATMOSPHER
             STACK
             BLOWER
                                 JET
                               SCRUBBER
             QUENCH
              TANK
                  SAMPLE LOCATION
                               SCRUBBER
                                 SUMP
                                                                CAUSTIC
                                                               RESERVOIR
FIGURE 2
                              SCHEMATIC  OF SAMPLING LOCATONS
                              FOR  THE  DEMONSTRATION TESTS

-------
00
VO
             REMOVABLE LID
               FEEDER PORT
              24 INCH I.D.
              CENTRIFUGE
             SLAG COLLECTION
             MOLD
                   ROTARY
                   DRIVE
                                                          RP-75T PLASMA
                                                          TORCH
                                                      OFF-GAS TO SCC
                                                       4 INCH DIA. THROAT
       FULL DOOR ACCESS
       FOR CLEANING AND
       SLAG  REMOVAL
ROTARY WATER
JOINT
                           FIGURE 3: SCHEMATIC OF PCF-2

-------
            THERMAL DESORPTION OF PCB-CONTAMINATED WASTE
                AT THE WAUKEGAN HARBOR SUPERFUND SITE
                              (A CASE STUDY)
                  Joseph H. Hutton, P.E., Operations Manager
                     Robert Shanks, Technical Supervisor

                         SoilTech ATP Systems, Inc.
                    6300 South Syracuse Way, Suite 300
                        Englewood, Colorado  80111
                              (303) 290-8336
 Abstract
 In June 1992, SoilTech ATP Systems, Inc. (SoilTech) completed the soil treatment
 phase of the Waukegan Harbor Superfund Project in Waukegan, Illinois, after approxi-
 mately five months of operation.  SoilTech successfully treated  12,700 tons of
 polychlorinated biphenyl (PCB)-contaminated sediments using a transportable SoilTech
 Anaerobic Thermal Processor (ATP) System nominally rated at 10 tons-per-hour (tph)
 throughput capacity. The SoilTech ATP Technology anaerobically desorbs contami-
 nants such as PCBs from solids and sludges at temperatures over 1,000 degrees
 Fahrenheit (°F).  Principal products of the process are clean treated  solids and an oil
 condensate containing the hydrocarbon contaminants.

 At  the Waukegan Harbor Superfund site, PCB  concentrations in the sediments
 excavated and dredged from a ditch, lagoon and harbor slip averaged 10,400 parts
 per million  (ppm) (1.04 percent) and were as high as 23,000 ppm  (2.3  percent).
 Treated soil contained less than 2 ppm PCBs and was backfilled in an on-site contain-
 ment cell.  The removal efficiency of PCBs from the soil averaged 99.98 percent,
 relative to the project performance specification of 97 percent.  Approximately 30,000
 gallons of PCB oil, desorbed from the feed material,  were returned to the potentially
 responsible party (PRP) trust for subsequent off-site  disposal. After  modifications to
 the emissions control equipment, compliance with the 99.9999 percent destruction
 and removal efficiency (ORE) for  PCBs  in stack emissions required by the United
 States Environmental Protection Agency was achieved.  SoilTech  demonstrated
compliance with the ORE requirement in eleven consecutive stack sampling events.
Feed rate  averaged 8 tons-per-hour at a mechanical availability  of 85  percent.
SoilTech revenues for the project were $700,000 in  fixed costs and  $185 per ton of
soil processed.
                                    90

-------
           THERMAL DESORPTION OF PCS-CONTAMINATED WASTE
               AT THE WAUKEGAN HARBOR SUPERFUND SITE
                              (A CASE STUDY)
                                    By
     Joseph H. Hutton, P.E., Operations Manager, SoilTech ATP Systems, Inc.
        Robert Shanks, Technical Supervisor, SoilTech ATP Systems, Inc.
1.0 INTRODUCTION

This technical paper outlines SoilTech's role in the Waukegan Harbor Superfund
Project. SoilTech was responsible for the soils processing phase of the project using
its unique rotary kiln  known as the SoilTech Anaerobic Thermal Processor (ATP)
Technology to remediate the PCB-contaminated soils and sediments.  The Waukegan
Harbor Project is the second commercial application of the SoilTech ATP Technology.
42,000 tons of PCB contaminated soils were successfully treated to nondetect levels
at the Wide Beach Superfund site in western New York in 1990 and 1991. The ATP
distills organic contaminants out of a  solid matrix  in an oxygen-free  environment.
Oxidative degradation of contaminants such as PCBs into more harmful reaction
products is therefore prevented.  Contaminants are collected in an oily condensate
which can then be economically disposed of.

Section 2.0 of  this paper describes  the technology and  Section 3.0  discusses
treatment  costs.  Section 4.0 summarizes SoilTech's activities at the Waukegan
Harbor site.
                                     91

-------
 2.0  SOILTECH ATP TECHNOLOGY DESCRIPTION

 The ATP Technology was originally conceived as a means to perform primary refining
 of tar sands and oil shales to crude oil in the early 1970s.  UMATAC Industrial
 Processes (UMATAC) developed and tested the technology over a period of more than
 15 years with  funding from the Alberta Oil Sands Technology Research Authority
 (AOSTRA). The technology has been further developed to handle a wide range of
 organic contaminants such as PCBs.  In 1988, Canonie Environmental Services Corp.
 (Canonie) entered into  an exclusive  license agreement to utilize the technology for
 waste treatment in the United States.  Together, Canonie and UMATAC,  as equal
 partners, formed SoilTech.  The  technology,  now  known  as the  SoilTech ATP
 Technology, is  described below.   Specific fundamental elements of an ATP plant
 include the ATP System, the vapor handling equipment,  the air pollution control
 equipment, and the control room.  Each  of these elements is described below.

 2.1  Anaerobic Thermal Processor

 The central element of the ATP Technology is the  processor itself which resembles
 a rotary  kiln from its exterior.  However, inside the processor are three physically
 distinct zones and four zones characterized by different physical processes. The four
 process zones are described as the following:

      1.     Preheat zone;
      2.     Retort zone;
      3.     Combustion zone; and
      4.     Cooling zone.

Figure 1 is a cross section of the ATP which depicts each of the four zones. The form
and function of each of  the zones are described below.
                                     92

-------
2.1.1  Preheat Zone

As indicated on Figure 1, the preheat zone is the first zone of the ATP into which feed
material is conveyed. The preheat zone is charged with feed material by a system of
variable-speed material conveyors and feed hoppers as indicated on the process flow
diagram shown on Figure 2. The preheat zone is a cylindrical compartment oriented
centrally along the axis of the ATP. It is indirectly heated by conduction of thermal
energy from hot treated material exiting the processor through the cooling zone.

The  material fed to  the preheat zone is  raised  from ambient temperature  to
approximately 500°F  as the material travels through the preheat zone.  Any light
hydrocarbons and moisture present are distilled out of the feed material and evacuated
through a vapor  pipe  mounted in the fixed  endframe of the preheat zone (see
Figure 1).  The blowers and dampers used to evacuate the vapors from the preheat
zone are configured to maintain a negative pressure of approximately 0.10-inch water
column  relative to atmospheric pressure.

When material reaches the end of the preheat zone, it exits  by passing through, a
circumferential grizzly and entering the proprietary sand seals before being conveyed
into the retort zone.  The proprietary sand seals allow conveyance of  solid material
while prohibiting the passage of vapors. This allows the retort  zone to be maintained
in an anaerobic condition.

2.1.2 Retort Zone

The ATP retort zone is a cylindrical chamber oriented along the axis of the processor
at the discharge end of the preheat zone. Material enters the retort zone from two
sets of sand seals.  The first set conveys preheated material from the preheat zone
into the retort zone, as previously described. The second set conveys treated material
from the combustion  zone back into the retort zone.  This  hot recycled material
                                      93

-------
  provides the necessary thermal energy to raise the temperature of the bulk solid
                              i
  mixture in the retort zone to temperatures of 950°F to 1,150°F.  The sand seals
  prevent the  passage of oxygen into the retort zone and also prevent contaminated
  vapors from passing into the combustion zone from the retort zone.

  Solid  material exits the retort zone through a third  set of sand seals into the
  combustion zone, as indicated on Figure 1.  Heavy hydrocarbons, which enter the
  retort zone on the material entering from the preheat  zone, are quickly vaporized,
  producing hot, decontaminated, inert solids. The distilled hydrocarbon vapors are
 evacuated from the retort zone through a vapor pipe mounted through the stationary
 endframe at the combustion end  of the processor.   The evacuated vapors are
 conveyed to a system of condensing equipment. The blowers and dampers employed
 to evacuate vapor from the retort zone are configured to maintain a negative pressure
 of approximately  0.12-inch water column relative to atmospheric pressure.

 2.1.3  Combustion Zone

 The combustion zone of the ATP makes up the annular space between the inner wall
 of the processor  and the outer  wall of the retort zone  bordered by the stationary
 combustion endframe and the cooling zone.  As shown on Figure 1,  there is no
 physical barrier between the combustion zone and the cooling zone.

 Treated material enters the combustion zone from the discharge sand seal of the retort
 zone. Accelerator flights reverse the direction of material flow back toward the feed
 end of the ATP unit. Lifting elements lift and drop the treated soil into the combustion
 flue  gases.   Temperatures of around  1,400°F are  maintained by two natural
 gas/propane burners that are fired into the combustion zone through the processor
 endframe.  A portion of the heat requirements of the system are provided by the
oxidation of any coke present on the decontaminated solids and light, noncondensable
hydrocarbons returned to the combustion zone from the  vapor condensing systems.
                                     94

-------
As the hot treated  material nears the entrance to the cooling zone, a fraction is
diverted back into the retort zone through the recycle sand seals described previously.
This recycled treated material provides the thermal energy necessary to achieve the
required desorption temperatures in the retort zone.  The material that is not diverted
into the retort zone passes through to the ATP cooling zone.

2.1.4 Cooling Zone

The cooling zone is  a continuation of the combustion zone. It is the annular space
between the inner wall of the processor and the outer wall of the preheat zone. In the
cooling zone, treated material continues to be lifted and dropped into the flue gases
and onto the exterior wall of the preheat zone. Thermal energy is thereby transferred
by conduction through the preheat zone wall to the incoming feed material in the
preheat zone while cooling  treated material in the cooling zone.  When material
reaches the end of the cooling zone, it falls through a stationary grizzly into a screw
conveyor.

2.2  Auxiliary Systems

The ATP is supported by various auxiliary systems which employ well-established
technology, most of which have been proven over many years in a wide range of
industrial applications.  A simplified process flowsheet for the ATP Technology is
provided on Figure 2.  A more detailed process flowsheet is included as Figure 3.

Each of these auxiliary systems is described in detail below.
                                     95

-------
 2.2.1  Vaoor Handling Equipment

 The ATP  plant contains two separate gas-handfing trains to process the vapors
 generated in the preheat and retort zones. The basic function of each of the trains
 is to remove particulate matter and cool the vapors such that water and hydrocarbons
 are condensed.

 2.2.2  Preheat Vapor Handling Equipment

 Steam and light hydrocarbons are removed from the preheat zone through a cyclone
 separator to remove particulate matter and then through a water-cooled shell and tube
 heat exchanger.  The liquids condensed in the shell and tube heat exchanger gravity
 drain into an oil/water separator.  Water collected in the separator is pumped to stor-
 age and subsequent treatment.  Oil collected in the separator  is pumped to storage
 for subsequent disposal off-site or can be used for reflux in the retort zone's vapor
 handling equipment.  Any noncondensable gases are passed through activated carbon
 and then into the combustion zone.

 2.2.3 Retort Zone Vapor Handling Equipment

 The hydrocarbon vapors and contaminants, such as PCBs, are removed from the retort
 zone and pulled through a vapor scrubber which removes particulate material too fine
 for removal in the upstream cyclones and cools the vapors from over 1,000°F to
 approximately 400°F. The vapor scrubber is a direct-contact condensing tower which
 relies on closed-circuit pumping to circulate oil countercurrent to the vapor flow.  The
 condensed oils and contaminants are pumped to storage for subsequent disposal off-
 site.  Most of the PCBs are condensed in the vapor scrubber.

Vapors exiting the vapor scrubber are pulled through the fractionater.  The f ractionater
is identical to the  vapor scrubber but operates at lower temperatures and cools the
                                     96

-------
vapors to approximately 250°F. Condensed oil can either be pumped to the vapor
scrubber or directly to storage.

The vapors which exit the fractionater are pulled through a  water-cooled shell and
tube heat exchanger identical to the one used in the preheat system. The condensed
liquids gravity drain into a two-phase separator. The water phase is pumped from the
separator to storage for subsequent treatment. The oil phase is pumped into the
fractionater  or vapor scrubber as a means of temperature and oil level control.
Noncondensable vapors are passed through activated carbon to remove residual heavy
hydrocarbons and then onto the combustion zone of the ATP through the combustion
endframe.

2.2.4 Air Emissions Control Equipment

The  combustion gases are pulled under vacuum from  the cooling zone by two
induced-draft fans through a cyclone separator, a quench tower, two  baghouses
operated in parallel, and an activated carbon bed. The relatively low gas flow rate and
temperature assure thorough removal of particulates and residual hydrocarbons.

2.2.5 Material Handling Systems

The ATP Plant includes two independent solid handling systems for feed handling and
treated soil  handling.

The feed handling system is designed to provide solid feed material to the ATP during
steady-state waste treatment operations and to  provide clean sand feed to the ATP
during startup, shutdown or upset conditions.  As  indicated on  Figure 3, the feed
system consists of two  hoppers equipped with apron belt conveyors, either of which
can feed a belt  conveyor that transports contaminated feed or clean sand into the
preheat zone.
                                     97

-------
 The treated soil handling system conveys treated soil from the base of the cooling
 plenum via a soil conditioner equipped with water sprays and onto a conveyor belt.

 2.2.6  Process Waste water Treatment System

 The ATP Plant produces aqueous condensate at a rate equal to the rate that moisture
 enters  the processor with the feed soil.  At the Waukegan Harbor site, the ATP
 System produced  6 gallons-per-minute (gpm) or less of aqueous condensate which
 required treatment.  The water was pumped from the phase separators into storage
 tanks and then through  the  on-site treatment system operated by Canonie.  The
 treatment system  is designed to process up to 25 gpm.

 The treatment  system relies  on filtration, oxidation,  and adsorption operations to
 remove contaminants from the aqueous condensate. Specific treatment steps include
 sand filtration, Klensorb® filtration, ultraviolet enhanced oxidation, cartridge filtration
 to 0.5 microns, and activated carbon filtration through multiple beds in  series.  The
 treated  water can be pumped  to the process water storage tanks for subsequent use
 in the plant. However, at Waukegan Harbor treated water was discharged to the sani-
 tary sewer.

 2.2.7 The Control Room

The plant control systems and electrical hook-ups are housed in  the control room
trailer.  Extensive  use of microprocessor technology for process  control and  logic
functions allows one lead operator to run and monitor the plant from the control panel
with the assistance of two technicians in the field. All process parameters such as
temperatures, flowrates, pressures, tank levels, equipment current draws,  and flue gas
quality are monitored and recorded.
                                     98

-------
2.3 ATP Plant Utility Requirements

The  ATP Plant requires electricity, fuel (natural gas or propane) and nitrogen in
addition to process water. Specifically, the plant consumes an average of 210 KVA
electricity, eight million BTU-per-hour of fuel, and  15,000 cubic feet-per-week of
nitrogen.

2.4  Staffing

The ATP System runs on a three-shift, 24-hour-per-day basis with an operating crew
of three or four people comprised of one lead operator and two or three assistant
operators working in the field.  Three or four maintenance personnel work day shift
five days a week. A Project Superintendent, Project Engineer, Site Safety Officer, and
Receptionist are also on-site during day shift  hours.

2.5  Health and Safety

At the Waukegan Harbor site the ATP System was located inside the exclusion zone.
Level D personal  protective equipment (PPE) was  used during plant set-up.  An
upgrade to Level C was required as soon as processing began.  The control room was
positioned in the support zone adjacent to the exclusion zone, allowing the lead
operator to work in street clothes.
                                     99

-------
 3.0  TREATMENT COSTS

 Costs for treatment using the 10-tph SoilTech ATP System can vary depending on the
 character of the waste material, ranging from $150 per ton to $250 per ton. Costs
 depend mainly upon the following variables:

       •     Moisture content of feed material;

       •     Particle size;

       •     Hydrocarbon content;

       •     Material handling characteristics; and

       •     Chemical characteristics.

 Mobilization and demobilization costs for the 10-tph SoilTech ATP System range from
 $700,000 to $1.5 million.

 As illustrated on Figure 4, the smaller 5-tph SoilTech ATP is generally more cost-
 effective for  projects of 5,000 tons or less, while the larger 25-tph ATP System is
 more cost-effective for project of greater than 45,000 tons.

 3.1  Moisture Content of Feed

 Feed moisture content is a rate-limiting parameter related linearly to plant throughput
 and therefore to the price of treatment. Ideal moisture content in feed material is in
the range of 5 to 10 percent.  At  lower moisture  contents, entrained dust can create
 problems in the vapor condensing systems. At moisture contents above 10 percent,
the latent heat required to distill the moisture from the feed in the preheat zone of the
                                      100

-------
ATP System can become rate limiting.  SoilTech evaluates, and when appropriate,
incorporates predrying of feed material into a project work plan when waste moisture
content might affect treatment economics.

3.2  Particle Size

Material should be smaller than 2 inches; if not, pre-screening and/or crushing may be
necessary. Where a material contains a  high percentage of very fine material such as
clay, co-feeding of coarse material may  be necessary. Fine material tends to reduce
the efficiency of the sand seals and will be drawn out of the processor with the flue
gases, reducing the quantity of material available for heat transfer in the combustion
and cooling zones.

3.3  Hydrocarbon Content

The SoilTech ATP System is designed to handle feed materials containing up to  10
percent hydrocarbons.  With higher levels or where a substantial proportion of those
hydrocarbons have a high molecular weight and amenable to cracking, the amount of
coke formed in the retort zone may be excessive. Elevated levels of  carbon monoxide
in the flue gas stream  may then result.   Reducing  feed  rate  or  incorporating  an
additional flue gas treatment step, such as catalytic oxidation, mitigates the problem
but increases treatment costs.

3.4  Chemical Characteristics

A feedstock exhibiting  unusual reactivity, corrosivity or toxicity  may require pre-
treatment  or special  handling  precautions.   Treatment cost would therefore  be
affected adversely.
                                      10.1

-------
3.5  SoilTech Bench-Scale Unit

SoilTech is able to accurately predict treatment performance and costs based on
bench-scale testing results.  The bench test unit accurately simulates conditions in the
full-scale system.  Over 2,000 bench-scale tests have been conducted on a variety
of contaminated materials. Strict procedures and protocols have been developed and
are rigorously applied.
                                     102

-------
 4.0 WAUKEGAN HARBOR SUPERFUND PROJECT DESCRIPTION

 The Waukegan Harbor Superfund site was listed on the EPA's Superfund Priority List.
 Contamination resulted from leakage of PCBs - used as an aluminum casting lubricant
 and machine tool lubricant - through floor drains from a major manufacturing facility
 into an adjacent stream and Waukegan Harbor.  PCB concentrations in excess of
 20,000 ppm were found in the harbor sediments and stream bed. Remediation of the
 site was complicated by the fact that the harbor was being used for both commercial
 and recreational boating activities and was adjacent  to a public beach.  The client
 insisted that operations not impact the day-to-day activities of the local community.
 It was therefore necessary for Canonie, the prime contractor, to build a new boat slip
 prior to isolating the old one. Sediments with PCB concentrations greater than 500
 ppm were hydraulically dredged from the old slip and pumped to a containment cell
 where dewatering was achieved before they were blended with soils from  the conta-
 minated stream bed.

 The 1984  Record of Decision (ROD)  called for stabilization of these  soils and
 sediments.  However, SoilTech  and the PRPs were able to convince the EPA that the
 use  of SoilTech's ATP Technology offered a  more environmentally acceptable and
 cost-effective solution.  The EPA accepted that the SoilTech ATP System was an
 innovative treatment technology that provided a  significant and permanent reduction
 in the toxicity and volume of PCB wastes at the Waukegan Harbor site.  SoilTech's
 approach fulfilled the requirements of the Superfund Amendments and Reauthorization
 Act (SARA). Under the 1984 ROD, there would  have been a potential future liability
 posed by stabilized stored wastes.

 SoilTech was contracted by Canonie to process the soils and sediments for $700,000
 in fixed costs and $185  per ton  of material processed.  SoilTech was not responsible
for providing utilities (water, gas and electricity). Site  preparation and excavation of
contaminated material were carried out by Canonie. The PRPs took full  responsibility
                                    103

-------
for disposing of the PCB condensate produced. Had SoilTech been responsible for any
or all of these additional activities, the unit price for soils processing would have been
correspondingly higher.

The project clean-up criteria called for a treated soil residual PCB level of 500 ppm,
or 97 percent removal efficiency from the soils, whichever was more stringent. The
average target clean-up level turned  out to be approximately 310 ppm.  Total soils
requiring treatment amounted to about 12,700 tons. Treated soils were to be placed
in one of two containment cells along with untreated soils containing less than 500
ppm PCBs before being closed with a Toxic Substances Control Act (TSCA)-approved
impermeable cap.  PCB-contaminated oils extracted from the soils were  returned to
the PRPs for subsequent disposal.

Since no applicable air emissions  standard exists for thermal desorption, the EPA
required that SoilTech meet the 99.9999 percent (6 nines) destruction and removal
efficiency (ORE). SoilTech was also required to meet the municipal waste incineration
standard of 30 nanograms per dry standard cubic meter (ng/dscm) total dioxins and
furans applied to incineration of PCB-contaminated wastes.  Discharged water was
to be less than 15  parts-per-billion (ppb) PCBs.

4.1  Sequence of Events

SoilTech arrived at the Waukegan Harbor Superf und site in the last week of November
1991. Modification and repair of key process components  continued at a shop in
Indiana while the site was prepared, and support equipment was  assembled and
erected. Shakedown and troubleshooting of the SoilTech ATP System was conducted
in mid-January.   SoilTech  began treating contaminated soils and sediments on
January 22, 1992. This date also marked the start of the 30-day proof-of-process
period.
                                     104

-------
 Between startup and March 6, 1992, the SoilTech ATP System averaged a feed rate
 of 10 tph, outperforming initial projections of 9 tph.  Net plant availability was over
 85 percent.  PCB concentrations in the feed averaged 14,000 ppm and treated soil
 concentrations less than 2 ppm PCBs were attained. The average removal efficiency
 from the treated soil was 99.96 percent.

 During  this  time period,  SoilTech performed  seven stack tests.   Although the
 emissions criteria for dioxins and furans were met, SoilTech was unable to meet the
 6 nines ORE for PCBs and consequently stopped operations at the direction of the
 EPA.

 During the next two months, SoilTech operated the ATP System intermittently to test
 various modifications to the ATP and to the emission control  system.  Specific
 modifications included:

      •  Increasing the volume of carbon in flue gas carbon bed;
      •  Removing the wet scrubber from the flue gas handling system;
      *  Adding carbon beds to internal gas recycle streams;
      •  Testing of continuous addition of powdered activated carbon upstream of
         the baghouses to facilitate adsorption of PCBs;
      •  Reducing flue gas carbon bed temperature; and
      •  Testing of continuous addition  of sodium bicarbonate in the combustion
         zone of the ATP in an attempt  to induce  catalytic destruction of residual
         PCBs.

In  the  latter phase of this period of  intermittent operation and testing, SoilTech
discovered a gap in the flue gas carbon bed seal which was allowing 70  percent of
the flue gas stream to bypass the carbon bed.  SoilTech corrected the poor seal before
stack testing on May 12, 13, and 14,  1992.
                                     105

-------
Each of four stack tests conducted on those three consecutive days demonstrated
performance superior to the 6 nines ORE required. A summary of the stack testing
conducted throughout the project is given in Table 1.

It  appeared that  the  final three tests, using powdered activated carbon in the
baghouse and then soda ash in the combustion zone, produced slightly better results
than the test conducted  without  these  additives.   The  results are,  however,
statistically inconclusive.

With EPA's approval SoilTech was then able to resume processing the contaminated
soils and sediments having met all of the project performance criteria.

To insure that this performance was maintained, SoilTech  instituted several new
procedures and operating conditions.

      •  Soda ash was fed to combustion zone at 200 Ibs/hour;
      •  Stack temperature was maintained at 170°F with automatic  waste feed
         shut off  at 190° F;
      •  The stack gas carbon bed was sampled on a daily basis. The plant was to
         be shut  down and the carbon bed changed when the PCB concentration
         exceeds  10 ppm;
      •  Weekly stack testing were performed by Clean Air Engineering.  Failure to
         meet 6-nines ORE would necessitate immediate shutdown until the cause
         had been identified and remedied; and
      • The pressure drop across the  stack carbon bed was maintained. The bed
        was to be inspected if it drops below 5-inches water column.
                                    106

-------
4.2  EPA Site Demonstration

On June 16,  1992 the EPA began a Superfund Innovative Technology Evaluation
(SITE) demonstration of the SoilTech ATP System.  This involved three days of
rigorous sampling and testing.  The test runs were conducted at typical operating
conditions. The fourth and final test was, however, run without the addition of soda
ash to the combustion zone. The other three tests were run while 200 Ibs/hr of soda
ash was added to the combustion zone.  Each test run consisted of 8.5 hours of
solids and liquids  sampling and 8  hours of stack  sampling.    During  the  site
demonstration, 224 tons of PCB-contaminated soils and sediments were processed.
In  addition to sampling, data on critical operating parameters were collected during
each test.

Based on the preliminary results at the site demonstration, the EPA concluded the
following:
      «  PCS concentrations were reduced from an average of 9,761  ppm in the
        untreated soil and sediment to an average concentration of 2 ppm in the
        treated soil and sediment.
      •  Approximately 0.12 milligram  (mg) of PCBs was discharged from the ATP
        System's stack per kilogram of PCBs fed to the ATP.
      *  The majority of PCBs removed from the untreated soil and  sediment were
        accumulated in the waste oil discharge from the vapor cooling system.
      •  No dioxins, other than a low concentration 0.1 ng/dscm of octachlorinated
        dibenzo-p-dioxin in one stack gas sample, were detected in the stack gas
        from the ATP System. Tetrachlorinated dibenzofurans were found in both
        the untreated soil and sediment (88 ng/g) and treated soil and sediment (5
        ng/g) and the stack gas (0.07  ng/dscm).
      9  Leachabie volatile organic compounds, semivolatile organic compounds, and
        metals in the treated soil and sediment were below Resource Conservation
        and Recovery Act (RCRA) toxicity characteristic standards.
                                     107

-------
       •  No  operational  problems affecting  the  ATP's  ability  to  treat  the
          contaminated soil and sediment were observed.

 Selected  results provided by the EPA are presented in Table 2.  As expected, the
 results confirm that SoilTech met all of the project performance criteria.  Removal
 efficiency for PCBs in  the treated soil averaged 99.98 percent.   ORE for the stack
 emissions exceeded the 6 nines criteria in each test and bettered 7 nines in two of the
 tests.

 The results also demonstrated that dioxins and furans are not produced in the ATP.
 Trace amounts detected in the waste oil, the treated soil and stack gas can be traced
 back to the untreated soils and sediments.

 The highest DRE for PCBs in the stack was achieved for the test in which no soda ash
 was added to the combustion zone of the ATP. This indicates that the addition of
 soda ash  provides no reduction in the flue gas emissions produced by the SoilTech
 ATP  System.  Further  bench testing might, however, provide a more statistically
 conclusive evaluation.

 Processing of  soils and sediments at the Waukegan Harbor site  continued without
 further incident and was successfully completed on  June 23, 1992 when a total of
 12,700 tons  had been processed.  A summary of production data  is provided in
Table 3. Approximately 30,000 gallons of PCB-contaminated oil was desorbed from
the soils and sediments  and returned to the PRPs for subsequent off-site disposal. An
average plant throughput of 8 tph had  been achieved with a mechanical availability
of 85 percent.

Process water generated  during operations was discharged to  Canonie's  water
treatment  system at approximately  5  gpm and released to the sewer at a PCB
concentration of less than 15 ppb PCBs.
                                     108

-------
5.0  CONCLUSIONS

The Waukegan Harbor Superfund Project was the SoilTech ATP Technology's greatest
challenge to date due to the high concentrations of PCBs in the feed material. System
modifications allowed the strict project performance criteria to be met and surpassed.
The SoilTech ATP Technology has demonstrated the capability to:

      •  Provide a significant and permanent reduction in the toxicity and volume of
         PCB wastes at  Waukegan Harbor,
      •  Reduce PCB concentrations in the soils and sediments from as high as 2.3
         percent to  below 2 ppm,
      •  Achieve a 7-nines PCB ORE in the stack gases,
      •  Meet or exceed all project performance criteria, and
      •  Provide a cost-effective  and more environmentally favorable alternative to
         established technologies.
                                     109

-------
References
Montgomery, A. H., C. J. Rogers, and A. Kernel, 1992, "Thermal and Dechlorination
Processes for the Destruction of Chlorinated Pollutants in Liquid and Solid Matrices",
AlChE 1992 Summer Annual Meeting, August 9-12. Minneapolis, Minnesota.

Vorum, M., 1991, "SoilTech Anaerobic Thermal Process (ATP): Rigorous and Cost
Effective Remediation of Organic Contaminated Solid and Sludge Wastes", AWMA
Conference. Kansas City, Kansas, June.

Vorum, M., and A.  H. Montgomery, "The Taciuk Process Technology for Anaerobic
Pyrolysis of Solid Wastes and Sludges". Conference on Hazardous Waste Research,
Kansas State University, Manhattan, Kansas.
                                    110

-------
                                 TABLE 1

                          STACK TEST SUMMARY
                           WAUKEGAN HARBOR
Date
01-28-92
02-04-92
02-11-92
02-18-92
03-04-92
03-05-92
03-05-92
03-18-92
04-09-92
04-10-92
05-12-92
05-13-92
05-13-92
05-14-92
06-02-92
06-02-92
06-09-92
06-16-92
Feed PCBs
(#/hr)
192.50
215.50
213.40
155.70
105.60
100.80
100.80
174.90
148.60
298.50
230.40
264.60
183.00
205.70
167.40
213.60
159.15
140.80
Stack PCBs
(#/hr)
0.0144000
0.0932000
0.0690000
0.0087600
i
0.0039700
0.0012700
0.0007720
0.0009890
0.0034500
0.0009430
0.0002000
0.0000735
0.0000464
0.0000448
0.0000942
0.0001850
0.0000432
0.0000050
PCB ORE
•(%)
99.9925195
99.9567517
99.9676664
99.9943738
99.9962405
99.9987401
99.9992341
99.9994345
99.9976783
99.9996841
99.9999132
99.9999722
99.9999746
99.9999782
99.9999437
99.9999134
99.9999729
99.9999964
NOTE:

1. PCB Destruction Removal Efficiency (ORE) criterion is 99.9999 percent.
                                    Ill

-------
                      TABLE 2
EPA SUPERFUND INNOVATIVE TECHNOLOGY EVALUATION
                  DRAFT RESULTS
                    Waste Feed
             Contaminant Concentrations
        Contaminant
   Total PCBs
   Dioxin/Furan:
        TCDF
        PeCDF
 Average Concentration
      9231 ppm

       86 ppb
       16 ppb
                    Treated Soil
             Contaminant Concentrations
        Contaminant
  Total PCBs
  Dioxin/Furan:
        TCDF
Average Concentration
     1.972 ppm

       5.4 ppb
                Waste Discharge Oil
             Contaminant Concentrations
        Contaminant
  Total PCBs
  Dioxin/Furan:
       TCDF
       PeCDF
Average Concentration
        32%

      136 ppb
       14 ppb
                       112

-------
                                 TABLE 2
          EPA SUPERFUND INNOVATIVE TECHNOLOGY EVALUATION
                             DRAFT RESULTS
                                (Continued)
                                Stack Gas
                        Contaminant Concentrations
       Contaminant
  Particulates
  Total PCBs
  Total TCDF
  HCL Gas
  Total Hydrocarbons
Average Concentration
   0.0039 jjg/dscm
   0.8330 //g/dscm
   0.0790 ng/dscm
   23.000 //g/dscm
         ND
 Output
 (Ibs/hr)
 0.071
 1.7E-5
 1.4E-9
0.00042
   0
Emissions
 Criteria
8.7 Ib/hr
30 ng/m3
0.2 Ib/hr
                           Stack Gas Emissions
                    Destruction and Removal Efficiencies
Test Run No.
1
2
3
4
PCBs Fed to ATP
(Ib/hr)
140.60
136.87
153.73
181.71
PCBs Exiting Stack
(Ib/hr)
18.9E-6
14.27E-6
17.15E-6
8.09E-6
ORE
(%)
99.999987
99.999990
99.999989
99.999996
NOTE:
1. Project emissions criteria for PCBs is 99.9999% ORE.
                                    113

-------
                                   TABLE 3

                             PRODUCTION DATA
                             WAUKEGAN HARBOR
Week
Ending
1-25-92
2-01-92
2-08-92
2-1 5-92
2-22-92
2-29-92
3-07-92
3-21-92
4-11-92
4-18-92
5-16-92
5-30-92
6-06-92
6-13-92
6-20-92
6-27-92
TOTAL
Production
(tons)
561
1,476
1,202
1,205
1,493
606
841
245
483
592
377
46
1,284
1,000
887
402
12,700
Average PCS in
Feed Soil
(ppm)
15,500
9,243
11,657
13,143
9,571
7,025
7,060
9,950
8,350
13,740
8,800
1 2,000
10,486
9,450
9,917
9,300
10,400
Average PCB
Removal
Efficiency (%)
99.97
99.95
99.98
99.98
99.98
99.98
99.99
99.97
99.97
99.99
99.99
99.99
99.98
99.98
99.99
99.95
99.98
NOTE:

1.     Soil treatment criterion is 97 percent removal efficiency.
                                       114

-------
MINIMAL
EMISSIONS
SYSTEM
   LOW TEMP. STEAM
   AND HYDROCARBON
   VAPORS FLOW
   FEED STOCKS
                                                                             SAND SEAL
                                                                             PRODUCT
                                                                             RECOVERY
                                                                             SYSTEM FOR
                                                                             RECYCLE
                                                                            HYDROCARBON
                                                                    — ——•*• AND STEAM
                                                                            VAPORS FLOW
COMPATABILITY
WITH MIXED
FEEDS

CONTINUOUS
PROCESS
SYSTEM
    SPENT SOLIDS
    TAILINGS
    KILN END SEALS (IYP.)
                                                                           AUXILIARY
                                                                           BURNERS
                                                                           COMBUSTION
                                                                           AIR FLOW
                    SINGLE STEP
                    FEED TREATMENT
                                            SELF-SUSTAINING
                                            HEAT SYSTEM
HEAT TRANSFER
MECHANISM
                                      FIGURE 1

                        SOILTECH  ATP  CROSS  SECTION

-------
STACK
             GAS
        TREATMENT
       PCB
   CONTAMINATED
       FEED
                   PREHEAT
                    VAPOR
ANAEROBIC
 THERMAL
PROCESSOR
                  TREATED
                  SOLIDS
                PCB
            CONTAMINATED
                OIL
                                 i
              VAPOR
            TREATMENT
AIR
                                  FUEL
         WATER
                    WATER
                  TREATMENT
                            NON
                            CONDENSABLE
                            GAS
ORGANIC
 VAPOR
                       TREATED
                        WATER
                                         SOUR
                                         WATER
                     OIL
                  RECOVERY
  PCB
PRODUCT
                        OIL
                                 FIGURE 2
      SOILTECH ATP  TECHNOLOGY SIMPLIFIED FLOW  DIAGRAM

-------
                                 FLUE GAS

                                 SCRUBSfR
                   T-400TF
J     I  T-«T,
1 FUE C*S I
I BACHOUSE I




vvv
                                 CYCUONE
T-100O"F
r~





SCRUBBER
CONDENSED

r ^
$


TMUNGS PILE
                          FIGURE 3


    SOILTECH  ATP TECHNOLOGY PROCESS FLOW SHEET

-------
00
                5 TPH
          o
          3
             12
             10
8
                  MOST  ECONOMICALLY  FAVORABLE ATP
                       10 TPH 	                 25 TPH
          O
          O
          UJ
          o:
          a.
             0
                                           T
                                       T
           .SEE    MOB/DEMOB  $AON
           5 TPH   $500.000    $300
           10 TPH  $1.000.000   $200
           25 TPH  $3,250.000   $150
                                                           • » -'   -  -. .  ",•••. • »•• : . t-
                                                              ..  -    »«•»*.   . .  •. •
                                                          - ..    >  ;,.-   • •,  .  r . . .
                        10
                   20
30
                                                    40
                                                50
                            60
                               QUANTITY OF CONTAMINATED MATERIAL
                                      (THOUSANDS  OF TONS)
                                                                                70
                                             FIGURE 4
                        EVALUATION  OF  OPTIMUM  SIZE ATP UNIT

-------
              THE  BABCOCK & WILCOX CYCLONE VITRIFICATION
                   TECHNOLOGY  FOR CONTAMINATED  SOIL
 J. M.  Czuczwa,  H.  Farzan, W.  F. Musiol, J. J. Warchol, S. J. Vecci
   Babcock & Wilcox R&DD, 1562 Beeson Street,  Alliance, Ohio 44601


INTRODUCTION

Technology Development at Babcock & Wilcox

     The Babcock & Wilcox cyclone furnace is a well-established
design  (over 26,000 MW installed electrical capacity) for the
combustion of high inorganic (ash)  coal.  The combination of high
heat release rates  (450,000 Btu/cuft for coal) and high turbulence in
cyclones assures the high temperatures required for melting the high
ash fuels.  The inert ash exits the cyclone furnace as a vitrified
slag.

     Taking advantage of the ability of the cyclone furnace to form a
vitrified slag from waste inorganics, the cyclone furnace was used in
a research and development project to vitrify municipal solid waste
(MSW) ash containing heavy metals.   The cyclone furnace produced a
vitrified MSW ash which was below EPA leachability limits for all 8
RCRA metals.  The successful treatment of MSW ash suggested that the
cyclone vitrification technology would be applicable to high
inorganic content hazardous wastes and contaminated soils that also
contain organic constituents.  These types of materials exist at many
Superfund sites, as well as sites where petrochemical and chemical
sludges have been disposed.  Our approach for establishing the
suitability of the cyclone vitrification technology relies on the
premise that for acceptable performance in treating contaminated
soils containing organic and heavy metals constituents, the cyclone
furnace must melt the soil matrix while producing a non-leachable
slag and must achieve the destruction and removal efficiencies (DREs,
currently 99.99%) for organic contaminants normally required for RCRA
hazardous waste incinerators.  The high temperature  (>2,500 to
3,OOOoF), turbulence, and residence time in the cyclone and main
furnace are expected to result in high organics destruction and
removal efficiencies (DREs).

Process Description

     The Babcock & Wilcox four to six million Btu/h cyclone furnace
located in Alliance, Ohio, was used to perform all pilot-scale
vitrification tests.  The furnace is water-cooled and simulates the
geometry of B&W's single cyclone, front-wall fired cyclone coal-fired
                                119

-------
 boilers.'  This cyclone facility has  been proven  to  simulate  typical
 full-scale cyclone units  in regard to furnace/convection  gas
 temperature profiles  and  residence times,  NOX levels,  cyclone
 slagging potential, ash retention in the slag, unburned carbon  and
 flyash particle size.

      The pilot cyclone furnace,  shown in Figure  1,  is  fired  by  a
 single,  scaled-down version of  a commercial  coal combustion  cyclone
 furnace.   The furnace  geometry  is a  horizontal cylinder  (barrel).  A
 summary of the process is illustrated in Figure  2.  For the  present
 application,  natural gas  and preheated primary combustion air  (820«F)
 enter tangentially into the cyclone  burner.  In  dry soil  processing,
 preheated secondary air (820QF),  the soil  matrix, and  natural gas
 enter tangentially along  the cyclone furnace barrel.   For wet soil
 processing,  an atomizer is  used to spray the soil paste directly into
 the  furnace.   The  soil is captured and melted and organics are
 destroyed in the gas phase  or in the molten  slag layer formed and
 retained on the furnace barrel  wall  by centrifugal  action.   The soil
 melts,  exits  the cyclone  furnace from the  tap at the cyclone throat,
 and  is  dropped into a  water-filled slag  tank where  it  solidifies.  A
 small quantity of  soil also exits as flyash  with the flue gas from
 the  furnace and is collected in a baghouse.  In  principle, this
 flyash could be recycled  to the furnace  as indicated in Figure  2 to
 increase the capture of metals  and to minimize the  volume of the
 potentially hazardous  waste stream.

      Particulate control  is achieved by  way  of a baghouse.   To
 maximize the  capture of metals,  a heat exchanger is used  to  cool the
 stack gases  to approximately 2000F before  entering  the baghouse.
 Although the  cyclone facility is  equipped with an acid gas scrubber,
 it was  not used for these tests  because  acid gas generation  (e.g.,
 HC1)  from the vitrification of  the SSM was expected to be low.

 Applicable Wastes  and  Soils and  Possible Technology Configurations

     An  advantage  of vitrification over  other thermal  destruction
 processes  is  that  in addition to  the  destruction of organic
 constituents,  the  resulting vitrified product captures and does not
 leach heavy metals or  radionuclides.   The cyclone vitrification
 technology would be applicable to high inorganic content  hazardous
 wastes, sludges  and contaminated  soils that contain heavy metals and
 organic constituents.    The wastes may be in the  form of solids, a
 soil slurry  (wet soil)  or liquids.  To be treated in the  cyclone
 furnace, the  ash or solid matrix must melt and flow at cyclone
 furnace temperatures (2400 to 3000QF).  Because of the technology's
 ability to capture heavy metals in the slag and render these non-
 leachable, an important application of the technology  is  contaminated
 soils which contain non-volatile radionuclides  (e.g.,   strontium,
transuranics).

     The cyclone furnace can be operated with gas,  oil or coal as the
 supplemental  fuel.   The waste may also supply a significant portion
of the required heat input.  Additional air pollution control
devices, such  as NOx reduction technologies,  can be applied as
                                120

-------
                 STACK PARTICULATE
                 SAMPLING LOCATION
               CONTINUOUS EMISSIONS
               MONITOR (CEM)
               SAMPLING LOCATION
           SSM FEED
           SYSTEM
     SSM
     SAMPLING
     LOCATION
      SLAG AND
      QUENCH
      WATER
      SAMPLING
      LOCATION
                                 ID FAN
                           SCRUBBER
                           (NOT IN USE)
                                                                FURNACE
                                                                STACK
                                              SLAKER
                                              (NOT IN USE)
COMBUSTION
AIR
                                            NATURAL GAS
                                           [INJECTORS
                                                     NATURAL
                                                     GAS
                                                    SOIL
                                                    INJECTOR
                INSIDE FURNACE
                 SLAG
                 TRAP
                                           \CYCl
          SLAG    ^CYCLONE
          SPOUT    BARREL
                                SLAG
                                QUENCHING
                                TANK
FIGURE l.  The pilot cyclone test facility at the fi&bcock & wilcox Research and Development
         Division.  The insert of the furnace shows the wet  soil feed configuration.
                                      121

-------
          Cyclone Vitrification Process
                                                         STACK
                                            EMISSION

                                            CONTROL

                                            DEVICE
FIGURE 2.
S^^S^^^l^tS1tS"firn2: J
possible, but was not demonstrated during this p?Sjfct
                                                               devices
                                                           metals capture is

-------
needed.  An acid gas scrubber would be required, for example, when
chlorinated wastes are treated.

Description of the B&W SITE Emerging Technologies Project

     The EPA Office of Solid Waste and Emergency Response and the
Office of Research and Development established a formal program
called the Superfund Innovative Technology Evaluation (SITE) Program
to accelerate the development and use of innovative cleanup
technologies at hazardous waste sites across the country.  This
project was sponsored under the SITE Emerging Technologies Program.

     The goal of the SITE Emerging Technologies project, is to
perform a pilot-scale test of the Babcock & Wilcox 6 million Btu/h
cyclone furnace for vitrification of an EPA-supplied synthetic soil
matrix (SSM) spiked with three heavy metals (lead, cadmium and
chromium).

     This paper will present the results of both the Phase I (1990-
1991) and Phase II  (1991-1992) Emerging Technologies efforts.  Both
Phase I and II used an EPA synthetic soil matrix spiked with lead,
cadmium and chromium.  The most important goal for both Phase I and
II was to produce a vitrified soil that passes the Toxicity
Characteristic Leaching Procedure (TCLP) limits for lead, cadmium and
chromium.  Secondary goals were the measurement of volume reduction
after treatment and heavy metals mass balance.  The most significant
difference between the Phase I and II is the use of dry and wet soil
feed system, respectively.  The wet soil feed system was used for the
SITE Demonstration, and thus this paper emphasizes the results for
this final system configuration.

     Measurement of organics destruction efficiencies, thought to be
less of a technical challenge compared with metals capture, was
reserved for a SITE Demonstration performed in November 1991.

The SITE Demonstration

     A SITE demonstration was performed on the pilot cyclone furnace
in November of 1991.  An EPA-supplied synthetic soil matrix spiked
with heavy metals  (cadmium, chromium and lead), organics  (anthracene
and dimethyl phthalate), and simulated radionuclides  (cold strontium,
bismuth, and zirconium) was used.

     The demonstration was designed to test the performance of the
cyclone furnace in  (1) producing a non-leachable slag for metals and
simulated radionuclides;  (2) destruction and removal efficiencies for
the organics,  (3) producing a greater than 10 to 1 ratio of slag to
flyash;  (4) capturing at least 60 percent of the least volatile
metals; and  (5) reducing the volume of the treated waste by at least
25%.  Results  from  the demonstration are expected to be published by
EPA in fall of 1992, but preliminary results are summarized here.

MATERIALS AND METHODS
                                123

-------
 Typical Run Conditions

      Typical run conditions for the Phase I and Phase II tests are
 given in Table I, below.  The SITE Demonstration was run under
 conditions similar to those of Phase II.

      Table I.  Typical Cyclone Furnace Tests Conditions.
      Condition
Typical Range of Values
      Heat Input (natural gas fuel)
      SSM Feed Rate
      Excess Oxygen
      1°  and 2° Air Temperature
      Total Air Split-Phase I
           Primary Air
           Secondary Air
           Feed Air
      Total Air Split-Phase II
           Primary Air
           Secondary Air
           Soil Atomizer Air
      Slag Temperature
      Cyclone Furnace Gas Temperature
      Flue Gas Exit Temperature
      Baghouse Temperature
       5 Million Btu/h
       50 to 300 Ib/h
       1.0%
       8300F

       25%
       72%
       3%

       not used
       96.5%
       3.5%
       2370-24600F
       2800-3000QF
       2100-2200QF
       200°F
RESULTS: SITE DEMONSTRATION VENDOR'S CLAIMS

     The effectiveness of the B&W Cyclone Furnace Vitrification
Technology at _ destroying organics and immobilizing heavy metals and
simulated radionuclides in a non-leachable slag was evaluated during
the SITE Demonstration.  To perform this evaluation, the following
critical and non-critical Vendor's Claims were developed by Babcock &

demonstration     '^ ^ ^ U'S* *™' *" 6Valuation in the
     These claims are compared with performance data for the cyclone
vitrification technology below.
                               124

-------
Parameter

TCLP
Slag to Flyash Ratio
Non-Volatile Metals
Capture (Cr) in the
Slag
Volume Reduction
DREs
CO, THC, Particulates
Claim
Critical Claims
Produce a vitrified slag that does not
exceed Toxicity Characteristic Leaching
Procedure (TCLP) regulatory levels for
cadmium (i.e., <1 mg/L), lead (<5
mg/L), and chromium (<5 mg/L).
Achieve at least a 10 to 1 ratio (dry
weight basis) of slag to flyash.
Capture at least 60% (by weight) of the
non-volatile metal chromium from the
dry, untreated SSM in the vitrified
slag.
Achieve at least a 25 percent volume
reduction in solids when comparing
product solid to untreated SSM.
Achieve a 99.99% destruction and
removal efficiencies DREs for each
organic contaminant spike (anthracene
and dime thylphthal ate ) .
Comply with emission limits for CO,
total hydrocarbons ( THC ) , and
particulates from the stack as
stipulated by 40 CFR 264 (i.e., CO of
<100 ppm, THC of <20 ppm, and
particulates of <0.08 gr/dscf at 7%
oxygen ) .
Non-Critical Claims
ANS 16.1 Simulated
Radionuc 1 ide
Leachability
Non-Volatile
Radionuclide Capture
in the Slag
Produce a slag that immobilizes (passes
leaching standards ) radionuclides as
measured by the American Nuclear
Society test (ANS) 16.1 (i.e., ANS 16.1
calculated leachability index (LI) >6).
Capture at least 60% (by weight) of the
non-volatile metals strontium and
zirconium in the vitrified slag.
COMPARISON OF PERFORMANCE RESULTS FROM THE TWO SITE EMERGING
TECHNOLOGY PROJECTS WITH THE VENDORS CLAIMS

Synthetic Soil Matrix and Feed Conditions
                               125

-------
     Two Superfund  Innovative Technology Evaluation  (SITE) Emerging
Technology projects were conducted prior to the SITE Demonstration.
These two projects, Phase I and Phase II, were conducted to establish
the feasibility of  the cyclone vitrification process for dry soil
(Phase I) and wet soil  (Phase II) treatment.  In each project,
measurements were made to evaluate TCLP leachabilities, volume
reduction, and materials and heavy metals mass balances.

     A synthetic soil matrix formulated by EPA was used for all
cyclone testing.  Both clean and spiked SSM were obtained from the
EPA Risk Reduction  Engineering Laboratory (RREL) Releases Control
Branch in Edison, NJ.  SSM, used by EPA for treatment technology
evaluations, has been well-characterized in previous studies [ 1 ].
Clean soil was used for furnace conditions optimization.  The spiked
SSM used in the Emerging Technologies projects contained 7,000 ppm
(0.7%) lead, 1,000  ppm (0.1%) cadmium, and 1,500 ppm (0.15%)
chromium.

     In Phase I, dry SSM was processed at feed rates of 50 to 150
Ib/hr.  In Phase II, wet SSM was processed at feed rates of 100 to
300 Ib/hr (dry basis).  Approximately 11 tons of spiked and unspiked
SSM were processed  during each of the two project Phases.

Performance Results

     A comparison of Phase I and II results against the Vendor's
Claims developed for the Demonstration is presented below.  Not all
of the Demonstration claims were tested during these projects (e.g.,
DRE, ANS 16.1 were omitted).  All claims tested were met or exceeded
during these Emerging Technology projects (the Claims were finalized
on the basis of these results).
Parameter
TCLP-Cadmium
TCLP-Lead
TCLP-Chromium
Slag to Flyash
Ratio
Non-Volatile Metal
(Cr) Capture in the
Slag
Volume Reduction
Performance
Criterion
in Vendor's
Claim
1.0 mg/L
5.0 mg/L
5.0 mg/L
10:1
60%
25%
Performanc
e Measured
in
Phase I*
0.13 mg/L
0.20 mg/L
0.11 mg/L
14.6:1
80-95%
35%
Performan
ce
Measured
in Phase
II*
0.07 mg/L
0.20 mg/L
0.04 mg/L
34:1
78-95%
25%
*Average results where several measurements were made.
                               126

-------
COMPARISON OF PERFORMANCE RESULTS FROM THE SITE DEMONSTRATION WJ2W
THE VENDORS CLAIMS

Synthetic Soil Matrix and Feed Conditions

     On the basis of the Phase I and II Emerging Technology projects,
Babcock & Wilcox was asked to perform a SITE Demonstration.  For the
Demonstration, a wet SSM was used.  Demonstration goals included the
vendor's claims given above.

     The SSM used in the SITE Demonstration contained 7,000 ppm lead,
1,000 ppm cadmium, and 4,500 ppm chromium; 6,500 ppm anthracene;
8,000 ppm dimethylphthalate; and the three simulated radionuclides:
4,500 ppm bismuth, 4,500 ppm strontium, and 4,500 ppm zirconium.  The
rationale for B&W's choice of radionuclide surrogates is as follows:
Bismuth was used as a surrogate for volatile radionuclides important
at DOE/DOD sites such as cesium (cold cesium was originally proposed
but found to be excessively expensive).  Cold strontium was used as a
surrogate for radioactive strontium (the cold version of the
radionuclide is the best possible surrogate).  Zirconium was
considered an excellent surrogate for radioactive thorium and uranium
from the standpoint of both volatility and chemical behavior (all are
oxophillic and tend to be in the +4 oxidation state).

     A total of 3 tons of SSM were processed during the Demonstration
at a feed rate of 170 Ib/hr.


Performance Results

     A comparison of the Demonstration results against the Vendor's
Claims developed for the Demonstration is presented below.  All
claims tested were exceeded during the Demonstration»

SUMMARY

     The Babcock & Wilcox 6 million Btu/hr pilot cyclone furnace met
or exceeded all critical and non-critical Vendor's Claims.  Because
these performance results were measured on a pilot cyclone furnace
configured as a utility boiler, and by no means optimized for soil
vitrification, a unit designed for dedicated soil vitrification may
improve process performance and throughput.


REFERENCE

1.   P. Esposito, J. Hessling, B. Locke, M. Taylor, M. Szabo, R.
     Trumau, C. Rogers, R. Traver, and E. Earth, "Results of
     Treatment Evaluations of a Contaminated Synthetic Soil," JAPCA,
     39: 294  (1989).
                               127

-------
Parameter
TCLP-Cadmium
TCLP-Lead
TCLP-Chromium
Slag to Flyash
Ratio
Non-Volatile Metal
(Cr) Capture in the
Slag
Non-Volatile Metal
(Sr) Capture in the
Slag**
Non-Volatile Metal
(Zr) Capture in the
Slag**
Volume Reduction
DRE-Anthracene
DRE-
Dimethylphthalate
CO
THC
Particulates
ANS 16.1
Leachability-
Bismuth**
ANS 16.1
Leachability-
Strontium**
ANS 16.1
Leachability-
Zirconium**
Performance
Criterion in
Vendor's Claim
1.0 mg/L
5.0 mg/L
5.0 mg/L
10:1
60%
60%
60%
25%
99.99%
99.99%
<100 ppm
<20 ppm
0.08 gr/dscf***
LI > 6
LI > 6
LI > 6
Performance
Measured
in Demonstration*
0.12 mg/L
0.29 mg/L
0.30 mg/L
15.6:1
76%
88%
96%
28.1%
>99.996%
>99.998%
4.8-54.1 ppm
<5.9-18.2 ppm
0.001 gr/dscf***
LI = 13.4
LI = 13.1
LI = 8.7
*Average results where several measurements were made.
**Non-critical parameter.
***Corrected to 7% oxygen.
128

-------
  THE  ECO LOGIC  PROCESS
          D.J. HALLETT, Ph.D., President

       K.R. CAMPBELL, P.Eng., Vice President

       ELI ECO LOGIC INTERNATIONAL INC.
143 Dennis Street
Rockwood, Ontario
Canada
NOB 2KO
2395 Huron Parkway
Ann Arbor, MI
U.S.A.
48104
                    129

-------
                              THE ECO LOGIC PROCESS
                                 Douglas Hallett, Ph.D.
                                Kelvin Campbell, P.Eng.
       The ECO LOGIC Process is a high-efficiency destruction alternative to incineration
which is particularly suitable for hazardous wastes with a  substantial water content.  The
process is  based on the gas-phase thermo-chemical reaction of hydrogen with organic and
chlorinated  organic  compounds.   At  850°C or  higher, hydrogen combines with organic
compounds  in  a  very  efficient reaction  known as  reduction to form smaller,  lighter
hydrocarbons, primarily methane.  For chlorinated organic  compounds,  such as PCBs, the
reduction products include methane and hydrogen  chloride.  This reaction is enhanced by the
presence of water, which can act as a reducing agent and a hydrogen source. Bench-scale,
lab-scale, and pilot-scale testing has  shown  that destruction  removal efficiencies (DREs)  of
99.9999% can be achieved.

       The main features of the patented ECO LOGIC Process include:

             1.  High destruction efficiency
             2.  No possibility of dioxin or furan emissions
             3.  Continuous monitoring and process control suitability
             4.  Suitability for aqueous wastes
             5.  Mobility
             6.  Reasonable cost

       The  high efficiency  of the  process has  been demonstrated  with  PCBs, PAHs,
chlorobenzenes,  and organochlorine pesticides.  During the demonstration testing of the pilot-
scale system at Hamilton Harbour in the summer of 1991, coal-tar PAHs at levels of 30% (dry
basis)  in the  harbour sediment were effectively eliminated.   DREs of 99.99999%  were
calculated (see Table 1), based on the total organic input and the PAHs analysed in the stack
emission.  During one test, the liquid waste  input was spiked with 500 ppm PCBs, with the
result that there were no detectable levels of PCB in the air emission, the processed solids,  or
the liquid effluents.  Based on detection limits for the stack sampling trains, a PCB DRE of at
least 99.9999% was achieved.
                                         130

-------
                        HAMILTON HARBOUR SEDIMENT



                        ANALYTICAL RESULTS SUMMARY



                            3% ORGANIC COAL TAR


Run
PI
P2
P3
P3

Target
Analyses
PAHs
PAHs
PAHs
PCBs
Waste
Influent
Cone.
(ppm)
21,000
30,000
30,000
500
Decant
Water
Cone.
(ppb)
483
680
423
ND

Grit
Cone.
(ppm)
1.67
7.76
0.37
ND

Sludge
Cone.
(ppm)
32.8
56.1
4.3
ND
Stack
Gas
Cone.
Oig/m3)
0.27
0.23
0.14
ND
Destructk
Removal
Efficienc
(%)
99.99999
99.99999
99.99999
99.9999
*   DRE = (Total Input - Stack Emissions) / (Total Input)

-------
       One of the distinct differences between the  ECO LOGIC Process and  incineration
processes is that there is no possibility of dioxin or furan formation with the reduction process.
These compounds can .be formed by the incomplete oxidation of PCBs,  but in an actively
reducing atmosphere with an  abundance of free hydrogen and no free  oxygen, the possibility
of their formation is eliminated.

       The use of hydrogen as an active reducing agent in the process also creates a product
gas  of low molecular weight without the formation of heavier hydrocarbons  common to
pyrolysis processes.  This product gas is suitable for continuous monitoring, and ECO LOGIC
has included in the process control design a very sophisticated on-line mass spectrometer which
can  monitor organic compounds continuously in the parts per billion range.  This allows the
operator and  the process control  system to  effectively  monitor  destruction efficiency  by
selectively analyzing for trace concentrations of known breakdown products of the hazardous
waste.

       Another feature of the ECO LOGIC  Process is its ability to process aqueous wastes.
Water does not inhibit, but actually enhances  the reduction process, and also provides a source
of hydrogen through the water shift reaction with methane.  After the destruction of the organic
contaminants in  the reduction reactor, the  water and the HC1 produced  are removed in a
scrubber and the clean, dry product gas can be recirculated back to the reactor or used as
supplementary fuel in the propane boiler.  At Hamilton Harbour, approximately  97%  of the
product gas was recirculated to take advantage of the hydrogen content, and about 3% was sent
to the boiler.   The  boiler produced steam which  was used to pre-heat the waste prior to
injection into the reactor.

       Other features of the  ECO  LOGIC Process are the high degree of mobility and the
reasonable operating cost.  The process is mounted on two standard drop-deck highway trailers
and is easily set up and taken down.  The relatively small size  and capital cost help keep the
operating cost per tonne fairly low.  The price for  processing high moisture-content matrices
such as contaminated soils, sludges, and sediments  is projected to start at $400 per tonne for
the first 100-tonne/day commercial unit.  High-strength PCB liquids can be processed  at the
same time as environmental wastes  with only a limited price increase.
                                         132

-------
  WASTES SUITABLE FOR DESTRUi

CHEMICALS:
Non-halogenated / halogenated biphenyls
Non-halogenated / halogenated benzenes
Non-halogenated / halogenated phenols
Non-halogenated / halogenated cycloalkanes
Non-halogenated / halogenated alkanes
Non-halogenated / halogenated dioxins
Non-halogenated / halogenated dibenzofurans
Polyaromatic hydrocarbons

* Note:  Halogenated means: Chlorinated
                          Brominated
                          Fluorinated

TYPICAL WASTES;
PCBs
Pulp mill wastes
Chlorinated solvent waste
Contaminated coal tars
Solvent still bottoms
Chlorophenols / Wood treatment waste
Pesticide wastes
Landfill leachates
Lagoon bottoms
                           133

-------
       The ECO LOGIC Process has been accepted into the US EPA Superfund Innovative
 Technology Evaluation (SITE) Program, following the successful demonstration at Hamilton
 Harbour last summer.  A demonstration program  in the autumn of 1992 at a landfill site in
 Michigan will be funded by ECO LOGIC,  Environment Canada, and Environment Ontario,
 and the evaluation will be performed by the US EPA.  The site is contaminated by a dense
 oil containing approximately 40% PCBs.  This oil will be processed along with contaminated
 groundwater and contaminated soil, in separate streams, as in Table 2.

       ECO LOGIC began operation in 1986 when Dr. Hallett proceeded to patent the gas
 phase  thermo-chemical  reduction process for the destruction of high hazardous organic
 chemicals.  Research has led to subsequent inventions in on-line continuous emission/process
 monitoring using chemical ionization mass  spectrometry, and computerized  process  control
 systems and software for hazardous waste. The company's mission is to manufacture, supply,
 operate, and service high technology for the destruction or control of hazardous chemicals and
 the remediation of hazardous waste landfills, lagoons, soils, and sediments.

       ECO LOGIC will enter the market to supply hazardous waste destruction services itself
 with its own machines,  and to license and sell equipment to companies that already supply
 services, or to licence and sell  to large chemical producers  or  users where ownership is
 economically advantageous. The  service market is the initial focus in order to demonstrate the
 machine and obtain approvals in jurisdictions where units will be licensed and sold when the
 technology has buyer (versus user) acceptance.

       We would encourage interested parties with organic hazardous waste problems to contact
 our offices in Rockwood and Ann Arbor.
       This offers  potential  clients  the  opportunity to obtain direct information on  the
application of this technology  to the resolution of their hazardous waste problems.
Wayland R. Swain, Ph.D.
Vice President
U.S. Operations
ELI Eco Logic International Inc.
2395 Huron Parkway
Ann Arbor, Michigan 48104
313-973-2780
Jim Nash, B.A.
Manager
Sales and Business Development
ELI Eco Logic International Inc.
143 Dennis Street
Rockwood, ON  Canada  NOB 2KO
519-856-9591
                                        134

-------
     During September and October, 1992, The U.S. Environmental Protection Agency, in
co-operation with the City of Bay City, Environment Canada, and the Ontario Ministry of the
Environment will sponsor a demonstration and evaluation of the EU Eco Logic Process at Bay
City's Middleground Landfill. This demonstration will occur under the auspices of the U.S.
EPA's Superfund Innovative Technology Evaluation (SITE) Program.
    HAZARDOUS WASTE DESTRUCTION DEMONSTRATION
                      BAY CITY, MICHIGAN
     OILY WATER  (7 T/DAY)
     5 kg/min contaminated water (TCE)
     .05 kg/min PCB oil (40%  PCBs)
     (feed concentration of4000ppm PCBs)
1.
•
•
2.   SOIL  (15 T/DAY)
•    10 kg/min contaminated soil (1000 ppm PCBs)
•    1  kg/min  contaminated water (TCE)

3.   OIL/WATER (3 T/DAY)
•    1  kg/min  PCB oil (40% PCBs)
•    1  kg/min  contaminated water (TCE)
     (feed concentration 20% PCBs)

4.   72-HOUR RUN OF  1 OR MORE OF 3 CASES ABOVE
                              135

-------
                              SM
                BioGenesis   Soil Washing Process
                        Technology  Summary

GENERAL D^CRIPTION
°   The BioGenesis   soil remediation technology is capable of extracting volatile and
    non-volatile oils including crude oil, chlorinated hydrocarbons, pesticides, and other
    organics from most types of soil and clay.  A single "Soil Remediation Unit," using
    a complex bioremediating surfactant and  water, washes 25 to 30 tons of soil an
    hour, depending on the pollutant, contamination level, and soil type.

DEVELOPMENT STATUS
°   Being introduced to the United States as part of the EPA's Superfund Innovative
    Technology Evaluation (SITE) Program.
n   First commercial use hi October, 1992, removing crude oil from contaminated soil
    at mid-west refineries; work in progress to complete cleaning of over 5,000 cubic
    yards by early 1993.

SPECIAL FEATURES
°   Wastes are reduced to recyclable oil, treatable water, and active soil suitable for
    onsite backfill.
°   Treats both volatile and non-volatile oils.
°   Complex surfactant used is safe, 100% rapidly biodegradable, and has low toxicity.
n   Processing rate is 25-35 tons per hour for one washer unit.
n   Organics do not combine with the surfactant.
n   Surfactant enhances the biodegradation of residual contamination.
°   Potentially effective on broad range of chlorinated hydrocarbons, pesticides, PCBs,
    and other organic pollutants.
°   Treats soils with high clay content and heavily weathered soils.
°   No air pollution except that connected with excavation.
n   No toxic by-products produced besides the contaminant removed.

PLANNING FACTORS
°   Variables.  Contaminant, amount of contamination, soil type, job size, and cleanup
    target.
°   Effectiveness. Heavy hydrocarbons, one wash-95%, two washes-99%+; light
    hydrocarbons, one wash-97 to 99%. Residuals polished through biodegradation.
°   Cost $40-180 per ton depending on the variables; average range $60-100 per ton.

EQUIPMENT & MANPOWER
n   Equipment.
       Earth moving equipment.
       Truck mounted washer unit, 20 cubic yards capacity.
       Hydrocyclones, centrifuge, and flotation units for the separation of fines.
       Gravity separators and coalescing filters for oil water separation.
n   Manpower. The mobile system is operated by five personnel: a system supervisor, a
    test director, two operators, and a materials handler.

 __	BioGenesis Enterprises, Inc.
 CHICAGO, ILLINOIS' TEL (708) 827-0024 • WASHINGTON, DC' TEL(703)250-3442
                                  136

-------
FULL SCALE FIELD RESULTS
3% CRUDE OIL CONTAMINATION, MINNESOTA REFINERY,
ONE WASH, METHOD 9073 TESTING, mg/fcg
- - Soil Fraction
Large Particles > 40 Mesh
Fines < 40 Mesh
TRPH Level
Before
30,800
30,800
TKPH Level
After
1,620
4,520
Extraction
Percent, One Wash,
95%
85%
BIODEGRADATIONOF RESIDUAL CONTAMINATION AFTER
, ONE WASH, METHOD 9073 TESTING, mg/kg
/, f~' ^ s
' Days After Washing
Zero
Seven
Fourteen
TRPH, Fines
< 40 Mesh
4,550
2,200
1,300
* N Cumulative
Effectiveness, Fines
85%
93%
96%
s BTEX REMOVAL EFFECTIVENESS, FINES < A
ONE WASH, METHOD 8920 TESTING, m
Contaminant
Benzene
Toluene
Ethylbenzene
Xylenes
Total BTEX
Unwashed
Soil
<0.24
<0.25
1.2
4.3
5.99
Fines
< 40 Mesh
<- ?"•
< 0.024
0.034
< 0.016
0.073
0.147
K) MESH,
g/kg
Removal
Effectiveness,
Fines, One Wash
90.0%
86.4%
98.7%
98.3%
97.6%
SIEVE ANALYSIS, UNWASHED SOIL AND FINES
Percent Passing •»-»••
Unwashed Soil
Fines
10 Mesh
72
100
40 Mesh
18
94
120 Mesh
4
33
200 Mesh
3
11
CONTACT POINT
o  Charles Wilde, TEL: (703) 250-3442, FAX: (703) 250-3559
                            137

-------
           BioGenesis"" Soil Washing  Process
Contaminated Soli
                  Vacuum         Oil Skimmer
                  Hood\        / System
'\     /
 Treated Soli
                                      J.
Atmosphere
    Soil Washer
        Unit
    BioGenesis
      Cleaner
   Holding Tank 1
                                             ' Effluent from ClL
                                              Washer Unit
                                      Strainer
                                  Water Calmer
                                                      ^ Activated Carbon
                                                          Filters
                                                            Solids
                          Oil Skimmer
                                        Holding Tank 2
                                                     Solids
                                             Baffle
                                           Separator
                       Recycle to'" •
                      Washer Unit
                   Holding Tank 3
 Clean water to Disposal
                             138

-------
              8M
 BioGenesis  Soil Washing Process
              8M
The BioGenesis  process is designed to treat
soils and other types of solids contaminated
with organic compounds, including volatile
and non-volatile hydrocarbons. The treat-
ment is a two-stage process; first, extract
contaminants from soil using water and Bio-
Genesis  cleaner; second, accelerate the
biodegradation of residual soil contamination
and contaminant-rich wastewater.

The major components of the treatment
system include the following:

° Soil washing unit. This is the principal
  component of the treatment system. The
  unit holds 20 cubic yards, about 25 tons.
  The washer has a perforated bottom to
  introduce air for mixing and to drain
  wastewater. A poly-canvas hood with
  vacuum extraction covers the top of the
  wash unit to collect any organic compounds
  volatilized during treatment.
  Oil skimmers. Oil is skimmed from the
  surface of the soil and water mixture. Oil
  can be skimmed two ways. A manual
  method requires workers to skim oil from
  the top of the water to the Baffle Separator.
  A mechanical method uses rising water to
  push the oil/water into a system that runs
   through an elastometer belt. Oil clings to
   the belt while water flows through.

 Q Strainers. Strainers are located at the ends
   of the oil skimmer troughs on the wash
   unit. The strainers prevent floating debris
   from damaging the transfer pump.

 a Two 7.5-horsepower(hp) transfer pumps.
   These pumps are used to transfer waste-
   water from the wash unit to the baffle
   separator.

 ° Baffle Separator. This unit is used as a
   primary separator to separate oil from the
   waste water through a series of baffles.

 a Gravity Separator and Oil Coalesces
   These units are used as a secondary sepa-
   rators to separate the oil phase from the
   wastewater. The unit is equipped with an
   infrared (IR) detector.  The detector con-
   trols a diversion valve that depending on oil
   concentration in the water, either returns
   the water to the influent line and to the
   gravity separator or to the bioreactor.

a  Bioreactor. The bioreactor is a cylindrical
   shaped tank with a holding capacity of ap-
   proximately 5,000 gallons. At the end of
   operations, wastewater from the oil/water
   separators is transferred to the bioreactor.
   A specially formulated BioGenesis8"
   solution is added to the bioreactor to stimu-
   late biodegradation of residual contamina-
   tion in the wastewater. Within the
   bioreactor, water is mixed by pumping it
  through a spray aerator fitted above the
  liquid phase. The bioreactor is covered with
  an activated charcoal filter to minimize
  losses due to volatilization.

   One 48-foot flat bed trailer. This trailer
  houses a 200-ampere (amp), 480 VAC,
                                      139

-------
  cooled air compressors; a vacuum source;
  and a gravity separator and oil coalescer.
              SM
The BioGenesis  process begins by introduc-
ing approximately 1,000 gallons of water into
the wash unit. Next, contaminated soil is
added to the wash unit, normally with a front-
end loader. The wash unit is capable of
treating 20 cubic yards per batch.  If volatile
organic compounds (VOC) are present, the
wash unit is covered with a retractable canvas
cover. A positive air flow is drawn through
the back of the wash unit, creating a negative
pressure within the unit. Air flow is directed
to strip away any VOCs. Volatile emissions, if
any, are passed through a granular activated
carbon filter before being vented to ambient
air.
                    EM
Water and BioGenesis  cleaner are premixed
in a 4,800-gallon holding tank (Holding Tank
1) and pumped into the wash unit. A typical
wash requires approximately 2,500gallons of
water and 8 gallons of BioGenesis  cleaner.
The resulting slurry is agitated by a series of
aerators in the bottom of the wash unit. After
the soil slurry is mixed for a period of time, air
is turned off, and water is added to raise the
fluid level to allow floating oil product to flow
out of the unit via ports located 8 inches from
the top of the unit into a holding tank
(Holding Tank 2). After the floating product
is removed, the soil slurry is agitated again for
a period determined by the operator. The
fluid level is again raised to allow for oil/water
to be removed through the ports. Soil settles
to the bottom of the wash unit.
Wash water from the bottom of the wash unit
and oil/water exiting through the ports are
pumped to Holding Tank 2 which is equipped
with an oil skimmer. After the water has
drained from the treated soil, the operator
inverts one end of the wash unit, dumping the
soil onto the ground whence it is moved by
front loader to a clean soil holding area
pending test and final clearance.
In Holding Tank 2, the oily material removed
by the skimmer is pumped to 55-gallon drums.
Material not removed by skimming is pumped
to a baffle separator. The use of the baffle
separator is optional based on the operator's
discretion. Any oily material recovered from
                                        140

-------
 the baffle separator is pumped to 55-gallon
 drums. Water from the baffle separator is
 then directed to a 4,800-gallon holding tank
 (Holding Tank 3) for storage prior to reuse in
 the wash unit. Approximately 15 percent of
 the wash water is retained in the soil;
                                      SM
 therefore, make-up water and BioGenesis
 cleaner must be added to the recycled water as
 needed. Any make-up water for the next
 batch of soil is supplied from Holding Tank 1.
                          SM
 Soil washing, the BioGenesis " method,
 excavates soil with earth-moving equipment,
 separates contaminants in a washing process,
 and returns the soil to the excavated area. It
 positively removes the contaminant, works on-
 site, and produces no toxic byproducts except
 the contaminant removed.

 Historically, difficulties with washing
 chemicals, equipment efficiency, and inability
 to treat soils with more than 30% clay have
 impeded implementation of soil washing.
 BioGenesis  solves these problems with its
 complex surfactant mixtures and efficient
 washing equipment. BioGenesis8** soil
 washing is competitive with other technologies
 in almost all situations. The determining
factors of price are soil type, contaminant
 type, dirtiness, job size, and cleanup target.
 Once all runs are complete, the water in
 Holding Tank 3 is processed through the
 oil/water separation unit, which includes a
 ring chamber gravity oil/water separator and
 an oil coalescing filter. Water from the coales-
 cer is monitored by an infrared (IR) detector
 and is directed to a bioreactorif the oil
 concentration is below 10 ppm. If the oil
 concentration is above 10 ppm, the water is
 recycled through the gravity separator and
 coalescer until the oil concentration is below
 10 ppm. Oily material from the gravity
 separator and the coalescer is pumped to 55-
 gallon drums. Sediments from the wash unit,
 Holding Tank 2, the baffle  separator,and the
 bioreactbr are stored in storage bins and
 covered with plastic sheets. Samples from the
 sediment storage and treated soil storage are
 collected over a period of tune and analyzed
for chemical composition.
                                       141

-------
          Steam   Enhanced   Recovery
                               (SERF)
                          William R. Van Sickle
                             (310)536-6547
Process
   ABSTRACT

   In-situ  steam enhanced  recovery,  utilizes  formation  tailored injected
   low  pressure steam  to  establish temperature and  pressure  gradients
   to improve  the  migration  of  hydrocarbons through  the  soil. Vapor,
   free-liquid  hydrocarbons and  condensate  are removed using  specially
   designed   vapor  extraction  wells.    Steam enhances  removal   of
   hydrocarbons  and allows extraction of heavier petroleum products  by
   elevating   the  vapor  pressure  and  reducing the  viscosity, thereby
   inducing greater hydrocarbon  mobility.  The time required  to remove
   hydrocarbon  vapors  and liquids  is reduced  and  removal  efficiency  is
   increased,  with  consequent reductions in project cost.   Vapors and
   liquids  are treated at the surface  using  thermal oxidation  and  waste
   water  treatment respectively.
   INTRODUCTION

   Steam injection  remediation was a  diversification  of  the Enhanced Oil
   Recovery (EOR) process  in  use  by the  petroleum  industry since the
   1950's (Hong 1989).  Brought  to  the environmental industry  by the
   Dutch  firm  of Heidemij  Reststoffendiensten,  laboratory  studies,
   research  and  pilot  projects  in   the  United  States  (Hunt,  et  al,
   1988/1989)  followed.    As  a  result  of  available  data and  site
   requirements,  the  Steam Enhanced Remediation  Process  was  selected
   for use  at a  large  industrial site in Southern  California (Dablow,
   1991).   This paper addresses  unique aspects of the  project as well as
   provided an update  of  the current status.

   PROJECT  SITE

   The  Rainbow  project resulted  from  the  inadvertent  puncture  of a
   diesel fueling  manifold  during  construction activities.   Undiscovered
1240 Rosecrans Ave., P.O. Box 10011, Manhattan Beach, CA 90266-8511 • (800) 262-HESI • FAX (310) 536-5434
                                  142

-------
for about two  years, estimates  of 20,000 -  100,000+ gallons  of fuel
leached down  to  the  groundwater and  migrated offsite  to  the  north
and  west.  Sediments  of  the  contaminated  formation consists  of
several  feet of fill material, partially cemented sands,  and clayey silts
with lenses  of  fine to medium  sand.  The fuel  leak formed  a column
approximately   100  feet  in  diameter  that  migrated  downward,
spreading laterally as less  permeable mud's were  encountered.  Upon
penetrating   the   mud's,  downward  migration  continued,  stopping
where  ground  water  was encountered.

The  plume  of  greater than  two acres  was  located primarily under  a
trash  recycling  and   transfer   station.  Expansion  requirements
required  an accelerated  remediation,  as  did  the  need to  prevent
further  offsite migration.  A high  volume of traffic (500 -  700 trucks
per day/6 days,  two shifts) required  an  in-situ  process  that .did  not
impede  operations  or  interfere  with  traffic.  Construction  was
completed  and   all   operating  permits received     to commence
remediation  in  April  1992.
SYSTEM DESIGN AND DEVELOPMENT

System  design  was  based  upon  soil characteristics,  estimated soil
volume,   well  spacing  and   quantity,  and   temperature/pressures
necessary for vaporization.  The  design was selected  to  accelerate the
remediation  of    the  large area  impacted.  Extensive  sampling and
characterization  of the  site were  combined  with  bench  scale  testing
and  modeling  to determine  injection and  extraction  well  influence
radius  and well placement.   37  injection wells  were  placed  to totally
surround   the   plume   to   stop  offsite   migration  and   to   move
contaminant  towards  the  center of  the  plume.   Oversized  boilers
were specified  to provide the  capacity for  rapid heating  of  the entire
site.   39 extraction  wells  were designed  and  installed  for  the dual
purpose   for vapor  and  product/condensate  recovery.     It  was
anticipated that 90%  of the contaminant recovered  would  be  liquid
phase  which  would  be treated  by  an oily water  separator,  micron
filtration  system and polished by  activated  carbon.   Vapor treatment
utilized a packed bed  thermal oxidizer.  80%  of the injection  wells,
extraction  wells  and  system piping  were  placed  in  trenches and
covered.     All  work  on  the  trash  transfer  station   site  was
accomplished on  the  third  shift  Monday  through  Saturday  or  on
Sundays.
                                 143

-------
 PROJECT  STATUS

 Remediation is in full  swing.   Significant  delays  were encountered as
 a  result  of under  estimations of  the  permitting processes and  the
 need  to  throttle down key  injection wells while an  alternative for the
 underground fueling system  was  designed,  permitted  and installed.
 At  this point only  small traces liquid phase  hydrocarbons have been
 recovered.  Four thousand  four hundred  gallons of  free product were
 recovered and  fourteen thousand gallons have  been recovered by  the
 vapor extraction system  in  five months of effective system operation.
 Approximately  1200 CFM of vapor  are  flowing  into the  thermal
 pxidizer  with  Total  Petroleum  Hydrocarbons  (TPH)   gradually
 increasing from 700  ppm to a current level  of  about  3,000 ppm.  As  a
 result  more  than  two  thousand   gallons   per  month   are  being
 recovered.

 CONCLUSIONS

 The  Steam  Enhanced  Recovery  Process  appears to  be  effectively
 removing  diesel fuel from  the soil  and perched aquifer.   The  plume
 has  been  reduced  to 20-30%  of its original  volume.   System design
 has  allowed monitoring,  control  and  manipulation  of pressure  and
 vacuum  gradients  to  protect  underground tanks  and  to  focus  on
 specific  areas to be  remediated.   Monitoring wells have not detected
 unwanted  migration  of hydrocarbons  off site  or into  lower aquifers.
 To date no  interruption  of site operation  has  occurred.

 The use  of the steam  enhanced  recovery process  when   applied  to
 appropriate  substances,  offers  the  following  advantages  over other
 techniques:
            - Enhances removal  of volatile  and  semivolatile
                hydrocarbons
            - Effective for heavy hydrocarbons
            - Not limited by soil concentrations
            - Process  easily  monitored and  verified
            - In-situ  process with minimal surface  impact

      The  following  disadvantages  of steam enhanced  recovery
process should be  considered:
           - Capital  costs
           - A  cap may be  necessary for remediation on sites with
              high vertical  permeability surface conditions
           - Extensive  permitting maybe required
                                 144

-------
           -  Trained boiler technicians  may be needed as part  of the
             O&M team
REFERENCES

Dablow,  J.F.,  1991 "Steam Injection to Enhance Removal of Diesel
Fuel"   presented at HAZMACON 91,  Santa Clara, California, April 15-
17, 1991.

Hong,  K.C.,  1989  "Recent Advances  in  Steamflood Technology",.
represented  at   the   1989   Indonesian  Petroleum  association
Convention.

Hunt,  J.R. Sitar, N. and Udell, K.S.,  1988  "Non aqueous Phase Liquid
Transport  and  Cleanup", Water Resources Research, volume 24.

Udell,  K.S.,  and  Stewart,  L..D.,  1989  "Mechanism  of  In  Situ
Remediation of Soil  and Groundwater  Contamination by  Combined
Steam  Injection and Vacuum  extraction",  Paper No.  119d presented
at  the  Symposium  on  thermal  Treatment  of  Radioactive   and
Hazardous Waste  at  the  AICHE  annual meeting,  San  Francisco,
California, November 6, 1989.
                                  145

-------
  SOIL VENTING TO REMOVE DBCP FROM SUBSURFACE SOIL:
                              A CASE HISTORY

            M.B. Bennedsen, F. Gomez, C.R. Kneiblher, J.D. Dean, J. Scott
                            Woodward-Clyde Consultants
                                   ABSTRACT

 Dibromochloropropane (DBCP) was widely used  as a  soil fumigant  for the control of
 nematodes during the 1970s.  In 1979, the EPA restricted its use, primarily because of its
 toxicity to mammals.  In the early 1970s, a spill of about 3,100 gallons  of DBCP occurred
 at an agricultural chemicals distribution facility in the San Joaquin Valley of California.
 Groundwater at the site is at a depth of about 100 feet. The vadose zone soils are alluvial,
 moderately stratified, and vary from silty fine sands to coarse sands with some gravel. In the
 1980s DBCP was detected in the area groundwater,  and investigations in the vicinity of the
 spill found DBCP throughout the soil column above the water table.  Because of the depth
 of the soils contamination, and the existing facilities at the  surface, only in situ remedial
 technologies were considered feasible. DBCP has a melting point of 6.7°C and low vapor
 pressure at ambient soil temperatures.  However, the decision was made to try soil  venting
 for removing the vapor phase component and  possibly stimulating biodegradation of the
 liquid phase component of  the material.  The  system installed for the purpose has been
 operating since January 1992, and at this time has removed several hundred pounds of DBCP
 plus many thousands of pounds  of CO2 from the area soil.  The focus  of this paper is to
 provide additional information on the design  and operation of the installed soil venting
 system. The project is considered of interest to  this conference because of the nature of the
contaminant, the use of an on-line G.C. to provide a continuous and current record of system
performance, relatively low costs for operations and because it does not interfere with  on-
going facility operations.
                                     146

-------
During the 1970s an estimated 3,100 gallons, or about 60,000 pounds, of 1,2-dibromo, 3-
chloropropane (DBCP) were accidentally released onto the ground surface during a rail tank
car unloading operation.  The material is known  to be toxic to mammals.  The release
occurred at an agricultural chemicals distribution facility, in the Central Valley of California.
Soils in the area of the spill are alluvial, varying from  silty fine  sands to coarse sand with
some gravel,  and some stratification.   Soil samples, collected in  the  late 1980s from
throughout the vadose zone, contained  detectable DBCP concentrations.  Also, DBCP has
been detected in the area groundwater, which is at a variable depth, but generally in the range
of about 90 to 110 feet below the  ground surface.  Limiting or preventing further migration
of DBCP from the soil to the groundwater was clearly desirable.

Because of the depth of the soils contamination, and  the physical constraints created by
existing facilities surrounding the spill site, only in situ remedial techniques were considered
applicable. In consultation with the State of California, Environmental Protection Agency,
Department of Toxic  Substances Control, Alternative Technologies Division,  and  with the
concurrence of the California Regional Water Quality Control Board-Central Valley Region,
it was agreed that soil venting might be an appropriate technology in this instance. A short
term venting test (a few hours) demonstrated that venting could remove DBCP from the soils
at a significant rate. Based on the test results, and a proposed venting system design  that
would  prevent discharge of extracted DBCP to the atmosphere, the Department  of Toxic
Substance Control granted a Variance  From  Permit Requirements, for a two-month,  trial
program operation.  At the conclusion of the trial program operation, and based on the results
of the  operation,  an Extension of the Variance From Permit Requirements was granted to
allow an additional year of operation.  The venting process was considered experimental in
this application because the vapor pressure of DBCP, at ambient soil temperature, is about
1 mmHg, and near the lower limit generally considered appropriate for soil venting.

Three  soil borings had been  drilled in the  area  of the spill to assess the extent of the
contamination. The borings were located within  the areas of maximum contamination as
determined by a  soil gas survey. One of the borings was  completed as a groundwater
monitoring well.  For the short-term venting  test, three vapor extraction/soil venting wells
were installed with 2-inch-diameter PVC casing and with screened intervals extending from
20 to 40 feet, 40 to  60 feet, and 60 to 80 feet, respectively, below the ground surface.
Figure 1 presents simplified stratigraphic data for the soils penetrated by these wells, and data
on contaminant concentrations in  soil samples collected during the drilling of the wells.

The borings were on a line paralleling, and about 10 feet from the  rail spur on which the tank
car was  set when the spill occurred.  The in-line separations between the wells were about
13 feet  and 40  feet,  respectively.  The  borings were advanced using dual-tube reverse-
circulation drilling techniques, and the boring diameters were about 10 inches. Because of
the toxic properties of DBCP,. the  drill crew worked in Level B protective equipment — i.e.,
fully covered  and with supplied air.
                                       147

-------
 To make a preliminary assessment of the applicability of soil venting at this site,  and to
 obtain design information in case the technique was judged applicable, the short term test,
 mentioned above, was performed.  The test demonstrated that it was possible to extract at
 least 100 cubic feet per minute (cfm) of soil gas from each of the three wells, at an applied
 well head vacuum of about one inch, mercury gage (1 in. Hg).  Also, based on the soil gas
 extracted during the test, it was estimated that a production VES would initially extract a gas
 containing 10 to 15 parts per million, volumetric  (ppmV) DBCP. This extraction rate was
 considered sufficient to justify installation and operation of a production type soil venting
 system.  Again, it is noted that the short-term test was performed with the test crew utilizing
 Level B protective equipment.

 Based on the results of the short-term test, a  vapor extraction system was designed and
 installed at the site. The major system components, in the order of flow were: wells, valved
 manifold, air/water separator drum (knockout  tank), two carbon tanks in series, blower,
 silencer,  and discharge stack.   To  monitor  the  system  performance,  an on-line gas
 chromatograph (GC) was included.  The GC was equipped with an electron capture detector
 and an automatic, programmable sampling pump and a multiport sampling valve connected
 to the gas stream at the inlets to the first and second carbon canisters, the discharge from the
 second canister, and to the atmosphere as a reference point for system blanks. The carbon
 canisters were located upstream from the blower, where they would operate under negative
 pressure relative to atmospheric.  The objective was to prevent the possible  release of
 untreated, extracted soil gas to the atmosphere.

 In November 1989 the system was initially operated.  However, it operated for only about
 100 hours before excessive pressure drops occurred in the flow through the carbon canisters,
 and the lid  of the  number  two canister collapsed.   Investigation of the cause of failure
 revealed that ammonia salts had plugged the carbon. Low ambient temperatures — about 30°
 to 40°F — at the time of the operation, probably contributed to solids formation. During the
 initial operation, a strong ammonia odor was noticed in the gas discharged from the system.
 It is known that large quantities of ammonia based fertilizers are handled at this facility, and
 it is expected that leaks or spills thereof resulted in ammonia being present in the extracted
 soil gas.  Because all work associated with installing the VES wells, and performing the
 initial short-term test, had been done with the workers in Level B protective equipment, they
 had had no opportunity to  smell the presence of ammonia in the area soils, or in the gas
 extracted from the wells during the test.  In addition, the laboratory analyses on the associated
 soil  and  gas  samples were not  programmed  for,  nor capable  of, detecting  ammonia.
 Therefore, the initial system design included no provisions for accommodating ammonia.

 Following this initial failure, the system was modified by addition of an air stream heater,
 installed between the water knockout tank and the carbon tanks.  Regulatory approval of the
revised design was received and authorization to operate was based on the original Variance
From Permit Requirements.  A schematic diagram  of the system following the modification
is included in Figure 1.
                                     148

-------
         SCHEMATIC OF SOIL VENTING SYSTEM FOR
          REMEDIATING DBCP CONTAMINATED SOIL
                                                         TO
                                                       ATMOSPHERE
  t,
  UJ
  u.
  2
d j J WATER WATER
it * FILLED FILLED
WATER U-TUBE U-TUBE



800 '

9H '
43 ,
1300




__2'°°J
13'




560 '
k 260,000




. 2100'

	

1300
	 -t



280 ,
53' 	 ,


SANDY SOIL

MEDIUM -FINE '
SAND - NO SILT
SILTY FINE
SAND '
MEDIUM -FINE
SAND
SILTY SAND
SAND - MEDIUM'
TO COARSE
. 	
FINE SILTY
SAND
, 	 4
KNOCKOUT
60 GAL. CAP. HEATER 0 | | | |©| j | | @©
1 	 1 1 T 1 1 ,ITI, 1 .ITT
VALVED 1 1 * • 1 * •
MANIFOLD iC\f~\ 1 1

, ) , .bKiHI ItMP |l |2
3 - r. PVC PIPES 1G^SS CONTROLLER ^ ^

l/U»t? Ht3lkS //w/*f /«s«?
1


750 3 - t/8"0 S.S. TUBES
TO AUTO SAMPLING PUMP
AND ON-LINE GAS
J9 CHROMATOCRAPH WITH
A



BIOWER T t
^99 U

u&it*
. '




ELECTRON CAPTURE DETECTOR


'NO
__ — SCREENED INTERVAL




-- °F WELL (TYP) LEGEND




•T,-— SC SAMPLING COCK
MEDIUM ""

SAND VA VACUUM CAGE
PR PRESSURE GAGE
. F|NE SAND T TEMPERATURE GAC
                                                            CALIBRATED
                                                            ORIFICE PLATE
                                                            AND U-TUBE FOR
                                                            FLOW QUANIIFICATION
   40	/luuj .
   60	_.. _i .	J	
  100. _
         FINE TO COARSE
           SAND
                                                   750« OBCP CONCENTRATION
                                                      ON SOIL MQ/Kq BASED ON
                                                      SOIL SAMPLING AND ANALYSES
                                                      IN 1987, AND 1988
                ,-,
               , V
                 APPROX. W.S. - VARIES
(O
c

-------
 The addition of an electric heater, immediately downstream from the knockout tank, was
 intended to maintain the gas stream at a temperature above  the point at which water could
 accumulate or solids form in the carbon.  Also, the revised design included addition of water
 filled, U-tube manometers for monitoring the pressure drops through  the carbon canisters.
 Start-up of the reconfigured VES occurred January 14,  1992, and the operation has been
 essentially continuous since then. The balance of this report is devoted to the information
 learned from the continued operation.

                  PROPERTIES OF DEBROMOCHLOROPROPANE

 The compound l,2-dibromo-3-chloropropane (DBCP) formerly was widely used in California,
 in the fruit and vegetable growing industry, as a soil fumigant for the control of nematodes.
 However,  since 1979, its use has been restricted to Hawaii, for cultivation of pineapples only,
 primarily  because of its toxicity to  mammals.   Acute toxicity studies have identified the
 kidneys, liver, and testes as the primary target organs of DBCP. Other properties of DBCP
 are:
                     Melting point
                     Boiling point
                     Molecular weight
                     Specific gravity at 20°C/20°C
                     Solubility in water at 19°C
6.7°C
195.5°C
236.33
2.5
<1 mg/ml
 When used as a fumigant, DBCP is considered to have a half-life in the soil of about 2 years,
 but of course, this factor is a variable, depending in large measure on site specific conditions,'
 such as soil texture and climate.  Probably it is not appropriate to extrapolate this half-life
 value to concentrated releases, such as occurred at this site, and where the  material has
 penetrated tens of feet below the soil surface. It is likely that the material has been toxic to
 all or most soil bacteria, at least in the areas where it is or was most concentrated. However,
 it is noted that, based on data from investigations conducted at the site in the late 1980s, it
 was possible to account for only about 1,000 pounds of DBCP in the site soils. It is further
 noted that the time that has elapsed since the spill  occurred (about 20 years), the quantity
 spilled (about 60,000 pounds), and the residual contamination accounted for in the soil (less
 than  1,000 pounds) are qualitatively consistent values for a material having a half-life of 2
 to 3 years and a time frame corresponding to about 6 half-lives (64,000 -•- 26 = 1000).

                         PERFORMANCE MONITORING

To assess the performance of a vapor extraction system it is necessary to measure the
quantity of gas extracted from the soil and the concentrations of target compounds in the
extracted gas.  On this project, flow quantification is accomplished by monitoring differential
pressures across a calibrated orifice plate inserted  in the system discharge-to-atmosphere
stack.  Differential pressures are measured, in inches water column (in. H2O) using a U-tube
manometer connected to  the pressure taps provided with the orifice plate.  A calibration
                                      150

-------
curve, provided by the plate manufacturer, indicates flows, in cubic feet per minute, as a
function of differential pressures. The VES is equipped with a positive displacement blower
and operated at constant speed.  The only significant changes in flow occur as a result of
adjusting the control valves on the pipelines to the extraction wells and thereby the vacuum
against which the blower operates.  System flow to date has been within the range of 220 to
260 ambient cubic feet per minute (acfm). The estimated average flow is approximately one
million cubic feet every three days.

The concentration of DBCP in the extracted gas stream is considered the primary variable to
be monitored, for evaluating the performance of the VES. For this purpose, the system is
equipped with an on-line GC that periodically analyses the gas extracted from the ground.
For most of the operation to date, the system has been programmed to sample and analyze
the extracted gas  stream every eight hours.  In addition, as part of each monitoring cycle,
samples are collected and analyzed from the discharge of each carbon canister, to check for
DBCP saturation  in the  first carbon canister, and to document  that DBCP is  not being
released to the atmosphere in the gas discharged from the second canister.   Also, in each
analytical cycle, a sample of atmospheric air is analyzed to serve as a system blank.  In
Figure 2 is presented a copy of a typical strip chart produced by the GC during one complete
analytical cycle.   It may  be seen in the figure that a peak for DBCP appeared only in the
analysis of gas as extracted from the soil.

Backup checks to the data generated  by the on-line GC are provided by passing calibrated
flows of gas, from the same four  sources, through sampling tubes  containing granulated
activated carbon (GAC), and submitting the samples to an independent, certified laboratory
for extraction and  analyses.  These checks are routinely made each time the WCC project
chemist visits the project site to check and adjust the overall system operation (i.e.,  once
every two weeks).  During these site visits, calibration checks are run on the on-line GC. If
the calibration checks indicate that the results have drifted outside the range of acceptable
accuracy the unit  is recalibrated and  the integrator reprogrammed with  a revised response
factor.

Secondary constituents of the extracted gas stream that are routinely monitored are carbon
dioxide, oxygen and ammonia (CO2, O2, and NH3).  The concentrations of those three gases
are checked by the project chemist  during each  site visit.   The checks are made using
compound specific, color reactive cartridges (Gastec™) and a hand-held, calibrated, piston-
type  pump  (Sensidyne™) to draw  measured quantities of gas through the  cartridges, in
accordance with the manufacturer's specifications.   The concentration is  indicated by the
length of analyte in the cartridge that changes color, and is read directly from a scale printed
on the cartridge.
                                      151

-------
TYPICAL ON-LINE G.C. PRINTOUT
5TJ3TI
/ Carrier Gas
5-'s» < 5? /DBCP
f.
ST As extracted
from soil
RUM t 2645 SEP/16/92 82=49:36
UORKFILE 10= Al
UORKFILE HftHE-
ID- 1
ESTD
RT AREft TYPE CAL 1 AHOUIH
3.51 9779168 DSBB 6.866
3 74 748938 D PV 6.868
3 89 662346 D VE 6.866
4,86 44657 6V 6.666
4,87 37W56 PP 8.866
S.2&* 126888 PB 6 686
6.41 2 2S64E+87 SBB 1 1.P58
9.46 464868 PB 6.866
TOTAL AREP= 3.4992E+67
MUL FACTOR- 1 .6686E+68
?1 Discharge from
Carbon #1
RVH 1 2646 SEF/1C/92 H4:4%-3<;
UOftKFJLE /O- Al
UORKFUE HAKE
ID 1
NO CALT "-•:: FOUND
RT Af'EA TYPf Aft/HT ftREn^
3 7? "438«6 OTBD 6 881 4'288
1 68 I8.°5Ce D B6 6.664 1.072
TOTAL IfceAs 1 *£24E+fl7
MUI. FACTOR 1 e88eE+«6
5TJKT
-


"" 3.5*
ST Discharge from
Carbon #2
RUN i 2847 SEP/16/92 66=47=42
UORKFILE 10= AJ
UORKFILE NAME:
ID: 1
NO CAL IB PEAKS FOUND
AREA*
RT AREA TYPE Aft/HT AREiV<
3.56 953^466 DSBB 8.866 186.666
TOTAL AREA* 9535486
MIL FACTOR* 1.86eeE+68
•


s
ST Ambient Air
-Blank-
BAfflo. Al SEP/1^92 «'«'«
UORKFILE NAME:
ID: 1
NO CAL IB PEAKS FOUND
AREA*
RT AREA TYPE AR/HT AREA*
3.51 9433286 DSBB 6.666 186. 66f
TOTAL AREA= 9433286
KUL FACTOR= 1.6888E+b6
Figure 2
152

-------
                                DBCP REMOVAL

The concentration of DBCP in the extracted gas  stream is measured by the on-line gas
chromatograph.  During the course of the operation, the concentration has varied from about
1 to 13 ppmV. A typical printout from the GC is presented in Figure 2. Figure 3 is a plot
of the average daily DBCP concentration in the gas stream as extracted from the soil. The
plot of Figure 3 is based, generally, on 3 analyses per day.  Also shown in the figure is a
schematic representation of extraction history by well. Much of the variation in concentra-
tions may be  explained by the well, or combination of wells, being pumped.  Clearly, well
VES-1 — screened at 20 to 40 feet below the ground surface — is the most productive of
DBCP and  well VES-2 — screened at 60 to 80 feet below  ground surface — is the least
productive of the three wells in the system. But none of the wells in the system has exhibited
the type of concentration trend that might be expected for this type of operation (i.e., starting
at a relatively high value, and a uniform progressive  decrease in concentration). For example,
well VES-1 started production in mid-January at about 3 ppmV, and this concentration was
sustained for about the first month of operation. Then the concentration increased relatively
quickly (i.e., over about two weeks) by a factor of about 3, to an average of about 10 ppmV,
and this average was generally sustained for about the next two months. This was followed
by a relatively rapid decrease in concentration over the next two weeks — from about 10
ppmV to about 7 ppmV.  At that time (April 9) the system was shut down to make piping
changes to allow reversing the order of flow through  the two carbon canisters,  as would be
required following DBCP breakthrough in the first  carbon canister and carbon changeout in
that canister.  The system was restarted with extraction simultaneously from Wells VES-1 and
VES-3, and it was no longer possible to track the trend in DBCP concentration for well VES-
1 by itself.

Based on the DBCP concentrations data, summarized in Figure 3, the quantity of DBCP
extracted over time has been estimated (see Figure  4). For example, during the two months
trial program operation the estimated extraction was about 60 pounds  DBCP; the next three
months of production operation resulted in an additional  180 pounds DBCP extracted, and
during the  next  three  months of production  operation  a further 50 pounds DBCP were
removed from the ground. In summary, during the first nine months of operation a total of
about 310  pounds DBCP were  removed from  the  area soils.  It appears that the VES
operation to date has made substantial progress towards removing the approximately 800
pounds of DBCP that were estimated, in 1988, as being present in the area soils.

                 OTHER GASES IN EXTRACTED GAS STREAM

Approximate measurements (Gastec™ tube) of the concentrations of carbon dioxide, oxygen
and ammonia present in the extracted soil gas have been made at about two-week intervals
since the start of the VES operation. Values that are considered to be representative of the
measurements for the three  gases are as follows:
                                     153

-------
e am6y
DBCP CONCENTRATION IN EXTRACTED SOIL GAS. DAILY AVERAGE. PPMV
h
i
i
i
... !
- •
"1
i
:-
Weft;-
»M2i-
r |:
	 p—
i»:n
| .
r •
i " ;
i
'
... ..L.
•
• IT.".*
,
;
' i
S; :
il
•Months Trial
30«m OpolaUo
.. s. i ^
Lull!
• i ' '
J..i.i.'i „
j - 1 !
i -1 !•
...f..; ..!.._;„
... .-.;_ j „ _
.__ 	 :... 	
"
: i
: i
	 1. J 	
w: ......... ... .
; "-I
r '-I
i
•
-f-— i 	
i [ :
! • - i I
VES
cor
*
•% V
...
..
„._,_„_
"***"
—
~
'

-
	
i
..
-
-
i
-•-
-I—
1

-
-
--


-
—

-
-_
_.





—
. i.
owir'i
-

-
i-
H-i_
-HwiWe^fia^

-i. J
. 1

:1
-t
.
F



r—

»













v-
j.
n
-
-

-



-
,


-r

>
4
'it

I *1
t
^J.


-
—I
-




-





c
c
*•*
'.1
du

p
I
_L




-



-
-r


i:

)F
E
sti
bsi.

i
-

i
J.
I


..
-1

i

:|;
"•;
Notes 1 Blower shul off several tim*s. lor • lew minutes
each uns-not indicated in operating diagram
2. DBCP concentration data from on-Bne O C.
3 Total How horn we*s: 220 10 260 arobtonl
cubic feel per minuta
,.,...i.j,j...i j..j
"*" DATE
»E
N
ar
3f
•t
._

1


•

:l
1
»r
W
t
-

1
"I
I."
Jtt
	

T
-•!
4*

i.
ii
s!
II
S
A
=?
>
n
— :
-
._.
L
TING DIAC
AT1ON IN E
»•**
.-
•
7
—
-
::.

-••

r-.



.-.
! . .
...:.. h 	
irl
si:
i!




•
• • •• .

;;j
i
:19

i
i
:
-~r
.1
^L:Ll
92""



	 i...
i


.1
.L
"~T
r-:;



1:-




:•.
K i^r



• ;l
-
?~
.:;.
:...
=.-'
....


—

-
i:
'T •
::!.!
!..-. L
^






	
i i.

•'r


-f
"""-""
_._
r|

-•-
r*r










••ir.
....
Li.


."





..„

J
...








-

W
EX
i '2
;nxi>
i
• i
.1
4
n:;

'Ai
TF
MQu
at(Jn(
l
r
..j.
-••i
-
-
•-
"T": 	
i i '





•
_,

-
1
._|_j_.


-
. .. ,. .
j.
. 1
-
:
i



i i

T"




•"1"™."~1

VI AND DBCP
i-ACTED GAS
!.rt*r: .; . r . .:,.. • | j
JptfraJJoo .. .
" ' " =
' ' .
. . Li i .
! r -.
1 . r
: i : • s
I i 1 i
j i i i
	 i -;••;-;—
! i H
	 r ~ ","~
I
• ; '
— i -~-: •'• -
If1 : : •

i 1 M-
! j 1 :
; I ; •
. '*•„,.
; i
: i
: i
— - 	 i -
.j
i
. ! '.:. . . . L.
'.

1
. . ....

! ' !
• ; -\ • I'-
i ; . . !
: ; i

1 =
.,. --. . .^ 	
i
i iii.
T
1~
j«.
i
,._!_ „
	
..jl



jJl,*
Phxlud
44-
M
,!.j:

>,


"

_j
... .,;_





«.*
— -
—
.




ao
J0f
--
T
•\-
....i "....
~JL-
« 	
—

--
— -~—
- . - ...i
i
	

;




i
i

I





--t— - - - • •• • •
iaiwl ;
OpeteSoH. ..
i i ,,
;c
.. .

	
I
~! -
'.\~.
i
.. :....:.. 1
= M 1
' I i 1 "
1 " 1
i ! 1
I-l
1
• !
i !
: :•
i f
j
! :
i •
! ' ''
i~t
•j j •
• j;;j;
_-J_|_
J i
i— r. r _,- 	 j 	
_ _ j LrT]_
:Llr-:;-
1. I
1-
1
— | 	 .
1

I

1
i
i • i
1 1
l —

	 ,

t '
— *-"-!
1
1
I I
I
4~L_
i


-------
500
480 i
480
440
420
400
380
1
 DATE 1992 "
i
i
i
1
_

.
*
-

\
\
- 1
1
i
!
-1
i

*
~_
—
-
-
•


-


—




1

-

-




-







*•




•

•-


j
|
_.
...

.:..

-


-•
J;:


_
-
'"


~
L
j~-
L

-
t

-




,
~
-
..



1
Bl
Rl
]•••






:
•?•









^





-
,





-i-
;L



rT*









1
CP EXTRACTED FROM SOIL,
ODUCTION OPERATION
, . , . M' ii .-••)»• 	 hi " NOWMHI >


L





•i M






1 -


....




ji

^7


i
'
\ * in i«. f,» ;«,
KINf

.

! !
_r.



-•
•:

t
i'
;
j


	
:

... „ ,J _ .....
'- LJ



1


'-
-


^i.

-4-
""^"^T:
J 1
. i
.^.L._
'i
i
j
:;
i~
I
r
1
,1


; :
--i "tlr
. . ..! ,.:....
• i .
i i ]„
\ \ \
..j. |...
1 1 *.
. j !
•••Hh-
j i

:
i


^^r— TT

L "



L







\ \
! ! i
1 I
1
1 1
1~i ~|T
.1.1 L
» M.Ult
- • \- -', •
! t i-

..__ flr^-.
-±-:j:

_. -r-+- 1 -y-
:, I. . ••[ '
•i


. • i
i
! ! j

- H-I-.-+-
.win ' *"~ :
1 j
• "t
!

- i L
	 •; r
i
.......
i-.
=1 ..
':..., .,1:
_j.. ^..ld_
" 1 	 : i""j :'!"
1
! ' - L L
-1.
' !- -14
-l;L,tl-
__j'.i....r i .
..!... .!.. . i 	
! _i I i
: 	 j . .

, ..; .J.L.
i .
_U~1 — 	 J 	
1 :
	 J — 	
:



i
: . --
S^~~*"
S '
" " ! i '.
.1 i
. .

•
Ij
;
.... | ...
.., ...L. - . i. :
! i ' - .
.J..:|.J_4_^
1
1
1 	 — , 	 . _
t 1
tr
JIJIL
1 :
	 	 . j 	 ! .
'.'. !"" .1 i
.. ;r 	 i_....

....... ': ...:. -

'
rET."H~:~
: -I..:. '. .. .::: . !
• |:: 'I. IT!"
irrTTti
1 : !
I ! : i
!"!


:
	 ._._, 	
•
|i I \irtt W
. .-.._!: .-i"-..:-.
'i i
-;---! i
! ! i
1 !
- -~j -H-
'-i-i-
, i i
" i i i
-1!
- i hi
iii
1 ' ' !
„ _j— -.
; , i i
i i i
	 [ . i ' !
; i ! . !

	 1 	 i 	
	 	
i - ,
i
t ~
I , L i
I 1
1 1.
i , ;
~"\ i
! i -
1 ! i
1 . t
" ' r i
;_ 1
i > '
j 4 i-'-
Till
rid ofH*


-------
              Constituent
 Approximate Concentrations, ppmV
                                       Low
              Median
High
            Carbon dioxide

            Oxygen

            Ammonia
10,000   20,000 to 30,000  70,000

 8,000   17,000 to 18,000  20,000

   100     1,000 to 3,000    20,000
In general,  the carbon dioxide concentration increases as the depth from which gas is
extracted increases, and the oxygen concentration decreases with depth.  So far, no other
trends have  been detected in the data on these two gases.  Ammonia has consistently been
most concentrated in the gas extracted from well VES-3, which is screened from 40 to 60 feet
below the ground surface. There has been a substantial, progressive decrease in the ammonia
concentration in the gas extracted from all three wells in the system.

Throughout  the VES operation,  the CO2 and O2 measurements have indicated general
agreement, on a mass balance basis, between the oxygen present in the extracted CO2 and the
oxygen deficiency in the gas stream, relative to ambient air. This indicates that the CO2 is
being produced, within the soil mass being vented, at about  the same rate as  it is being
extracted. It is conjectured that the CO2 is being produced mainly by on-going biological
processes in  the soil rather than by purely chemical reactions. The quantity of CO2 extracted
by the VES during the  first  nine months of operation  is  indicated by  the  following
approximate calculation:

       •  Average soil gas extraction rate:
           about 230 cfm; or about O.SSxlO6 ftVday

       •  Assumed average CO2 concentration  in soil gas:
           about 25,000 ppmV, or 25,000 ftVlO6 ft3

       •  Average CO2 extraction rate:
           about  8,000 frVday

       •  Assumed 400 ftVpound  mole, at  site  average temperature  and pressure  (359
         frYmole  at STP):
          about 20 Ib mole CO2/day

       •  One Ib mole CO2 = 44 Ib, or 12 Ib C/lb mole CO2
                                     156

-------
       • After 270 days of VES operation:
           20 Ib mole/day x 44 Ib COj/lb mole x 270 day's
           = about 240,000 Ib CO2
           or about 65,000 Ib carbon removed from soil

The  souice(s) of the extracted carbon is not known.   Two candidate sources have been
identified:  organic carbon, naturally present in the soil, and the carbon component of the
spilled DBCP. The estimated 3,100 gallons of spilled DBCP would have contained on the
order of about 10,000 pounds of carbon, so clearly DBCP cannot account for all of the carbon
that has already  been extracted by the VES.  But this does not exclude the possibility that
some fraction of the extracted carbon  in the CO2 is the result of in situ biodegradation of
DBCP and or intermediate degradation products of DBCP.                 ;

In another, unrelated, site investigation, conducted near the subject site, a total of 58 soil
samples were collected and analyzed for total organic carbon. The samples were collected
at depths ranging from 1 to 80 feet below the ground surface, and it is believed the results
of the analyses are generally representative of conditions within the soil mass being affected
by the  VES operation.  The average organic carbon content of the 58 soil samples was 0.12
percent.

To provide a qualitative appreciation of the quantity of carbon naturally present in the area
soils, the following calculation is provided:

       • Assume a cylinder of soil having a radius of 100 feet and a height of 80 feet (these
         dimensions may approximate the volume, but not the shape, of the soil mass being
         vented):

          •   Volume of soil in cylinder:
                100 x  100 x  n x 80 = about 2.5xl06 ft3

          •   Assumed average dry wt. of soil:
                1001b/ft3

              Wt. of soil in cylinder:
                100 lb/ft3 x 2.5xl06 ft3 = 250xl06 Ib

          •   Wt. of carbon in the cylindrical mass of soil:
                0.0012 Ib C/lb soil x  250 x  106 Ib soil =
                about 300,000 Ib carbon

The above calculation clearly indicates that organic carbon, naturally present in the area soils,
can account for all of the carbon that has been extracted by the VES to this  time, without
having to make  unreasonable assumptions concerning the outlines or dimensions of the soil
mass being vented.  It is noted that when only well VES-1 was being pumped, a vacuum of
                                       157

-------
 about 3.8 inches H2O was measured in the riser pipe of well VES-3. The distance between
 the two wells is  about, 53 feet.  This would seem  to indicate  that in some strata and
 directions, the horizontal dimension of the soil mass being vented is significantly more than
 50 feet from the  extraction well.  A semi-log plot of vacuum versus  distance from the
 extraction well could be extrapolated to indicate a measurable effect probably occurs at 100
 feet from the well.

 The significance of the ammonia present in the extracted soil gas cannot  be quantified with
 the presently available information. In general, on-going biological processes in soil, as seem
 to be occurring in  this case, require more than simply the presence of free oxygen, which is
 being  supplied by the VES operation. An adequate supply of water and nutrients also is
 required. These may be either naturally present in the soil or must be supplied. Because free
 water generally is being collected in the knockout tank, it is believed that the air being moved
 through the soil generally is at about 100 percent relative humidity and capable of supplying
 the water requirements of in situ  biological processes. Also, it seems likely that the nitrogen
 available in the ammonia is serving as a supplied  nutrient.  On  the assumption that the
 nitrogen  is necessary for continued,  stimulated biological activity in  the  soil,  and that
 continuing  the  activity  may  be beneficial,  with  respect to  remediating the DBCP
 contamination in the soil, we have limited our use of extraction from well VES-3 because it
 is the most productive of NH3 in the system.  It is expected that pumping of wells VES-1 and
 VES-2 causes ammonia present  in the vicinity of well VES-3 to flow towards the two
pumped wells.  If the ammonia supply is depleted at some time during the operation it will
be interesting to see if it has an effect on the CO2 production rate.

                              COST INFORMATION

Costs incurred on  this  project  to date, and estimated  to the end of the  Extension of the
Variance From Permit Requirements — i.e. May 1, 1993 —  are summarized  as follows:

       •  Investigation to assess the extent of the contamination, including drilling, sampling,
         laboratory analyses, VES feasibility test, and reporting, but not including an on-
         going groundwater monitoring program
                                                                     about $150,000
         Cost to design, permit, purchase and install VES
                                                                     about $200,000
         Cost to operate VES for two months trial program, including maintenance, power,
         sampling and analyses, and reporting
                                                                      about $60,000
                                      158

-------
         Estimated costs to operate the VES for one year, (mid-march 1992 to mid-march
         1993) including maintenance, power, carbon replacements and disposal, sampling
         and analyses, and reporting
                                                                    about $200,000

         Estimated costs  to complete closure,  assuming VES  operation is terminated in
         March 1993, including: verification soil borings  and soil sampling in Level C
         protective equipment, sample analyses, and a final report, but  not including any
         post-closure monitoring
                                                                     about $30,000
                                                              Total about $640,000
                        SUMMARY AND CONCLUSIONS
A soil venting/vapor extraction system (VES) was installed to remediate vadose zone soils
contaminated with DBCP.   The  system causes  atmospheric  air to flow  through the
contaminated soil mass, where it picks up vapor phase DBCP, and is then withdrawn from
the  soil through wells installed for the purpose.  The  system has been in  essentially
continuous operation since January 14,  1992.  The estimated quantity  of DBCP present in
the  soil, in 1988, was about 800 pounds. The DBCP spill occurred about 1971, and the
estimated quantity spilled was about 3,100 gallons, or about 60,000 pounds.  The maximum
concentration of DBCP found in the soil, in the 1988 investigation, was 260 milligrams per
kilogram.  Ammonia is also present in the soil mass being vented, and has been found in the
extracted soil gas at a  concentration of up to 2 percent, volumetric.

The installed system includes three wells, each with a 20-foot-long screened section, and with
screened intervals spanning  60 feet, or from 20 to 80 feet below the ground surface. The
contaminated soils are of alluvial origin, moderately stratified, and vary from silty fine sands
to coarse sands with gravel.  The depth to groundwater is variable, but generally 90 to 110
feet below the ground surface.

The installed system uses a positive displacement blower,  capable of pumping about 220 to
260 ambient cubic feet per minute,  with the  actual flow dependent  on  the well, or
combination of wells,  being pumped. In  general, the blower operates at a vacuum of about
4 to 6 inches mercury column, and discharges to atmosphere.

The extracted soil gas has contained DBCP at varying concentrations, but within the range
of  about 1 to 13 parts per million, volumetric (ppmV).   The DBCP is removed from the
extracted  gas stream  by  passing it through two canisters of granulated activated  carbon,
connected in series, and located on  the vacuum side of the blower.  The cumulative DBCP
extraction, over the first nine  months of system operation, was  about 310  pounds.  No
estimate is provided of the remaining time the VES will be operated, nor of the concentration
of DBCP in the extracted gas that will  serve as a criterion for system shutdown.  However,
                                      159

-------
 for cost estimating purposes, we have assumed operation until about March  15, 1993, as
 presently covered by the Variance From Permit Requirements, granted by the Department of
 Toxic  Substance  Control, Alternative Technologies Division.   It is planned that a  soil
 sampling and analysis program will be implemented, following system shutdown, to allow
 comparison of before and after DBCP concentrations in the soil.

 In addition to removing DBCP from the site soils, the VES has removed a large quantity of
 carbon dioxide.  Through the first nine months of operation, at least 200,000 pounds have
 been removed. It appears that the carbon dioxide is being generated at  the rate it is extracted
 because there is a balance between the oxygen present in the CO2 and the oxygen deficiency
 in the gas stream relative to atmospheric air. The carbon in the CO2 must be present in the
 soil, because only an insignificant fraction of it can  be attributed to the CO2 present in the
 air being circulated  through  the soil.  Two possible sources of CO2  have been  identified.
 They are naturally occurring organic matter and the spilled DBCP. It is possible, but not
 documented, that the soil venting has caused stimulation of microorganisms present in the soil
 that are capable of biodegrading DBCP and or some of the intermediate degradation products
 of DBCP.  The ammonia present in the extracted gas stream may be providing nitrogen to
 the soil bacteria.

 The only waste stream generated by  the project is the carbon in the air emissions control
 system. The extracted gas stream has, at times, contained water that has collected in an air/
 water separator tank located immediately downstream from the valved manifold that controls
 the flow from the wells.  Initially, the collected water had a pH of about 11, but after about
 9  months of operation it had declined to about pH 8  or 9.  The decline may be attributed to
 the decrease in the ammonia present in the soil as a result of the VES operation.  The
 collected water has been transferred to a holding tank, from where it has been reinjected into
 the extracted gas stream, at a controlled rate, immediately ahead of an air stream heater.  The
 heater maintains the air stream entering the carbon canisters at a temperature of at least 90°F,
 in  order to prevent  the  possible accumulation of  water in the carbon  canisters or the'
 deposition of solids within the carbon, as occurred with the initial design of the system. The
 water reinjection process has resulted in all collected water being treated and disposed.

 The existing system is being operated under a Variance From Permit Requirements, granted
 by the California Department of Toxic  Substance Control, Alternative Technologies Division,
 and an Extension of the Variance, to May 1, 1993.

 The total estimated costs associated with this DBCP remediation project are about $640,000;
 including $150,000 for site investigation; $200,000 for design and construction; $60,000 for
 a two month  trial program operation; $200,000 for one year of production operation;  and
$30,000 for a post-operation, verification soils investigation.  Post-closure monitoring costs
may be  additional.
                                      160

-------
                                                       n
                                                     n L
iNGENIEURBURO FUR UMWELT7ECHNIK
1ARBAUER GMBH & CO KG HEERSTRASSE 16 1000 BERLIN 19
               5 YEARS OPERATIONAL EXPERIENCE
                            WITHTHE
                HARBAUER SOIL WASHING PLANTS
                         Winfried Groschel
                      Harbauer GmbH & Co. KG
                       Dr. Hansgeorg Balthaus
                        Philipp Holzmann AG
                               161

-------
  /.       Abstract

  Soil washing has become a well established soil cleaning process that offers broad applicability, a
  high technological standard and consistent cleaning results.  In addition, soil washing provides
  production reliability and general acceptance by the public.

  The first large scale Harbauer soil washing plant began operation in early 1987 in Berlin. Since that
  time a wide range of soil contaminants have been successfully treated. More than 120,000 tons of
  soils have been brought from 70 sites to the plant for treatment. In 1990 a mobile HARBAUER soil
  washing plant started operation in Vienna.

  The high efficiency of the Harbauer soil cleaning process is achieved by using a broad spectrum of
  processes partly similar to those commonly used in the mining and mineral processing industries. We
  can guarantee optimum cleaning results with relatively low energy consumption by selecting our
  mechanical agitation processes based on the grain size fraction.

 This year a  third generation of soil washing plants is under construction in Berlin with completion
 planned for March 1993. In mid  1993 a sophisticated soil washing plant for the removal of mercury
 from soil and rubble will begin operation in Bavaria. This plant will use a special vacuum distillation
 system to treat fines and other volatile and semi-volatile contaminants.

 The main emphasis in the further developments of the HARBAUER soil washing system  lies in the
 minimization of the residues from the washing process. While we have tested various techniques to
 date, vacuum distillation has proven to be the most promising additional  step in the soil washing
 process. It has proven successfull for a large number of volatile or semi-volatile contaminants such as
 mercury, PAHs or other organics.
 2.
Introduction
 The first large scale HARBAUER soil washing plant began operation in early 1987 in Berlin. This
 plant was designed for the treatment of contaminated soil from the former Pintsch used oil recycling
 site. This soil was heavily polluted by residues from the oil refining activities. During the operation of
 the soil washing plant from 1987 to 1991, the plant engineering was optimized, several treatment
 stages were added and the permitted uses of the plant were expanded. This made it possible to treat
 soil from other contaminated sites with different contaminants. Up to now more than 120,000 tons of
 contaminated soil from approximately 70 different sites have been cleaned successfully.

 The gained operational experience has been used for the design of new soil treatment plants. In 1990
 a mobile HARBAUER soil washing plant started operation in Vienna. This plant was designed for
 the treatment of cyanide contaminated soil and rubble.

 This year a third generation of soil washing plants is under construction in Berlin with completion
 slated for March 1993. This plant will be a soil washing center for Berlin and its surrounding areas. It
 represents the latest developments in soil washing techniques  with a broad application spectrum for
 different soils and pollutants.

 Another soil treatment plant for removal of mercury is under construction in Marktredwitz, Bavaria.
This plant consists of a soil washing unit and a newly developed vaccum distillation unit. It will start
operation in mid 1993.
                                             162

-------
3.       Process Description
To achieve a high cleaning efficiency for different types of soils and contaminants, the HARBAUER
soil washing system uses a multi-step process. Some of the process units are similar to those used in
the mining and mineral processing industry. Others have been developed especially for treatment of
contaminated soils. The primary operational steps  of the HARBAUER soil cleaning  system that
determine the cleaning efficiency are the following:
Process
Process Equipment
1.   Dislodging of contaminants from soil
     particles using multi-step mechanical
     agitation
2.   Separation of highly contaminated grain-size
     fractions by means of classification
3.   Separation of highly contaminated organic
     fractions by means of sorting

4.   Exchange of contaminated interstitial water
     by means of rinsing and dewatering
- Bladewasher
- Stirring Reactor
- Vibration Reactor
- Hydrocyclones
- Classification Screens
- Mineral Jig
- Fluidized Bed Sorter
- Flotation
- High Pressure Water Sprays
- Countercurrent Hydrocyclone Unit
- Fluidized Bed Sorter
- Dewatering Screens
A simplified flow chart of the HARBAUER soil cleaning system is shown in figure 1.  At first,
particles > 16 mm are separated using two classification screens. The oversize material has to be
crushed and can then be added again to the material feed. Usually the amount of material > 16 mm is
less than 1 %. Therefore crushing can be done in batch quantities using a mobile crusher.

The mechanical agitation takes place in three main stages (mechanical agitation I - III). Due to the
fact that the required energy for the cleaning of coarse particles is much smaller than for the fines,
the specific energy input increases in the direction of the material flow. In all agitation stages mainly
shear forces between the particles or between the liquid  and the particles are activated. This
guarantees an optimum cleaning efficiency at a relatively low level of energy consumption.

After the dry classification the contaminated soil is conveyed to the first washing unit, a two-stage
blade washer (mechanical  agitation I).  Before entering  the  blade washer metal scrap will  be
separated. The  blade washers are followed by a classification screen (classification stage III)  where
the gravel and stone fraction is separated. Before this  fraction is discharged as cleaned material, light
materials like tar, wood, etc. are separated using a mineral jig (sorting stage I).

Following the classification of the coarse  fraction a multi-stage hydrocyclone unit (classification
stage IV - V)  is used to separate the fines < 15 ^m. These  particles are concentrated  through
flocculation/sedimentation/thickening. Finally they are dewatered mechanically. The second stage for
mechanical agitation is a stirring reactor followed by a fluidized bed sorter for the separation of light
materials from the sand fraction (sorting/rinsing stage  n).
                                           163

-------
 The fine particle fraction between
 15  jam  and 150 \*m is  treated in
 sorting  stage III. This  is a two-
 stage  pneumatic  flotation  step
 where very fine organic particles
 are   separated.    Finally    the
 HARBAUER vibration  screw is
 used  to dislodge  the  remaining
 contaminants. In this machine the
 particles are subjected to intense
 vibrations using high amplitudes.
 The  previous   and   following
 dewatering screen ensure  that an
 intensive rinsing process  of  the
 cleaned material is achieved.

 To  avoid any emissions the soil
 treatment system is equipped with
 a process water and  process air
 treatment unit.  The total  amount
 of process water which  is needed
 for  an operational capacity of 30
 t/h  of contaminated soil  is  120
 m3/h.  The water runs in different
 circuits. The first treatment stage
 for  the  process water is  a  large
 flocculation/sedimentation     unit
 consisting of 6 thickeners followed
 by a  tilted plate separator. This
 unit needs to be so large because
 the soil  treatment plant is  able to
 treat soil with up to 30 % fines
 below 63 [Jim.


Transport •*— — — — __-
Charging |
Material Feed



Classification
Staee I / 200 mm
—i
1
Classification
SHeeIT/60mm
•
-:_,J | Crushing Stage j fW?~V
T (mobile) 	 * \ia£/
Oversize Material
Mech. Aguanon
Staee I

r-» Sorting Stage 1 — » Etewaf™S 	 ». /iSHpy,
Gravel
Clamif. Stage in
2.0-8.0 turn
i
__ ,, Dewnenne i. i
Staee n ...'.. '.
Gravel Fraction
Classification 1 dasaficaaon
Stage IV | Staee V T* Floccuiaaon


Mech. Agitation
Staee II




Sorting / Rinsing
Staeell


Dewitenng
Staee in
\
_ . Classificanon I t v /
Staee VT 	 * ' • ' ' -J

, , Sand traction
Sorting Stage m J Mechanical j 	 ^ ^........y
Dewatenn? I W* 	 f

f
Mech. Agitation
Rinsine Staee 01




Dewitenng
Staee IV
1
Cleaned Soil
Res dues
FlowChan ^^" 	 ££^ 	 _
Harbauer Fa. Harbauer H
foil Wishing hgeaiaatamKr ft
— — 	 , 	 ; 	 1 Sy!ittm 11'm^lr.x.hn.fc H_
                                   Fig. 1  Simplified Flow Chart
Light substances such as oil flow separately to a dissolved air flotation unit where they are separated.
The cleaned overflow of the tilted plate separator is split into three streams. Stream I is directly
reused in the first washing step of the soil treatment (circuit 1). Stream II is filtered using sand filters
and then reused (circuit 2). The third stream of the process water is cleaned using the treatment
stages of oxidation/precipitation and activated carbon adsorption. The operation of the oxidation and
precipitation stages is optional and only used for heavy metals and cyanides. The cleaned water
leaving the activated carbon adsorption stage is  reused in the final rinsing processes of the soil
treatment.  Approximately 15 m3/h of this water  is discharged and replaced  by clean water. The
sedimented solids are dewatered using chamber filter presses.

In the process air treatment unit 30,000 m3/h of air is cleaned. 15,000 m3/h are collected from the
resource material feed/dry classification/intermediate storage areas. Dust is first removed from this
Stream and then together with the other collected air stream finally cleaned using activated carbon
adsorption. The quality of the discharged air is controlled on-line using a TOC-monitor.  Furthermore
the cleaned air stream will be regularly tested in a laboratory.
                                                164

-------
4.      Design
HARBAUER soil washing plants are subdivided in four main areas. Area I consists of a 1,000 m3
intermediate storage area for contaminated soil, the material feed and screens for dry classification at
200  and at 60  mm. If necessary a crusher can be integrated into the process here. To avoid
emissions,  this area is a  closed hall provided with material locks and rolling  bars. The soil is
transported from  here to  area  II using an enclosed belt-conveyor. Area n consists of all wet
mechanical separation processes and the dewatering stages for the residues. Areas III and IV are the
process water and process air treatment units. Rgure 2 shows a top view of the Berlin soil washing
plant with the required infra-structure. The minimum required space for a soil treatment plant of that
size is 8,000 m2. The whole site is subdivided into sections:
   intermediate storage and material feed:
   soil washing plant:
   intermediate storage for cleaned soil:
   storage area for containers:
   roads, offices, etc.:
  700m2
1,600 m2
  700m2
1,000 m2
4,000 m2
n
>-
1



GBCHZC G«*OtSTB*nl





PROCESS AIR
1^ n n n n n
c
1&\-
D HALL FOR n
THICKENERS/TANKS U
-
n
O~ SOtt. TREATMENT pi pi
CeH UNIT
(£__ 	 ,
F™"*""*"™!
SANDySILT
LJ t
l^-l GRAVEL
^S • i »r-lr.-i
JL CLEANED MATERIAL
(T) DISCHARGE

(
r -- r — i — i 	 \
' INTERMEDIATE STORAGE 'l
| FOR CLEANED SOIL •'
= \ \ I I
". 	 m 	 1 	 i 	 1
juDDDDQuLUi
I


— , 	 ,
PROCESS WATER
TREATMENT UNIT
n n n a
a a a
RESIDUES
DISCHARGE
JUUU
UUUU
SPECIAL CONTA
CONTAMINATE]
SSSB
D i
i
~T


(t
;
,
I



BELT CONVEYOR (T; (C
JZT-, =t j.
/^

=
^w-
JNEF
3 SO

)
INT
FOB
SFOR
L/RESIDUl
ERMEDIATE STORAGE
CONTAMINATED SOIL


'
£S







MATERIAL
*
l


t
D
^
i
FEED



_-

-T)

s
1
Fig. 2  Top View of the Soil Washing Plant Berlin

                                           165

-------
 An example for the modular design of one of our plants is shown in figure 3. The  whole plant
 consists of 50 containers that are 3 m wide, 3 m high and with a length of 10 or 14 m. The individual
 containers are prepiped and prewired with quick interconnections. This makes a rapid  mobilization
 or demobilization possible which is very important for timely delivery and cost effective use during
 on-site treatment.
Fig. 3 Example for the Modular System Design
                                            166

-------
 5.      Contaminant Specific Suitability
 The most operational  experience gained  thus far has been gained with organic contaminants
 specifically TPHs and PAHs. For TPHs the input concentrations varied between 500 and 20,000
 mg/kg and for PAHs between 500 and 5,000 mg/kg. The residual concentration in the cleaned soil
 has typically been < 100 mg/kg for TPHs and < 5 to 10 mg/kg for PAHs. In some unusual cases of
 very high input concentrations, the maximum residual concentrations have been 300 mg/kg for TPHs
 and 20 mg/kg for PAHs. Figure 4 shows typical examples.
                10000
                 1000
                    Example!: PETROLEUM CONTAMINATED SOIL
                             (PROJECT: RUNGIUSSTR, BERLIN)
                     CONCENTRATION TPH (mg/kg]                     REMOVAL [%]
                                                                     = 100,0
                                                                       80,0
          REMOVAL [%]H
          Input (mg/kg]»
         Output (mg/kg] d
              10000,00
                     Example 2: TPH, PAH CONTAMINATED SOIL
                                (PROJECT: Plntsch, BERLIN)
                     CONCENTRATION [mg/kg]
REMOVAL [%]
                                                                       100,0
               1000,00 fe
                                                                       60,0
          REMOVAL (%]H
          Input (mg/kg] •
         Output (mg/kg] d
Fig. 4 Examples for Treatment Results of Soils Contaminated with Organics
                                          167

-------
The clean-up efficiency for volatile organic contaminants like chlorinated solvents or BTEX is very
good. The residual concentrations always have been below the detection limits. For PCBs the input
concentrations have been relatively low ranging from 5 to 25 mg/kg. The residual concentration has
consistently been < 0.1 mg/kg or below the detection limit

Soil washing is  also an effective treatment method for heavy metal or cyanide contaminated soil.
Examples are given in figure 5.
                   Example 3: HEAVY-METAL CONTAMINATED SOIL
                             (PROJECT: TRE1DELWEG, BERLIN)
                10000
                    CONCENTRATION [mg/kg]
REMOVAL (%]
                                                                       100,0
                                                                       80,0
         REMOVAL [%]H
          Input [mg/kg] •
        Output [mg/kg] E3
                     EXAMPLE 4: CYANIDE CONTAMINATED SOIL
                           PROJECT GASAG (FORMER MGP-SITE)
             1000
                 CONCENTRATION [mg/kg]
REMOVAL [%]
Fig. 5 Examples for Treatment Results of Soils Contaminated with Heavy Metals and Cyanides

                                       168

-------
 6.      Technical/Operational Data
 The most relevant technical and operational data of the HARBAUER soil cleaning system is shown
 in figure 6.1. The operational capacity ranges between 25 and 40 t/h depending on the amount of
 fines in the soil. The annual capacity is 100,000 tons for two-shift-operation and 160,000 tons for
 three-shift-operation. We have proven that this capacity is sufficient for a stationary plant
OPERATIONAL
CAPACITY
WASHING CAPACITY
TREATMENT
RESIDUES
LIMIT FOR
APPLICATION
SPEC. ENERGY
CONSUMPTION
SPEC. WATER
CONSUMPTION
OPERATIONAL
MATERIAL
QUANTITY OF WASTE
WATER/
OUTGOING AIR
FLOOR SPACE
REQUIRED
25 - 40 t/h depending on the portion of fines < 63 \m
The maximum washing capacity depends on the texture of the soil, the
concentration and the kind of bindings of contaminants. Treatability tests
are necessary to determine the minimum achievable residual concen-
tration. The following concentrations for cleaned material are guide
values refering to our operational experiences.
TPH PETROL. HYDROCARB. 10.0 - 200.0 ppm
BTEX 0.5 - 5.0 ppm
PAH (EPA) 1.0 - 20.0 ppm
HALOGEN. VOL. ORGANICS 0.5 - 5.0 ppm
PCB 0.5 - 10.0 ppm
PHENOLE(EPA) 0.5 - 5.0 ppm
CN (total) 1.0 - 20.0 ppm
Hg 1.0 - 50.0 ppm
Pb, Cu, Ni 50.0 - 300.0 ppm
Cd 1.0 - 20.0 ppm
- FINES < 15 Mm
- SEPARATED ORGANICS
- LOADED ACTIVATED CARBON
PORTION OF CLAY/SILT: max. 30 - 40 %
ca. 12 kWh/t
ca. 200 - 1,000 1/t
- FLOCCULANTS
- HC1, NaOH (pH-adjustment if required)
- SURFACTANTS (if required)
- ACTIVATED CARBON
INTEGRATED PROCESS WATER AND AIR TREATMENT UNIT:
CLEANED PROCESS AIR: 15,000 - 30,000 mYh
CLEANED PROCESS WATER: 200 - 1,000 1A
SOIL WASHING PLANT: 1,000 - 2,000 m2
MAX. HEIGHT: 20 m
ROADS, INTERMEDIATE
DEPOSITS FOR CLEANED AND
UNTREATED MATERIAL, ETC.: 8,000 - 10,000 m2
Fig. 6 Technical/Operational Data of the HARBAUER Soil Washing System
                                        169

-------
 The energy and water consumption of the treatment process is relatively low. The specific energy
 consumption is 15 kWhA or approx. 450 kWh/h. The specific water consumption ranges between
 200 and 1,0001/t and depends mainly on the type of soil and the input water content.

 Additives used in the process are flocculants for the  sedimentation, HC1 and  NaOH for  pH-
 adjustment and  activated carbon for the process water treatment. Surfactants are only used in special
 cases when otherwise the required treatment results cannot be reached.

 Soil washing is  a volume reduction process which concentrates the contaminants  in the fine particle
 fraction of the soil. The economic limit for the application of soil washing is established generally by
 the portion of clay or silt in the soil.  The clay/soil fraction should not exceed 30 to  40 % of the
 volume. This limitation exists because the amount of residues which remain following washing of soil
 with more than 40 % clays will increase to such a point that the additional landfill and post treatment
 costs plus the reduced operating capacity will push the  soil washing per ton costs into the non-
 competitive range. To help minimize the amount of residues the cut-off point of the HARBAUER
 soil washing process is as low as 15 \un. In additional to the fines, solid organic matter such as roots,
 leaves, humus, etc. with a high affinity to contaminants must be separated. The soil washing process
 will generate a small quantity of loaded activated carbon as a waste by-product.

 The quantity of filtered air ranges  between 15,000 and 30,000 m3/h and the  cleaned process water
 discharge ranges between 200 and 1,000 1A.
 7.
Costs
 The investment cost for a large scale stationary soil washing plant is approximately 19 millions DM
 or 12 millions US$. The costs can be broken out as follows:
    Earthworks, foundations, subsurface sealings, etc.
    Steel construction, buildings
    Equipment
    Infrastructure
                                                  2.5 millions US$
                                                  3.0 millions US$
                                                  6.0 millions US$
                                                  0.5 millions US$
It should be noted that these are the investment costs for a stationary, multi-functional plant which is
able to clean many soils with different textures and contaminant characteristics.
The specific operating costs are shown below:

•   Power:
•   Water/waste water
•   Additives, activated carbon
•   Analytics
•   Labor
                                                  4.00 US$/t
                                                  2.80 US$/t
                                                  5.30 US$/t
                                                  3.00 US$A
                                                 15.00 US$/t
These numbers are specific for German costs and conditions. Most important for the specific cost per
ton is the rate of utilization. Figure 7 shows the specific cost depending on the rate of utilisation for a
plant with an annual capacity of 100,000 tons. If the plant is running at full capacity, the total costs
are 70.— US$/t. If the rate of utilisation is only 50 %, the cost will be as high as 120.-- US$/t. These
prices do  not include any cost for the  post treatment or deposition of residues. The most  cost
effective way to run a soil washing plant however is the  treatment of big quantities of soil from just
one remediation site with a specially designed plant
                                         170

-------
                        TREATMENT COSTS SOIL WASHING [ US $ /1 ]
                                      STATIONAL PLANT:    .
                     MAX. OPERATIONAL CAPACITY: 100.000 Va in 2 SHIFTS
                 300
            CO

            is
            8
            O
            o
            0.
            CO
200-
                 100-
                      0   10   20   30   40   50   60  70  80  90  100

                                    RATE OF UTILIZATION [%]
Fig. 7.1 Soil Washing Cost Relative to the Rate of Utilization

8.      New Developments
The main thrusts in our soil washing research is to reduce the amount of residues which have to be
post treated and to increase the application spectrum. One method is to use a flexible separation limit
for the fines which goes down even below 15 nm in grain-size. By using this method, it is possible to
adjust the optimum separation point for each type of soil which has to be cleaned. By doing this, the
amount of fine particle residues is minimized.

In some cases of heavy metal contaminated soil it is possible to reduce the level of contamination
significantly by soil washing but the required limits may not be reached. Then a chemical extraction
using low concentrated acids  is applicable. Results of tests which  have  been done with a lead
contaminated soil are shown in figure 8.1. The input concentration in the range of 40,000 to 60,000
mg/kg was reduced  under the required level of 150 mg/kg. After the treatment the fine particle
fraction showed very high lead concentrations so that this material was suitable for metallurgical
recycling. Chemical extraction without soil washing was also tested but was not successful.
                                            171

-------
The most promising new techology is the vacuum distillation system. This technology can be used as
additional treatment step for the fines following the soil washing process or directly for soil which
cannot be washed due to very high concentrations of silt/clay or contaminants. It has proven that
soils contaminated with volatiles  like  BTEX,  solvents or mercury as well  as  semi-volatile
contaminants like TPHs and PAHs can be cleaned

The process uses pressures in the range of 50 to 150 hPa and temperatures between 300 and 400 °C.
To ensure that no dioxines are produced, the process works under inert conditions.

Figure 8.2 shows the boiling points of the EPA PAHs under normal pressure. It can be seen that the
vacuum is necessary to reduce the boiling points of such semi-volatile substances. Figure 8.3 shows
the results of laboratory scale tests for soil from a former coking plant. The  removal of the  input
concentration has been > 99.9 % for PAHs, > 99 % for TPHs and even 99 % for cyanides.
    CONCENTRATION OF LEAD (g/kg)
                INPUT
AFTER SOIL WASHING
AFTER CHEM. EXTRACTION
Fig. 8.1 Treatment Results - Combined Soil Washing and Acid Leaching Process
                                             172

-------
IU
ee
I
UJ
500
480
460
440
420
400
380
360
340
320
300
280
260
240
220
    200
Fig. 8.2  Boiling Point of PAHs (EPA standard) at 1,013 hPa
      100000
Fig. 8.3 Treatment Results Vacuum Distillation
                                              173

-------
 9.
Outlook
 By using the HARBAUER soil washing system it is possible to clean soils with multiple complex
 pollutants and problematic particulate composition. The system allows the soil to be rapidly cleaned
 at site and immediately be used as backfill. As a result of the removal of clays and silt the cleaned
 backfill material has a greater loadbearing capacity than the original soil prior to the washing.

 We have experience from over 70 sites producing more than 160,000 tons of soil. The sites include
 refineries, chemical plants, oil recovery facilities, process plants, landfills, and marine environments.
 The range of contaminants includes hydrocarbons, cyanide, heavy materials, mercury,  lead,  VOC,
 PAH, PCB, and others. In many of the soils more than one contaminant was present. Inspite  of the
 field experience we have  gained over the years we still believe that a treatability study must be done
 with  each type of soil to be cleaned. Harbauer performs inhause the full range of services from
 treatability studies  to  equipment  fabrication and from  plant operation to the  design of special
 treatment plants for unusual contaminated sites.

 By 1993 there  will be three large  scale HARBAUER soil washing plants  in operation plus other
 smaller scale operations.  HARBAUER is also presently involved in the  planning of four other soil
 treatment centers in Germany. Much of this growth and extensive research and  development is
 possible as a result of the  acquisition of HARBAUER by the Philipp Holzmann Company. Holzmann
 is Germany's largest civil  engineering company and ranks in the top 10 in the world with over $ 10
 billion in revenues. During  the  lat  12  years Holzmann  has been  aggressively  entering the
 environmental remediation field through both acquisitions and  internal development. Engineering
News Record recently named Holzmann  as one of the largest environmental contractors  in the
world.

If you have soil washing  problems  that you would like to discuss with us,  we operate  in the U.S.
through J.A. Jones Construction Company of Charlotte, North Carolina. Jones is also a member of
the Holzmann Family of companies.

We believe that soil washing will become a more acceptable method of  remediation in the U.S. as
landfill space decreases and the costs for incineration rise. Soil washing  is not a panacea for  every
site but when applied to the proper set of circumstances is  a quick and cost effective method for
remediation of soil contamination.
                                           174

-------
       HELD DEMONSTRATION
                 of
           SOIL WASHING
                at the
  KING OF PRUSSIA SUPERFUND SITE
         MICHAEL J. MANN, P.E.
ALTERNATIVE REMEDIAL TECHNOLOGIES, INC.

                 and

           HANS VAN DORD
     HEIDEMIJ RESTSTOFFENDIENSTEN
              175

-------
     FIELD DEMONSTRATION OF SOIL WASHING AT THE KING OF
                        PRUSSIA SUPERFUND  SITE

 Overview

 A soil washing demonstration run was conducted at Moerdijk, The Netherlands, on July
 22,  1992.  The demonstration run was conducted on soils from the King of Prussia
 Technical  Corporation Site (KOP) in Winslow Township, New Jersey by Alternative
 Remedial Technologies, Inc.  (ART) a Joint Venture Company of Geraghty & Miller,
 Inc. of the  U.S.A. and Heidemij Reststoffendiensten of The Netherlands.

 Site investigations had been accomplished through a Remedial Investigation (RI) and
 expanded by a Supplemental RI. The RI was further enhanced by the excavation of a
 number of  test pits.  The test pits provided an excellent  means to physically examine the
 disposition of the target soils. Soil washing was selected as the remedy for the metals-
 contaminated  soil at the site in the United States Environmental Protection Agency's
 (USEPA's) Record of Decision (ROD) dated September 28, 1990.

 To define the soil washing system, a detailed treatability study was performed under the
 EPA's "Guidance  Document for the Performance  of Treatability  Studies under
 CERCLA".  The study was  completed in March, 1992 and  later  approved by  the
 USEPA.  The treatability  study  confirmed the  process and  optimized the system
 configuration.  The study did,  however,  leave some questions unanswered, and  to
 further confirm the  process, a "Demonstration  Run" was proposed to be conducted at
 the full-scale Heidemij plant in The Netherlands.

 First, approvals to ship the demonstration soils were obtained from the USEPA under
 the Resource Conservation and Recovery Act (RCRA) Export Rule and then from the
 Dutch VROM, the  equivalent USEPA organization. For  the demonstration run,
 representative  soils were excavated, field-quantified using the Environmental Resources
 Management,  Inc. (ERM) X-Ray  Fluorescence (XRF) equipment,  bagged in "super
 sacks", and placed in sea-going shipping containers. The containers were transported to
 the Port  of Newark and then, by  sea, to the Port of  Rotterdam.  In Rotterdam, the
 shipment cleared customs and was transported by land to  the Moerdijk treatment
 facility.

 At Moerdijk, the site soils were screened,  blended, and processed in a state-of-the-art
 soil washing facility using an integrated system of screens, hydrocyclones, flotation
 cells, dewatering equipment, thickeners, and a  belt filter  press. The  run lasted
 approximately seven hours  during which  time samples were collected from  various
 intermediate stages, and, most importantly, from the process residuals, namely: sand
product, sludge cake, and oversize materials.

The collected samples were first analyzed  by Daniel C. Griffith (Holland) B.V. (D.C.
 Griffith), a multi-national environmental  laboratory operating in The Netherlands.
Split samples were analyzed by IEA Labs, a U.S. Contract Lab Program (CLP) Lab.
                                    176

-------
These  data were consistent and confirmed  achievement of the soil cleanup levels
specified in KOP ROD.

The process residuals were returned to the U.S. on October 17, 1992. The oversize and
the product sands were returned to the KOP  site as clean, while the sludge cake was
disposed at the GSX Pinewood Treatment, Storage and Disposal Facility (TSDF).

Preparation  Activities

Excavation Plan

The  excavation of target  soils  at the  site was  performed in  accordance with a
demonstration run plan.  The  original plan  was to collect 200 short tons of target
materials to be placed in 200 super sacks, but, due to lower than expected soil densities,
the total amount of material collected was 164 short tons.

The  material was shipped from the site to the Port of Newark, then to the Port of
Rotterdam, and downloaded for the trip to the soil washing facility located at Moerdijk
(near Dordrecht), The Netherlands.

At the  Moerdijk site the soil was first rough screened in its  received state  at 4
centimeters (cm). A "gross oversize" fraction was defined as that material >4 cm, and
consisted of less than one ton, comprised mostly of plants, tree roots, and some  hard
lumps of sludge.  The gross oversize material was bagged for return to the site where
debris-washing studies will be conducted.  The underfall material, that is everything
<4cm, was blended and staged as the process feed material.

Process Plant Configuration

The  process plant was configured exactly as recommended in the KOP Treatability
Study.

The plant consisted of the following primary subsystems:

•  The Feed Hopper and Conveyor
•  The Wet Screening System
•  The Separation System, including Hydrocyclones
•  The Flotation System, including the Contact Scrubber and Surfactant Addition
•  Sand Dewatering and Conveying
•  Lamella Clarifiers and Thickeners
•  A Belt Filter Press and Conveyer

Systems available at the Moerdijk Plant  that were not used for the demonstration run
included the spiral concentrators,  the 500 micron  screen, and the sieve bend.  These
subsystems were intentionally  bypassed in order to match the proposed configuration
determined from the  Treatability Study.
                                     177

-------
 Chronology of Demonstration  Run Activities
 The key activities of the demonstration run were:
 May 24, 1992
 May 30, 1992
 May 31, 1992
 June 14, 1992
 June 24, 1992
 July 18, 1992
 July 22, 1992
 July 24, 1992
 August 21, 1992
 October 17, 1992

 Data Collected
Soil excavation began.
Soil bagging and containerizing complete.
The SV Sextant sailed for Rotterdam.
Soil arrived in Rotterdam.
Soil downloaded and arrived in Moerdijk.
Soil screened and plant cleaned out.
Demonstration run performed.
Preliminary data presented to USEPA and discussed.
Demonstration Run report submitted to USEPA Region II.
Treated soils returned to the U.S.
 Detailed  data were collected to  evaluate the process and the effectiveness  of the
 treatment as measured by the ROD Soil Cleanup Standards. The data were verified by
 split sample analyses using a CLP-qualified laboratory.  Based upon averages from data
 developed, successful compliance with the required standards was achieved.

                         Treatment Products. Tmg/kg)
Source
Cleanup Goal
Process Feed
Oversize
Sand
Sludge Cake
£E
483
770
420
170
4750
Qi
3571
1500
790
350
8030
m
1935
460
540
70
2630
Detailed data is included in Tables 1 through 12
                                    178

-------
Disposition  of Process  Residuals

The oversize material and the product sand meet the treatment standards and will be
staged onsite and later incorporated into the restoration of the site.  The gross oversize
material  will be returned to the site  and staged for additional  studies to  determine
whether a debris-washing system is appropriate for use in  attaining additional volume
reductions by removing fines from the material. The sludge cake has been disposed at
an appropriate off-site facility.

Conclusions

Key conclusions drawn from the performance  of the demonstration run were:

1.  The system as recommended in the Treatability Study can treat the KOP soils to
    levels in compliance with the requirements  of the ROD.

2.  The distribution of the  process residuals indicates that, on a dry solids  basis,
    approximately  13%  of the material introduced into the plant will require off-site
    disposal.

3.  That several design modifications for the KOP plant will improve the operations of
    the facility on-site; specifically improvements in wet screening  and in sludge cake
    production.

The Gross Oversize, the Oversize, and the clean sand will be returned to the site. The
sludge cake consisting of approximately 55% dry solids or now approximately 20%  of
the total soil treated will be disposed at an off-site facility.

For more information about this project and  soil washing, contact Michael J. Mann,
P.E., Alternative Remedial  Technologies. Inc..  14497 North Dale Mabry  Highway,
Suite 140, Tampa, Florida 33618.  Telephone  (813) 264-3506.
                                      179

-------
                                     TABLE 1
                          MATERIAL TO BE TREATED

                            King of Prussia Technical Site
                                Demonstration Run
                             Moerdiik. The Netherlands
                                   July 22. 1992
                                     (All tons)
Source
Sludge Band
Swale
Lagoon 1
Lagoon 6
Total
. Target *
145 (72.5%)
5 (2.5%)
25 (12.5%)
25 (12.5%)
200
Actual
122 (74.4%)
3 (1.8%)
21 (12.8%)
18 (11.0%)
164
Gross Oversize
0.5
<0.1
-0.25
-0.25
-1.0
Process Feed
121.5
3
20.5
18
163
A nominal 200 U.S. tons of material were targeted for treatment in the soil washing process.
Actual quantities received were 164 tons with a percentage relationship very similar to the
targeted ratios.  All source materials were dry-screened at 4 cm and yielded a gross oversize
totaling nearly one ton. Lagoon 1 and 6 sludge, planned to be pre-screened dry as 1 cm,
were found to be too wet and approval was given by the KOP Remedial Project Manager
to conduct 1 cm screening on the 2 mm process oversize at the completion of the run. That
data is contained in Table 3.
                                    180

-------
                                     TABLE 2
                           PROCESS FEED MATERIAL

                           King of Prussia Technical Site
                                Demonstration Run
                             Moerdijk. The Netherlands
                                   July 22. 1992
Influent Concentrations
                                    (All mg/kg)
Saiople *
;•
1
2
3
4
5
6
7
8
9
Average
Cr
DCG
790
745
705
705
910
815
855
710
735
770
IEA
872

759

982

1080

675
870
Cu
DCG,
1600
1600
1300
1400
1850
1900
1500
1250
1250
1500
IEA,
1470

1080

2170

1310

1110
1430
Ni
D€G
433
415
408
420
660
473
460
393
435
460
-IEA
409

357

639

368

378
430
Dry Solids
<%)
83.5
83
85.5
85
82
85
83.5
86
86
84.4
Per the agreed plan,  all discrete process materials were mixed into a feed blend pile.
Results of this activity were captured on video tape.

Efficiency of the blending operation and feed to the plant was measured via a series of nine
(9) radial hollow stem auger borings, analyzed for contaminant metals chromium, copper
and nickel.  In addition, five (5) samples were split  for CLP analysis by IEA Laboratories
in the United States.

Analysis of the  nine samples by D.C. Griffith showed good consistency with averages and
ranges for each metal.  CLP analysis by IEA on five split samples showed similar consistency
and close agreement to the results generated by the Dutch laboratory. From these data, it
was concluded that the feed pile was sufficiently blended to introduce a consistent feed to
the process.
                                       181

-------
                                    TABLES
                              PROCESS OVERSIZE

                           King of Prussia Technical Site
                                Demonstration Run
                             Moerdijk. The Netherlands
                                  July 22. 1992
                                   (All mg/kg)

Material
s *•
"
<4 cm >2 mm
Initial Feed (10%
misplacement)
<4 cm >2 mm
Reprocess O/S
< 4 cm > 1 cm
Screened
< 1 cm > 2 mm Screened
Concentrations""
Cr "

DCG
890


420

270

540
IEA
696


305

185

408
Cii

DCG
1400


790

335

800
IEA
1220


617

300

700
Ni
..A"* ;$ f
DCG
1200


540

970

730
IEA
932


345

730

498

%Dry
 2 mm which meets the treatment requirements.
                                     182

-------
                                  TABLE 4
                         PRE-FLOTATION PRODUCT

                         King of Prussia Technical Site
                              Demonstration Run
                           Moerdiik. The Netherlands
                                July 22. 1992
                                                      mg/kg
- — / 'Cr \ :' :, ':., '
. ] r^ .c^il T * :
- ""Ni '- '
% DJ^ Solids
375
680
170
76%
*No splits on this sample.

This sample was a composite of samples taken every 30 minutes during processing.  This
sample was taken from the underflow of Cyclone #1.
                                    183

-------
                                    TABLES
                              FLOTATION IMPACT

                           King of Prussia Technical Site
                                Demonstration Run
                             Moerdijk. The Netherlands
                                   July 22. 1992
 Sand Product
                                                   mg/kg
Process Step v" " ' x"
* -S** " K • f v-> s
s ' % >
Pre-Flotation Composite
Final Sand Product (Avg. composite of
15 Data Points)
Contamination Reduction
Percent of Flotation Input
Cr
375
170
45%
Ott -
680
350
51%
. NT'
_, s
170
70
41%
%£>ry
Solids
76
84

This table demonstrates that the use of the flotation step provides a significant reduction
in contamination residing on the sand. Flotation provides an important protection to make
sure that the sand returning to the site meets the clean-up standards.
                                     184

-------
                                   TABLE 6
                               PRODUCT SAND
                          King of Prussia Technical Site
                               Demonstration Run
                           Moerdijk. The Netherlands
                                 July 22. 1992

                                  (All mg/kg)
Sample
1 - 0900
2 - 0930
3 - 1000
4 - 1030
5 - 1100
6 - 1130
7 - 1200
8 - 1230
9 - 1300
10 - 1330
11 - 1400
12 - 1430
13 - 1500
14 - 1530
15 - 1600
16 - 1630
Average
Treatment
Requirement
- Cr
DCG
IEA
Oi
DCG
IEA
Ni "
DCG
IEA,
,%Dry
Solids
No sample taken, sand not discharging
98
250
185
130
115
155
76
150
140
140
235
185
205
220
205
170
483

266

97

161

129

183



195

170

195
465
370
270
240
315
145
305
280
310
520
455
465
445
430
350
3571

668

187

353

258

428



429

390

41
105
73
53
46
67
33
63
54
65
120
87
97
91
89
70
1935

119

43

77

66

98



99

80

90
81
83
84
84
83
84
84
84
84
81
83
86
83
83
84

ERM Split Sample #13 was broken during shipment. This
be placed back on the site.  It clearly meets the treatment
product sand is the material to
requirements.
                                     185

-------
                                     TABLE?
                             SLUDGE CAKE RESULTS

                            King of Prussia Technical Site
                                 Demonstration Run
                             Moerdiik. The Netherlands
                                   July 22. 1992
                                  (All mg/kg)
Sample
1
2
3
4
Average
'••• --.•:• Or
DCG
4400
4400
4700
5500
4750
IEA

4470
4760

4615
Cu
DCG
7300
7400
8100
9300
8030
TEA

7330
7950

7640
Ni
DCG
2300
2300
2700
3200
2630
JEA

2360
2670

2515
% -Dry Solids
44
46
46
44
45
This table tabulates the results of the produced sludge cake. The sludge cake contains the
treated contaminants and will be disposed at an appropriate off-site facility. Work will be
focused on  improving  (increasing) the produced sludge  dry solids concentration.   A
reasonable goal is 55% dry solids.
                                     186

-------
                                   TABLES
                  SLUDGE CAKE RESULTS - TCLP METALS

                          King of Prussia Technical Site
                               Demonstration Run
                           Moerdijk. The Netherlands
                                 July 22. 1992
                               IEA Analyses Only
TCLP
1 Metal
Arsenic
Barium
Cadmium
Chromium
Mercury
Lead
Selenium
Silver
Regulatory-
Standard
(mg/1)
5.0
100
1.0
5.0
0.2
5.0
1.0
5.0
Results
Sample Number (i»g/l)
, 1
<0.61
<14
<0.12
2.1
<0.02
<0.65
<0.11
<0.60
2
<0.61
<17
<0.12
1.8
<0.03
<0.71
<0.11
<0.60
3
<0.62
<48
<0.12
<0.65
<0.02
<1.0
<0.11
<0.60
4 '
<0.63
<37
<0.12
<0.67
<0.02
<0.96
<0.11
<0.63
The TCLP Metal Analyses confirm that the produced sludge cake does not exceed TCLP
regulatory standards.  The sludge cake is not the product of the treatment of any listed
RCRA hazardous waste and does not demonstrate any hazardous characteristics. Therefore,
the sludge cake is not a RCRA hazardous waste.
                                    187

-------
                                    TABLE 9
                        PROCESS WEIGHT COMPARISON

                            King of Prussia Technical Site
                                Demonstration Run
                             Moerdiik. The Netherlands
                                   July 22.  1992

Feed
In/Out as Weighed
' ' (tons) '
163
% Dry Solids
x\
0.844
Dry Solids in Tons
137.5
Products
> 1 cm O/S
< 1 cm O/S
Sand Product
Sludge
Total
Recovery
9.92
12.56
104.20
36.56
163.0
—
0.96'
0.90'
0.953"
0.479"
...
—
9.5
11.3
99.3
17.5
137.5
100%
*_- Measured by DCG and adjusted for overnight drying
" - Measured by HRD

This table documents a mass balance for the demonstration run. The recovery is excellent,
and the mass distribution of produced sludge  is reasonable.
                                    188

-------
                                   TABLE 10
                         CHROMIUM MASS BALANCE

                          King of Prussia Technical Site
                               Demonstration Run
                            Moerdijk. The Netherlands
                                  July 22.  1992

Feed
Dry Solids in Tons
137.5
Cr mg/kg
DCG
770
IEA/
870
Cribs
DCG '
212
IEA
239
Products
> 1 cm O/S
< 1 cm O/S
Sand Product
Sludge
Total
Recovery
9.5
11.3
99.3
17.5
137.5
—
270
540
170
4750
—
...
185
408
172
4615
—
—
5
12
34
166
217
102%
4
9
34
162
209
87%
This table demonstrates the mass balance and recovery for chromium.  The concentrations
for  DCG use all data, while the IEA uses the split sample results.  In both cases,  the
recovery was excellent.
                                     189

-------
                                   TABLE 11
                           COPPER MASS BALANCE

                           King of Prussia Technical Site
                                Demonstration Run
                            Moerdijk. The Netherlands
                                  July 22. 1992

Feed
Dry Solids in Tons
137.5
Cu mg/kg
DCG
1500
IEA
1430
Culfas
DCG
413
IEA
393
Products
> 1 cm O/S
< 1 cm O/S
Sand Product
Sludge
Total
Recovery
9.5
11.3
99.3
17.5
—
—
335
800
350
8030
—
—
300
700
387
7640
...
—
6
18
70
281
375
91%
6
16
77
267
366
93%
This table demonstrates the mass balance and recovery for copper.  The concentrations for
DCG use all data, while the IEA uses the split sample results. In both cases, the recovery
was excellent.
                                   190

-------
                                  TABLE 12
                           NICKEL  MASS BALANCE

                          King of Prussia Technical Site
                               Demonstration Run
                            Moerdiik. The Netherlands
                                 July 22. 1992

Feed
' Dry Solids In TORS
137.5
Ni mg/kg
DCG
460
IBA
430
Niibs
DCG
127
IEA
118
Products
> 1 cm O/S
< 1 cm O/S
Sand Product
Sludge
Total
Recovery
9.5
11.3
99.3
17.5
—
—
970
730
70
2630
—
—
730
498
84
2515
—
—
18
16
14
92
140
110%
14
11
17
88
130
110%
This table demonstrates the mass balance and recovery for nickel. The concentrations for
DCG use all data, while the IEA uses the split sample results.  In both cases, the recovery
was excellent.
                                     191

-------
  The Lurgi-DECONTERRA  Process
              Soil Washing
 Commercml DECONTERRA® plant under construction
Dr. Eckart Kilmer
Ted J. PoUaert
                    192

-------
The DECONTERRA® process
is a multi-stage wet
mechanical separation
process for the remediation of
contaminated soil (organics
and inorganic compounds),
building rubble and railway
ballast.

As extraction agent, water is
used commonly without any
addition of detergents,
solvents, emulsifiers or acids.
      In addition to the size
      separation procedure, it uses
      sorting techniques for
      separating contaminated
      particles from the clean bulk.
      This is why soil types with
      extremely high silt or clay
      contents can still be treated
      economically.
The development of this
process is based on Lurgi's
know-how in the field of wet
mechanical processing of
minerals, especially of ores
and salts, as well as of sludges
from stagnant and running
waters, and on the equipment
experience, which Lurgi has
gained building such plants.
DECONTERRA® Process
       CONTAMINATED SOL (FEED)
       LIGHT MATERIAL
       (CONTAMINATED)
HEAVY MATERIAL
  (CLEAN)
                                                            CONTAMINATION
                                             193

-------
                                Process Description
  The contaminated soil is
  reclaimed, fed to a grizzly,
'  followed by a vibratory
  screen.

  The coarse fraction is crushed
  and, together with the fine
  fraction, is transferred to a wet
  attritioner drum.

  In the attritioner drum, the
  bulk of the adhering pollutants
  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 is matched to the type
  of soil and the character of the
  contamination, up to 10
  kWh/ton being transferred to
  the material. The material
  discharged from the drum is
 screened in several stages.
 The fraction over 20mm is
 returned via the crushing
 cycle back to the attrition
 stage. The l-20mm 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 wood particles, being
 obtained as a pollutant
 concentrate.

 Of the fraction under 1mm, the
finest fraction is cleared from
mud and transferred to a
hydrocyclone, the overflow of
which is concentrated.
 The cyclone underflow,
 together with the deslimed
 coarse fraction of the
 classifier, is introduced into
 the 2nd attrition step and
 subjected to a further
 purification.

 The material discharged from
 the 2nd attrition step reaches a
 flotation system, the special
 reagents of which are
 matched to particular
 requirements of the pollutants
 involved.

 The flotation underflow
 contains the purified material,
 which is subsequently
 dewatered to a residual
 moisture content of about 15-
 20% and can be disposed or
 piled as cleaned product.

 The pollutants are
 concentrated in the froth
 discharged during flotation.
 This froth reaches the
 thickener for pre-dewatering.
 The thickener underflow is
 brought to a moisture content
 of about 20% in a dewatering
 step and being discharged as
 filter cake.

 The pollutant concentrate is
 composed out of the following
process streams:

 1. Light fraction from the
  gravimetric sorting (such as
  wood, tar fragments, coal,
  coke, etc.).
 2. Overflow from the
   hydrocyclone, essentially
   the fines portion of the soil
   with the adsorbed pollutants.

 3. Flotation froth.

 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
 composition of the soil and the
 nature of the contamination,
 the yield of decontaminated
 soil is between  70% and 90%
 of the untreated soil. The
 residual pollutant content in
 the purified soil is clearly
 below the B values of the
 "Dutch list."

 The process water used in the
 Lurgi-DECONTERRA® plant is
 recycled.  The input of fresh
 water, therefore, is limited to
 about 25 gallons/ton,
 depending on feed moisture.
 In order to prevent an
 accumulation of contaminants,
the process water passes  a
water treatment unit, where
fines as well as dissolved
compounds are precipitated
and removed from the circuit.
                                            194

-------
Batch Scale Tests and Pilot Plant
Lurgi has investigated and
cleaned laboratory scale
samples of 50 kg of many
different contaminated soils
from coking plants, gas works
and other locations of
contaminated waste.
The knowledge gained led to
the development of the
DECONTERRA® -process and
to the construction of a pilot
plant with a throughput of
about 1 ton/hour.
The soils tested on a
laboratory scale were also
treated in the pilot plant, in
which the decontamination
results were confirmed.
    Transportable Pilot Plant
                                           195

-------
 Some of the materials tested so far:
PROJECT: Town Gas Plant Site
Material Characteristics:
Silt content: 30%
All values given in ppm
Contaminants
PAH, sum (EPA)
Naphthalene
Anthracene
Pheno anthrene
Pluoro Anthrene
Pyrene
3, 4 Benzo pyrene
Hyrdocarbons
Raw Soil Clean Soil
1,150 16
16 1.6
23 <1
180 1.7
150 2
88 <1
18 <1
1,900 50
Limit
20
200
Reduction (%)
98.6
90
99
99
98.6
99
99
97.3
Clean soil recovered: 85%
PROJECT: Non-Ferrous Smelter Site
Material Characteristics:
Silt content: 28-47%
All values are in mg/kg
Contaminants Raw Soil Clean Soil
Pb
Cu
Zn
Hg
840 42
2,615 28
1,970 87
265 36
Limit
150
100
500
50
Reduction (%)
95
98.9
95.5
86
Clean soil recovered: 75-86%
PROJECT: Ammunition Plant Site /Stadtallendorf
Material Characteristics:
Silt content: 20%
Type of soil: Residue of weathered sandstone
All values are in ppm
Contaminants Raw Soil Clean Soil
PAH
Hydrocarbon
2, 4, 6 TNT
6,400 (max)
1,500 (middle) „, 0.7
Limit
1
Reduction (%)
99.9
Clean soil recovered: 79%
PROJECT: Coking Plant Site
Material Characteristics:
Silt content: 62%
Type of soil: Sandy-clayey, with some coal and coke rests
All values are in ppm, except as noted
Contaminants Raw Soil Clean Soil
PAH
Phenol (ges.)
Cyanide
Total Carbon (%)
Cu
Hg
1,030 13
1.5 0.4
8.9 0.4
5.9 1.8
16 7.2
2 <0.5
Limit
20
0.5
1.0
10
1.0
Reduction (%)
98.8
73
95.5
70
55
75
Clean soil recovered: 76%
Many other materials have been tested. So far we have not run pesticide and/or arsenic
contaminated material, but we are set up to do so.
                                          196

-------
                           Commercial Experience
The commercial plants have a
capacity of 25 MTPH. The first
one started up in June of this
year. It is cleaning up a former
smelter site which is being
converted into an office park.
Its performance confirms the
pilot plant  results.
Another commercial plant with
a capacity of 40 MTPH has
been in operation since June
1990. It removes and recovers
mercury from soil removed
from a former wood treating
factory; now a shopping mall.
A third commercial plant is
cleaning soil at a former coke
oven site. It is in operation
since October 1992.
  Commercial plant under construction. Attritioner drum and clarifier at the right side of the picture.
                                           197

-------
A fourth one, to clean up the
Stadtallendorf TNT, is in
fabrication.

In general these plants are
built up of transportable
modules.
  Commercial plant in operation.
                                               198

-------
Dutch List
     Since the question is asked so often, a copy of the Dutch list is
     reprinted, with permission. In this list

     — A, B, C are indicative values.

     — A —Reference category, natural background

     — B —Threshold value, postulate further investigation.

     — C — Anomalous value, which indicates remediation activities.

' Vorkommen in:
StoH / Konzentration
I.Metalle
Cr
Co
Ni
Cll
Zn
As
Mo
Cd
Sn
Ba
Hg
Pb

II. Anorganische Verbindungen
WlalsN)
F(gosamt)
CNtgesamt-frei)
CNfgesamt-kompl.)
S(gesamt)
Br(gesamt)
PO'(alsP)

HI. Aromatische Verbindungen
Benzol
Athylbenzol
: Toluol
Xylole
Phenole
Aromaten (gesamt)

IV. Polycyclische aromatisclie
Kohlenwasserstoffe (PAKI
Naphthalin
Anthracen
Phenanlhren
Fluoranthen
Boden (nig/kg TS)
ABC

100 250 800
20 50 300
50 100 500
50 100 500
200 500 3000
20 30 50
10 40 200
1 5 20
20 50 300
200 400 2000
0.5 2 10
50 150 600



200 400 2000
1 10 100
5 SO 500
2 20 200
20 50 300



0.05 0.5 5
0,05 5 50
0.05 3 30
0.05 5 50
0.05 1 10
0.1 7 70



0.1 5 50
0,1 10 100
0.1 10 100
0.1 10 100

Grundwasser (ji/ll
ABC

20 50 200
20 50 200
20 50 200
20 50 200
50 200 HOO
10 30 100
5 20 100
1 2.5 10
10 30 150
50 100 500
0.2 0.5 2
20 50 200


200 1000 3000
300 1200 4000
5 30 100
10 50 200
10 100 300
100 500 2000
50 200 700


0.2 1 5
0.2 20 60
0.2 15 50
0.2 20 60
0.2 15 50
1 30 10



0,2 7 30
0,1 2 10
0.12 2 10
0,02 1 5
Vorkommen in:
Stolf / Konzentration

Cryscn
Benzolnl.-iiilhrnzttn
Bcnzo(a)pyren
Bcnzo (k) fluuronlhCMi
n(c|lH)|>
-------
                  Subsurface Remediation Technologies
                                                                                                     VflC
                                                                                           Subsufoce Aorwdiotlon Spccioliso
           Efficient and Cost-Effective Programs for Soils and
                                Groundwater Restoration
Joseph A. Pezzullo, P.E., Terra Vac, Princeton, New Jersey, U.S.A. and Bretton E. Trowbridge, P.E., Terra Vac, Costa
Mesa, CA, U.S.A.
ABSTRACT
  As the nations of the world struggle with the complexities of
cleanup of the hazardous waste which already exists and attempt
to control and manage the discharge of more wastes, they are
caughtin a dilemmabetween a plethora of regulatory guidelines
and the limitations of technology development It is difficult to
enforceregulationsonindustry if thereisno technology available
to remediate the waste.
  UsuaUy.themostseriousformofhazardouswasteisthatwhich
cannot be seen; waste that is in the soils and groundwater.
  The wastes, commonly in theform of petroleum hydrocarbons,
volatile organic compounds, and inorganic compounds, slowly
leach through the soils and into the groundwater where conven-
tional methods of remediation may take decades, and the costs
may become outrageous. More effective means of subsurface
remediationmaybeaccomplishedbytreatmentofthecontamina-
tion at the source and managing the migration of the contami-
nants away from the source.
  Complex subsurface contamination problems require innova-
tive solutions, and fortunately, for the most part, the advances of
technology have kept pace with the regulatory environment
  This paper presents three common, yet complex,  cases  of
subsurface contamination where both soils and groundwater
were restored by the application of innovative and cost effective
remedial measures.

INTRODUCTION
  To citizens worldwide, environmental topics pose major con-
cerns. Industrializednationsbattlethesevere, polluting effects of
hazardous waste discharges while the developing nations tor-
ment with ideas on how not follow in the footsteps of their
brethren while building their own economies.
  How then to manage such a task; to clean up the hazardous
wastewWch already exists andmanagemedischargeofaU future
hazardous waste?
  During the recent Earth Summit in Rio de Janeiro, nations all
over the world took it upon themselves to make a cooperative
effort toward a cleaner environment Some of the time, such
"cooperative efforts" mean individual efforts and voluntary con-
straints, but most of the time "cooperative efforts" result  in
regulations.
  Regulationsmaybethekey.butonefactorishowwekeeppace
between regulations and technology. After all, how can nations
enforce regulations if there are no suitable and cost effective
technologies available to solve to the problem? Hazardous waste
problems are complex and they require innovative solutions.
  Notwithstanding these complexities, most casesof hazardous
waste can be grouped into three general categories; petroleum
hydrocarbons, volatile organic compounds, and inorganic com-
pounds.
  One way to mitigate the problems associated with hazardous
waste in surface water is to treat the waste at the source by
controlling and treating discharges. This is an obvious, healthy,
and aesthetic move toward making the environment pristine.
  However, the silent killers rest beneath the ground surface, in
contaminated soils and groundwater. Contamination which en-
ters thesoileventuallyleachesintothegroundwater; groundwater
which eventually finds its way to fresh water supply wells and also
some surface waters. Subsurface contamination is a major prob-
lem, and to treat it you do the same thing as you do for surface
water. That is, treat the problem at the source and control its
migration.
  This paper presents  three cases where cost effective and
innovative solutions were applied to solve a problem in each of the
three categories of subsurface contamination.
  These cases represent only a sampling of the many technolo-
gies which are available, nevertheless they depict three situations
where cost effective, innovative techniques of source treatment
and migration control were utilized to solve the problems of
subsurface contamination.

SUBSURFACE CONTAMINATION
  Typically, the source of subsurface contamination begins in the
soils. Surface spills, leaking underground storage tanks, heap
leachingfields, and wornorruptured pipelines arejustafewways
in which soils can become contaminated.
  The contamination may exist in several different phases in the
subsurface soils as pictured in Figure 1 (next page). The aqueous
phase is that which is dissolved in the soil's water saturation. If the
contaminantis volatile, the vapor phase fills the soil's air porosity
fraction. Lesser amounts are adsorbed to the soil matrix. In
addition, there also  may be free phase liquid product With all
                                Copyright 1991 by Terra Vac, Costa Mesa. CA, USA

                                                     200

-------
    Hydrocarbon Vapor
Soil
Adsorption
   Moistu
   Droplet
   with
   Dissolv.
   Hydr
  Residual
  Liquid
  Hydro-
  carbon
                                                           :loating
                                                           iquid
                                                          Hydro-
                                                           :arbon
      FIGURE 1: HYDROCARBONS IN THE SUBSURFACE
 these fractions describing the source of soil and groundwater
 contamination, itis essential thatremedial techniques address all
 phases of the contamination.

 PETROLEUM HYDROCARBONS
  Some of the more common contaminants in the subsurface are
 petroleum hydrocarbons. Petroleum production, refining, trans-
 fer, and marketing operations all contribute to the oil, gas, diesel,
 and jet fuel, amongst some contaminants, which are known to be
 in the soils, floating on the groundwater, or migrating off some-
 where in a groundwater plume. Nearly everyone knows of a case
 like this, howeverwhatmany do notrealizeis thatittakes just one
 liter of gasoline to render one billion liters of water non potable.
  One particular case reported by Trowbridge and Malot, 1991,
 involves a residential subdivision  in the northwestern U.SA
 where in 1987, petroleum hydrocarbons were found in an under-
 lying aquifer. The subdivision is located down gradient from
 several petroleum refining, pipeline, pumping, and storage facili-
 ties. The discovery of liquid phase hydrocarbons directly beneath
 a residential area wat, of immediate concern.
  Figure 2 is a site map that shows the location of the hydrocar-
bon plume and the apparent hydrocarbon thickness contours.
The measured liquid phase hydrocarbon thickness varies from
0.06 to 0.2 meters. The unsaturated soils above the water table
were tested, but their low hydrocarbon content confirmed that
the liquid hydrocarbons had indeed migrated to the residential
area along the surface of the water table. Groundwater levels in
the area fluctuate about one meter per year.
  The geology of the site is unconsolidated alluvial deposits
overlying a sandstone and shale bedrock. The alluvium is charac-
terized by discontinuous clayey silts and silty clays to a depth of
            approximately 1.5 meters. The next20 meters are a
            fairly homogeneous medium to course grained
            sand with gravels and cobbles. The groundwater
            table begins at about 10 meters, and the aquifer is
            believed to have a thickness of about 10 meters.
              Initial efforts to recover free floating product
            included aproductpumpingsystem; however, after
            several weeks of operation and obviously much
            more product in  the ground, the product pumps
            were collectively delivering a mere four liters per
            day of product. The radius of influence of the prod-
            uct pumps was insignificant, and the system was
            incapable of drawing the long plume of floating
            hydrocarbons into the recovery wells.
              In the fall of 1988, a pilot test of the innovative
            vacuum extraction system was installed to evaluate
            the effectiveness of the technology for removing
            liquid phase hydrocarbons from the water table.
            Four vacuum extractionwells were installed forthe
            pilot test Design specifications for the test were
            14m3/min vapor flow, and the process system in-
cluded a vapor/water separator, extractionblowerandacatalytic
oxidizer for vapor treatment Figure 3 (next page) is a process
diagram of the pilot system.
  The recovery rate of the four-well pilot test was approximately
50 gallons per day, with a radius of influence of over 30 meters per
well. Over a period of four weeks 4,500 Kg of hydrocarbons, or
the equivalent of about 5,300 liters of gasoline, were removed.
  The success of the pilot system led to expansion into full scale
operations. Several more wells were installed and the process
equipment was upgraded. Interestingly enough,  many wells
showed no evidence of free floating product until vacuum was
applied to the wells, indicating the effectiveness of the vacuum
extraction system in capturing and vaporizing free productmuch
faster than other pumping schemes.
  Figure 4 (next page) is a crossplot of typical wellhead vapor
      FIGURE 2: SITE MAP WITH HYDROCARBON
                          PLUME
                                                      201

-------
                                        1500SCFM
                                      Catalytic Oxkllz«r
  FIGURE 3: PILOT SYSTEM PROCESS DIAGRAM
concentrations for one well during the pilot test The plot is
indicative of the rapid vaporization and recovery of petroleum
hydrocarbons by the vacuum extraction process.
  In all, the expanded vacuum extraction system recovered
approximately 87,400 liters (23,000 gallons) of hydrocarbons
over a 350-day period. Figure 5 shows the removal trend and
cumulative amount of hydrocarbons extracted. These high ex-
traction rates and cumulative removal trends support the effec-
tiveness of vacuum extraction to vaporize liquid phase hydrocar-
bons which were otherwise not removable with traditional pump-
ingtechniques.After350days,thevacuumextractionsystemwas







I,.
i
1




o














i

Q m
D
300 '
Si
^ tf, 200 •
Da
D
100 •
13 cnfl




j


D 0 1 o
(3 _ I U
m a B@

^ D D

INSET

























20



D
Q
D
0 100 200 300
Run Time (days)





400
still recovering hydrocarbons ata rate of 190 to 230 liters per day.

CHLORINATED HYDROCARBONS
  Another common form of subsurface contamination are vola-
tile organic compounds (VOCs). Virtually every manufacturing
process of raw materials, from high tech industries to machine
shops and dry cleaning establishments, have used some sort of
chlorinated solvents for de-greasingor cleaningpurposes. In fact,
of the 1,700 sites on the United States National Priority list (i.e.
Superfund Sites), more than 60% have VOCs as the primary
pollutant
  At one Superfund Site in Michigan, U.SA,  (Malmanis et al,
1989, Malmanis et al. 1990, Trowbridge and Malot, 1991) the
    8000
                                                          S 6000 -
                                                        I
                     100           200
                         Run Time (days)
                                                                                                         300
 FIGURE 4: WELLHEAD VAPOR CONCENTRATION
                      TREND
                                                           FIGURE 5: HYDROCARBON REMOVAL TREND
facility was used as a transfer and storage plant for handling
industrial solvents. The site had 21 buried underground storage
tanks (USTs), and a variety ofVOCs werefoundin the subsurface
soils and groundwater,  including trichloroethylene (TCE),
tetrachloroethylene (PCE), trichloroethane (TCA), methylene
chloride, xylenes, acetone, toluene, ethyl benzene, and 1,1,-
                                                     202

-------
dichloroethylene. Soil concentrations were as high as 1,800 mg/
kg (ppm/, and the area of contamination covered approximately
3,250 ma. The groundwater table is situated at a depth of about 7
m. The site plan is shown in Figure 6.
  The U.S. EPA designated vacuum extraction as the remedial
technology to clean the soils. The cleanup criteria were for soil
samples to be less than 10 mg/kg, with no more than 15% of the
samples above 1 mg/kg total VOCs.
  A vacuum extraction system similar to the schematic in Figure
7 was installed at  the site. The system included 23 vacuum
extraction wells strategically placed throughout the area of the
buried tanks. The system began operations with activated carbon
for vapor treatment; however,  with extraction rates as high as
2,000 kg/day, preparations were made to switch to a catalytic
oxidizer. Since catalytic oxidation of chlorinated compounds had
previously never been permitted at a Superfund Site with VOCs,
                                 Oroundwaur Monitoring
                                    BUM Ing
               FIGURE 6: SITE PLAN
 a source test was completed which demonstrated the effective
 ness of the catalytic oxidizer to destroy the chlorinated hydrocar-
 bons and other compounds. Figure 8 shows the results of the
 source test which achieved an overall destruction rate of 99.8%.
   Free floating non aqueous phase liquids (NAPLs) were also
 identified at the site, but as in case 1, many wells did not contain
 any free product until vacuum was applied. Nearly 0.6 meter of
 free product was identified in selected locations. Once gain, the
 innovative vacuum extraction remedy showed unparalleled per-
 formance for removing free floating contaminants from the
 groundwater table, and Figure 9 (next page) illustrates that the
 vacuum extraction system completely vaporized the free floating
 product within a period of 40 days.
   In all, over 20,000 kg of VOCs were removed form the soils
 within two years. The soils have been declared clean by EPA to 0.1
 mg/kg, yet groundwater treatment continues.  A separate
 groundwater treatment system has been modified to allow Dual
CATOX EFFICIENCY FOR
CHLORINATED HYDROCARBONS

COMPONENT
Acetone
Methylene Chloride
TCA
Benzene
TCE
Toluene
PCE
Xylene
Other
Total
INLET
(ug/l)
7.7
0.8
8.3
0.1
21
18
54
25
1M
319
OUTLET
(ug/l)
ND
ND
ND
ND
ND
ND
0.6
ND
ND
0.6
PERCENT
REMOVAL
	
	
___
	

..
^^
--.
	
99.8
    FIGURE 7: VACUUM EXTRACTION SYSTEM
                    SCHEMATIC
         FIGURE 8: SOURCE TEST RESULTS


Extraction™, or the  simultaneous recovery of vapors and
groundwater fromthesame wells. The Dual Extraction™ system
significantly increased VOC recovery rates from the saturated
zone, and impact of applied vacuum in the wells concurrent with
groundwater pumpinggreatiy accelerated vapor phase partition-
ing of VOCs. Figure 10 (next page) is a schematic of the Dual
Extraction™ system.
  To further accelerate in the groundwater cleanup,  an innova-
tive nitrogen sparging system was installed in October 1991.
Known as SpargeVAC™, this in-situ water treatment functions
just like an in-ground air stripper. As fresh air or nitrogen is
injected into the saturated zone, the air rises through the
groundwater. The dissolved, adsorbed, and liquid phase VOCs
are partitioned into the vapor phase, rise to the unsaturated zone
(i.e. vadose zone), and are recovered by the vacuum  extraction
                                                     203

-------
process. Figure 11 represents the SpargeVAC™ process. Tie
sparge gas used at this Superfund Site was nitrogen, as the high
iron concentrations in the groundwater raised concerns regard-
ing the formation of ferric oxide in the pore space of the aquifer
if regular compressed air was used.
 The sparging system resulted in a 21-fold increase in extracted
              10      20      30
                   Run Time (Days)
40
  FIGURE 9: NAPLTHICKNESS VS. VES RUN TIME


toluene concentrations, and 300% increase in the PCE and TCE
concentrationsbeing extracted in thevaporphasebythevacuum
system.TheEPA'sgroundwaterextractionsystemhad operated
for Syears at200 m3 per day and itwas only effective in reducing
the groundwater concentrations from an initial level of about 19
ppmtoaboutSppm.
  Using the SpargeVAC™ technology, the groundwater concen-
trations were reduced 93% to less than  150 ppb within four
months.There was no comparison between the effectiveness nor
the costs of conventional groundwater pumping and the superior
                    DUAL EXTRACTION™
       FIGURE 11: VACUUM EXTRACTION/AIR
                    SPARGING

performance and lower costs noted with the innovative vacuum
extraction, DualExtraction™ andSpargVAC™ technologies. The
EPAhad spent more money studying the site than the innovative
remedial technologies cost to clean it

INORGANIC COMPOUNDS
  The generation and release of inorganic compounds into the
soils and groundwater are yet more cases of hazardous waste
contamination which have fallen under intense scrutiny by regu-
latory agencies worldwide. Industries such as mineral explora-
tion and metal working industries are vital contributors to an
economy, yet they are also major causes of contaminated water-
ways and estuaries by heavy metals and other compounds.
  However, inorganic compounds undergo several different
geochemical and biological interactions in the subsurface, and
many situations which might appear disastrous at first can be
controlled  and mitigated by innovative  techniques  involving
           chemical oxidation,  natural attenuation, and en-
           hanced biodegradation.
             Oneinorganiccompound, cyanide, iswidelyused
           in theminingindustry to leachgold silver and other
           ores from processed rock. It is well documented
           that cyanide undergoes numerous reactions in the
           subsurface by a variety of geochemical and biologi-
           cal mechanisms. Rouse and Pyrih, 1991 describe
           several different physio-biochemical mechanisms
           in the subsurface in which cyanide may participate.
             In 1991, routine ground and surface water moni-
                                                                   toring activities at a heap leaching mine facility
        FIGURE 10: DUAL EXTRACTION™ REMEDIATION         discovered cyanide contamination in the ground
                                                        204

-------
 and surface water (Rouse and Gochnour, 1992). A schematic
- diagram of the site is depicted in Figure 12.
   Although  the leach pad was constructed with triple liners,
 apparently a silicone seal in one of the drain pipes failed and
 allowed the cyanide-bearing leachate solution to migrate out of
 thepadandinto theunderlyingpadfillmaterialandsoils beneath.
 Thecyarudesolutionsubsequentiymovedalongashallowburied
 alluvial channel and was detected in seep springs near a monitor-
 ing well. The cyanide seepage was found to contain approxi-
        Pump-back trench,    , H, O, Injection trench
                  FIGURE 12: SITE MAP
 mately 80 mg/1 CN WAD (weak acid dissociable). There were
 also elevated quantities of copper, gold, silver, and zinc.
  The mine owner responded immediately,  and a series of
 trenches and ponds were constructed to capture and control
 some of the migration of the cyanide and to treat the contami-
 nated media. Treatment of the seepage was accomplished by
 adding hydrogen peroxide to oxidize cyanide, and a recirculating
sprinkler system which enhanced the ultraviolet decomposition
of cyanide in the water.
  Upon the addition of hydrogen peroxide, the water in the
trenches turned green as the copper cyanide complexes de-
graded and a copper precipitate was formed. In addition, biologi-
cal studies on the soils revealed that they contained indigenous
microorganisms which were capable of biologically degrading
the residual cyanide. The addition of phosphate fertilizer en-
hanced the process, and bacteriological counts confirmed that
      the biodegradation of residual cyanide was successful.
        The result was a rapid decline in cyanide concentrations
      in the seepwater and soils in response to the containment,
      chemical oxidation, and biodegradation processes. Total
      cyanide concentrations were reduced to < 0.1 mg/1 within
      one month.

      CONCLUSION
        The case studies presented herein are representative of
      only some of the many technologies which are available to
      remediate hazardous waste in the subsurface and restore
      the environment Environmental problems involving sub-
      surface contamination are complex; however, innovative
      technologies do exist  to control, mitigate and cleanup
      subsurface contamination.
        Moreover, an aggressive approach toward cleanup of
      thesourcesofcontaminationisnearly always the mostcost
      effective.

      ACKNOWLEDGMENTS
        The authors would like to thank SteveThompson for his
      efforts in preparing the graphics for this paper.

      REFERENCES
        Malmanis, E.; Fuerst, D.W.; and Piniewski, RJ.; 1989,
      "Superfund Site Soil Remediation Using Large-Scale
      Vacuum Extraction," Proceedings HMCRI Conference,
      New Orleans, U.S A, pp. 538-541.
        Malmanis, E.; McClenahan, R., 1990, "Full Scale Vacuum
      Extraction Remediation—Michigan Superfund Site," Air
      and Waste Management Association, 83rd Annual Meet-
      ing and Exhibition, Pittsburgh, PA, June 24-29,1990.
        Trowbridge,B.E.,andMalot,JJ.,1991,"SoURemediation
      and Free Product Removal Using In-Situ Vacuum Extrac-
      tion with Catalytic Oxidation," National Groundwater As-
      sociation, Dublin, Ohio, U.SA
  Rouse, J.V., and Phrih, R.Z., 1991, "Geochemical Attenuation
and Natural Biodegradation of Cyanide Compounds in the Sub-
surface," 	
  Rouse, J.V., and Gochnour, P., "Remediation of Soil and Water
Contaminated By Cyanide Using Peroxide and Biodegradation,"
Proceedings Randol Gold Forum '92, Vancouver, British Colum-
bia, Canada, March 25-27,1992.
                                                      205

-------
             RADIOLYTIC REMEDIATION OF A TCE GROUND SPELL
                      USING AN ELECTRON ACCELERATOR*

               S.M. Matthews, A.J. Boegel, J.A. Loftis and R.A. Caufield,

                        Lawrence Livermore National Laboratory
               P.O. Box 808, Mail Code L-629,  Livermore, CA 95550 USA
 ABSTRACT
       Radiolytic decomposition of chlorinated hydrocarbons and other toxic compounds has
 been experimentally measured using ionizing radiation produced by electron accelerator and
 nuclear  isotope sources.   Decomposition products have  been  identified.   A  portable,
 commercially available electron accelerator was set up at a Superfund site where vapor extraction
 wells  were removing trichloroethylene (TCE)  from a spiU  into the unsaturated soil.  The
 extraction vapor was passed through the accelerator beam to decompose the TCE.  On site
 radiolytic decomposition of TCE vapor using an accelerator is shown to be significantly less
 expensive than filtration of TCE vapor using activated charcoal.
 KEY WORDS
       Ionizing radiation, bremsstrahlung, electron beam, chlorinated hydrocarbons, Superfund
 site,  radiolytic decomposition, VOC, TCE.
 INTRODUCTION
       Toxic compounds and hazardous materials  can be radiolytically decomposed when
 exposed to ionizing radiation.  We have experimentally demonstrated this effect on samples of
 volatile  organic compounds (VOCs) and high explosives dissolved in groundwater and on
 polychlorinated biphenyls (PCBs) dissolved in transformer oil. These results have been reported
 elsewhere.  This report describes results of a trial demonstration at a Superfund site where an
 electron accelerator was used to radiolytically decompose trichloroethylene (TCE) vapor.
 FIELD DEMONSTRATION
       Site-300 is on the National Priority List (NPL) due to a TCE ground spill that permeated
 the unsaturated subsurface soil. This Superfund site was in remediation process using vacuum
 extraction wells that removed subsurface air enriched with TCE vapor. The air was passed
 through an activated charcoal filter to remove TCE before release to the atmosphere.  An
 industrially available, portable, 2 MeV electron accelerator with a beam power of 1 kW was
 brought to the site and connected into the remediation system upstream of the activated charcoal
 filter so that all vapor  extracted from the soil was passed through  the electron beam before
*Work performed under the auspices of the U.S. Department of Energy, DOE Contract Nos.
W-7405-ENG-48.1
                                        206

-------
 reaching  the  filter.   Electric power to the  accelerator was  provided by  a 10 kW input.
 Retrofitting the accelerator and related equipment into the Site-300 system required one week.
 Vapor  samples  upstream and downstream of the electron accelerator were  analyzed  and
 compared with the beam power set at six different levels.  The TCE concentration in the vapor
 was monitored using a field portable gas chromatograph (GC).   Vapor  samples were also
 collected in steel spheres and sent to four independent chemical laboratories for analysis of TCE
 content and reaction products.  Two of these labs were EPA approved to perform a TO-14
 chemical analysis scan for forty different VOCs.
       The hardware that was installed at the site  to radiolytically decompose the TCE vapor
 is schematically shown in Fig. 1.
                                              Electron accelerator
                                              Electron beam
                                              Inflow plpea
                                              Exhauat plpea
                                              Expoaure plenum
                                                     20 fe»t
             Fig.  1.       Schematic of field hardware setup to radiolytically decompose TCE
                          vapor from vacuum extraction wells.

       The exposure plenum (5) was fabricated from sheet steel as a cylindrical chamber with
a 45° conical section at one end.  The plenum diameter was 213 cm with an overall length of
610 cm.  This length was chosen because it is approximately the range of a 2 MeV electron
through atmospheric air.  The plenum structure was placed within a 610 cm deep hole with an
oversized diameter to allow clearance between the sides of the plenum and hole. The apex of
the plenum conical section pointed vertically upwards giving an appearance similar to a missile
in an underground silo.  The electron accelerator was  placed on the ground surface  with  the
beam pointed vertically downward, entering the plenum through a  foil window placed at  the
conical apex. A small shielded structure made from cement blocks  was constructed about  the
accelerator.  This arrangement minimized background radiation exposure to personnel while the
accelerator was  operating.   Background  exposure outside the  shielded  structure was 5
micrograys/hr with beam power at 40 watts.  The  accelerator was operated by two workers
located at a control module placed inside a truck approximately  18 meters away. The control
                                         207

-------
module was connected to the accelerator by a cable.
       The TCE enriched vapor was pumped from the wells by a vacuum extraction pump that
exhausted the vapor at atmospheric pressure.  Plastic PVC pipe was connected to the vacuum
pump exhaust and brought the vapor to the two inflow pipes (3) where it entered the plenum.
The entering vapor was deposited at the chamber floor at atmospheric  pressure and moved
vertically upward toward the two exhaust pipes (4) at the top of the plenum where the vapor
exited.  The vapor was exposed to the electron beam (2) during its passage through the plenum.
The vapor exited the plenum and was routed through PVC pipes to the activated charcoal filter.
Vapor circulation through the system was provided by the vacuum extraction pump.
       Maximum electron beam power was restricted to 400 watts because the  accelerator was
originally designed as an industrial x-ray unit and had to be modified for this demonstration.
The vacuum extraction pump provided vapor circulation through the system at a fixed rate of
270 CFM (127 1/s).  Vapor travel time through the plenum was approximately 2 min. Monte
carlo calculations of the electron beam transport through  the plenum indicated, that at this
circulation rate, the vapor received a dose of 0.9 kGy with the beam power set at 400 watts.
The electron beam power was increased from zero to 400 watts in 80 watt increments to provide
data at six power levels. The  accelerator ran  for one hour at each power level while vapor
circulated through the system so that equilibrium could be established before gas samples  were
collected.
       The TCE concentration  in the unirradiated vapor was measured to be 60 ppmv.  The
TCE concentration in the irradiated vapor is plotted as a function of accelerator beam power in
Fig. 2.
                      04-
                        0
50  100  150  200  250  300   350  400  450

      Electron Beam Power - Watts
             Fig. 2.        TCE concentration in irradiated soil extraction vapor as a function of
                          electron beam power.
                                        208

-------
        These TCE concentration data were gathered using the on-site portable GC and are in
 good agreement with the measurements made on the  collected vapor samples by the  off-site
 analytical laboratories.  These data show that the TCE concentration in the vapor extracted from
 the soil is reduced from 50 ppmv to less than  1 ppmv when the electron beam is operating at
 the 400 watt power level.  Data obtained from off-site analysis of the collected vapor samples
 are shown in Fig.: 3. where  the concentrations  of  organic reaction products are  plotted
 logarithmically as a function of exposure dose.
                Q.
                Q.
                    1.000
                   0.100
                c
               _o
               *J£
                2
               "g  0.010-
                o
                c
                o
               o
                   0.001
                             Trfm«thylb«nz«n«	Q
                          PCE
             Chloroform  	— •
    Carbon Titrachloridc

DlchtorerMthano
                       -10      10      20       50       70

                                  Radiation Dose — kRads
                                         90
              Fig. 3.        Concentration of organic reaction products in irradiated soil vapor as a
                           function of exposure dose.

       The 90 kRad (0.9 kGy) dose shown in Fig. 3 corresponds to 400 watt beam power. The
figure shows that the TCE concentration is  reduced to approximately 0.2 ppmv.  A small
concentration of  perchloroethylene (PCE), initially present in the  vapor at 0.1 ppmv, was
reduced to 0.002 ppmv while carbon tetrachloride, initially at a concentration of 0.01 ppmv was
reduced by only 10%. Three VOCs were produced in low concentration from the TCE reaction
products.  These were chloromethane and dichloromethane which were increased to a maximum
concentration of 0.008 ppmv and 0.020 ppmv respectively and chloroform which was initially
present in the vapor at a concentration of 0.01 ppmv and increased to 0.1 ppmv by exposure to
the electron beam. Trimethylbenzene was produced in the vapor at a maximum concentration
of 0.4 ppmv when irradiated to a dose of 0.35 kGy but decreased in concentration with increased
dose.  It is suspected that the trimethylbenzene was  formed from trace amounts of diesel fuel
that accompanied  the TCE spill.  Small concentrations of acetone at less than 0.1 ppmv were
also produced and then partially destroyed with increasing dose.  It is not understood how the
acetone was formed.
       The net concentration of organic reaction products shown in Fig. 3 amount to much less
that the initial 50 ppmv TCE concentration present in the unirradiated  vapor. Most of the TCE
                                        209

-------
decomposition products are suspected to be inorganic chloride and carbon dioxide.  Indications
of hydrochloric acid corrosion in the plenum were present.  Unfortunately, the collected gas
samples could not be analyzed for hydrochloric acid because of chemical reaction of the acid
with the walls of the stainless steel collection spheres.  The increase in carbon dioxide was
calculated to be only a  small fraction of the normal concentration of CO2 in air and therefore
was not measured.  An  analysis of other inorganic compounds in the irradiated vapor found no
oxides of nitrogen or ozone.  Only oxygen, nitrogen, argon, carbon dioxide and water vapor
were found at nominal concentrations.
       This demonstration was  terminated after six  weeks at the site because the electron
accelerator, which was  borrowed from  another program, had to be returned.
ECONOMICS
       The experience gained with the Site-300 demonstration provided information on hardware
requirements,  fabrication costs, retrofit installation problems, performance parameters,  and
operating expenses.  We were able to estimate the economic performance of an electron beam
remediation system and compare costs with remediation using activated charcoal filtration.  Our
estimates were based upon a system sized to the TCE spill at the Savannah River Site (SRS)
where vacuum extraction wells withdraw subsoil vapor at a TCE concentration of 500 ppmv and
extraction rate of 500 CFM (2361/s).  If the extraction wells operate with an average up-time
of 80%, then 17.5 metric tons of TCE would be removed annually.
       A portable, 2 MeV electron accelerator with a 3 kW beam power can be coupled with
an exposure plenum similar to that used at the Site-300 demonstration.  This system could
operate with an annual up-time of 80% and provide a radiation dose of 3.6 kGy to a TCE vapor
stream flowing at a rate of 500 CFM. This dose is sufficient to reduce TCE concentration from
500 ppmv to below 0.2 ppmv.  Trace organic reaction products formed in the vapor are also
expected at concentrations below this level.  The hardware required to construct this type of
system consists of the accelerator, exposure plenum, and shielded structure.  Operating expenses
consist of salaries for two employees working 24 hours per day, annual equipment maintenance
at 10% of capitalization, electricity costs, and miscellaneous.  The estimated capital costs and
annual operating expenses are summarized in Table 1.
       These data indicate that the capital expenditure for an electron beam remediation system
is approximately $975,000 with an annual operating expense of approximately $465,000.
                                         210

-------
              Table 1.       Costs for electron beam remediation system.
                           TCE vapor concentration - 500 ppmv.
                           Vapor flow rate - 500 CFM (236 1/s)
            Capital Equipment
            3 kW electron accelerator
            Exposure plenum
            Shielded facility
            Retrofit installation
            Miscellaneous
            TOTAL CAPITAL COST

            Annual Operation
            Labor - two workers 24 hours per day
            Maintenance at 10% capitalization
            Electricity at $0.09 per kWH
            Miscellaneous
            TOTAL OPERATING EXPENSE
Capital Costs
 $750 K
  75
  25
  50
  75
 $975 K

Annual Operating Expense
 $300 K
  98
  19
  50
 $467 K
       The  TCE  could  be  removed  from  the vapor  using  activated  charcoal filters.
Approximately 623 metric tons of activated charcoal would be needed annually to remove 17.5
metric tons of TCE from  the vapor.  This estimate assumes that the charcoal is recycled  twice
before disposal. The TCE concentration in the filtered vapor would be 1 ppmv which is higher
than with the accelerator system.  The cost for activated charcoal and disposal when loaded with
TCE is estimated at $12.50 per kg.  The annual charcoal cost for the system is $7.78 million
or more than an order of magnitude higher than the  annual operating cost of the accelerator
remediation system.
REFERENCE
Matthews,  S.M.,  Boegel,  A.J., et.al.,  (1991),  High-Energy Irradiation  of  Chlorinated
Hydrocarbons.   ANS Topical Conference MARC-H, Kona.
                                        211

-------
Rethinking the hydraulic cage concept for hazardous
waste disposal facilities  and mitigation of  contaminated
sites1
Charles Voss
Associate, Colder Associates Inc., Redmond, Washington, USA
Abstract

In comparison with conventional hazardous waste disposal facilities, a hydraulic cage facility
relies on a zone of increased hydraulic conductivity around the control region to redirect
groundwater flow around the disposed waste.  Waste containment is achieved by decreasing
the hydraulic gradient through the disposal facility. This approach is potentially more reliable
than conventional engineered barriers, such as a clay layer, because the consequence of a
breach in a low conductivity barrier is much greater than a hydraulic cage.

This paper discusses a series of analyses that were performed to investigate the performance
of a hydraulic cage in a number of rock types.  Three different methods for constructing a
hydraulic cage are considered. The percentage of the groundwater a cage can divert around
a disposal facility is very sensitive to the contrast between the hydraulic conductivity of the
cage and the conductivity of the surrounding rock mass. The greater the contrast, the better
the performance. The analyses demonstrate that cages constructed using surface-based
methods can successfully improve the performance of a waste disposal site but are inferior to
underground construction techniques in most rock types.  Surface-based methods are most
appropriate in rocks having a low hydraulic conductivity.


1  Introduction

A "hydraulic cage" is an engineered structure that short-circuits groundwater flow within a
site by providing a flow path with much lower resistance to flow than the natural system.
The groundwater is redirected around a central region, isolating the enclosed zone from the
regional groundwater flow system.

The hydraulic cage concept has potential application as part of waste disposal and
   1  This work was supported by the U.S. Department of Energy under contract number 87-
3497.
                                       212

-------
 contaminated site remediation. In both instances, movement of hazardous substances
 through a saturated zone takes place by advection processes. Advection, or groundwater
 flow, is governed by the hydraulic conductivity of the medium (rock or soil) and the pressure
 gradient over the site. The relationship is defined by the simple equation


                                        Q=KiA
 where Q is the volumetric flow rate, K is the hydraulic conductivity of the medium, i is the
 gradient, and A is the cross-sectional area of flow. In order to reduce the movement of
 materials either from a disposal site or a contaminated area either the conductivity or the
 gradient must be reduced.

 Conventional engineering methods typically rely on low conductivity barriers such as grout
 or bentonite slurry walls.  The difficulty with this approach is ensuring low hydraulic
 conductivity for all possible pathways through the area over long-time periods. A hydraulic
 cage produces a low gradient within the cage boundaries creating a stagnant groundwater
 flow zone. The head (pressure) in the cage will be approximately equal to the mean value of
 the heads at the upstream and downstream portions of the cage.  As a result, the water in
 the cage adjacent to the upstream portion will be lower than the site and water will drain
 into the cage.  The opposite is true at the downstream part of the cage where the water will
 be discharged back into the surrounding medium. Without the pressure gradient,
 contaminant transport within the cage will be dominated by diffusion, a very slow process.

 The original concept of the hydraulic cage originated in the early 1980's in the Swedish High-
 Level Nuclear Waste Program  (SKB, Svensk Karnbranslehantering AB) (Figure 1).  The nuclear
 waste is stored in an array of inclined  boreholes within a central storage chamber. The
 chamber is surrounded by an excavated slot containing compacted bentonite and sand to seal
 any fractures and provide a low conductivity barrier to flow and transport.  Outside the
 bentonite-sand barrier is the hydraulic cage consisting of a series of tunnel and boreholes that
 reduce the gradient within the storage chamber by short-circuiting the groundwater flow
 around the repository. The WP-Cave design employs a redundant barrier approach for
 reducing the potential for groundwater flow and transport by providing a conventional
 engineered barrier to reduce the hydraulic conductivity and a hydraulic cage to reduce the
 pressure gradient.

The approach was eventually abandoned for high-level radioactive waste  disposal
 applications because the heat produced by the decaying waste would create thermal
 gradients inside the hydraulic cage. The resulting groundwater flux would potentially negate
 the isolating capability of the hydraulic cage. The concept is still attractive for other types of
waste disposal facilities that do not generate heat. A hydraulic cage could also be used to
enhance the performance of some in situ remediation technologies such as microbial
                                       213

-------
           Figure 1. Original WP-Core Concept
                                                      ,-Tunnel
                                                       Waste
                                                       Emplacement
                                                       Boreholes

                                                       Bentonite


                                                       Boreholes
degradation by increasing the resident time of the contaminant plume.

Some of the additional advantages provided by the hydraulic cage include:

       «      The reliability of a hydraulic cage should be greater than a low permeability
              barrier.  The consequence of a breach in a traditional, low permeability barrier
              is potentially more significant than a partial plugging of the hydraulic cage.

       •      A more favorable environment for engineered barriers such as waste canisters
              and buffer materials.

       •      The cage could be used for dewatering of the waste containment facility during
              construction, waste emplacement, and backfilling of the facility.

       •      The behavior of a hydraulic cage can be measured directly and with more
                                         214

-------
             confidence than low hydraulic conductivity barriers.

This paper discusses a series of analyses that were performed to investigate the performance
of a hydraulic cage in a number of rock types. Three different methods for constructing a
hydraulic cage are considered. Section 2 reviews earlier analyses that were performed to
evaluate the performance of the WP-Cave design.  Section 3 presents two alternative
approaches for constructing a hydraulic cage using surface-based methods. The performance
of the alternative cage designs is evaluated in Section 4. Section 5 discusses additional
applications of the hydraulic cage concept in waste disposal facilities.
2  WP-Cave Performance

Earlier analyses of the WP-Cave design for nuclear waste disposal were performed by Strack
(1985), and Axelsson and Olsson (1985). Two- and three-dimensional groundwater flow
models were developed to assess the relative contribution of the two barriers (bentonite-sand
and hydraulic cage) and the sensitivity of system performance to various design assumptions
such as the number of boreholes in the cage and the thickness of the bentonite-sand barrier.
The hydrologic conditions at the site were assumed to be homogeneous and isotropic. A
relatively low hydraulic conductivity was assigned to the site based on measurements in
crystalline rock (the candidate repository sites in Sweden are located in crystalline rocks).
hydraulic cage was assumed to have a diameter of 150 m and, in the three-dimensional
model, a height of 300 m.
The
Several borehole spacing designs were considered. The minimum spacing considered was
less than 2 m with a maximum of 23.5 m.  The hydraulic head in the tunnels connecting the
borehole drains was assumed to be constant (i.e., there is no resistance to flow inside the
cage).

The contributions of the hydraulic cage and the bentonite barriers to controlling groundwater
flow were assessed separately. The results of the hydraulic cage model were as follows:

      •      The percentage of the groundwater flow that passes through the zone within
             the hydraulic cage is reduced by 70 to 99% depending on the borehole spacing.
             The closer the spacing the better the performance.

      •      Cage performance decreases as the resistance to flow within the cage
             (hydraulic conductivity) increases. Plugging of boreholes or backfilled tunnels
             would reduce the amount of water diverted by the cage.

      •      The effectiveness of a hydraulic cage will vary depending on the site
             conditions. A hydraulic cage will provide a more effective barrier in low
                                        215

-------
              conductivity sites.

 The numerical model of the bentonite-sand barrier assumed a barrier radius of 50 m.
 Variables in the analyses included the thickness and hydraulic conductivity of the barrier.
 Barrier thicknesses of 1 m and 5 m were modelled with hydraulic conductivities varying by
 one order of magnitude.  The results are summarized in Table 1.

 Table 1.  Results of the Bentonite-Sand Barrier Analysis
Thickness (m)
5
5
1
Kbuife/Ktock
0.1
0.01
0.01
% Passing Barrier
63
12
52
The analyses demonstrate that the percentage of groundwater flow that passes through the
interior of the bentonite-sand barrier is sensitive to the hydraulic conductivity contrast
between the barrier and the host medium.  An order of magnitude decrease in the Kb^^/K.^
ratio improves the performance by a factor of 5. The thicker barriers would be required in
low conductivity rock (e.g. crystalline rock containing few fractures) because of the low
conductivity contrast that can be achieved.  The barriers would be more effective for sites
with relatively high hydraulic conductivities where the contrast would be greater.

The relative contributions of the hydraulic cage and bentonite-sand barriers depends on their
conductivity contrast with the site and the efficiency of the hydraulic cage.  The contribution
of the bentonite-sand barrier would be minimal in a crystalline site with a relatively low
hydraulic conductivity (10"' to 10"12 m/s) and a cage with 3 m borehole spacing.  As the
borehole spacing increases, less water is diverted by the cage and the contribution of the
barrier increases significantly. The performance of the two barriers will vary significantly
depending on site conditions. For sites with higher conductivities, such as sandstones or
fractured systems, the isolation capability of the bentonite-sand barrier would be similar or
better than the hydraulic cage.


Alternative Methods for Constructing a Hydraulic Cage

The earlier  analyses  summarized above demonstrated that the WP-Cave design is capable of
isolating disposal facilities from the groundwater flow field.  The analyses were limited to
crystalline sites and  assumed the connectivity of the cage to be ideal (i.e., there was no
resistance to flow). This section discusses two alternative methods for constructing a
hydraulic cage: hydraulic fracturing and blast fragmentation. The designs employ vertical
boreholes, similar to the WP-Cave concept, but rely on surface-based methods for
                                        216

-------
 hydraulically connecting the boreholes. The alternative approaches are intended as possible
 means for enhancing clean-up activities at contaminated sites or as part of waste disposal
 facilities where the original design may be prohibitively expensive.

 Blast Fragmentation

 The effectiveness of the hydraulic cage relies on developing a highly conductive and
 continuous region within the rock mass. The WP-Cave design employs a series of tunnels to
 connect the drainage holes within the cage and provide a low resistance pathway for
 diverting the groundwater flow. An alternative method for connecting the boreholes would
 be to fragment the rock between the boreholes by detonating explosives in the  boreholes.

 The feasibility of successfully constructing a network of interconnected boreholes using
 explosives depends on several factors:

       •      The ability to fragment the rock over considerable distances.

       •      The magnitude and direction of the in situ stresses at the site.

 The ability to create and propagate fractures with explosives decreases as the distance to a
 free face or relief surface increases. For the case under consideration, the lithostatic and
 tectonic stresses will tend to increase with depth with a corresponding increase in the rock
 mass strength.  In general, blast fragmentation zones having high hydraulic conductivities
 will be difficult to construct at depths below 100 m.

 The size and shape of fragmentation zones will also be influenced by the rock quality.
Fractures  propagated during a blast will tend to terminate when they intersect existing
 discontinuities such as joints or bedding planes. The impact of such features will depend on
 their orientation and frequency.  Vertical discontinuities oriented perpendicular to the
 direction between adjacent boreholes would limit the extent of the blast induced fractures.
The impact of discontinuities will be less important in sites where the mean spacing of
 discontinuities is greater than the distance between boreholes.

The in situ stress state will also influence the shape of the fragmentation field. In highly
anisotropic stress fields, fractures tend to propagate preferentially in the direction parallel to
the maximum principal stress.  Notched boreholes or shaped charges could be used to focus
the energy between boreholes or the spacing between boreholes parallel to the minimum
principal stress could be decreased.

In addition to a well connected set of fractures, the hydraulic conductivity of the network
must be significantly  greater than the surrounding rock mass in order to  be an effective cage.
The conductivity of a blast fracture will depend on the shear dilation that occurs and the
intact strength of the rock.  Dilation  resulting from the dislocation of the fracture surface
                                          217

-------
causes the conductivity of the fracture to increase. This effect is less pronounced in weaker
rocks where crushing of asperities would reduce the dilation of the fracture.

Hydraulic Fracturing

Hydraulic fracturing has been used extensively in the petroleum industry for improving
production in oil and  gas reservoirs. A segment of the  borehole containing the formation is
isolated with inflatable packers and the zone is pressured until the stresses exceed the
strength of the surrounding rock.  The resulting fracture is propagated by pumping
additional fluid into the region. The fracture can be propped open to increase its aperture
(and, therefore, the conductivity) by including sand, bauxite, or ceramic particles in the
pressurizing fluid.

The use of hydraulic fracturing for constructing a hydraulic cage is  limited by many of the
same constraints as blasting. Existing discontinuities will tend to limit the distance a
hydraulic fracture can be propagated and the significance of these features will depend on
fracture orientation and frequency.

The in situ stress field has a large influence on the orientation of the induced fracture plane.
Technology is available for controlling the initial geometry of hydraulic fractures although the
fracture will tend to realign parallel to the maximum principal stress as it propagates.
Nevertheless, hydraulic fracturing experts from the petroleum industry estimate that existing
hydraulic fracturing techniques could reliable connect boreholes 5 m apart even in sites with
high deviatoric stresses.


Performance  of Alternative Designs

In order to consider the performance of hydraulic fracture and blast fragmentation cages,
additional analyses were performed. Four different rock types were considered based on
conditions at a number of DOE sites.  A probabilistic approach was used in order to account
for the uncertainty in the site conditions and the hydraulic properties of the cages.  The
assessment consisted of the following:

       1.      Two-dimensional finite element analyses  to develop a response surface
              describing the relationship  between the cage performance and the hydraulic
              conductivity of the cage and rock mass.

       2.      A literature review to develop probability distribution functions for the
              hydraulic properties at different sites.

       3.      A probabilistic analysis of cage performance using the Monte Carlo method
              and the results from steps 1 and 2.
                                        218

-------
 This approach reduced the number of numerical models required and provided an estimate
 of the range in performance that can be expected.

 Numerical Analysis

 A two-dimensional finite element flow model was used to evaluate the relationship between
 cage performance and the hydraulic conductivity contrast between the cage and the rock
 mass.  A two-dimensional model was chosen based on the results of Strack (1985) who
 demonstrated that two-dimensional and three-dimensional models give very similar results
 especially when the groundwater flow is assumed to be horizontal (perpendicular to the
 cage).

 The conceptual model assumed a cage diameter of 125 m with a uniform hydraulic
 conductivity. The model did not include other types of barriers. The hydraulic conductivity
 of the site was assumed to be 10ru m/s with a horizontal hydraulic gradient equal to 0.002.

 The response surface was constructed by substituting a range of hydraulic conductivity values
 for the cage and monitoring  the amount of water diverted by the cage. The greater the
 amount of water diverted by the cage, the better its performance. Cage performance Pa  is
 given by
                                       cage
where CL,^ is the flow through the interior of the cage and Q^ is the flow through the same
region if the cage is not present. The range of hydraulic conductivity values for the cage was
10.]2 to 10., m/s. The results of the analyses are summarized in Figure 2.

The results illustrate that the amount of groundwater passing through the cage is markedly
reduced as the conductivity contrast between the cage and rock mass is increased. An order
of magnitude difference reduces the flow through the cage by only 15% while the
corresponding reduction for a cage with a two order of magnitude difference is more than
50%.  The rate of change in performance slows considerably with additional increases in the
conductivity contrast  The analyses by Strack (1985), and Axelsson and Olsson (1983),
assumed the head was equal at all points within the cage (i.e., the cage offered no resistance
to flow).  Their results correspond to K^ values > 104 m/s.
                                       219

-------
                         100%

                         90%

                         80%

                         70%
             Hydraulic Cage QQO^
             Performance
             Index P      50%

                         40%

                         30%

                         20%

                         10%

                          0%
        site
Based on Continuum
Simulation with K roct<
10'12 m/s
Rtted Polynomial

    Simulation Result
                               10'12 10'11  ID'10 10'9  10'8  10'7
              10"'
                                    Hydraulic Conductivity of Hydraulic Cage (m/s)
                                                  'cage
            Figure 2. Results from 2-D Model of Hydraulic Cage
A simple relationship between cage performance and the ratio of conductivities
was developed from a regression analysis of the results.  The expression is given by:
Values for constants c, through cs are provided in Table 2.  The parameter R is the ratio of
conductivities and is defined as:
                                              220

-------
 Table 2.  Constants for Pcage Relationship.
Constant
cl
c2
c3
c4
c5
c6
Fitted Value from
Regression Analysis
6.94
-0.514
0.0176
-42.52
-14.32
8.873
The equation for Page was used to evaluate the performance of hydraulic cages in a range of
rock types. Four rock types, limestone, sandstone, basalt, and weathered limestone, were
chosen based on literature reviews of documents describing the site conditions at several
Department of Energy (DOE) defense sites.  The dependent variable K,ite was estimated from
DOE site documents and conversations with hydrogeologists familiar with the areas. Because
of the uncertainty related to these estimates, the parameter was represented by a probability
distribution function that reflects the uncertainty about the expected value.  Triangular
distribution functions were assumed.   The minimum, maximum and expected values for K^
are listed in Table 3.

The hydraulic conductivity of a cage was assumed to be a multiple of the site conductivity.
The increase varies according to the rock type and construction method. In general, fracture
dilation in hard rocks, e.g., basalt, is greater than in sandstones or other soft rock.  Therefore,
the conductivity of a hydraulic cage in the basalt site was assigned a higher range of values
than the limestone  and sandstone rock types. The expected increase in conductivity for blast
fragmentation cages was assumed to be greater than hydraulic fracture cages because of the
greater number of fractures involved.  The probability distributions for K^g. are also
triangular. Table 4 contains the minimum, expected, and maximum values for calculating the
cage conductivity.
                                         221

-------
Table 3. K,te Values for Monte Carlo Analysis
Rock Type
Limestone
Sandstone
Basalt
Weathered
Limestone
K*. (nVs)
minimum
5 x 1010
10*
10-io
to-7
expected
3.2x10*
10s
10-9
10s
maximum
SxlO"7
lo-4
lO"7
lo-4
Table 4.  Distributions for K^ multiplication factor.
Rock Type
Limestone
Sandstone
Basalt
Weathered
Limestone
Blast Fragmentation
(x'sK.J
min
75
10
75
75
exp
750
100
1000
500
max
1000
1000
105
5000
Hydraulic Fracture
(x's K,, J
min
10
10
10
10
exp
100
50
250
100
max
1000
500
5000
750
Monte Carlo Analysis

The relative performance of hydraulic fracture and blast fragmentation cages in the four rock
types was assessed using the polynomial relationship for Pagt. Values for R were calculated
by sampling from the distributions for K,,te and K^. The Monte Carlo method was used to
sample from the distributions and the process was repeated 1000 times to obtain a probability
distribution of cage performance for each of the rock types.

The results of the analysis are summarized in Figure 3.  Each of the box plots represents the
results for a different combination of cage type and rock type. The box represents the .25 to
.75 percentile values for Pagt and the horizontal line the .50 percentile value. The vertical lines
                                          222

-------
         Qcage /Qsite
                          Blast   Frac  Blast   Frac   Blast   Frac   Blast   Frac
                          Limestone    Sandstone       Basalt      Weathered
                                                                  Limestone
          Figure 3. Results of Monte Carlo Analysis of
          Hydraulic Cage Performance
extending above and below the box represent the range of values between the .05 and .95
percentile values.

The magnitude and range of the hydraulic cage performance results varies considerably
depending on the rock type and construction method.  The blast fragmentation cage
performed significantly better than the hydraulic fracture cage due to the higher conductivity
that could be developed. The difference in performance is less noticeable in the basalt rock
because of the greater contrast in conductivities.  The results show that there is a significant
probability that a hydraulic fracture cage would divert a relatively small percentage of the
groundwater flow in most of the rock types. The performance of the blast fragmentation
cages is much better and fairly consistent within the different rock types with the exception
of the sandstone which has the lowest conductivity of the rock types considered.
                                         223

-------
 5  Discussion

 The expected performance of the alternative hydraulic cage designs is considerably poorer
 than the original design. This reflects the difficulty in developing highly conductive fracture
 networks in situ and the optimum conditions that can be engineered in an underground
 facility. Nevertheless, the results indicate that the surface based methods could divert a
 significant portion of the groundwater flow around an area and contribute to the isolation
 capability of a disposal site retarding the movement of contaminants.  Therefore, the
 hydraulic cage represents an additional engineered barrier concept that should be considered
 as part of an underground disposal facility.

 Underground  disposal facilities could be constructed using old underground mines. The
 hydraulic cage could be developed using existing openings that are adjacent to a disposal
 room or by developing new tunnels. An alternative would be to construct high conductivity
 panels that could be placed inside the waste disposal rooms. The high conductivity panels
 would be attached to the walls of the disposal rooms and low conductivity materials, such as
 bentonite clays, would be added to separate the panels from the waste. The combination of
 low and high conductivity barriers would improve the reliability of the inner clay barrier by
 decreasing the gradient inside the storage area and reducing the significance of a breach.  A
 similar approach could be used for surface disposal facilities (Figure 4).


 6  References

 Axelsson, C-L.  and T. Olsson, 1985. "Modeling of groundwater flow and streamlines through
 a WP-Cave." Geosystems AB, Uppsala, Sweden.

Strack, O.D.  1985. "Report on Phases 1  ad 2. WP-Cave ground water model."  Itasca
Consulting Group Inc., Minneapolis, MN.
                                         224

-------
  Cover

  Hydraulic Cage

  Slurry Wall-
  HOPE Liner

  Clay Liner
Figure 4. Conceptual Design of Surface Waste Disposal
Facility Incorporating Hydraulic Cage Barrier
                               225

-------
             Innovative Contracting Strategies for Equipment Procurement
                             Bofors Nobel Superfund Site
                                    Muskegon, MI

                          Ted H. Streckfuss and Craig Olson
                    U.S. Army Corps of Engineers, Omaha District
I.  Site History:  The Bofors Nobel Superfund Site is located six miles east  of downtown

Muskegon, Michigan along  Evanston Road in Edelston Township.  The site is located in a

chemical manufacturing area and is currently owned and operated by Lomac, Incorporated. The

superfund site has been divided by Region V of the Environmental Protection Agency (EPA) into

two operable units, designated as Lagoon Area Operable Unit (LOU), and the Groundwater/Plant

Area Operable Unit (GOU).  The LOU is irregularly shaped and includes ten abandoned sludge

lagoons  and associated soils south  of the Lomac facility.   The POU  area consists  of the

contaminated groundwater located beneath the entire 85 acre site. The southern boundary of the

site is bordered by Big Black Creek, which has been chosen as the selected point of discharge.

       Lakeway Chemicals, Incorporated began producing industrial chemicals in 1960. During

this period, the manufacturing facility produced such chemicals as 3,3-dichlorobenzidine (DCB),

benzidine, and azobenzene for use in alcohol based detergents and as die intermediates.  Raw

materials used in the production of these components 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 late 1960's and  1970's, DCB, zinc oxide and other process wastes were

discharged through open trenches to  the lagoons located south of the plant area.  The lagoons

functioned as either direct discharge basins or overflow settling ponds.  The basins do not
                                         226

-------
 themseleves contain a homogeneous waste material.  During the 1970's, lagoon berm failures



 resulted in the discharge of lagoon sludge into Big Black Creek.  The use of lagoons for waste




 disposal was discontinued in  1976  when the facility was connected to the Muskegon County



 Wastewater Treatment System.




       In 1977, Lakeway merged with Bofors Industries,  Incorporated and the company name



 was changed to Bofors Lakeway, Incorporated.  In 1978, the State of Michigan filed a Consent




 Judgement against the company. The consent judgement required Bofors Lakeway to remediate



 the contamination which had occurred both on and off site. In 1981, a Consent Judgement was



 entered into between the State of Michigan and Bofors Lakeway, Inc. calling for full restoration




 of the air, land, and waters of the State contaminated through previous operations at the Bofors



 site.




       In 1981, Bofors Lakeway, Inc. merged with Nobel Industries of Sweden and the company



 name was changed to Bofors Nobel, Incorporated.  In order to conform with the requirements



 stipulated   within  the  consent judgement,  Bofors and  other  investors  established  the



 Environmental Systems Corporation of Michigan (ESCM)  to own and operate the PACT™




 (Powdered Activated Carbon Treatment) system and a clay  lined waste disposal landfill required



 by the judgement.  The disposal landfill was constructed  on site to accept ash from a sludge



pyrolysis process.  Before the sludge pyrolysis system was implemented, however, Bofors Nobel



and ESCM filed for bankruptcy.




      The Bankruptcy Court granted permission to Bofors Nobel and ESCM to procure a



broker and pursue a potential buyer.  Sale of the Bofors assets to Lomac was completed in 1987.




As a part of the sales agreement, an "Agreement and Covenant Not to Sue"* was also entered
                                          227

-------
into between the State and Bofors Nobel for the previous contamination of air,  ground,  and




water caused by facility operators. These agreements and covenants allowed Lomac to operate




the facility independently of previous site activities.



       Because of the nature and extent of the contamination on the Bofors Nobel site, the State



of Michigan contacted the U.S. EPA to evaluate the site for possible placement on the National




Priorities List (NPL), pursuant to the Comprehensive Environmental Response, Compensation



and Liability Act (CERCLA)  as amended by the Superfund Amendments and Reauthorization



Act (SARA).  The EPA conducted their site evaluation, with the results calling for nomination



to the  NPL in  July of  1988,  with  listing  occurring  in March  1989.   A Remedial



Investigation/Feasibility Study (RI/FS) was initiated in August 1987 and the Record of Decision



(ROD)  for the Bofors Nobel site  was  signed in September 1990.  The ROD addressed




remediation of the lagoons, as well as restoration of the groundwater aquifer.








n. Problem Background: One of the primary concerns associated with the contracting effort



on the Bofors Nobel  Superfund Site was the procurement of equipment  required to treat the



contaminated groundwater. The previously referenced Feasibility Study and the project ROD



called for Ultraviolet Oxidation (UV Oxidation) as the selected technology for destruction of



organic contaminants  contained in the groundwater.  The option for effluent discharge spelled



out in the ROD mandated a restrictive discharge requirement which had not been considered in



the Feasibility Study.  The Feasibility Study had considered the groundwater treatment system



as necessary only for "pretreatment" prior to discharge to the sanitary sewer.  The final ROD,



however,  directed  final effluent discharge to a cold water trout stream located on site.  Rapid
                                         228

-------
 and complete evaluation of the chosen technology was performed to determine whether the low




 discharge levels  for  the  organic  contaminants  contained within the groundwater could be




 achieved. The results of this study indicated that the technology was suitable for application on




 a waste of this type.




        The relatively recent development  of the UV Oxidation  technology  has limited the




 corporate competition within this field.   To date, only three  vendors have exhibited the




 capabilities necessary to effectively treat a groundwater contaminated with both volatile and




 semivolatile constituents on a full scale basis.  An important factor to consider  when reviewing




 this treatment technology is the high operational and maintenance costs (O&M) associated with




 its use.  In the case of the  Bofors Nobel groundwater treatment system, the O&M costs for the




 UV Oxidation system were projected to comprise the majority of the annual operational expense.




 Because of this fact, consideration  was given to life cycle cost rather than simply capital cost




 during the design and predesign phase of the project.   By promoting a  predesign effort that




 emphasized treatment capability  as well as life  cycle  cost, the most effective system could




 ultimately be procured for the contract. Based upon the outcome of the predesign testing, a sole




 source justification would be initiated to procure that UV Oxidation system deemed to be most




 appropriate for this specific groundwater application. Justification would be contingent upon the




 capability of the system to treat the contaminated groundwater to acceptable levels, and life cycle



 cost to the end user.









HI.  Nature of the Corporate Technologies:  At the time contractors  were being  solicited for




the treatability  and predesign  work associated with this superfund  site, there were essentially
                                            229

-------
three vendors who had demonstrated proven  technologies at non-related  sites.   The three




corporations have expended a great deal of energy in the promotion of their specific technology




as the "best" for implementation under conditions similar to those present at the Bofors site.




       The  traditional  method  for  determining full  scale  operational parameters  for  UV




Oxidation has been to provide a specific vendor with samples of the contaminated water.  The




vendor  is responsible  for  sampling  and testing the water sample for various  water quality




parameters,  and to then use laboratory scale equipment to determine  the optimum operational




parameters for full scale equipment.  An alternative that each of the vendors has available is use




of a pilot scale treatment system which is mobilized  to an individual site.  An on site pilot scale




system yields more accurate results regarding upsizing criteria for full scale implementation than




does bench scale treatability testing.




       Following the gathering of this predesign data,  the project design is normally developed




in a generic  fashion, whereby any one of the three vendors would have an equal opportunity to




bid on a project and potentially be awarded the  contract for supply,  based upon the lowest  bid.




The plans and specifications are normally prepared with a performance requirement, so that the




successful bidder is contractually obligated  to perform to predetermined levels.  The train of




thought on the Bofors project was to provide an opportunity to all vendors to travel to the site,




treat the contaminated groundwater,  and propose a  full scale system suitable for the treatment




of the contaminants. Based upon the validated laboratory data and the vendor reports, a decision




regarding  the  most appropriate vendor would be made, and the project design would proceed




around a specific vendor, rather than around a generic performance requirement.  Life cycle  cost




would figure highly in the decision making  process.  It was felt that the costs associated with
                                           230

-------
 the predesign effort would be more than offset by the savings associated with annual O&M over




 the life of the project.









 IV. Lessons Learned Guidance:  At the same time that the Bofors Nobel project was planning




 its predesign requirements, a nearby superfund site was performing treatability testing on a waste




 stream of similar quality and clarity.  The choice of a UV  Oxidation vendor for the other




 superfund site's treatability testing was based upon which vendor provided the lowest quote for




 performing the work.  The results of the bench scale treatability testing by that specific vendor




 indicated that the O&M requirements for a UV Oxidation system would exceed the initial capital




 cost of a complete system within  approximately two years.   Because of the excessive cost




 associated with the UV Oxidation technology at this site, this treatment alternative was removed




 from  consideration.  Other UV Oxidation vendors claimed they could treat the waste stream at




 a significantly lower cost but were not considered because of time constraints.




       Based on the results of the bench scale testing at the other superfund site, it was felt that




 the alternate contracting  techniques previously discussed should be investigated for application




 at Bofors Nobel.  Because each  individual  UV Oxidation  vendor  holds  patents on  their




 equipment and all market their equipment as "significantly different" from the other competitors,




 it was determined that the most appropriate course of action would be to compare each of the




 three  available UV Oxidation  vendor's products  under identical conditions.  This decision led




 to a "treat-off between the vendors at the Bofors Nobel site.  All  vendors were contracted to




be  on site during a specific  time  frame,  and all vendors  were  provided  the contaminated




groundwater from a  common header.   The  results  of the testing  period  are discussed in
                                            231

-------
subsequent paragraphs.








V.  Buy American Act Applicability: One of the proven UV Oxidation companies selected for




the on site treatability testing is wholly owned and manufactured outside the United States.  This



was a concern because of potential conflicts with the Buy American Act restrictions contained



within the federal procurement regulations. If the non-American produced equipment was found



to be most competitive, on the basis of treatment capability and cost effectiveness, the ability



and legality of a government contract to purchase the process equipment was questionable.  This



issue was quickly resolved upon further investigation within the Corps.  Reference to Part 25,



subpart 25.1, section 25.101 of the Federal Acquisition Regulations indicated that "Components,




as used in this subpart, means those articles, materials, and supplies incorporated directly into



the end products."  The section continues "End Products, as used in this subpart, means those



articles, materials, and  supplies to be acquired for public use under the contract."   Further



review of the section  indicated that "On  acquisitions above $25,000 in value, components of



Canadian origin are treated as domestic."   These issues specifically  applied to the concern



regarding the use of a Canadian manufactured and supplied UV Oxidation system. With  these



passages  in mind and according to the FAR, the UV system marketed by the Canadian supplier



would qualify as a Buy American product if procured through a supply contract, or provided as




Government Furnished Equipment.



       The equipment could be legally provided under a construction contract, but the approval



process would be lengthy and the product would have to qualify for a waiver under the FAR.



Waiver to the FAR involves one of three potential scenarios:
                                          232

-------
         a) If the cost of the Canadian product is less than the American product by ten percent



 after all tariffs, taxes, and fees, the product would qualify;




         b) If the usage of any other product is "impracticable", the foreign product  would



 qualify; and,




        c) If there are no American products which meet  the job requirements,  the foreign



 product would be acceptable.




        The DoD further determined that  application  of the  Buy American Act  to  the  end




 products of Canada and other specific countries (under subpart 225.103) would be inconsistent



 with the public interest.








 VI. Presentation of Concept to Applicable Entities: The concept of providing an avenue for




 all competent and capable vendors to treat the Bofors groundwater from a common header had



 previously not been used within the Omaha District Corps of Engineers.  The task remained to



 coordinate with the other parties involved in the preparation or approval of the completed set



 of plans and specifications.  Concurrence from the Contracting  Division of the Corps, the state



 of Michigan  and the U.S. EPA, Region V were all necessary prior to proceeding with  the



 development  of a workplan suitable for the on site treatability testing. By emphasizing  the



 expected recurrent annual O&M costs associated with a UV Oxidation system, it was shown that



 the overall cost associated with the up-front (predesign) testing could be recouped, with both the



 state and the government potentially saving a significant amount of money.   All appropriate



agencies concurred with the concept of treatability testing with all three vendors brought on site



during  the same time period.
                                            233

-------
       The capital costs associated with the unit process were also a primary consideration, but




overall life cycle costs would be the ultimate vehicle used in the selection of the successful



equipment supplier. The first and foremost criterion within the predesign treatability testing was



the capability of the individual vendor to successfully treat the combined purge well flow to



acceptable discharge levels, as defined by the state of Michigan. Upon successfully meeting this



criterion, the vendor was  subsequently  reviewed  on the basis  of expected life cycle cost.



Following completion  of the treatability testing and receipt of the individual vendors final



reports, the Corps made the decision as to the most appropriate vendor to be incorporated within




the final design of the groundwater treatment plant.








VII. Performance of On-site Treatability Testing:  The UV Oxidation vendors were brought




on site to determine the efficacy of their individual treatment processes. The specific testing



objectives, based upon the workplan prepared by the Corps,  were as follows:



       a)  Evaluate the effectiveness of the individual vendor treatment systems at treating the



combined purge well flow.  See Table 1  for the analytical results of the raw groundwater.



       b)    Determine specific  design  parameters such  as required  oxidation  time and




pretreatment requirements necessitated by the site specific groundwater.



       c) Monitor and define the individual vendor's electrical consumption and chemical usage



(if applicable) during the system optimization and performance periods.  This information was



subsequently used to estimate and establish the requirements for the full scale systems.



       d)  Determine any interferences caused by the native raw  water quality, and define any



pretreatment requirements necessary for adequate treatment of the contaminated groundwater in
                                           234

-------
 the proposed full scale system.
TABLE 1
DETECTED ORGANIC COMPOUNDS - UV OXIDATION PILOT TESTING
COMBINED PURGE WELL MANIFOLD
Bofors Nobel Superfund Site, Muskegon, MI
Concentration: ug/1
Volatile Organic
Compounds
Benzene
Chlorobenzene
1,2- Dichloroethane
trans 1,2-Dichloroethene
Ethylbenzene
Methylene Chloride
Tetrachloroethylene
Toluene
1,1,2- Trichloroethane
Trichloroethylene
Vinyl Chloride
Xylene
..^.i ..." "->;« •• ' ' '
Concentration: ug/1
Semi-Volatile Organic
Compounds
Benzidine
1 , 2-Dichlorobenzene
1 , 4-Dichlorobenzene
DCB
Isophorone
1,2, 4-Trichlorobenzene
2-Chloroaniline
Aniline
Dichlorobenzidine Isomer
Benzidine Isomer
Sampling Date
2/89 -
5/91(A)
ND-1800
ND-120
ND-6
ND-200""
ND-70
ND-90'0
66 - 340
ND-1500
ND-40
ND-50
ND-80
ND-40
14 Aug
1991
1400
140
ND
170m
34
15
320
750
ND
77
120
43
27 Sep
1991
1400
160
ND
170™
51
47
320
1500
ND
91
110
73
1 Oct
1991
1900
220
ND
200""
61
53
470
2000
ND
120
120
88
2 Oct
1991
1700
190
ND
ISO"'
67
200
360
1700
ND
110
150
91
" J, " ' '' , ,"'"' '" " "'''' ",""*_ '<"',' :: ,»*" *'"(' "'; "
Sampling Date
2/89 -
5/91(A)
ND-970
ND-12
ND
ND-210
ND-19
ND
71-1500
83-250
320-680
ND-1300
14 Aug
1991
360
30
ND
120
7
ND
1400'°'
140
NR
NR
27 Sep
1991
930
28
2
180
ND
4
760™
230
NR
NR
1 Oct
1991
390
18
ND
100
ND
ND
1100™
110
NR
NR
2 Oct
1991
590
23
1
110
14
3
740™
160
NR
NR
*** Data from Michigan Department of Natural Resources
<"> Reported as total 1,2-DCE
10 Methylene Chloride data from Groundwater FS (GZA/Donohue, 1989)

-------
semivolatile constituents compare favorably with results from previous sampling activities.



       The existing groundwater manifold transporting contaminated groundwater to the Lomac



PACT™ facility was tapped to provide water to the treatment area established for the purpose



of the predesign testing.  The four inch manifold line was then split into three feeds to service



each individual vendor's equipment.  The data from the 14 August 1991 sampling round was



sent to each of the individual vendors prior to  their coming to the Bofors site  for equipment



setup purposes.  This information allowed the vendors to tentatively establish and initial starting



point, based upon institutional knowledge gained from previous operating experience.  Specific



mention was made to the vendors that well cleaning activities in early September had generated




a groundwater that was noticeably  colored, and which  carried a higher  suspended  solids



concentration  than  had previously been experienced.  Each vendor was instructed that the



responsibility  for any pretreatment activity  was strictly their own.  Non-potable water for use



in equipment  cooling  and decontamination was  brought  to  the treatment  area  through the



construction of a temporary pipeline from an existing well located  on site.   Each vendor was



supplied electricity for their individual process through the use of a 150 kW portable generator.



The electrical  usage was monitored through a microprocessor based meter on each connection.








       Each vendor was instructed  that the intent  of the on  site testing, was  to treat the



contaminated groundwater to a level suitable for discharge to Big Black Creek. The preliminary



discharge standards were established  by the state of Michigan,  and  are contained in Table 2.
                                          236

-------
UV/OXIDATI01
PRELIMI
Bofor
1
Compound

Aniline
Azobenzene
Benzene
Benzidine
3,3' -Dichlorobenzidine
Methylene Chloride
Acetone
Butyl Benzyl Phthalate
Chlorobenzene
Dichloroazobenzene
3, 3-Dichlorobenzidine Isomer
Dichlorobiphenyl
TBD = To Be Determined
TAB]
V ON SITE
NARY DIS(
B Nobel .
luskegon,
Cone.
(ug/1)
4.0
TBD
5.0
0.04
0.06
5.0
500.0
5.0
5.0
5.0
0.06
5.0

LE 2
PILOT SCALE TESTING
2HARGE STANDARDS
Superfund Site
Michigan
Compound

1 , 2-Dichloroethane
1 , 2-Dichloroethylene
Trans-1 , 2-Dichloroethylene
Ethylbenzene
Isophorone
Phenols
Tetrachloroethylene
Toluene
Trichloroethylene
1,2, 4-Trichlorobenzene
Vinyl Chloride
Xylene



Cone
(ug/i)
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
, 5.0
5.0
5.0
5.0

       Each  vendor erected their equipment on the 110'x 30' asphalt pad constructed for the




pilot test.  A three inch lip was  constructed around the  periphery of the pad to contain any




potential leakage.  Any water captured  on the pad was drained to the effluent tank to be




subsequently discharged to the PACT™ treatment system.




       The initial three days of the on site treatability testing period were devoted to equipment




unloading and setup.  Approximately two days were expended making the necessary plumbing



and electrical connections.




       Two  separate sampling and  analysis activities were performed  during  the on  site




treatability testing.  The initial round of samples was taken by the vendors during the system




adjustment period.  Each  individual vendor was responsible for the  sampling, shipping, and




receipt of the sample analysis in the required turnaround time.  A one to three day turnaround




time was contracted during the system adjustment period.  This was necessary due to the nature
                                           237

-------
of the on site testing.  This stringent turnaround time was not required for the optimization




monitoring period. The results obtained from the initial system adjustment period were used to




determine the appropriate unit dosages for the optimization period.  The specific analytes were




chosen by the vendor based upon the most critical compound which dictated the chemical usage




and flow rates from their equipment.




       The second sampling event occurred during the  optimization monitoring period.   A




specified number of samples were taken by the vendors at  common sampling times.  The




optimization  monitoring period was a two day cycle during which time the vendors operated




their equipment ten to twelve hours a day.  Each of the three systems was allowed to equilibrate




prior  to being sampled.








VIII.  Results and Conclusions:




       A. Metals and Water Quality Data: Tables 3, 4, and 5 show the results for specific




metals and water quality parameters which were monitored during the performance period.  The




data were taken to verify raw water quality and  to monitor the pretreatment of the individual




vendors.  All three vendors utilized  a pretreatment filter on  the influent line  supplying  the




process equipment, although the filters were of different mesh sizes.  Vendor #1  utilized a




pretreatment tank prior to the pressure filters to remove iron and manganese in the raw water.




Sodium hydroxide and hydrogen peroxide were added to the holding tank to form floe particles




prior to filtration.  Vendors #2 and #3 utilized conventional cartridge filters to remove suspended




solids.  Table 3 indicates that the measures taken by Vendor #1 were relatively ineffective in




terms of overall removal of the iron and manganese.  The pretreatment system used by Vendors
                                          238

-------
#2 and #3 had no appreciable affect on the water quality parameters.  Acid addition to improve

the performance of their individual processes was performed by both Vendors #1 and #2.  The

use of acid in the process train significantly lowered the alkalinity of the groundwater, and

elevated the Total Dissolved Solids of the flow. Other water quality changes worthy of mention

include the oxidation of nitrite to nitrate and the reduction of the total COD by Vendors #2 and

#3. Vendor #1 did not experience either of these changes.  The inability of Vendor #1 to affect

these changes may be attributed to a non-optimized system.   All vendors experienced only a

slight reduction in the Total Organic Carbon (TOC) as the groundwater was treated.  This may

be attributed to the formation of daughter products during the  oxidation process.




       B. Volatile Organic Compound  Analysis:  Tables  6, 7, and 8 provide information

relating to the performance of the vendor equipment on treating  volatile organics (VOCs)

contained within the contaminated groundwater.  Based upon the final reports submitted by the

vendors, tetrachloroethylene (PCE) was indicated as the most recalcitrant of the VOCs in this

water matrix.  The influent concentrations of toluene and benzene are typically more than four

times higher than the PCE concentration, however, these compounds  are rapidly reduced  in all

of the vendor's systems.   In  general, the chlorinated compounds such as 1,2-Dichloroethane

(1,2-DCA), PCE and Trichloroethylene (TCE) are the most difficult to degrade due to the higher

energy associated with the molecular bonds  in these saturated organic compounds.  Other

compounds such as acetone and methylene chloride were present in the effluent, but these VOCs

are common laboratory contaminants, and  the results may not be reflective of actual conditions.
                                                                   *
In general, vendor's #2 and #3 were able to effectively treat the contaminated groundwater to
                                           239

-------
 the required discharge levels.  The removal of PCE was typically down to the discharge limit



 of 5 ppb.  The final effluent for Vendor #3 did contain PCE and benzene above the discharge



 standards during one sampling event.  The reason for exceeding the discharge level is uncertain,



 since the system was capable of meeting the discharge levels during the other sampling events.



 Both vendors #2 and  #3 had occasional hits on methylene chloride, 1,2-DCA, and TCE,  but



 these values were near detection limits and are within the bounds of the analytical accuracy.



 Vendor #1 experienced difficulty in meeting the discharge limits for several VOCs, indicating



 that their system may  not have been optimized because the vendor #1 system had sampled and



 met the discharge criteria during the system adjustment period.




       C.  Semivolatile Organic Chemical Analysis:  The sample results for the semivolatiles



 (SVOCs) contained  within the process flow are shown in Tables  9, 10, and 11.   These data



 indicate that the pretreatment filters used by the vendors appeared  to remove small  amounts of



 aniline, while other SVOC concentrations did  not appear to be affected.  Vendors #2 and #3



 were able to reduce the levels of the SVOCs to the required levels, while Vendor #1 experienced



 difficulties in the removal of several  SVOCs.  The Vendor #3 system did experience hits on



 benzidine and isophorone for a portion of the pilot testing. Vendor #3 indicated that the timing



 of those infractions was at the same time the systems hydrogen peroxide feed was experiencing



 mechanical problems.  Both vendors #2 and #3 were effective in reducing the concentrations of



 tentatively identified compounds (TICs), while vendor  #1 experienced limited success.








IX. Sole Source Procurement Strategy and Design Preparation:  Based upon the data and the



reports provided b,y  the vendors,  the Corps of Engineers  selected  a vendor around which the
                                          240

-------
design package would be prepared. Selection of a vendor was based upon a two phase screening




process. The screening process relied upon two criterion to define the most appropriate vendor




for the full scale system.   The first criterion  was defined as the ability  of  the vendors to




effectively treat the waste constituents.  Following this primary screening process, the second




screening criterion was used to select the vendor to be used in the project design.  The second




criterion was cost related.  The vendor which indicated the lowest life cycle cost was selected




over the other vendors which met the first criteria.   See Table 2A for data regarding cost data




generated by the vendors during the testing period.




       Following the report review, the vendors were contacted regarding the selection results.




The groundwater treatment facility footprint was defined based upon dimensions provided by the




UV Oxidation vendor.  The selected vendor was also contacted periodically during the design




phase to coordinate mechanical, electrical, and structural requirements.




       A number of benefits associated with this contracting methodology became apparent as




the process design continued. Specifically, the following items were streamlined or clarified as




a result of this innovative contracting  attempt:




       a) The building footprint becomes easier to develop when the exact  dimensions  of the




required equipment are obtained from the vendor.




       b) The electrical loads, known to vary considerably depending upon the selected vendor,




may be obtained early  in the design.   This facilitates the entire design and control sequence.




       c) The process piping may be clearly  defined and shown.  Specific elevations may be




coordinated with  the vendor to insure compatibility with the remainder of the process train.




       d)  Pretreatment  requirements, which may vary depending upon the vendor, can be
                                           241

-------
defined and incorporated into the process design.




      e) The contract specifications are stream-lined when a specific vendor has been selected.



This allows the designer to call for specific pieces of process equipment, provided that the



appropriate approvals have been obtained prior to contract advertisement.



                 TABLE 2A:  VENDOR PROJECTED COST DATA
Annual Cost:
Category
$/year
UV Energy1
Replace Lamps
Peroxide Use2
Acid
Base
Catalyst3
Ozone Elec.
Liquid Oxygen
Total Annual
Total Capital
Life Cycle"
Vendor #1

772,660
90,000
192,300
17,000
74,000
Not Used
Not Used
Not Used
1,145,000
1,325,000
14.22 M
Vendor #2
Alt. #1 Alt. #2
with
catalyst
735,000
189,000
448,000
33,000
147,000
19,000
Not Used
Not Used
1,571,000
1,575,000
19.27 M
without
catalyst
572,000
147,000
1,120,000
17,000
74,000
Not Used
Not Used
Not Used
1,930,000
1,575,000
23.30 M
Vendor #3
Alt. #1 Alt. #2
with O2
feed
44,300
22,600
65,700
Not Used
Not Used
2,000
113,900
441,700
690,200
1,510,000
9.28 M
with air
feed
44,300
22,600
65,700
Not Used
Not Used
2,000
219,000
Not Used
353,600
1,695,000
5.68 M
' Power costs based upon $0.06/kWh
2 Hydrogen Peroxide costs ($/gal) vary with usage rate
3 Catalyst costs represent proprietary additives or vapor phase treatment
units.
4 Life Cycle costs assume 30 year operating period at 8 percent interest.
                                        242

-------
       Within the Army Corps of Engineers, special contracting efforts were required to insure




 that all appropriate FARs were complied with.  The sole source justification  to procure the




 selected vendor began shortly after receipt  of the final vendor reports.  A justification and




 approval  (J&A) action was necessary to insure  that the government was obtaining a system




 capable of meeting all  necessary  requirements.  After approval is granted for sole source




 procurement,  the contract documents are simplified by calling for a specific manufacturer. To




 insure that there  were  no other  vendors capable of providing suitable equipment  for this




 installation, a "Sources Sought" effort was initiated.  The results of this request conclusively




 indicated  that the three vendors targeted for  the "treat-off were the  only vendors suitable for




 this application.









 X.  Conclusions: The predesign, design and contracting efforts performed for the procurement




 of the UV  Oxidation  equipment in the  Bofors Nobel Groundwater  Treatment  Facility  were




 shown to be extremely effective. This contracting technique allowed the government to attempt




 to procure proven process equipment, capable of meeting stringent discharge standards, and still




provide a  cost effective treatment system.   The ability  to  allow  all  available vendors an




opportunity to test their equipment  concurrently has been shown  to be an effective tool which




facilitates the procurement of high cost equipment items.  Although perhaps not suitable for all




superfund sites because of project specific constraints, the "treat-off technique proved to be




very worthy of the high expectations originally targeted.  In this specific instance, it is estimated




that the use of the "treat-off technique will save the government in excess of five million dollars




during an estimated thirty year operating cycle, with the cost of the predesign treatability testing
                                            243

-------
recouped  within  six  months.   If traditional contacting  techniques  had been  used in the



procurement of a UV Oxidation vendor, the selected vendor would not have been awarded the



contract because of higher initial capital cost. Although the selected vendor's equipment carries



a slightly higher capital expense, the life cycle cost of the process has been shown to make this



innovative contacting mechanism eminently suitable for the Bofors Nobel  Superfund site.
                                         244

-------
TABLE 3
UV OXIDATION TESTING: METALS AND UATER QUALITY
VENDOR f 1
Bofors Nobel Superfund Site, Muskegon, Michigan
Predesign Testing
ANALYTE CONC.
(ug/l)

Arsenic
Calcium
Iron
Manganese
Magnesium
Zinc
Combined Purge Well Manifold
9/27
1991
4.6
.
1530
405
.
38
ANALYTE
Concentration (mg/l)
COO (Total)
Alkalinity
Chloride
Sulfate
Ammonia Nitrogen
Nitrate Nitrogen
Nitrite Nitrogen
Total Kjeldahl
Nitrogen
Phosphorus
IDS
TSS
TOC I
TOG II (Dup)
pH, Laboratory
52
248
97.5
283
9.35
0.24
0.091
11.8
0.16
743
3
14.2
15.0
7.34
10/1
1991
3.9
-
1440
408
.
44
10/2
1991
4.3
108,000
1380
401
19,300
38

52
236
92.8
210
7.26
0.17
0.077
11.6
0.16
717
<2
12.6
12.8
7.17
50
230
94.8
230
7.79
0.18
0.058
12.6
0.18
740
<2
14.0
14.0
7.20
After Pretreatment
UV Oxidation Influent
9/27
1991
4.7
-
1070
288
-
<15

55
11.8
99.5
429
10.9
0.22
0.079
12.3
0.14
1130
1
14.8
14.0
5.01
10/1
1991
4.5
-
1120
293
-
85
10/2
1991
3.9
86,600
1220
279
18,700
279

52
68.2
95.6
277
10.5
0.20
0.051
10.5
0.14
897
<1
12.5
12.4
5.85
52
74.9
97.4
334
7.89
0.18
0.064
12.0
0.15
917
<2
12.2
12.2
5.94
Intermediate
9/27
1991'"
4.5
-
1110
288
-
305
10/2
1991181
4.3
-
1050
269
-
54
UV Oxidation System
Effluent
9/27
1991ICI
5.1
-
1050
303
-
<15

76
58.3
96.9
376
9.50
0.14
0.094
11.3
0.17
910
<2
12.4
12.6
5.93
67
63.3
99.0
353
7.78
0.14
0.096
11.9
0.14
890
<2
11.9
11.9
5.92
58
9.06
99.4
396
10.9
0.17
0.097
12.4
0.15
960
0
13.2
13.1
4.95
10/1
1991
4.8
-
1180
291
-
517

71
51.0
97.3
457
6.93
0.17
0.066
11.3
0.14
930
<2
12.0
12.0
5.82
10/1
1991
4.5
-
1330
320
-
37

77
87.5
97.7
392
8.15
0.15
0.088
11.3
0.16
820
<2
12.0
12.1
6.09
10/2
1991
4.5
87,000
992
268
19,400
54

67
63.4
98.8
332
8.63
0.13
0.082
13.3
0.14
912
<2
11.9
11.8
5.85

10/2
1991
4.4
-
1420
317
-
26

70
71.7
97.6
461
7.60
0.16
0.089
11.9
0.15
983
<2
12.7
13.0
5.96
'" Intermediate Sampling Point after Lamp #3 (50% Retention Time)
"" Intermediate Sampling Point after Lamp #4 (67% Retention Time)
'C1 Effluent Sample taken on 27 September 1991 during equipment adjustment period.
- Not Analyzed For

-------
NJ
-fc.
ON

ANALYTE COMC.
(ug/l)

Arsenic
Calcium
Iron
Manganese
Magnesium
Zinc
-,
TABLE 4
UV OXIDATION TESTING: HETALS AND UHTER QUALITY
VENDOR *2
Bofors Kobel Superfund Site, Huskegon, Hichigan
Predesign Testing
Combined
9/27
1991
4.6
-
1530
405
-
38
Purge Well Manifold
10/1
1991
3.9
-
1440
408
-
44
10/2
1991
4.3
108,000
1380
401
19,300
38
After Pretreatment
UV Oxidation Influent
9/27
1991
4.5
-
1780
404
-
17
10/1
1991
4.2
-
1320
413
-
78
10/2
1991
4.3
113,000
1320
410
18,400
26
Intermediate
9/27
1991'"
5.0
.
13,700
487
.
<15
10/2
1991'"
4.5
.
13,600
500
_
41
UV Oxidation System
Effluent
9/27
1991KI
5.0
_
1590
386
_
58
10/1
1991
4.9

13,300
491

45
10/1
1991
5.1

18,800
523

78
ANALYTE
Concentration (mg/l)
1 COD (Total)
Alkalinity
Chloride
Sulfate
Ammonia Nitrogen
Nitrate Nitrogen
Nitrite Nitrogen
Total Kjeldahl
Nitrogen
Phosphorus
TDS
TSS
TOC I
TOC II (Oup)
pH, Laboratory
52
248
97.5
283
9.35
0.24
0.091
11.8
0.16
743
3
14.2
15.0
7.34
52
236
92.8
210
7.26
0.17
0.077
11.6
0.16
717
<2
12.6
12.8
7.17
50
230
94.8
230
7.79
0.18
0.058
12.6
0.18
740
<2
14.0
14.0
7.20
49
248
98.0
123
7.60
0.20
0.072
12.3
0.19
750
<4
13.0
13.3
7.37
46
236
93.4
167
7.58
0.18
0.069
11.7
0.17
714
<2
13.0
13.2
7.24
49
237
94.7
238
8.64
0.19
0.051
12.2
0.15
723
<2
13.4
13.6
7.26
28
<5
95.3
658
8.80
0.27
<0.050
11.1
0.59
977
<2
9.6
9.5
2.55
39
<5
98.2
492
8.98
0.27
<0.050
11.3
0.66
967
1
10.5
10.4
2.41
67
229
97.7
267
9.72
0.27
0.096
12.2
0.85
813
1
10.7
10.8
7.00
'" Intermediate Sampling Point after Lamp #2 (67% Retention Time)
" Intermediate Sampling Point after Lamp #1 (33% Retention Time)
IC' Effluent Sample taken on 27 September 1991 during equipment adjustment period.
- Not Analyzed For
18
<5
94.8
568
7.97
0.26
<0.050
11.5
0.59
937
3
8.5
8.7
2.53

27
<5
96.9
623
8.08
0.27
<0.050
11.6
0.66
1040
3
8.4
8.1
2.57

10/2
1991
4.1
112,000
13,300
492
18,300
140

22
<5
97.8
518
9.00
0.29
<0.050
11.1
0.67
965
6
8.2
8.6
2.42

10/2
1991
4.8

16,000
509

35

39
<5
97.1
192
8.66
0.31
<0.050
13.5
0.80
1010
4
8.4
8.5
2 47


-------
TABLE 5
UV OXIDATION TESTING: METALS AND WATER QUALITY
VENDOR *3
Bofors Nobel Superfund Site. Huskegon, Michigan
Predesign Testing
ANALYTE COHC.
(ug/l)

Arsenic
Calcium
Iron
Manganese
Magnesium
Zinc
Combined Purge Well
Manifold
9/27
1991
4.6
-
1530
405
-
38
10/1
1991
3.9
-
1440
408
-
44
10/2
1991
4.3
108,000
1380
401
19,300
38
After Pretreatment
UV Oxidation Influent
9/27
1991
4.6
-
1600
400
-
51
10/1
1991
4.2
-
1290
405
-
32
10/2
1991
3.94.3
113,000
1520
410
18,400
88
Intermediate
9/27
1991'"
4.8
-
1250
408
-
<15
10/2
1991""
4.0
.
1240
406
-
49
UV Oxidation System
Effluent
9/27
1991"'
5.5
.
1470
391
-
29
10/1
1991
4.6
.
1190
394
-
243
10/1
1991
5.0
.
1260
399
-
649
10/2
1991
4.3
117,000
1350
417
19,200
93
10/2
1991
4.4
_
1200
416
_
61
ANALYTE
Concentration (mg/l)
COD (Total)
Alkalinity
Chloride
Sulfate
Ammonia Nitrogen
Nitrate Nitrogen
Nitrite Nitrogen
Total Kjeldahl
Nitrogen
Phosphorus
TOS
TSS
TOC I
TOC 11 (Oup)
pH, Laboratory
52
248
97.5
283
9.35
0.24
0.091
11.8
0.16
743
3
14.2
15.0
7.34
52
236
92.8
210
7.26
0:17
0.077
11.6
0.16
717
<2
12.6
12.8
7.17
50
230
94.8
230
7.79
0.18
0.058
12.6
0.18
740
<2
14.0
14.0
7.20
47
246
97.5
239
11.20
0.21
0.075
12.2
0.18
796
<4
13.5
13.3
7.30
50
240
96.3
200
8.65
0.19
0.053
11.2
0.17
740
<2
13.0
12.7
7.18
52
233
94.8
207
7.73
0.18
0.065
11.3
0.18
738
<2
13.4
13.0
7.19
33
225
93.4
259
8.52
0.31
<0.050
11.3
0.18
738
<2
13.6
13.6
7.07
18
206
95.7
265
9.08
0.39
<0.050
11.5
0.18
724
11
10.7
11.0
7.30
<8
212
97.1
280
9.84
0.67
<0.050
11.9
0.20
770
14
7.6
7.7
7.55
16
208
91.8
252
9.50
0.37
<0.050
11.1
0.19
764
7
11.0
10.9
7.09
21
214
94.0
135
7.51
0.34
<0.050
11.6
0.18
767
7
11.2
11.2
7.16
15
207
95.3
279
8.30
0.40
<0.050
11.3
0.15
732
10
10.6
10.4
7.23
32
219
94.3
124
8.45
0.31
<0.050
10.6
0.17
748
<2
13.1
13.1
7.41
'*' Intermediate Sampling Point after Lamp #3 (50% Retention Time)
'• Intermediate Sampling Point after Lamp #5 (83% Retention Time)
10 Effluent Sample taken on 27 September 1991 during equipment adjustment period.
- Not Analyzed For

-------
00
TABLE 6
UV OXIDATION TESTING: VOLATILE ORGANIC COMPOUNDS
VENDOR *1
Bofors Kobe I Superfund Site, Huskegon, Michigan
Predesign Testing
AHALYTE COHC.
(ug/l)

Vinyl Chloride
Methylene Chloride
Acetone
1,2-Dichloroethene (total)
1,2-Dichloroethane
Trichloroethene
Benzene
Tetrachloroethene
Toluene
Chlorobenzene
Ethyl benzene
Xylene (total)
Corrfoined Purge Well
Manifold
9/27
1991
110
47
30
170
<50
91
1400
320
1500
160
51
73
10/1
1991
120
53
<140
200
<68
120
1900
470
2000
220
61
88
10/2
1991
150
200
<140
180
<72
110
1700
360
1700
190
67
91
After Pretreatment
UV Oxidation Influent
9/27
1991
77
24
150
140
<31
81
1100
290
910
120
31
33
10/1
1991
60
16
22
120
<38
77
1200
340
920
120
26
32
10/2
1991
34
73
69
91
<33
58
870
190
660
100
<33
28
Intermediate
9/27
19911*1
20
10
19
47
<10
30
310
140
200
35
6
6
10/2
1991'"
18
8
36
45
<5
27
220
110
110
26
4
4
UV Oxidation System
Effluent
9/27
1991*'
10
17
63
23
<5
16
120
71
57
14
2
<5
10/1
1991
<10
3
31
13
8
9
60
50
21
7
<5
<5
10/1
1991
8
4
25
19
8
13
87
68
34
10
1
<5
10/2
1991
4
4
25
21
10
12
83
65
31
10
1
<5

10/2
1991
13
5
24
30
10
18
120
96
47
15
2
<5
'" Intermediate Sampling Point after Lamp #3 (50% Retention Time)
'" Intermediate Sampling Point after Lamp #4 (67% Retention Time)
IC1 Effluent Sample taken on 27 September 1991 during equipment adjustment period.

-------
VO
TABLE 7
UV OXIDATION TESTING: VOLATILE ORGANIC COMPOUNDS
VENDOR *2
Bofors Nobel Superfund Site, Nuskegon, Michigan
Predesign Testing
ANALYTE CONC.
(ug/t)

Vinyl Chloride
Methylene Chloride
Acetone
1,2-Dichloroethene (total)
1,2-Dichloroethane
Trichloroethene
Benzene
Tetrachloroethene
Toluene
Chlorobenzene
Ethyl benzene
Xylene (total)
Combined Purge Well
Manifold
9/27
1991
110
47
30
170
<50
91
1400
320
1500
160
51
73
10/1
1991
120
53
<140
200
<68
120
1900
470
2000
220
61
88
10/2
1991
150
200
<140
180
<72
110
1700
360
1700
190
67
91
After Pretreatment
UV Oxidation Influent
9/27
1991
97
65
260
150
<50
92
1400
320
1500
160
50
70
10/1
1991
11
8
16
18
<5
10
170
40
170
19
5
8
10/2
1991
110
38
<37
150
<18
88
1400
320
1500
170
51
81
Intermediate
9/27
1991 '*'
<10
3
35
<5
5
<5
<5
10
<5
<5
<5
<5
10/2
1991™
<10
6
26
4
7
8
9
58
<5
<5
2
<5
UV Oxidation System
Effluent
9/27
19911C1
<10
7
73
<5
4
<5
<5
4
<5
<5
<5
<5
10/1
1991
<10
2
46
<5
4
<5
<5
2
<5
<5
<5
<5
10/1
1991
<10
4
31
<5
4
<5
<5
2
<5
<5
<5
<5
10/2
1991
<10
10
35
<5
4
4
<5
4
<5
<5
<5
<5
10/2
1991
<10
6
38
<5
3
<5
<5
<5
<5
<5
<5
<5
"' Intermediate Sampling Point after Lamp #2 (67% Retention Time)
'" Intermediate Sampling Point after Lamp #1 (33% Retention Time)
ICI Effluent Sample taken on 27 September 1991 during equipment adjustment period.

-------
to
TABLES
UV OXIDATION TESTING: VOLATILE ORGANIC COMPOUNDS
VENDOR *3
Bofors Nobel Superfund Site, Huskegon, Michigan
Predesign Testing
AHALYTE COHC.
(ug/l)

Vinyl Chloride
Hethylene Chloride
Acetone
1,2-Dichloroethene (total)
1,2-Dichloroethane
Trichloroethene
Benzene
Tetrach loroethene
Toluene
Chlorobenzene
Ethylbenzene
Xylene (total)
Combined Purge Well
Manifold
x/27
1991
110
47
30
170
<50
91
1400
320
1500
160
51
73
10/1
1991
120
53
<140
200
<68
120
1900
470
2000
220
61
88
10/2
1991
150
200
<140
180
<72
110
1700
360
1700
190
67
91
After Pretreatment
UV Oxidation Influent
9/27
1991
<100
47
180
180
<50
97
1500
350
1700
170
55
79
10/1
1991
120
23
130
200
<50
110
1800
440
1800
220
56
84
10/2
1991
97
18
<25
140
<12
86
1200
320
1500
170
48
71
Intermediate
9/27
1991'"
<10
3
41
9
8
7
35
63
3
5
<5
<5
10/2
1991"
<10
18
44
<5
3
<5
1
3
<5
<5
<5
<5
UV Oxidation System
Effluent
9/27
1991*'
<10
2
48
<5
<5
<5
<5
<5
<5
<5
<5
<5
10/1
1991
<10
6
63
<5
5
<5
1
6
1
<5
<5
<5
10/1
1991
<10
3
45
<5
4
<5
1
7
<5
<5
<5
<5
10/2
1991
<10
15
44
<5
3
<5
1
2
<5
<5
<5
<5
10/2
1991
<10
2
13
4
4
4
18
32
2
2
<5
<5
'*' Intermediate Sampling Point after Lamp #3 (50% Retention Time)
'" Intermediate Sampling Point after Lamp #5 (83% Retention Time)
ICI Effluent Sample taken on 27 September 1991 during equipment adjustment period.

-------
TABLE 9
UV OXIDATION TESTING: SEN I VOLATILE ORGANIC COMPOUNDS
VENDOR *1
Bofors Nobel Stperfund Site, Kuskegon, Michigan
Predesign Testing
ANALYTE CONC.
(ug/l)

Phenol
Aniline
2-Chlorophenol
Benzyl Alcohol
1,2-Dichlorobenzene
2-Hethylphenol
4-Methylphenol
N-Nitroso-Di-n-Propytamine
Isophorone
1,2,4-Trichlorobenzene
4-Chloroaniline
Benzidine
3,3'-Dichlorobenzidine
Combined Purge Well
Manifold
9/27
1991
8
230
3
<10
28
14
6
2
<10
4
10
930
180
10/1
1991
<100
110
<100
<100
18
<100
<100
<10
<100
<100
<100
390
100
10/2
1991
10
160
3
<10
23
12
5
1
14
3
7
590
110
After Pretreatment
UV Oxidation Influent
9/27
1991
130
64
19
6
18
48
83
<10
14
3
3
7
61
10/1
1991
51
85
7
1
24
23
30
1
14
3
7
140
130
10/2
1991
<10
<10
<10
3
28
19
23
<10
13
4
9
<50
130
Intermediate
9/27
1991'*'
100
17
20
5
9
89
77
<10
22
1
<10
<50
<20
10/2
1991""
<10
<10
<10
5
8
67
58
<10
23
1
<10
<50
<20
UV Oxidation System
Effluent
9/27
1991'°'
<10
<10
<10
<10
<10
<10
<10
<10
5
<10
<10
<52
<21
10/1
1991
32
2
8
2
1
16
16
<10
14
<10
<10
<50
<20
10/1
1991
39
5
10
3
3
25
26
<10
15
<10
<10
<50
<20
10/2
1991
35
4
10
3
3
33
29
<10
17
<10
<10
<50
<20
10/2
1991
<10
<10
14
4
5
52
43
<10
17
<10
<10
<50
<20
'" Intermediate Sampling Point after Lamp #3 (SOX Retention Time)
'" Intermediate Sampling Point after Lamp )W (67% Retention Time)
Kl Effluent Sample taken on 27 September 1991 during equipment adjustment period.
to

-------
TABLE 10
UV OXIDATION TESTING: SEMI VOLATILE ORGANIC COMPOUNDS
VEHDOR K
Bofors Hobel Superfund Site, Muskcgon, Michigan
Predesign Testing
ANALYTE COHC.
(ug/l)

Phenol
Aniline
2-Chlorophenol
Benzyl Alcohol
1 , 2-0 ich lorobenzene
2-Hethy I phenol
4-Methylphenol
N-Mitroso-Di-n-Propylamine
Isophorone
1,2,4-Trichlorobenzene
4-Chloroaniline
Benzidine
3,3'-Dichlorobenzidine
Combined Purge Well
Hani fold
9/27
1991
8
230
3
<10
28
14
5
2
<10
4
10
930
180
10/1
1991
<100
110
<100
<100
18
<100
<100
<100
<100
<100
<100
390
100
10/2
1991
10
160
3
<10
23
12
5
1
14
3
7
590
110
After Pretreatment
UV Oxidation Influent
9/27
1991
9
180
3
1
20
15
6
1
9
3
8
630
130
10/1
1991
8
170
2
<10
22
11
5
<10
<10
"3
7
540
130
10/2
1991
9
150
2
<10
22
11
6
1
<10
3
7
600
120
Intermediate
9/27
1991"'
<10
<10
<10
2
<10
<10
<10
<10
2
<10
<10
<50
<20
10/2
1991""
<10
<10
<10
2
<10
<10
<10
<10
<10
<10
<10
<50
<20
UV Oxidation System
Effluent
9/27
1991ICI
<11
<11
<11
<11
<11
<11
<11
<11
2
<11
<11
<50
<22
10/1
1991
<10
<10
<10
1
<10
<10
<10
<10
1
<10
<10
<50
<20
10/1
1991
<10
<10
<10
1
<10
<10
<10
<10
1
<10
<10
<50
<20
10/2
1991
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<50
<20
10/2
1991
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<50
<20
IAI Intermediate Sampling Point after Lamp #2 (67% Retention Time)
'" Intermediate Sampling Point after Lamp #1 (33% Retention Time)
IC' Effluent Sample taken on 27 September 1991 during equipment adjustment period.
K

-------
TABLE 11
UV OXIDATION TESTING: SEHIVOLATILE ORGANIC COMPOUNDS
VENDOR «
Bofors Nobel Superfund Site. Huskegon. Michigan
Predesign Testing
ANALYTE CONC.
(ug/l)

Phenol
Aniline
2-Chlorophenol
Benzyl Alcohol
1 , 2-D i ch I orobenzene
2-Hethylphenol
4-Methylphenol
N-Nitroso-Di-n-Propylamine
Isophorone
1 ,2,4-Trichlorobenzene
4-Chloroaniline
Benzidine
3,3'-Dichlorobenzidine
Combined Purge Well
Hani fold
9/27
1991
8
230
3
<10
28
14
6
2
<10
4
10
930
180
10/1
1991
<100
110
<100
<100
18
<100
<100
<100
<100
<100
<100
390
100
10/2
1991
10
160
3
<10
23
12
5
1
14
3
7
590
110
After Pretreatment
UV Oxidation Influent
9/27
1991
16
190
4
5
25
24
6
<10
<10
3
9
790
160
10/1
1991
<10
91
3
<10
22
12
7
3
12
3
7
620
150
10/2
1991
9
150
2
<10
21
11
5
1
<10
3
7
580
110
Intermediate
9/27
1991'"
6
<10
<10
1
2
2
4
<10
10
<10
<10
<50
<20
10/2
1991"
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<50
<20
UV Oxidation System
Effluent
9/27
19911CI
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<50
<20
10/1
1991
<10
<10
<10
<10
<10
<10
<10
<10
1
<10
<10
<50
<20
10/1
1991
<10
<10
<10
<10
<10
<10
<10
<10
2
<10
<10
<50
<20
10/2
1991
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
63
<20
10/2
1991
2
<10
<10
<10
<10
<10
<10
<10
8
<10
<10
<50
<20
'" Intermediate Sampling Point after Lamp #3 (SOX Retention Time)
"" Intermediate Sampling Point after Lamp #5 (83% Retention Time)
ICI Effluent Sample taken on 27 September 1991 during equipment adjustment period.
to
(Jl

-------
 Remediation of Groundwater Contaminated with Volatile Organic
Compounds in the Saturated Zone at a California Superfund Site in
                     Mountain View, California
             R F Battey, D A Burbank, M M Milani, A N Stessman


                     The Earth Technology Corporation
                          (formerly Aqua Resources Inc.)
                       2030 Addison Street, Suite 500
                        Berkeley, California  94704
                                   254

-------
 Remediation of Groundwater Contaminated with Volatile Organic Compounds
              in the Saturated Zone at a California Superfund Site
                           in Mountain View, California
Historical Site Summary

This site, located in California's "Silicon Valley", was used for the manufacture of printed
circuit boards from the mid-1960s to 1981.  Various rinse waters were collected in five
underground storage tanks along one side of a row of buildings. These washing and rinsing
operations reportedly used chlorinated and  other solvents.  These concrete tanks have been
recently removed and replaced with engineered fill.  The buildings on the site are now
occupied by a variety of small industries and service establishments.

The site is a California State Superfund Site.  Investigations completed to date include a
Remedial Investigation and a Feasibility Study. A Remedial Action Plan has been devel-
oped and approved by the California Environmental Protection Agency, Department of
Toxic Substances Control.  Regulatory oversight during the previous investigations and  the
implementation of the Remedial Action Plan has been provided by Department of Toxic
Substances Control.

This paper will cover the remedial investigation, feasibility study, and  engineering design of
remediation facilities for the contaminated groundwater using a relatively new treatment
method, which results in the destruction of contaminants, rather than  their transfer to
another medium.

Site Characterization

During the remedial  investigation, approximately 30 monitoring wells and 25 soil borings
have been completed at the site to characterize the extent of contamination. Figure 1
shows the site layout and locations of monitoring wells. Contamination extends through the
vadose zone into  two distinct aquifers, which are separated by an clay layer aquitard. There
are two separate plumes; this discussion will cover only the eastern plume.  Two extraction
wells have been installed and used to field test an ultraviolet light hydrogen peroxide
groundwater treatment process.  Wells have been monitored for more than two  years.   The
principal contaminants in the groundwater are shown in Table 1.

Setting

Mountain View is located about 40 miles southeast of San Francisco.  The temperate
climate is typical  of California coastal areas, with an annual precipitation of 15 to 20 inches,
predominantly in the period from November to March. Summer daytime temperatures are
in the 70s to 80s, with nighttime temperatures in the 60s.  Winter temperatures  rarely drop
below freezing.
                                           255

-------
The ground surface at the site is covered by buildings, pavement, and small landscape
improvement areas.  Surface topography is  relatively flat, and has been locally modified by
grading and other site improvements.  The  site is located geologically at the base of an
alluvial fan, with the regional ground surface sloping gently to the north toward San Fran-
cisco Bay.

The site geology is characteristic of an alluvial environment which was periodically inun-
dated by bay waters during interglacial periods.  The lithology is characterized by inter-
bedded unconsolidated sands, silts, and clays.  Three aquifers have been identified and are
referred to as  Zones A, B, and C in order of increasing depth.  Zone A is a shallow water
table aquifer with a low hydraulic conductivity typical of silty sands.  The present saturated
thickness of this zone is only 3.0 ft, down from 6.5 ft in 1987, due  to several years of
drought.

Zone B is separated from Zone A by a clay unit 5 to 15  ft in thickness first appearing at a
depth of about 25 ft.  The conductive interval of Zone B is characterized by 10 to 15 ft of
sand with varying amounts of clay and gravel,  which is locally interbedded with sandy clay.
The hydraulic  conductivity of this zone is roughly 5 times that of Zone A.  Pump  tests
characterize this aquifer as leaky with a hydraulic connection to Zone A.  The piezometric
surface of Zone B is approximately the same as the free groundwater surface of Zone A.

Zone C is separated from Zone B by 10 to  15 ft of clay which forms a hydraulic barrier
between the two aquifers. The piezometric surface of this zone is approximately  1 ft higher
than Zones A  and B.  Groundwater flow in  the upper zones is generally to the north toward
the San Francisco Bay with a gradient between 0.002 and 0.003 ft/ft.  Figure 2 is a cross
section showing conceptual lithology of the  site.

Feasibility Study

A Feasibility Study considered a number of cleanup scenarios for both soil and groundwa-
ter.  Chemical oxidation by means of ultraviolet light and hydrogen peroxide was selected as
the treatment  method. Extensive field testing has shown that a Perox-Pure LV-60 ultravio-
let light hydrogen peroxidation (UV/PX) unit, manufactured  by Peroxidation Systems Inc.
to be effective in reducing concentrations of VOCs  to less than 1 ppb and can meet the dis-
charge limits set by a NPDES permit.  The  treated  water will be discharged to the city
storm sewer which ultimately flows into San Francisco Bay.

The UV/PX unit was tested at flows up to 50  gpm,  hydrogen peroxide concentrations in the
feed as high as 300 ppm, and power inputs up to 50 MW. Selected initial conditions will be
50 gpm feed, 200 ppm hydrogen peroxide, and four lamps operating (50 MW power).

Hydrogeologic modeling studies have shown that pumping from the Zone B aquifer at a
sufficient rate  will capture the contaminated plume in Zone B and simultaneously dewater
Zone A. A parallel soil vapor extraction operation  will be carried out in the vadose zone
and the dewatered Zone A.
                                           256

-------
 Facility Design and Construction

 The criteria for design of the groundwater extraction and treatment system included:

   •   Use of the Perox-Pure UV/PX unit furnished by Peroxidation Systems Inc. which
       was used in the full-scale field tests.  The reactor vessel has been modified with a
       new design to improve water flow patterns and minimize bypassing.

   •   Discharge limits as set by a NPDES Permit; the discharge is ultimately to San
       Francisco Bay. Discharge limits for VOCs are 5.0 ug/L each for 1,2-dichloroethy-
       lene, trichloroethylene, tetrachloroethylene, benzene, ethylbenzene, and xylenes, and
       0.5 ug/L of toluene.

   •   Provision for discharge to the city sanitary sewer, ultimately to the Palo Alto Waste
       Water Treatment Plant. The Industrial Waste Discharge Permit with the City of
       Mountain View permits discharge of 750 ug/L of each VOC with a maximum of
       1,000 ug/L total VOCs.

   •   Secondary containment for tanks or equipment containing Untreated groundwater
       and 50% hydrogen peroxide solution, either double-walled piping or a  bermed area

   •   Provision to enable unattended  operation so that the facility may be serviced once a
       week, after the startup is complete

   •   Automatic shutdown for high water level in the untreated, groundwater storage tank,
       water on the floor of the bermed process area, and misoperation or malfunction of
       the Perox-Pure unit.  Conditions in the Perox-Pure unit causing  a shutdown include
       low groundwater feed  flow rate, low hydrogen peroxide solution  pump  discharge
       pressure, high temperature in any of several locations in the unit, and moisture in
       any of several locations in the unit.                       • •

A detailed design was prepared for the groundwater extraction and treatment system which
included  drawings and specifications. Contract documents were also prepared for a com-  '
plete bidding package. Bids were received from several local contractors. Construction of
the facility is essentially complete and the startup phase is beginning.

Estimated Costs

The estimated construction cost is $185,000; this does not include the UV/PX unit (it will
be leased under a service agreement), or any .engineering costs. Estimated operating and
maintenance costs for the first year of operation are shown  in Table 2 for the first year of
operation.
                                           257

-------
                   Table 1.  Principal Contaminants in Groundwater
                                 at the Plessey Site
                      Contaminant
Typical Concen-
    trations
    (ug/L)
                      Ethylbenzene
                      Perchloroethylene
                      Trichloroethylene
                      Xylenes
                      1,2-Dichloroethylene
         300
         400
       8,000
         700
         600
              Table 2.  Estimated O&M Costs at the Plessey Site (1st Year)
$1000/yr S/1000 gal
Startup (1)
Equipment Lease (2)
Utilities:
Electricity
Storm Sewer Fees
Other
Operating and Maintenance (3)
Building Rent
Regulatory Activities
Sampling/ Analysis (4)
Hazardous Waste Fees
Special Reports
Total
20.7
59.4

41.0
10.5
1.0
30.5
13.6

38.1
56.2
25.2
297
0.93
2.66

1.84
0.47
0.04
1.37
0.61

1.71
2.52
1.13
13.3
Notes   (1)  First year only
        (2)  For UV/PX unit; includes maintenance of the unit and hydrogen peroxide
             supply
        (3)  Routine checking each week and routine reports
        (4)  Includes sampling labor and analytical laboratory charges
                                          258

-------
                                  .D7
                                                                      .06
                                                                   S94.
IsJ
Ui
VO
                             ,T2
                                                                           S6.
                                                                              rD10
  D8
D2+X +S3
S8
                                                                               T3
                                                                                         D9
                                                                                   V—
                                                                                         S18 ,
	 7 	 1,
TANK D/

\TANK C

TftNK F/

f" 7
TANK B/

\TANK A
TANK S./

                                                                                               + T5
                                                                                          -1+.+SI5
                                                                                          —•  SI I
                                                                                         ,D3
             +S,3
                          D4'
                                                                                                S10
                                                                                                   ,-H-
                                                                                                     D5
                                  PROJECT
                          TRUE    NORTH
                          NORTH
                                                                                                                                  LEGEND
+ SI... - SHALLOW WELLS (A-ZONE)
+ Dl... - DEEP WELLS IB-ZONE)
x D8, D13... - DEEP WELLS  (OZONE)
+ Tl... - TEST WELLS (B-ZONE)
                    NOTES;

                    WELLS
S1-S4, Dl. D2
S5-S13. D3-D8. Tl
T2, T3, D9-DII
SI4
T4
T5. D12, SIS
                                                             INSTALLED
                                                             OCT-NOV 1987
                                                             MAR-MAY 1988
                                                             NOV-DEC 1988
                                                             APR J989
                                                             JUN 1989
                                                             AUG-SEP 1990
                                                                                                                          TANK E — REMOVED JUN 1989
                                                                                                                          TANKS A.  B, C. D. F --
                                                                                                                              REMOVED AUG 1991
                                                                                                                                   APPROX SCALE
                                                                                                                                                100 FT
                                              FIGURE  1.   MOUNTAIN  VIEW SITE  MONITORING  WELLS

-------
                                              MONITORING-
                                               PRODUCTION
                                                  WELL
  GROUND SURFACE
                                       SANDY SILT
                     SAND LENS
WATER LEVEL
ZONES A & B
                                     CONTAMINANT
                                        PLUME
SANDY SILT
              FIGURE 2. MOUNTAIN VIEW SITE CONCEPTUAL LITHOLOGY

-------
                       SOIL WASHING OF LEAD-CONTAMINATED
                          SOIL AT A FORMER GUN CLUB SITE
 Anthony Saracino, Director of Environmental Services
 Wallace - Kuhl & Associates, Inc.
 West Sacramento, CA 95691
 (916) 372-1434

 Christine Parent, Associate  Hazardous Materials Specialist
 Department of Toxic Substances Control
 10151 Croydon Way, Suite 3
 Sacramento, CA  95827
 (916) 255-3707
Abstract

The subject site operated as a rifle range, a pistol range and a trapshooting area from 1976 to 1985.
As a result of a preliminary site assessment, it was determined that surface soil contamination by
paniculate elemental lead in the form of spent bullets and shot existed in certain areas of the site.

Pilot studies performed on representative soil samples indicate that it would be feasible to separate and
remove the lead shot, slugs and fragments from the stockpiled soil using  a modified sand and gravel
plant as the treatment plant.  Approximately 15,000 cubic yards of contaminated soil were excavated
from various areas  of the site and field screened through a one-inch vibratory screen.   The finer-
grained material (approximately 9000 cubic yards) was processed in the treatment plant, which utilized
the difference in specific gravity between the paniculate lead and the soil to physically separate and
remove the lead.

The treatment plant consisted of a hopper feeder above a jigging box used to capture the larger lead
slugs and  shot, and two sand screws that fed the contaminated soil onto three stacked screens.  The
fine material passing the smallest screen was sent through a three-foot-diameter centrifugal separator.
The plant  utilized  approximately  300  gallons of water per  minute,  which was  stored  in two
recirculation settling ponds.  Clarified water from the settling ponds was recirculated as part of the soil
washing and material transport  water in the plant.  The treated soil flowed along a drag tank and was
discharged to ground (beached).
                                           261

-------
The goal of the treatment process was to remove the participate elemental lead from the fine-grained
soil so that a total lead concentration of less than 170 mg/kg for a given volume  of sample was
achieved.   The test results showed a high degree of effectiveness in the removal of the free lead,
although some lead remained bound to the silt, clay and organic fractions of the soil.  Elemental lead
particles ranging in size from coarse slugs down to approximately 100 mesh were recovered from the
contaminated soil.  The mean lead concentration for the treated soil was 181 mg/kg.

Twenty-one drums  (31,970 pounds) of recovered lead were transported to a recycling facility.  The
treated soil and the coarse-grained, uncontaminated material from the screening operation were used
as backfill beneath a road being constructed  on  site.  The total cost of the soil  washing operation,
including all earthwork operations, was approximately one million dollars.  The project team consisted
of two consultants and a contractor who worked in conjunction with each other to oversee the remedial
action, conduct the sampling, and to operate the  treatment plant.

Introduction

This paper summarizes the investigation and remediation of lead-contaminated soil at a former gun club
site  in  Folsom,  California.   The  investigation and  remediation were part  of a Preliminary
Endangerment Assessment (PEA) performed  for the site; the PEA legislation was enacted in 1989,
which allowed the State of California Department of Health Services (currently Cal EPA, Department
of Toxic Substances Control) to oversee PEA projects.

The study site comprised approximately 40 acres in Folsom, Sacramento County, California, and the
overall property is  to be developed  into a residential subdivision.  The  site was purchased by the
Orangevale Sportsman Club, Inc. in February, 1976. Three relatively isolated areas of the site were
used by the club  for target shooting; the shooting was directed into soil berms designed to trap the
spent bullets.  The berms were constructed of on-site materials consisting primarily  of sand,  gravel
and cobbles.   Another portion of the site was used for trapshooting.  The trapshooting  operations
resulted in particulate elemental lead contamination in the surface soil.

No specific measures were taken during the  time of gun club operation  to contain,  store, treat, or
dispose of the spent bullets and shot, except for construction of  soil  berms  in the target  area.
Approximately 1,850 cubic yards of soil from the berms containing visible  lead were removed from
the site in October 1987.

The proposed development of the site required several site studies, including an Environmental Impact
Report and a Preliminary Site Assessment.  As a result of these studies, a determination was made that
                                           262

-------
 surface soil contamination by particulate elemental lead in the form of spent bullets and shot remained
 in certain areas of the site.

 Regulatory Perspective

 The Department of Toxic Substances Control (Department) oversight  was requested by the City of
 Folsom to ensure that a health based soil cleanup level was achieved. The Preliminary Endangerment
 Assessment (PEA) prepared by Wallace Kuhl and Associates, Inc.  further characterized the site and
 defined the vertical and horizontal extent of the soil contamination after initial sampling confirmed
 elevated levels of lead in the soil. The PEA is an evaluation tool that assesses  what remedial action
 is  required to protect public health and the environment.  Information and determinations gained
 through conducting the PEA included:

        1)    Site Description and History;
        2)    Apparent Problem;
        3)    Environmental Setting;
        4)    Sampling Activities and Requirements;
        5)    Human Health and Environmental Threat Assessment; and,
        6)    Conclusions and Recommendations.

 The PEA also incorporated a description of the remedial action. The PEA presented information and
 data in a standard format to allow Departmental review  of the potential for lead exposure from on-site
 soils.

 The soil  cleanup level for lead was  determined for  unrestricted  use of the  site.   The  type of
 contamination at the Folsom Gun Club site and the land  use issue that designated change to residential
 development necessitated a treatment that was protective of human health.  A Health Risk Assessment
 was conducted to predict whether the estimated residual concentrations of lead would pose a potential
 negative impact on the health of residents who would live in the development.   A level of 170  ppm
 residual lead  for the treated soil to be placed beneath  an on-site  roadway was established  by a
 Department staff toxicologist based on risk factors to children and pregnant women, the most sensitive
 receptors.

Acceptable residual soil levels at the proposed residential lots from which the contaminated soil was
excavated was determined using the U.S. EPA's Uptake/Biokinetic Model (Model) to estimate the total
lead uptake (micrograms per day) in humans,  in particular for infants and young children.  This
approach was developed by the U.S. EPA because there  is a lack of empirical evidence for a threshold
for many of the noncarcinogenic  effects of lead in infants and young  children. Therefore, the use of
                                           263

-------
traditional toxicity criteria (the  reference dose; RfD) in assessing risk for lead is not considered
adequate by the U.S. EPA.  The Model provides a method for predicting blood lead levels in infants
and children exposed to lead in multiple environmental media (air, diet, soil, indoor dust, water, etc.).
The potential exposure pathways evaluated for the health risk assessment were:

        1)    ingestion of soil;
        2)    inhalation of fugitive dust originating from site soil;
        3)    ingestion of homegrown vegetables and fruit; and,
        4)    in utero exposure as result of maternal lead intake.

The Model, which used the residual lead concentration of 77 mg/kg for the residential lots, predicted
blood lead levels in infants and children aged 0 to 7 years to be 3.89 ug/dL, which is less than the
California Department of Health Services' threshold concentration of 5.0 ug/dL.  The site exposure
pathways contributing the greatest amount to blood levels in children are ingestion of homegrown
vegetables and fruit, and ingestion of soil.  Based on the results  of the health risk assessment, future
residents of the development will not be adversely impacted by residual concentrations of lead in site
soil.

Another regulatory issue was to determine if lead, in essentially an elemental form, constitutes a
hazardous waste.  A related issue was also addressed - do lead bullets  and shot meet criteria under
the Scrap Metal Exclusion? It was determined that the lead as a result of lead bullets and shot was
not generated as a hazardous waste but was a hazardous substance.  The lead in the surface soil,
therefore, constituted disposal, and would have had to have been handled as a hazardous waste if it
had been excavated and disposed of off site.  Therefore, any remedial activity that involved moving
the lead-contaminated soil triggered hazardous waste  requirements in Title 22, California Code of
Regulations (Title 22, CCR) and procedures mandated fay the Department.

The Department is required under Title  22,  CCR to implement regulations  to  manage hazardous
substances and wastes in order to protect the public health and  the environment.   In Title 22,  CCR
Section 66261.24, Characteristic of Toxicity, the Total Threshold Limit Concentration (TTLC) for lead
is 1000 parts per million (ppm) or mg/kg wet-weight.  The TTLC is defined in Title 22, CCR as, "the
concentration of a solubilized,  extractable and nonextractable  bioaccumulative  or persistent  toxic
substance which, if equaled or exceeded in a waste, renders the waste hazardous." Any material with
a lead concentration exceeding 1000 ppm TTLC is to be handled as a hazardous waste.

The  leachability  of  the lead  through the  soil  is  determined by  the  Soluble Threshold Limit
Concentration (STLC). The STLC is  an indication of how soluble a substance or compound is through
a medium.  The STLC for lead  is established at 5 ppm (milligrams per liter) in Title 22, CCR and
                                            264

-------
 defined as,  "the concentration of a solubilized and extractable bioaccumulative or persistent toxic
 substance which, if equaled or exceeded in a waste or waste extract...renders the waste hazardous."


 The solubility test to determine the teachability is the Waste Extraction Test (WET).  The WET was
 designed to reproduce leaching potential in a landfill. The WET is considered an aggressive solubility
 test for metals because it uses citric acid as a buffer. This tends to pull the metal out of solution and
 will not necessarily reflect what transpires  in non-landfill environments subjected only  to rainwater.
 Therefore, deionized water was substituted for the citric acid in a subsequent WET.  The WET using
 the citric acid buffer on soils from the site exceeded the STLC limit of 5 ppm, whereas the test using
 the deionized water did not exceed the limit.  Although neither test completely  reflects what  occurs
 in  the environment, a comparison of the two allowed  for a site management decision based  on the
 results  of both methods.

 Another regulatory  issue addressed whether or not lead bullets and shot classifies as  scrap  metal.
 Under Section 66260.10 Title 22, CCR, "scrap metal"  is defined as one or more of the following:

        1)   Manufactured, solid metal objects  and products;
        2)   Metal workings, including cuttings,  trimmings, stampings, grindings,  shavings,  and
             sandings;  or
        3)   Solid metal residues of metal production.

 The definition also lists exclusions  for metallic wastes that are not considered scrap  metal.   These
 exclusions include under Section 66260.10 subsection (b)(6), "Sludges, fine powders, semi-solids, and
 liquid solutions that are hazardous wastes." The argument has been made that based on the regulatory
 definition, lead bullets and  shot are manufactured,  solid metal objects or products and therefore are
 excluded  from regulation as a hazardous waste.   If the lead bullets and shot are found to have
 disintegrated into a fine powder (defined in Section 66260.10 as, "a metal in dry, solid form having
 a particle size smaller than 100 micrometers in diameter"), then based on the exclusion in subsection
 (b)(6), it would not be considered scrap metal.   If the lead bullets and shot are mixed with a fraction
of lead  as a fine powder, then testing is required to determine whether it has any characteristics of
hazardous waste.  The site soils at Folsom did not meet  the criteria for scrap metal exclusion because
fines of 100 micrometers or less were extractable from the soils. The larger question of whether lead
shot meets the definition of a scrap metal is  not  resolved.
                                           265

-------
Soil Sampling

A surface soil sampling plan was developed to assess the extent of the lead contamination. The results
of this sampling confirmed that the areas of the site containing elevated levels of paniculate elemental
lead were near the soil berms originally constructed for the purpose of intercepting spent bullets, and
in the trapshooting area that received spent shot.  Concentrated sampling  in these areas defined  the
limits of the surface soil contamination.  The following phases of sampling for characterization,
delineation, and confirmation were conducted:
                                       Initial Sampling

The first phase of sampling took place on January 23 and 25,  1990.  A total of 47 surface soil samples
were obtained over the entire site to identify areas of potentially elevated lead concentrations  and to
determine background levels of lead in soil in  areas of the site  away from the former gun club
operations. The samples were obtained from the upper six inches of soil.  Laboratory analysis of these
samples  indicated that four areas of the site contained lead  concentrations greater than background
levels: three former soil berm  areas and the trapshooting area.
                                    Delineation Sampling

The second phase of sampling took place on February 14,  1990, and March 7 and 19, 1990.  Fifty
surface and subsurface samples from depths ranging from zero to four feet were obtained in the areas
identified by the initial sampling as containing elevated levels of  paniculate elemental lead;  the
delineation samples were taken to determine the lateral and vertical extent of lead-contaminated soil
in these areas.

Based on the laboratory analyses of the delineation samples, the lateral and vertical limits  of elevated
levels of lead (above 170 mg/kg) were defined.
                           Post-Excavation Confirmation Sampling

Subsequent to delineation of the four areas of lead contamination, the contaminated soil was excavated
and stockpiled.  The former soil berm areas were then resampled to confirm that the lead-contaminated
soil had been removed.  When confirmation samples indicated remaining isolated areas of residual
contamination,  the process was repeated — the isolated areas were excavated, and the excavated
                                           266

-------
 material stockpiled. Upon completion of the excavation process, confirmation sampling indicated that
 the highest concentration of lead remaining in the former trapshooting  and soil berm areas was 77
 mg/kg.
 Stockpiled Soil

 The entire stockpile of lead-contaminated soil, approximately 15,000 cubic yards, was field screened
 through a one-inch vibratory screen.  The uncontaminated material retained on the screen (larger than
 one-inch diameter) was  stockpiled separately and  subsequently used as backfill  beneath an on-site
 asphalt-concrete roadway.  The finer-grained soil passing through the screen (less than one-inch in
 diameter and containing the paniculate elemental lead) comprised approximately 9,000 cubic yards and
 was processed in the treatment plant.

 The screened soil consisted of gravel, sand and silt.  Sieve analysis data revealed that the gravel
 comprised approximately half of the final weight of the soil; an average of 70 percent by weight was
 retained on the No.  12 sieve and an average of five percent by weight passed the No.  30 sieve.
Treatment Plant

A modified sand and gravel plant was used to treat ("wash") the lead-contaminated soil.  Sand and
gravel plants operate on the principle that alluvial sand and gravel have an average specific gravity of
approximately 2 to 3 grams per cubic centimeter (g/cc).  The difference in specific gravity between
the paniculate lead (11.4 g/cc) and the soil was utilized to physically concentrate, separate and remove
the lead from the soil.

The modified plant consisted of a hopper feeder above a jigging box used to capture the larger lead
slugs and shot, and two  sand screws which fed material onto three stacked screens  (3/4", 1/4" and
1/8"). The fine material passing the smallest screen was sent through a three-foot-diameter centrifugal
separator.  The remaining material (treated soil) flowed along a drag tank until it dropped off the end
of the plant; the treated soil was  "beached" from this point outward in a triangular shape onto a gently
sloping surface previously created by earthmoving equipment.  The treated soil was contained at the
furthest extent of the beached area by a three- to five-foot-high soil berm.  The concentrated lead in
the jigging box and centrifugal separator was removed and stored in sealed 55-gallon drums on site.
                                            267

-------
The treatment plant processed approximately 20 cubic yards of soil per hour and utilized approximately
300 gallons of water per minute; the water was stored in two recirculation settling ponds, described
below.
Recirculation Ponds

The ponds were constructed using large Caterpillar scrapers by creating two trenches approximately
150 feet long by 15 feet wide.  The excavated trenches were lined with 30-mil continuous  plastic
sheeting to form an impermeable barrier. Approximately 200,000 gallons of water from a hydrant on
site was used to fill both ponds to a depth of approximately five to six feet.

The pond water initially became turbid with solids  (silt and clay  sediment), which increased its
viscosity.  The increase in water viscosity  made it difficult for the lead to settle in the plant. This
problem was solved by the addition  of calcium chloride to the circulating water.   Calcium chloride
acted as a flocculent, causing suspended fines and clay particles in the processing water to lump
together and drop out of suspension.  Empirical observations in the field showed that approximately
40 pounds of calcium chloride added to the water were required per 100 yards of processed material.

When the level of settled fines in the lower  pond rose to a point near the water pump intake, the fines
were either pumped out with a trash pump, or removed using a large bucket excavator.  The fines
were placed against the soil berms constructed downslope from the ponds.

Results of lead analysis in water samples obtained from the recirculation ponds showed < 10 fig/1 lead
on October 8, 1990, and 46 fig/t lead on October 17, 1990. Lead concentrations in the recirculating
processing water were below the reporting  limit of 10 /zg/l in four water samples taken March 4, 7,
 15, and April 3, 1991.
 Reconsolidation Process

 Upon completion of the soil washing process, the fines from the recirculation ponds were combined
 with the treated soil (reconsolidated) using a Caterpillar D8 bulldozer equipped with three four-foot-
 long ripper teeth during early April, 1991.  The reconsolidation process involved continuous mixing
 with the bulldozer in a grid pattern for five days (42 hours). The mixing area was approximately 630
 feet long by 190 to 260 feet wide.
                                            268

-------
 After mixing  operations,  the  reconsolidated soil  was sampled in place.  The  grid for the  sample
 locations was  based upon a minimum of fifty samples.  At each sample location, a backhoe trench
 was excavated from the surface to the bottom of the reconsolidated soil, approximately three  to four
 feet deep.  Three discrete samples were collected from the top,  middle and bottom of the trench at
 each sample location.

 The treated soil, following the reconsolidation process, had a mean lead concentration of 181 mg/kg.
 The plastic pond liner was removed from the ponds by the excavator bucket.  Three soil samples were
 obtained from below each of the former ponds.  The test results for these samples showed lead levels
 in the native soil below the former pond sites well below 170 mg/kg.

 Recovered Lead

 Elemental lead particles ranging in size from coarse slugs  down to approximately 100 mesh were
 recovered from the stockpiled soil.  The recovered lead was  in the form of slugs,  shot  and small
 particles.  The slugs comprised eight to ten percent of the total weight of recovered lead.  Of the
 remaining lead shot and particles, 98.5 percent passed the No. 8 mesh screen.

 All recovered  lead (slugs, shot and particles) was stored on site in covered 55-gallon drums.  Twenty-
 one  drums were transported on May 13,  1991  to a recycling facility.  The  net weight of lead as
 delivered was  31,970 pounds.
Cost

The total cost for the project was approximately $1,000,000,  including excavation, field screening,
recirculation pond construction, soil treatment and reconsolidation, and placement of the treated soil
beneath the roadway.
Acknowledgments

The  authors would like to acknowledge the contributions of Eric Hubbard of Wallace Kuhl and
Associates,  Inc., who prepared  much of the Preliminary Endangerment Assessment,  Western
Environmental Science and Technology (WEST), who coordinated the treatment operations, Lee R.
Shull, Ph.D. of Western Environmental Health Associates, who prepared the health risk assessment,
and Richard Brausch of the California State Department of Toxic Substances Control, who contributed
to the scrap metal exclusion analysis.
                                           269

-------
Mr. Markku Aaltonen
EKOKEM Limited
P.O.BOX 181
SF-11101  RHHIMAKI
FINLAND
IMPROVEMENTS IN THE CONVENTIONAL HIGH TEMPERATURE
ROTARY KDLN INCINERATION TECHNOLOGY
1. INTRODUCTION
           Ekokem Ltd operates two rotary kiln incinerators for hazardous wastes in
           Finland. Since 1984, when the first kiln started, the company has been
           developing the technology with Outokumpu Ecoenergy, a Finnish
           engineering firm. As  a result of this co-operation,  a second kiln was
           introduced in 1991.

           The most significant difference compared to similar kilns,  is the extremely
           high temperature used in a rotary loin, 1.300 - 1.400  ' C (2.400 - 2.600 F)
           This results in  several advantages: high combustion efficiency, slag
           vitrification and increased capacity. Refractory lining will normally be
           damaged when using increased temperature. This problem has been solved
           by installing a special cooling system around the kiln.

           Another common source of disturbance in rotary kiln technology is
           inhomogeneity of waste feed. A shredder and  feed homogenizer have been
           designed which allow for more solid organic material to be fed, as  well as
           decreasing the need for auxiliary fuel (waste oil or solvents).
                                         270

-------
2. INCINERATION ONE H

            Ekokem designed a new rotary kiln incinerator which has been in operation
            since April 1991. An activated carbon injection system with additional bag
            filter was installed in October 1992. Compared to the original incinerator
            this system has several new features which give better result in combustion,
            gas cleaning and solid material pretreatment.

            The capacity of the new line is 25.000 tonnes/a of organic waste and 6.000
            tonnes/a of waste water.  The net value of the investment was
            approximately 30 MUSD.

            Ekokem has created its own philosophy in respect of hazardous waste
            combustion, where:
                       the temperature in the rotary kiln is extremely high (1.300 -
                       1.400  °C)
                       waste is homogenized and fed uniformly to the kiln
                       after burning chamber is designed to ensure complete
                       mixing and sufficient retention time for the whole
                       gas stream
                       the gas cleaning technology is the best available to
                       reach  the extremely tight emission limits

           The advantages of this method of operation are:
                       to reach the highest possible combustion efficiency
                       to have an insoluble vitreous  slag resulting in lower
                       costs of landfilling or utilization
                       increased capacity of the  kiln
                                          271

-------
2.1 The Kiln Cooling System
            If a temperature of < 1.100 °C is applied in the kiln, the after burning
            chamber must be operated at a higher temperature using auxiliary fuel.
            These are normal operational conditions in most European incinerators.
            There are some  disadvantages, such as, lower combustion efficiency in the
            kiln and the fact that some types of waste (for example PCB) cannot be
            incinerated efficiently. Also heavy metals will not evaporate from the slag.
            Moreover, the slag is not melted and a protective layer of molten slag is not
            formed.

            Ekokem operates both kilns at 1.300 - 1.400 °C temperature made possible
            only by a special cooling system.

            External cooling of the kiln  has been in succesful operation at Ekokem
            since October 1987. The improvements have been the following:
            -          Lifetime of refractory lining has more than doubled
            -          Capacity has increased  10 - 20 % on an annual basis
            -          Operating hours per year have increased
            -          The thickness of refractory lining can be reduced
            -          PCB waste can be incinerated due to increased temperature
            -          Heavy metals such as lead are  totally evaporated from the  slag
                       Steel barrels are oxidized and dissolved
            -          Thermal stress on refractory lining and shell is reduced
           The main objective of the cooling system is to keep all  parts of the kiln
           shell at a constant temperature. Cooling must also be maintained during
           power failures. Inside the refractory lining, the  formation of a layer of solid
           and high viscous slag takes place, which protects the lining  from erosion
           and corrosion.
                                           272

-------
            An important factor of the control of kiln operation is the understanding of
            molten slag chemistry. The viscosity and other properties can be influenced
            by proper feed and additives thereby avoiding disturbances caused by
            uncontrolled slag properties.

            The quality of slag is shown in Table 1. The low solubility means that the
            material is not classified hazardous or toxic and it can be used in
            construction of roads, etc.

2.2 Pre-Treatment and Feed to the Kiln

            Pre-treatment starts with the reception of waste and a rigid laboratory
            control on incoming materials. Waste oil and oily wastes should be treated
            to separate oil, water, paste and solid fractions and then individually fed to
            the incinerator. Chlorinated solvents should be separated from normal
            solvents.

            Barrels form the  most difficult handling problem in the pre-treatment area.
            The main difficulties are the disposal of empty barrels and  barrels that
            cannot be emptied because of solidified content or risk of polymerization of
            the contents.

           The trend is towards a highly automated barrel handling system which
            meets the requirements of the working environment and fire safety
            regulations. Handling includes a shredder system  for the steel barrels  which
           allows the steel scrap to be added to the bulk material.
                                           273

-------
The contents of the barrel are separated into  different fractions to be fed
separately to the incineration line. In the future, feeding of complete barrels
will virtually be unknown,  with the probable exception of infectious waste
and some special wastes.

The bulk material is usually fed to the incinerator with a grab crane,
typically in 1 ton batches. Even though the bulk material is well blended,
the batch introduced to the kiln will cause variations in the oxygen content
of the flue gases and consequently problems in CO emissions. A solid waste
feed homogenizer has been installed resulting in stable combustion and a
better control of emissions. Table 2 shows emission standards for both the
Ekokem kilns.
The patented cooling system has already been adapted to several European
rotary kilns.
                            Table 1
                    Slag Suitability After EI'A-Tcst
El'A limits ing.!
Il.i
A*
Cd
Cr
1'b
Sc
AH
UK
100
s
1
s
s
1
s
0,2
F.kokcm slag ing/I
0,2 - 0,3
0,005 - 0.018
0,001 - 0,004
0,001 - 0,004
0.01 - 0,03
< 0.01
< 0.002
< 0.0005
                                274

-------
                                                 'I able 2
                         Limit Values for Incineration Lines 1  & 2 (dry, 12 % O2)

Dust mg/ni3
HC1 nig/in3
HF mg/m3
CO mg/m3
Cx Hj. mg/m3
Hg //g/m3
TCDD-ekv ng/m3
Incinerator 1
30
50
2
70
20
50
1,0
Incinerator 2
10
20
2
70
20
50(I)
1,0 *>
(1)  Obligation to study the possibility to reach value 30 ,/ Obligation to study the possibility to reach value 0,1 ng/m3
                                                      275

-------
  EVALUATION OF GROUTING TECHNOLOGY! DOE PERSPECTIVE

                                         by
                                   Dr. W. Roberds
        Colder Associates Inc., 4104 148th Ave NE, Redmond WA 98052; (206)883-0777
                                        and
                                  Dr. A. DuCharme
  Sandia National Laboratories, Org 6301, POBox 5800, Albuquerque NM 87185; (505)844-5571

 ABSTRACT The US Department of Energy (DOE) must make a variety of decisions within their
 environmental management (EM) program  regarding (1) development and transfer of improved
 EM technologies and (2) application of the optimum EM technologies at specific sites.  To assist
 in  this decision-making  process,  a methodology  has  been  developed  for evaluating EM
 technologies in terms of (1) the degree to which they would accomplish specific EM activities and
 satisfy specific requirements, under various conditions, if they were available, and (2) the costs
 and risks (if any) associated with their becoming available. For demonstration purposes, this
 methodology was applied to (1) the preliminary evaluation of the generic application of a
 Ukrainian clay-based grouting technology to the DOE complex, and (2) the detailed evaluation
 of the specific application  of a French silicon-based grouting technology to DOE's underground
 storage tanks at Hanford, Washington. Based on these evaluations, justifiable recommendations
 regarding further development and transfer of these technologies were made.


                                  INTRODUCTION

 The US Department of Energy (DOE) manages an extensive complex of facilities throughout the
 United States which are concerned with radioactive and hazardous wastes. DOE is responsible
 for environmental management (EM) of these facilities. In this role, they must make significant
 decisions regarding (1) application of the optimum available EM technologies at each site, and
 (2)  improvements in available EM technologies. The potential consequences of poor decisions
 in these areas include unnecessary hazards (short- and long-term), expense, and/or delays, as
 well as possibly inadequate performance. Hence, a methodology is needed to help ensure that
 appropriate decisions are  made.  As discussed herein, this methodology consists essentially of
 evaluating EM technologies with respect to specific DOE needs, both for individual facilities and
 the complex as a whole.

The context of the methodology within the DOE EM program is first discussed. The application
of the methodology to the evaluation of two grouting technologies is then presented.


                                DOE EM PROGRAM

The US Department of Energy (DOE) is responsible  for environmental management (EM),
consisting of waste management (WM) and environmental restoration (ER), at a number of
facilities throughout the US.
                                       276

-------
EM Activities - The various types of EM activities have been identified through a simple EM
system process chart, as illustrated in Figure 1. Although broadly defined, this chart is believed
to represent a comprehensive set of EM activities (DuCharme et al, 1992a):
                Waste
            Characterization
   Waste
 Treatment
                                                                             WM
                Waste
              Generation/
             Minimization
   Waste
Trasportation/
  Storage
 Waste
Disposal
                                            Characterization
                      (Secondary Wastes)
       ER
                                 Monitoring
           In-Situ Treatment
            or Site Waste
              Removal
            In-Situ
          Containment
     (adapted from DuCharme et al, 1992a)
  Figure 1  DOE EM Activities
      Waste Management (WM)  - An initial  waste stream  is generated through various
      processes.  During generation, the amount of waste can be minimized through specific
      procedures.  The waste should be characterized and can possibly be treated, where
      treatment may include further waste minimization and require additional post-treatment
      characterization. Such treatment may be conducted off-site, thus requiring transportation
      of the waste to the treatment facility. Whether treated or not, the waste must ultimately
      be transported to a disposal facility. The disposal facility, in turn, must be characterized
      prior to disposal and monitored after disposal.


      Environmental Restoration (ER) - If an operating disposal facility is found by monitoring
      to not perform adequately, or if an existing contaminated site (e.g., a closed disposal
      "facility") is found by characterization to not perform adequately, it can be: (1) remediated
      (cleaned up) to various degrees by in situ treatment or by removal of the contaminated
                                        277

-------
       materials; and/or (2) mitigated (contained).  Containment may: (1) precede remediation,
       either as a stop gap measure until remediation can be conducted or to prevent further
       contamination  during the remediation process;  or  (2) follow  remediation  if the
       remediation process is incomplete and residual contamination remains on-site.   If
       mitigated, then  monitoring must  continue  to  ensure adequate  performance, with
       subsequent additional mitigation or remediation possible. If remediated, a secondary
       waste stream  is generated, which must be characterized,  treated, transported and
       disposed of in similar ways as the primary waste stream.

BM Conditions - EM activities will be conducted under specific conditions at each site.  These
conditions have been broadly but comprehensively categorized as  follows (DuCharme et al,
1992a):

  •    Waste
             specific types of waste (i.e., contaminants);
             amount of each type (e.g., in terms of an operating  period and  rate of newly
             generated wastes, and/or in terms  of current concentrations and volumes of
             previously disposed wastes); and
             waste form (e.g., dissolved in groundwater, packaged in barrels, vitrified, etc.).

  •    Site (see Figure 2)
             waste environment;
             transport characteristics from the waste to potential receptors, which in turn are
             a function of the engineered barriers, the geology, the geohydrology, and the
             climate; and
             potential receptors, including distance and demographics.

  *    Waste-site relationship (configuration, Figure 2)
             release rates; and
             pathways.

  •    Infrastructure (framework within which the waste/site system operates)
             regulatory;
             institutional;  and
             public.

EM Program Requirements - EM activities must be conducted within specific requirements:

  •    Technical - Functional requirements (eg, minimum capacity) may be established for
       specific activities (eg, waste disposal). Similarly, performance requirements (eg, maximum
       allowable dose) may be established for other activities  (eg,  in situ containment).  In
       addition, cost and schedule, as well as safety, will generally be of concern, and may be
       mandated (eg, by negotiated agreement).

  •    Institutional - As policy, there will be a minimum acceptable level of confidence, either
       implied or stated, that the technical requirements will be satisfied.  This  requires high
       confidence that the  proposed approach for conducting the EM activities will work as
       anticipated, which in turn suggests high confidence that the technologies involved will
                                        278

-------
         Meteorology
                                                                Demographics
                                                Form
                                                Concentration
                                                Volume
         (adapted from DuCharme et al, 1992a)
  Figure 2 ER Site Conditions
       be available when needed and moreover will be accepted (eg, by the regulators and the
       public).

EM Program Decisions - A wide range of technologies are currently available for conducting the
various EM activities and satisfying the specified requirements.  Improvements on  these
technologies, as well as new innovative technologies, can be developed and/or transferred (if
necessary), and thus become available in the future.  Significant decisions must be made by DOE
regarding: (1) programs for developing and/or transferring EM technologies; and (2) technologies
for conducting EM activities at specific sites.

Technology Attribtltes - The various technologies can accomplish specific activities  and satisfy
specific requirements to various degrees,  depending on the specific conditions involved.  For
example, advanced incineration can be  used to treat primary organic wastes, thereby reducing
the volume for disposal, while meeting air emission requirements. Each technology has a specific
set of relevant "attributes", which include:
  •    effectiveness in satisfying specific activities/requirements under various conditions;
  •    application cost (fixed and variable);
  •    rate of application;
  •    hazards associated with its application; and
  •    acceptability (ie, to regulatory agencies, institutions, public, etc.) of application.      "

If the technology is not yet available, then the cost and time associated with it becoming available
(eg, through research and development, and/or transfer), and the likelihood of it ever becoming
                                          279

-------
 available, must also be considered.  The likelihood of future availability is a function of the
 current status  of developmenl/transfer, the plans for completing development/transfer, and
 potential problems associated with developmenl/transfer.

 Evaluation  Process  -  Decisions  on  technologies  for site  specific  applications  and for
 developmenl/transfer programs can be made based on comparative evaluations of their attributes
 with those of "base case" technologies, in conjunction with tradeoffs made among the attributes.
 For example, one technology might be preferred over another even though it is less effective
 because it can be applied faster and cheaper.  As illustrated in Figure 3, the evaluation approach
 consists of  first screening  the  technologies with  respect to applicability  to specified EM
 activities/conditions. The potentially applicable technologies are then comparatively evaluated
 against Base Case alternatives  in terms of their relative effectiveness in solving the specific EM
 problems, and  the costs, schedule, safety issues and acceptability issues associated with their
 application, assuming that  they are available  (ie, fully developed and transferred).  These
 comparative evaluations can then be combined with value judgements regarding tradeoffs among
 the various  factors to establish a rating for the specific application of that technology.  If the
 application rating is positive, the technology (once available) would be preferred to the Base Case
 alternative.  If the application rating is not positive, the technology would not be preferred to the
 Base Case alternative, regardless of availability, so that any remaining developmenl/transfer
 should not be carried out. However, if the technology is preferred over the Base Case alternative
 but is not yet available, then the likelihood of successfully completing development and transfer
 of the technology,  and the associated cost  and schedule, must  also  be considered.  These
 evaluations  can then be combined  with the application rating, the priority of that application,
 and value judgements regarding tradeoffs among the various factors to establish an overall rating
 for the development, transfer, and application of the technology. In this evaluation, the cost of
 making such a technology available should not necessarily be borne by any one site, but spread
 among all those sites where it would be used. Based on the overall rating, a decision would be
 made whether to continue development and transfer of a specific technology.

 International Program - Within  EM and complementary to other programs in accomplishing
 DOE's EM mission,  DOE instituted the International Program Division (IPD), previously called
 International Technology Exchange Program (ITEP), with the following primary goals:

  •    Import - Transfer foreign EM technologies which are more effective ("better"),  cheaper,
       faster, safer or more acceptable to apply to DOE's EM activities/conditions/requirements
       than those technologies currently available in the US.

  •    Export - Identify foreign EM needs (ie, specific  site activities, with associated conditions
       and  requirements) for  which currently  available US technologies  are  more  effective
       ("better1^, cheaper, faster,  safer  or  more  acceptable to  apply than  those foreign
       technologies  currently available (ie, market opportunities).

Within IPD/ITEP, a  methodology has been  developed for definitively evaluating candidate
technologies, consistent with the above evaluation process (DuCharme et al, 1992a).  Such
evaluations are  done at two levels:  (1) preliminary  evaluations with respect to generic, system-
wide applications for screening  a  large  number of candidates for potential development  (if
necessary) and transfer; and (2) detailed evaluations with respect to site-specific applications.
                                        280

-------
                                                                          Specific
                                                                      WM/ER Problem
                                                                                  i
                                                  Potential
                                           Application to Specified
                                              WM/ER Activities/
                                                 Conditions
                                     Base Case
                                     Alternative
                                     Technology
Cost-

Schedule*
Safety
Issues *

Acceptability
Issues*
                   Effectiveness*
                                 Value
                              Judgements
                               (Tradeoffs)
        Do not apply or
        consider further
                                   Likelihood of
                                    Success**
                                                                              Value
                                                                          Judgements
                                                                           (Tradeoffs)
  Continue    /
Development/
  Transfer
                (Cancel
             Development/
               Transfer
          Attributes of application of candidate technology if available (ie., fully developed and
          transferred), compared to application of Base Case alternative technology.

         r For full development and transfer of candidate technology.

          (adapted from DuCharme et al, 1992a)
  Figure 3  Comparative Evaluation of EM Technologies

EnviroTRADE  -  Ultimately, the site-specific  evaluations will be accomplished  through  a
computerized relational data base (ie, EnviroTRADE) which is being developed within IPD/ITEP
                                                281

-------
  by Sandia National Laboratories for DOE (Harrington and Harlan, 1992).  EnviroTRADE will
  consist of the following compatible modules:

   •    World-wide EM needs data base, describing the relevant conditions and requirements for
        various EM  activities at sites around the world (including the responsible parties),
        categorized by country;

   •    World-wide EM technology data base, describing the relevant attributes of technologies
        (including application parameters, availability and contacts); and

   •    Decision module for matching EM technologies and EM needs, either to identify and rank
        appropriate technologies for user-specified needs, or to locate potential applications (ie,
        markets) for user-specified technologies.


                      GROUTING TECHNOLOGY EVALUATION

 In order to demonstrate the use of the methodology for evaluating EM technologies and thereby
 assisting DOE to make rational and informed decisions regarding (1) development and transfer
 of foreign technologies and (2) application of technologies at specific sites, similar but different
 technologies have been evaluated at various levels:
           Ukrainian clav-based groutine technology is adequately described in other papers
       (Spickelmier, 1992; Verma, 1992). Its attributes are briefly summarized, and then the
       technology is evaluated in a preliminary fashion for generic applications to determine
       whether its development and transfer should continue.

  •    A. French silicon-based gel grouting technology is described.  It is evaluated in a detailed
       fashion with respect to application at a specific site, ie, to provide containment of the
       single shell underground tank wastes at Hanford, Washington, to determine whether it
       should be considered further for this application.

Grouting in general can be used to create a low permeability barrier for in situ containment, eg,
prior to remediation as a stop gap measure, during remediation to prevent transport of mobilized
wastes, or for  the long-term.  Grouting can  also be  used to redirect and concentrate a
contaminant plume for remediation. The resulting seals are applicable to containment of various
contaminants from basins/ponds, storage  tanks,  trenches/cribs, containerized wastes,  and
leaks/spills.


UKRAINIAN CLAY-BASED GROUTING TECHNOLOGY

A specific clay-based grout which may  provide improved containment has  been identified
through ITEP puCharme et al, 1992a and b).  The clay-based grout was developed by the
Ukrainian company SPETSTAMPONAZHGEOLOGIA (STG), which translates to "grouting and
geological enterprise".  This grouting technology can and has been  applied  to a variety of
problems, especially to provide relatively impermeable barriers (eg, during construction of shafts)
in fractured or karstic water-bearing rocks. Of particular interest  is its potential application in
                                       282

-------
providing a barrier to contain hazardous wastes at similar sites managed by DOE.

Properties - The clay-based grout consists of the following components:
  •    89% clay slurry at densities up to 87 pcf, which in turn consists of kaolin (or possibly illite
       if kaolin is unavailable) mixed with water;
  •    less than 10% cement or processed tailings; and
  •    1 to 2% of structure-forming reagents, such as sodium silicate.

The resulting grout is reported to have the following beneficial properties:
  «    chemically stable in a variety  of conditions;
  •    long-term pliability so that small ground displacements can be accommodated without
       compromising barrier characteristics;
  •    low shrinkage and erodability so that barrier  characteristics are not compromised by
       changes in groundwater conditions;
  •    non-toxic; and
  •    low cost relative to other chemical- or cement-based grouts.

Evaluation - The clay-based grout is very viscous, and thus can only be applied effectively in
"open" geologic systems, ie, fractured or karstic rock. In such cases, compared to other chemical-
or cement-based grouts, application  of the clay-based grout is believed to have the following
attributes, based on preliminary subjective evaluations:
  •    more effective because it is more pliable and long-lasting;
  •    cheaper;
  •    similar rate of application;
  •    similar safety issues; and
  •    more acceptable (once transferred and demonstrated) due to its better performance and
       use of natural, geologic, non-toxic materials.

Due to a lack of experience  with this technology in the US, however, there are significant
uncertainties in these attributes, especially in the range of site and waste conditions for which
the grout would  be effective and the duration of such effectiveness.  Such issues must be
adequately resolved, eg, by  demonstration  testing in the US,  before development  of the
technology can be  considered  complete and it  can be used for  DOE projects.   Such a
demonstration has been proposed, and has been estimated to take on the order of two years to
complete.  In addition to completing technical  development, the technology must  also be
transferred to the US before it is available for application. This has been at least partially effected
by STG forming a joint venture with Morrison-Knudsen to market the technology in the US.

Recommendation - Based on the preliminary evaluation of the clay-based grouting technology,
it appears that it should be considered further, once it becomes available, where containment in
"open" geologic systems is needed at DOE sites.


FRENCH SILICON-BASED GROUTING TECHNOLOGY

A specific inorganic, silicon-based grout which may provide improved containment has also been
identified through ITEP (DuCharme et al, 1992a and b). The silicon-based grout ("KLEBOGEL")
was  developed by  the French  company Societe Franchise Hoecsht (SFH).   This grouting
                                        283

-------
technology has been applied to a variety of problems, primarily to provide additional strength
for granular soils during excavation, but it has  reportedly also been used successfully in oil
reservoir secondary recovery to plug a permeable zone around an oil reservoir, under extreme
conditions (60-80°C and 1000m depth).  It is now  being considered as a means to provide a
barrier to contain hazardous wastes  around  underground storage tanks (USTs) at Hanford
Washington for Westinghouse Hanford Company (WHQ, a DOE  contractor.


Evaluation Process - The detailed evaluation  of the silicon-based grout for this  specific
application involved significant interactions between the evaluators and the technology owner
(ie, SFH), and between the evaluators and the potential users (ie, DO^VVHC).  As illustrated in
Figure 4, this was accomplished through the following steps:
        SFH -*-
   Detailed
 Description of
 Attributes of
 Silicon-based
   Grouting
 Technology
      Technical
       Experts
      (Including
     Operators)
Description of
Hanford-UST
 Technology
Requirements
                                                               WHO
  Review of
Description of
 Attributes of
Silicon-based
  Grouting
 Technology
 Description of
  Attributes of
 Hanford-UST
  Base Case
  Technology
      Comparative Evaluation of Silicon-based
       Grouting Technology (w.r.L Base Case
          Technology) - Tradeoffs among
            Differences in Attributes
        SFH-*-



       WHO4-
Bring SFH and
WHO Together
  to Pursue
Application of
Silicon-based
  Grouting
Technology to
Hanford-UST
                 Silicon-
              based Grouting
                Technology
                Preferred?
  Terminate
   Further
 Consideration
of Application of
 Silicon-based
  Grouting
 Technology to
 Hanfard-DST
        UST - underground storage tanks
        WHO - Westinghouse Hanford Company, DOE contractor at Hanford, Washington
        SFH - Societd Francase Hoechst, French developer of silicon-based grouting technoloav
        (adapted from DuCharme et al, 1992a)                                     ay
 Figure 4 Detailed Site-Specific Evaluation of French Silicon-Based Grouting Technology
1.    The list of relevant technology attributes was sent to the technology developers (ie SFH)
      for them to fill out


2.    Technical experts (including operators) reviewed the foreign participants' description and
      assessment of their technology's attributes.  These reviews were then resolved with the
      foreign participants.
                                           284

-------
 3.    The end users  (ie,  WHC)  identified and described  (in similar  terms)  their specific
       technology requirements for the USTs, and the domestic base case for comparison.

 4.    A quantitative comparative evaluation (i.e., differences in attributes) of the silicon-based
       grouting technology with respect to the base case technology for the Hanford USTs was
       conducted, involving WHC in determining appropriate tradeoffs among the attributes.

 5.    If the evaluation of the silicon-based grouting technology is favorable (ie, it is potentially
       better,  safer,  cheaper,  faster,  and/or more acceptable  compared to the  base  case
       technology), then SFH will be brought together with WHC to pursue application of their
       technology to the Hanford USTs.  Otherwise, application of the silicon-based grouting
       technology would not be pursued further for the Hanford USTs, although it may still be
       applicable for USTs at other DOE sites or for other facilities (eg, cribs, trenches, or ponds)
       at various sites (including Hanford).

Generic Properties - The silicon-based grout consists of various portions of sodium silicate, "silica
sol" (a dispersive suspension of very fine silica particles), "glyfix" (acid-based hardener) and tap
water.  They produce an amorphous, silicon-dioxide gel, with about 5-10% solids. The recipe is
proprietary and  confidential,  and can be  varied  on-site to suit  observed conditions.  As
demonstrated in the laboratory, the resulting grout has low permeability (ie, less than 10"9 m/s),
is very flexible, has low shrinkage (even under high temperatures) and is chemically stable, as
long as it is not allowed to dry out.  The materials are non-toxic  and EPA/NEPA approved
(although for other purposes). Prior to setting, the grout has very low viscosity (ie, 5 cPoise,
similar to water), which allows it to be injected in relatively low permeability ground (including
very fine sands); conceptually, grouting in highly permeable ground may require a first stage
grouting using fast sets or  more coarse, viscous types of  grout (eg, bentonite).  The viscosity
increases rapidly at setting (to an amorphous gel).  The setting time can be controlled over a wide
range (from minutes up to 2 days), in order to achieve proper penetration during injection, by
changing the mixture. The cost of the grout  material is a function of the mixture, which in turn
is determined by the desired setting time. A fast set mix (up to 2 hours) does not include silica
sol and costs about $l/gal, whereas a slow set mix  (up to 2 days) includes silica sol and costs
about $2/gal.

Generic Procedures - Access and equipment for drilling boreholes, and mixing and injecting
grout on site are necessary. Boreholes are  drilled around and  possibly below the site at an
appropriate spacing. These boreholes must be sufficiently stable and porous to allow subsequent
grout injection, and spaced closely enough to ensure adequate overlap.  The thickness and
continuity of  the barrier will be a  function of  borehole  spacing and grout penetration
characteristics.   For lateral containment,  the  boreholes are vertical whereas  for vertical
containment (eg, below tanks), the boreholes must be inclined or even horizontal. Such inclined
or horizontal boreholes are typically much  more difficult to drill and especially stabilize.  A
promising technique for drilling horizontal  boreholes below the site is "micro-tunnelling". It
should be noted that drilling will produce secondary wastes (contaminated soil) which must be
disposed of.

The grout is mixed per specification (proprietary and confidential) in a standard batch plant, and
injected into specific zones of the boreholes.  The  penetration distance of the grout into the
ground around a borehole is a function primarily of the following:
                                         285

-------
  •     Grout Viscosity - As previously noted, the viscosity of the grout is initially very low until
        it sets, at which point it increases rapidly and penetration effectively ceases; hence, the
        penetration distance is a function of the setting time, which can be varied over a wide
        range. Injection conditions (eg, increased temperature, dissolved organics, aluminates and
        salt content) may modify the setting time.

  •     Ground Permeability - The ground must be sufficiently permeable (ie, more permeable
        than clay) to allow adequate grout penetration at injection pressures below those which
        cause hydraulic fracturing, and yet not too permeable (ie, less permeable than gravel)
        which would cause excessive grout loss.  If the ground is too permeable (ie, >10"3m/s),
        then a first stage of grouting with bentonite cement or clay suspension, for example, may
        be necessary. The permeability of the ground will be a function of a variety of factors,
        including the moisture content  in the vadose zone.  Variability in soil or moisture
        conditions can cause preferential penetration, resulting in incomplete barriers. Similarly,
        injection in the  saturated zone requires water displacement, which may be incomplete,
        and may result  in dilution of the  grout

  •     Injection Pressure - The injection pressure should be maintained below that which would
        cause hydraulic fracturing.

 The general penetration characteristics of the grout are known to be relatively  good from "soil
 consolidation" applications, although uniform penetration is not as important an issue in that
 case.  It is  anticipated that grout  penetration  of several meters can  be obtained with an
 appropriate mix.

 In order to be effective, the grouted zone must be continuous across the potential contaminant
 transport paths and maintain a low permeability (over the required time frame).  Hence, the
 resulting barrier must be verified and possibly monitored with time. During injection, this might
 be done by monitoring injection pressures and volumes. After injection, this might be done by
 geophysical testing, eg:
  •    by ground penetrating radar in the grout holes or in nearby monitoring wells; and/or
  •    by DC resistivity surveys using sensors embedded in the grout holes
        o    for cross-hole analyses in the plane of the barrier, and
        o    for borehole-to-surface analyses across the barrier at various locations, although
             this might not be appropriate for horizontal barriers below a site.

 Generic Issues - Grouting can be used in ground which is sufficiently permeable to  achieve
 adequate penetration (eg, silty fine sands through coarse sands), although it may be difficult to
 drill stable open holes in such ground especially if not vertical However, the sealant may not
 be chemically compatible with all wastes  or site conditions.  Also, uniform penetration  may be
 difficult to achieve in heterogeneous site conditions (ie, soils,  moisture, stresses), although two-
stage grouting might be effective. It may be difficult under any conditions to install and verify
a continuous grouted zone over large areas, especially below a  site.

The specific technical issues associated with the generic application of the silicon-based grouting
technology to containment of hazardous wastes include:
                                         286

-------
 •     Horizontal Drilling - Although vertical drilling procedures are well established, horizontal
       drilling techniques (including micro-tunneling) have not been applied in contaminated
       granular soils.  Hence, there is a question whether stable horizontal grout holes can be
       drilled. This is a generic grouting issue, rather than specific to the silicon-based grout

 •     Injection Conditions - The penetration distance for specific site conditions is  uncertain
       until field-tested. Similarly, the uniformity of penetration in heterogeneous conditions (ie,
       relating to soils, moisture/water, and stress) is also unknown.  Again, these are generic
       grouting issues, rather than specific to the silicon-based grouting technology.  However,
       the effects of various injection conditions  (eg,  presence of non-neutral  pH,  dissolved
       organics, aluminates and/or salts) on the gelation process have not yet been determined.
       Adverse injection conditions might be modified by neutralizing (eg, with other chemicals)
       or removing (eg, by sweeping  with  deionized water), although this might mobilize
       existing contamination.

 •     Verification - The geophysical methods required for verification have not yet been tested
       on the grout. Again, this is a generic grouting issue, rather than specific to the silicon-
       based grouting technology.

 •     Long-term Integrity - The potential chemical reactions of the grout with various wastes
       and injection conditions have not yet been  determined.  Although the grout properties,
       both short-term (including setting times) and long-term, are anticipated to be good, based
       on general chemical stability  considerations, they may be affected by the following: (1)
       interactions with the waste, either during injection or subsequently;  (2) biodegradation
       in the vadose  zone; (3)  dehydration; and (4) other chemical-physical  reactions (eg,
       diffusion). For example:
        o    If the site is highly acidic or alkaline, the gelation  process* could be affected or the
             gel could dissolve  over time.
        o    Water soluble organics might diffuse into the gel, or be contained  within the gel
             if  not  displaced  during injection, and  the gel  might then  be subject to
             biodegradation.
        o    The gel might diffuse  ionically into very low salt content water over time.
        o    The gel might dissipate if evaporation is possible, although this is considered to
             be a very slow process at the depths of interest
        o    The  characteristics might  change with time if subject to hydrolysis (eg, due to
             radiation).
        o    The gel may be affected by changes in temperature, although based on limited
             testing such effects (eg, shrinkage)  are not significant for moderate changes in
             temperature and as long as evaporation does not occur. High temperatures affect
             the setting time, but apparently not the other properties.
        o    Cations may react with the silicates to form salts, although if the salts are formed
             on the outer boundary only, they may be beneficial and inhibit any dissolution
             of the gel with time.
        o    The presence of aluminates, caustic soda, sodium hydroxide, or potassium oxide,
             for example, during injection may have significant negative effects.

Availability - The issues identified above must be adequately resolved before development can
be considered complete.
                                         287

-------
 SFH, which is a subsidiary of Hoechst AG (Germany), has developed the silicon-based grout
 mixture (KLEBOGEL) and considers it to be proprietary and confidential. Hoechst AG also has
 an American subsidiary, Hoechst Celanese Co. in Charlotte NC, which is thus affiliated with SFH.
 SFH has expressed the desire to work with DOE. Hence, transfer of the technology to the US
 should pose few problems.

 No special equipment or expertise is needed for installation, so that a competitive environment
 exists for contractor services. Special contractors may be needed, however, for drilling horizontal
 grout holes and for geophysical verification.

 Generic Alternatives - The alternative to KLEBOGEL is other clay-, paraffin-, chemical- or
 cement-based grouts. Depending on the specific grout being compared, KLEBOGEL is generally:
 more effective in reducing permeability (due to its low viscosity and extremely small particle size,
 which allows greater penetration) although it may be less stable over the long-term; more costly;
 possibly less acceptable, although it is NEPA-approved, due to long-term integrity issues; and
 equally fast and safe to apply.

 Hanford USTs - The DOE has 334 USTs for storing radioactive and hazardous wastes at their
 operations across the US, with 149 single shell tanks (SSTs)  and 28 double shell tanks at Hanford.
 A total of 79 of these tanks, 66 of which are located at Hanford, are either known or are assumed
 to leak, thereby contaminating the soil and groundwater. In addition to the USTs, other facilities
 (eg, cribs, piping and trenches), either at Hanford or elsewhere (eg, INEL), experience similar
 problems.

 The Hanford SSTs are single shell reinforced concrete with  carbon steel liners, which are of
 various dimensions and buried at shallow depth.  Some of them have corroded and otherwise
 deteriorated over the years  to  the point that they are  now leaking their  contents into the
 subsurface.  The contents are liquid mixed wastes, which were relatively well known at the time
 of placement (due to controlled procedures), with the possible exception of some additional
 undocumented solvents.  However, over time, (1) the supernates were pumped off, but not
 characterized,  so that some wastes  may have been selectively removed, (2) water was added
 (sometimes many 10,000s of gallons) when the contents  leaked, possibly removing the more
 soluble wastes, and (3) chemical reactions have occurred, changing the remaining wastes. For
 example, several tanks are of special concern: one tank has high heat, so that 8000-10000 gallons
 of water are added per month for cooling, which presumably evaporates; one tank experiences
 hydrogen gas buildup, which vents (or "burps");  and another gas  has relatively high
 concentrations of ferrocyanide which (if high enough) could be explosive.

The subsurface consists primarily of more than 100 feet of permeable soils (sands and gravels
with some fines), overlying fractured basalt. The soil around and to some depth (about 8 feet)
below the tanks was excavated and reworked during tank construction.  The groundwater is
relatively deep (about 200 feet), with a significant gradient towards the Columbia River several
miles away. Near the tanks, the pH is generally somewhat alkaline (pH of 8-10), with elevated
temperatures (up to 60-75°C). Below leaking tanks, there  is a vertical plume of contaminants,
extending to various depths depending on the extent of leakage but generally contained within
the vadose zone.
                                       288

-------
Requirements for Hanford USTs - The Hanford USTs are to be totally remediated according to
a specific schedule, per the Hanford Federal Facility Agreement and Consent Order (ie, the "Tri-
Party Agreement - TPA").  In the TPA, the DOE has committed to a demonstration of a full-scale
closure  method for the USTs by 2004.   This will  require  the  following  activities: (1)
characterization of waste, tanks, and the site; (2) treatment and disposal, either insitu or retrieval;
and (3) site closure and monitoring, depending on whether any wastes remain in-place. Options
for insitu treatmeni/disposal consist of stabilization, grouting or in situ vitrification, all of which
would  result in  residual  wastes in-place  when  closed.    Options for  retrieval  and
treatmeni/disposal include  mechanical excavation or hydraulic sluicing followed  by various
treatment/disposal options, which may result in mobilization of wastes during removal and some
residual wastes in-place when closed. Waste separation techniques could be implemented as part
of treatment, to reduce the amount of wastes to be disposed of.

Functional requirements for  remediation are currently being developed (Rouse et al, 1992; Boomer
et al, 1990).  The primary performance requirements for the technology include occupational
safety (e.g., dose), public safety (e.g., dose), and groundwater protection (e.g., concentrations).
The long-term performance  requirements are currently being considered for a period of 300-1000
years after closure.

Base Case Technology for  Hanford USTs - Consistent with the specified requirements, the
reference base case technology is as follows:

  •    Removal of the wastes within the tanks by hydraulic sluicing followed by treatment and
       disposal.  Treatment and disposal consist of mixing the wastes with cementitious grout
       to stabilize it, and  placing the grout in vaults in the 200 Area.

  •    No subsurface containment is currently planned at the tank sites, because all of the waste
       is supposed to be  removed. However, following treatmeni/disposal,  the site would be
       closed per RCRA requirements (ie, an engineered cover), unless complete remediation has
       been effected and  verified.

Potential Application of Technology to Hanford USTs - The candidate technology, as described
above, can be applied to help satisfy the Hanford UST requirements  by enhancing specific
treatmeni/disposal options, as well as closure. Specifically, the silicon-based grouting technology
can be applied to the following areas:

   •    Retrieval and Treatment/Disposal - Create barrier below site to reduce leakage into the
       groundwater to acceptable levels,  during retrieval (especially hydraulically).  Such  a
       barrier is an additional feature, since none is currently contemplated, and might make the
       hydraulic retrieval option much more attractive. Such a barrier could incorporate a sump
       to collect and remove any contaminants.                    ,

   •    In Situ Treatment/Disposal Options - Create barrier below site to reduce leakage into the
       groundwater to acceptable  levels, both during and after treatment  During treatment,
       such a barrier is an additional feature, since none is currently contemplated. Similar to
       above, such a barrier could incorporate a sump to collect and remove any contaminants.
                                         289

-------
   •   Site Closure - If residual wastes remain in-place from retrieval and treatment/disposal,
       then the site must be closed as a landfill Similarly, if the in situ treatment/disposal option
       is adopted, then the site must be closed as a landfill, assuming that residual wastes will
       remain in-place.  In either case, a barrier below the site will supplement a cover to reduce
       leakage into the groundwater to acceptable  levels.  Conversely, if no residual wastes
       remain  in-place  (above the potential subsurface barrier), then the  subsurface barrier
       would not contribute to meeting  system performance requirements,  and would be
       unnecessary.

Concerns with the base case (or any other) technology include increased waste volume (including
secondary waste such as drill cuttings or leachate) and increased mobilization of waste (eg, due
to drilling fluid), as well as other aspects of short- and long-term performance.

Application Costs - The primary cost components for application of the silicon-based grouting
technology are the material costs and the  drilling costs. These and the total costs have been
estimated for a bowl-shaped subsurface barrier with a nominal diameter of 75 feet and a nominal
depth of 50 feet (ie, similar to a barrier surrounding a typical individual tank at Hanford), as
illustrated in Figure 5, as a function of borehole spacing and grout type (ie, fast or slow set),
based on the following assumptions:
      ^ 05
      tn .2
      O *5
      O &
 3.2
  3
 2.8
 2.6
 2.4
 2.2
  2
 1.8
 1.6
 1.4
 1.2
  1
0.8
0.6
0.4
0.2
                                                          material fast ($1/gal)
                                                          drilling ($250/lf)
   0              10              20

                      Borehole Spacing, s (ft)

   (adapted from DuCharme et al, 1992a)
                                                             30
                                                                             40
 Figure 5 Costs of French Silicon-Based Grouting Technology Applied to Hanford USTs
                                         290

-------
 •     average soil porosity of 0.2;
 •     grouting overlap of 33% with the grouted zone on either side (along the plane of the
       barrier, assuming a cylindrical grouted zone) in order to ensure barrier continuity, so that
       the average barrier thickness is about 1.5 times the borehole spacing;

 •     average drilling costs of $250/lf (although the typical costs for conventionally drilling
       vertical 2.5-4 inch diameter boreholes in contaminated material at Hanford is reported to
       currently be approximately $1500-2000 per linear foot, the total footage involved and the
       use of streamlined techniques should bring this cost down);
 •     total material costs of $l/gallon for fast set grout and $2/gallon for slow set grout; and
 •     all other costs (eg, for shipping, mixing,  injecting and verifying) are insignificant
The optimal (ie, lowest total cost) borehole spacings would thus be 12 and 8 feet for the fast and
slow grouts, respectively. However, it must be recognized that there is a maximum allowable
borehole spacing for each grout type, depending on its maximum penetration distance.  For a
33% overlap between adjacent holes, the maximum spacing cannot exceed 6/5 of the maximum
penetration distance, or the designed overlap will not be achieved and "windows" in the barrier
may occur.   The penetration distance,  in  turn,  depends on the site conditions, but might
reasonably be expected to  be on  the order of 6 and  12 feet for the fast and slow grout,
respectively, in soils similar to those at the Hanford UST. Hence, although the optimal spacing
(ie, 12 feet) for the fast grout cannot be achieved, the maximum possible spacing (ie, 7 feet)
results in a lower total cost ($1.0 million vs.  $1.2 million) than the achievable optimum spacing
for the slow grout and would be preferred.  Different optimal spacings and total costs would
result for different maximum penetration distances.

Application Rate - The rate of application is determined primarily by the rate of drilling. The
grout can generally be mixed and injected in less time than it takes to  drill the borehole and the
drill  rig is not necessary for this procedure.  Hence, the drill  rig can be drilling boreholes
continuously while previously completed boreholes are grouted. This process may be accelerated
by using multiple drill rigs. The total time required to install the bowl-shaped subsurface barrier
described above, with a borehole spacing of  7 feet, would be about 93  days if the combined rate
of drilling is, for example, about 50  ft per day (ie, considering multiple shifts and multiple drill
rigs) and if other activities (eg, drill rig setup) result in about a one day delay per borehole. This
would increase to about 93 days if the borehole spacing decreased to 7 feet.

Safety of Application - There are no significant safety issues associated with the grout itself, since
it is non-toxic and US EPA National Environmental Policy Act (NEPA) approved. However, there
are the normal operating safety issues associated with drilling, mixing and injection operations,
especially at a hazardous waste  site, although these operations are generally carried out at the
periphery of the site in order to contain the waste.  The grout  materials  are nonflammable,
inorganic, alkaline or acid chemicals, which  after mixing are weakly alkaline.

Acceptability of Application - The technology is  relatively cost-effective, ie, potentially improving
site performance at reasonably low cost In addition, the grout material will be US supplied,  can
be installed by local contractors, has been used previously in the US (for other purposes) and is
EPA NEPA approved (although for other purposes).  However, it may be difficult to construct
and verify seal continuity over large areas and for long times. Hence, although such seals might
be acceptable for short-term containment (eg, prior to or during remediation), they might not be
acceptable as the sole barrier for long-term containment but only as  a supplement to other
                                         291

-------
technologies; it should be noted that it may not be acceptable in any case to leave any waste in-
place.  Natural analogues and especially successful field demonstrations (eg, for the UST-ID)
would make this technology more acceptable (from an institutional,  regulatory,  and public
standpoint).

Comparative Evaluation of Technology Applied to Hanford USTs - The candidate technology
has been evaluated with respect to the base case in terms of their likely differences in attributes.
These assessed differences in attributes are very approximate, based on judgement, and should
subsequently be analyzed in greater detail if found  to be important  These assessments are
summarized below and in Table 1:

      Table 1 Comparison of Candidate Technology with the Base Case Technology
ATTRIBUTES
Performance
Costs*
Schedule*
Safety*
Acceptability
1 ESTIMATED FRACTIONAL
IMPROVEMENT IN
ATTRIBUTES RELATIVE TO
BASE CASE
1+0.6
-0.2
-0.1
-0.1
+0.4
Total Score
RELATIVE WEIGHTS
Performance
0.6
0.1
0.1
0.1
0.1
+0.36
Cost
0.1
0.6
0.1
0.1
0.1
-0.04
Acceptability
0.1
0.1
0.1
0.1
0.6
+0.26
      Includes additional costs associated with development and transfer, but does not include
      additional savings associated with decreased likelihood of subsequent remediation.


      Performance - Releases to groundwater and thereby dose to the public will be reduced
      significantly (about 60%) over the long term due to the presence of a subsurface barrier,
      even considering issues regarding installation and long-term integrity.

      Costs - Costs will be increased slightly (about 20%) due to activities associated with in situ
      grouting (including development and transfer costs), although the expected total life cycle
      costs would not increase as much due to a decreased likelihood of subsequent expensive
      remediation.

      Schedule - The schedule will be increased slightly (about 10%) due to the additional
      activities associated with in situ grouting (including development and transfer  time),
      although the expected total life cycle schedule would not increase as  much  due to a
      decreased likelihood of subsequent remediation.
                                       292

-------
  •    Safety - Operating hazards will be increased slightly (about 10%) due to the additional
       activities associated with in situ grouting, although the expected total life cycle hazards
       would not increase  as much due to a decreased likelihood of subsequent hazardous
       remediation activities.

  •    Acceptability - Acceptability will  be increased moderately (about 40%)  due to the
       unproved  long-term performance and decreased likelihood of subsequent remediation
       provided by the supplementary subsurface barrier.

Tradeoffs among the various attributes can be made by assessing and applying relative weights
to the differences in  attributes, and then summing the weighted differences (DuCharme et al,
1992a).  The relative weights, however, will be a function of the absolute level of the attributes
of the Base Case technology, so that a specific percentage improvement in an attribute will have
greater weight if the initial value is larger. In the absence of such information for the Base Case
Technology, several sets of relative weights have been used to cover the possible range in values,
as summarized in Table 1:

  •    Performance Dominated - If improvement in performance is six times more important
       than improvement in any of the other factors, which are otherwise equally important, the
       total score for the  candidate technology is +0.36, and it is thus strongly preferred.

  •    Cost Dominated - If improvement in cost is six times more important than improvement
       in any of the other factors, which are otherwise equally important, the total score for the
       candidate technology is -0.04, and it is thus slightly not preferred.

  •    Acceptability Dominated - If improvement in acceptability is six times more important
       than improvement in any of the other factors, which are otherwise equally important, the
       total score for the  candidate technology is -f 0.26, and it is thus moderately preferred.

Hence, based on the assessed differences in attributes between the candidate technology and the
Base Case technology and on various sets of relative weights (where each set reflects an emphasis
on improving a different factor), it can be concluded that the candidate technology is preferred
unless cost improvement is more than about 5 times  as important as improvements in other
factors.

It should be recognized that this evaluation did  not include the additional savings associated
with a decreased  likelihood of subsequent remediation. The sensitivity of  these results to the
assessed differences in attributes (including consideration of such potential savings), as  well as
to relative weights, should also be evaluated.

Proposed Issue Resolution - Issues associated with the application  of the silicon-based grouting
technology relate primarily to waste compatibility, long-term integrity, grout penetration (distance
and uniformity), and barrier verifiability. In addition, for horizontal barriers, issues relate to the
ability to drill stable horizontal grout holes.

Laboratory and field tests for barrier construction are necessary  to resolve these issues and
complete development of this technology for this application, and are being proposed  for the
UST ID.  This would include three phases: (1) laboratory bench scale testing to determine waste
                                        293

-------
 compatibility, long-term integrity and post soil injection performance; (2) single borehole injection
 tests  to determine penetration characteristics under field conditions; and, if warranted, (3)
 multiple borehole injection tests to determine barrier characteristics.  The laboratory tests would
 include enhanced 9CV90 testing, as recommended by EPA guidelines, supplemented with radiation
 and high heat  Accelerated testing would use simulated wastes for the Hanford UST, which is
 a Battelle-PNL recipe and is itself hazardous.  All the testing phases would include consideration
 of verification, ie, testing of the electrical resistivity and  dielectric properties of the grout in the
 laboratory and calibration of the geophysical testing in the multi-hole field test (eg, by subsequent
 drilling and sampling, or even excavation and inspection, of the barrier). The field tests would
 be conducted, for safety and economics, at a similar but uncontaminated site. Additional work
 may be necessary to customize the material for specific soiVwaste conditions and for refining the
 horizontal drilling and verification techniques.  Natural  analogues (if any) for the grout will be
 investigated.

 The time to complete development of this technology  (ie, demonstration  at Hanford  UST) is
 estimated to be on the order of less than two years.  This  does not include development of
 horizontal drilling techniques or new geophysical verification techniques.

 Recommendation - The detailed evaluation suggests that the silicon-based grout represents an
 improved  technology for application to the  Hanford  USTs, and thus  warrants  further
 consideration.
                                   CONCLUSIONS

DOE must make a variety of decisions within their EM program regarding (1) development and
transfer of improved EM technologies and (2) application of the optimum EM technologies at
specific sites.  To assist in this decision-making process, a methodology has been developed for
evaluating EM technologies in terms of (1) the degree to which they would accomplish specific
EM activities and satisfy specific requirements, under various conditions, if they were available,
and (2) the costs and risks (if any) associated with their becoming available. For demonstration
purposes, this methodology was applied at different levels of detail to two grouting technologies:

 •    The generic application of a Ukrainian clay-based grouting technology was evaluated in
      a preliminary fashion. It was determined that this technology, if available, represents a
      potential improvement over other grouts in "open" geologic systems. Hence, development
      and transfer to the US should be completed if containment in such conditions is a large
      enough problem for DOE to justify the development/transfer costs.

 •    The specific application  of a French silicon-based grouting technology to the Hanford
      USTs was evaluated in a detailed fashion.  It was determined that this technology, if
      available, represents a potential improvement over the current base case technology for
      the Hanford USTs, even  when the costs  of  making the technology available  are
      considered.  Hence, development and transfer to the US should be completed, and if
      successful its application to the Hanford USTs should be strongly considered.
                                        294

-------
                                   REFERENCES

 Boomer et al, 1990, "Functional Requirements Baseline for the Closure of Single-Shell Tanks,"
 WHC-EP-0338, draft report by Westinghouse Hanford Company to US Department of Energy -
 Office of Environmental Restoration and Waste Management, June.

 DuCharme, A., W. Roberds, and R. Jimenez, 1992a, "Matching Leading DOE Waste Management
 and Environmental Restoration Needs with Foreign-Based Technologies", report prepared for US
 Department of Energy  - Office of Environmental Restoration  and Waste Management,
 International Technology Exchange Program, October.

 DuCharme, A., R. Jimenez, and W. Roberds, 1992b, "International Technology Transfer to Support
 the Environmental Restoration Needs of the DOE Complex", Proceedings of the Nuclear and
 Hazardous Waste Management International Topical Meeting - Spectrum 92, Boise ID, August

 Harrington, M., and C. Harlan, 1992, "EnviroTRADE: An Information System for Providing Data
 on  Environmental Technologies and Needs Worldwide", Proceedings of  the  Nuclear  and
 Hazardous Waste Management International Topical Meeting - Spectrum 92, Boise ID, August

 Rouse  et  al, 1992, "Underground  Storage Tank  - Integrated  Demonstration  Functional
 requirements",  WHC-EP-0566,  draft  report by Westinghouse  Hanford  Company  to US
 Department of Energy - Office of Environmental Restoration and Waste Management, April

Spickelmier, K., 1992, "Former USSR's Largest Grouting Firm Speaks to US Audience", Mining
Engineering. August

Verma,  T., 1992,  'Potential  Application  of the Ukrainian Grouting  Technology  in  the
Environmental Restoration Area", US EPA Fourth Forum on Innovative  Hazardous Waste
Treatment technologies: Domestic and International, San Francisco, California, November.
                                        295

-------
         DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY (SITE)
                             Dr. Ernest Mayer
                        E. I. du Pont de Nemours, Inc.
                     Du Pont Engineering, Louviers 1359
                              P. O. Box 6090
                            Newark, DE 19714
                              (302) 366-3652
                                 Abstract

The novel  Du  Pont/Oberlin  Microfiltration  Technology  has  recently  been
demonstrated in  EPA's Superfund Innovative  Technology  Evaluation (SITE)
program.  Its key features are fine microfiltration at low cost using Du Font's new
Tyvek®*  T-980 flashspun olefin filter media coupled with Oberlin's  reliable
automatic pressure filter (APF) and Enviroguard's PROFIX" filter aid for metals
stabilization.

This new  microfiltration technology is best suited for contaminated heavy  metal
wastewaters and groundwaters. The SITE demonstration (1990) actually removed
Zn, Cu, Cd, Se and  Pb from the Palmerton, 'PA Zinc smelting Superfund site.
Basically,  99.95% removal of Zinc and Total Suspended Solids  (TSS); and firm,
dry (41%  solids) cakes that passed both the "Paint Filter Test"  and TCLP were
demonstrated.   Thus,  this  new  technology  provides  low  cost  metals
removal/stabilization all in one simple operation.

This paper will describe this new technology in detail, and will present some typical
application results.  It will  also detail where it has been applied since the SITE
demonstration.
* Du Font's trademark for its flashspun HOPE nonwoven filtration media.

" Enviroguard's trademark for its patented filter aid/stabilization agent.
                                    296

-------
           DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY (SITE)

                                  Dr. Ernest Mayer
                            E. I. du Pont de Nemours, Inc.
                          Du Pont Engineering Louviers 1359
                                  P. O. Box 6090
                                 Newark, DE  19714
                                  (302) 366-3652
TYVEK'T-980 MEDIA

     Du Pont has commercialized a new filter media based on flash spinning technology for
Tyvek spunbonded olefin. It has an asymmetric pore structure, a greater number of submicron
pores, and a smaller average pore sizel. 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 medial. 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 the 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
T-980 grade was sufficient so the added cost for a higher basis weight grade was not
warranted.  For example, T-980 produced slightly poorer effluent quality than the 'standard'
0.45(1 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 2). 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 Plant2-3.  This
installation realized almost a $350M annual savings when T-980 media was used with a more
efficient filter aid. An added benefit is its superior strength compared to microporous
membranes and the PTFE laminate media. This strength permits its use in robust automatic
pressure filters 4. The high strength 5 coupled with the tight, ~1- micron pore structure 1 led
to its use in the EPA Superfund  SITE program 6~9.

OBERLIN AUTOMATIC PRESSURE FILTER (APF) **

     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 10. Their APF
has other advantages, namely:

     •   Completely automatic,  unattended operation save for disposable media roll
         replacement and chemical treatment makeup. Standard PLC control.

     •   Enclosed operation for safety and handling of hazardous wastes.

     •   Fairly high operating pressure (up to 60 psig).
                                         297

-------
     •   Complete 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); and, optional media reuse if desired.

DU PONT TYVEK®/OBERLIN MICROFELTRATION TECHNOLOGY

     Thus, the Du Pont Tyvek/Oberlin Microfiltration technology has some unique
advantages, especially its completely automatic submicron, low-cost filtration and its dry cake
discharge 6.  This dry cake discharge feature is precisely why the resultant cakes pass the
modified EPA "Paint Filter Test" for land disposal n and in some instances pass the new EPA
TCLP (Toxic Characteristic Leaching Procedure) test for hazardous components 12"15,
provided a stabilization agent is used (i.e., Profix*** was used in the SITE program 9). Dry
cake/submicron filtration in one operation is why the Microfiltration technology was selected at
Savannah River over conventional crossflow microfilters and ultrafilters. In plating-waste
treatment the simple, one-step Microfiltration technology 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 Microfiltration technology has been
successfully applied. The technology is most suitable for hazardous wastewater where the
solids loading is not too high (i.e., generally less than about 5000 ppm).  Examples include
contaminated grpundwater, plating wastewaters, low-level radioactive wastes, plant equip-
ment/floor washings, cyanidic wastes, plant wastewaters that contain heavy metals, and metal
grinding wastes 2,3 & 6_ jn most of these cases Profix was used per the SITE technology 9 to
accomplish automatic, submicron, low-cost microfiltration and stabilization all in one simple
operation.
  Enviroguaid's trademark for its patented filter aid/stabilization agent.
                                         298

-------
CASE HISTORIES

     •Many successful case histories have been presented previously 2, but a few are worth
repeating here, especially those involving heavy metals removal and stabilization.  Other new
applications are also presented.

Savannah River Operation

     Table n details actual Savannah River plant wastewater treatment specifically aimed at
aluminum and uranium removal3. The data were obtained from two T-980/Oberlin units
which had been operating for about four 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
Microfiltration technology at very high 1200 gfd (gallons/sq ft/day) rates16. This rate is about
six 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.
Plans are now underway to evaluate Profix in this operation.

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. Two Microfiltration systems were
installed instead of crossflow microfilters because their dry cake feature significantly reduced
waste volume^ These units repaid their cost in three months of operation based on disposal
cost savings alone 16.

     Table HI details actual plant operation and compares the Microfiltration effluent with the
plant's discharge limits. As shown, these units reduced the effluent Total Suspended Solids
(TSS) and lead levels to well below the plant's required discharge limits at high 800 gfd flux
rates. In addition, these units routinely achieved >50% solids dry cakes that could be disposed
of in a RCRA-approved landfill (at significant cost savings compared to the concentrate from
the crossflow microfilters). Subsequent evaluation of the-Profix filter aid/stabilization  agent
per the SITE technology 9 actually improved flux and resulted in stabilized cakes (passed
TCLP).

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 16. A Microfiltration system 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 system produced dry cakes suitable for landfilling off-site, and
significant labor savings resulted.

     Table IV demonstrates that the Microfiltration system easily met the TSS and lead
discharge limits at very high 1500 gfd flux. Cakes were also quite dry at ~50% solids.
                                           299

-------
 Clarifier  Underflow Heavy Metals Removal

      This plant was faced with a land ban of their main claiifier metal hydroxide underflow
 sludge (2-3% solids) because it did not pass the EPA "Paint Filter Test" n.  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 Microfiltration technology was installed
 about three years ago. Table V 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 Microfiltration system operates automatically at significant labor
 savings compared to the manual press. Profix has been evaluated here and was implemented
 recently since the resultant cakes pass TCLP particularly for lead (Table V). As a result, off-
 site disposal costs were reduced about $200M annually.

 Battery Manufacturing Heavy Metals Removal

      This plant consistently exceeded their permitted discharge limits from their conventional
 clarifier/underflow press system and considered an overflow polishing sand filter.
 Simultaneously, they heard of our Microfiltration technology and decided to evaluate it as a
 polishing filter. Testing showed that it could replace the entire clarifier/press installation.
 Table VI shows excellent  metals removal, TSS reduction, and excellent effluent turbidity
 (0.5 MTU).  Flux is quite  high at 2400 gfd and capacity from the single system exceeds
 60,000 gpd. These benefits are in addition to dry cakes that could be easily disposed of in a
 RCRA-approved landfill.

 Heavy  Metals  Removal from Chemical  Plant Wastewater

     This plant was faced with severe heavy metal discharge restrictions based on recent EPA
 chronic toxicity regulations (ie, 15 ppb Cu, 60 ppb Pb, 130 ppb Ni, 150 ppb Or, and 300 ppb
 Zn). In  addition, this mixed heavy metals waste had to be stabilized to meet the "third-third"
 EPA BDAT (Best Demonstrated Available Technology) regulations for land disposal
 2,12 & 13> jon exchange was first evaluated, but it could not meet the low ppb limits for all
 the metals and its regenerant required off-site stabilization at high cost As a result, the
 Microfiltration technology was next evaluated since it could produce dry, stabilized cakes, but
 previously it never achieved these low metals limits. An extensive pilot testing program was
 conducted and finally, it was demonstrated that these limits could be achieved quite routinely
 (Table VII). In addition, these low metals limits were obtained at a very high 3600 gfd flux;
 and the resultant "dry" cakes passed the TCLP test for all eight toxic metals (per EPA "third-
 third" regulations issued May  8,1990 12 & 13). As a consequence, the Microfiltration
technology was selected and one large system (along with a Profix addition system) has been
operating successfully for one year; and a second system is being installed now.

 Groundwater Lead  Removal

     A RCRA hazardous  site required lead removal from contaminated groundwater to a level
of less than 50 ppb to meet primary drinking water standards as well as the chronic toxicity
criteria.  The responsible party hired an outside consulting firm to study and  recommend a
suitable  treatment scheme. Unfortunately, their piloted treatment process could only achieve
                                         300

-------
4 ppm; and their recommended activated carbon polishing step also could not achieve the low
required 50 ppb lead discharge limit. The Microfiltration technology was then evaluated and
outstanding <10 ppb lead removals were obtained (Table Vin). As a result, this technology
will be used and a single large system is being installed. Profix was used here and the cakes
passed TGLP for lead.

Chemical Plant's Groundwater Metals  Removal

     This plant had .groundwater contaminated with metals and organics, and hired an outside
consulting firm to study and recommend a suitable treatment scheme.  Unfortunately, their
piloted sarid filter scheme couldn't achieve the required low metals levels; and as a result the
Microfiltration technology was evaluated. As can be seen from Table IX, it achieved
outstanding metals removal (i.e., below detection limits) at a very high 4000 gfd flux. In
addition, filtrate turbidity, TSS, and cake % solids were all excellent; and the cakes passed
TCLP.  As a result, the system is now being installed to treat this groundwater.

Groundwater Hexavalent  Chrome Removal

     This Superfund site required groundwater chrome removal to the 50 ppb level. Simple
sodium bisulfite chrome reduction followed by Microfiltration resulted in excellent chrome
removal to below the 50 ppb detection limit, as well as excellent filtrate TSS and turbidity
(Table X).  Flux was very high at 3600 gfd and the resultant cakes were stabilized. However,
the client opted for ion exchange (DC) since only one metal was present and it reduced solid
waste volume. In this case, IX was a better choice for the client; and illustrates that all metals
removal technologies have their niche.

Groundwater Mixed  Metals Removal

     This mixed organics/metals contaminated groundwater was chemically treated to destroy
the toxic organics and the resultant effluent was treated by the Microfiltration technology
(Table XI). As can be seen from Table XI, excellent metals removals were obtained basically
to the detection limits. Flux was reasonable (1750 gfd) considering the high  50 ppm Fe
content; cakes passed the new TCLP test; and filtrate turbidity and TSS were outstanding at
0.05 NTU and 0.03 ppm, respectively (i.e., better than distilled water).  In this case, the
regulatory limits were later relaxed so direct discharge was possible; and  the Microfiltration
technology was not required.

Wastewater Mixed Metals Removal

     This plant required stringent heavy metals removal to meet chronic  aquatic toxicity
criteria. A variety of technologies were evaluated, but the Microfiltration technology provided
the best overall metals removal of this waste (Table XII) as most were removed to below
detection limits. The resultant cakes also passed TCLP, which,was also  a big factor since other
technologies required additional post-treatment to produce stabilized solids.

Wastewater Zinc Removal

     This wastewater required zinc removal to 1 ppm prior to discharge; and in this case, only
the Microfiltration technology was evaluated because selective ion exchange could not achieve
the 1 ppm limit. Table Xin shows that the Microfiltration system easily achieved the 1 ppm Zn
                                          301

-------
 limit as well as excellent Cd, Pb, and Cu removals like in the SITE demonstration 9 where
 these metals are normally associated with Zn smelting. In addition, filtrate turbidity and TSS
 were excellent and flux was reasonably high (1000 gfd) for this high 3000 ppm TSS waste.
 Cakes again passed the TCLP test In this case, despite the promising microfiltration results
 the plant opted to incinerate this waste.

 Explosives Plant Pb Removal

 This plant sends their Pb wastewater off-site for treatment and stabilization at considerable
 expense.  They heard of the Du Pont/Oberlin Microfiltration technology and evaluations in
 Table XIV demonstrated complete feasibility as Pb was removed to well below the proposed
 discharge limits and cakes passed TCLP at a reasonably high 2000 gfd flux.  As a
 consequence, a small system has been started up recently.

 Chelated Wastewater  Cu  Removal

 This plant had two wastewater streams which required Cu removal to meet the 50 ppb toxic
 discharge limit as well as to protect the on-site biological waste treatment system. One stream
 contained chelated Cu (Case II) while the other was chelated as well as heavily contaminated
 with organics (Case II). Ion exchange and oxidation with iron were tried on  Case I stream
 with litSe success while hot sodium sulfide precipitation was successfully demonstrated for
 Case H after the organics were removed. However, it was later found that complete organics
 removal could not be assured so the Microfiltration technology was evaluated on  both streams.
 Table XV shows that Cu removal to 10 ppb was obtained at high 4200 gfd flux (Case I) and
 <200 ppb Cu at 300 gfd flux for Case H The higher -200 ppb Cu in Case II was due to the
 copious ~50% organics present, but combined waste discharge was below 50 ppb. In
 addition, both cakes passed TCLP and even organics were stabilized when a mixed
 Profix/organic stabilizer (FO1) was used.  Economics showed that Case I treatment costs
 amounted to low $1.60/1000 gallons while Case II about $130/1000 gallons, but Case H
 disposal cost savings amounted to about $1350/1000 gallons due to the high level of organics
 present that could be recovered and recycled. As a consequence, a project is being
 implemented to install a Microfiltration system.

 ACTUAL EPA SITE DEMONSTRATION

     The  Du Pont/Oberlin microfiltration technology was demonstrated by the EPA under the
 SITE program in April - May, 1990 at the Palmerton, PA Superfund site 7-8-9-& 16.  This
 site's groundwater is contaminated with heavy metals by runoff from a huge 2.5 mile long
 cinder bank as a result of a zinc smelting operation. As expected, the groundwater contained
 high concentrations of zinc as well as manganese, lead, cadmium, copper and selenium (Table
XVI). The Microfiltration system was able to remove 99.95% of the zinc, 99.95% of the TSS,
produce dry cakes that passed the "Paint Filter Liquids Test", 41% cake solids, filtrate that met'
NPDES discharge limits, and cakes that passed both the EP Toxicity and the new TCLP test,
because the new filter aid / stabilization agent (Profix) was used 7«8-9-& *6.

SUMMARY

     These cases and actual operating applications demonstrate the utility of the Du Pont/
Oberlin microfiltration technology to remove heavy metals at very high flux rates and to
simultaneously produce dry cakes that pass the EPA "Paint Filter Test", and in some cases, the
                                        302

-------
new EPA TCLP leaching test (if Prefix filter aid/stabilization agent is used per SITE
demonstration - Refs. 2,6,7,9). This technology is quite competitive when compared to
microfilter cartridges, crossflow microfilters, and ultrafilters.  In some instances, the system
can replace the conventional three-stage metal finishing 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 and stabilizing hazardous metal-bearing
wastewaters and groundwaters.


REFERENCES


1    H S. Lim and E. Mayer, 'Tyvek® for Microfiltration Media," Fluid/Particle Separation
     Journal. 2(1): 17, (March, 1989).

2.   E. Mayer, "Du Pont/Oberlin Microfiltration System for Hazardous Wastewaters," Paper
     presented at US EPA Second Forum on Innovative Hazardous Waste Treatment
     Technologies", Philadelphia, PA, May 15-17,1990.

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

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

5    Du Pont Tyvek® Bulletin E-24534, "Tyvek® Engineered Specifically for Filtration,"
     (1988).

6.   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).

7.   K. Topudurti, S. Labunski, and J. Martin, "Field Evaluation of a Microfiltration
     Technology to Treat Groundwater Contaminated with Metals" Proceedings of 11th
     National Conference, Superfund '90, Washington, DC, November 26-28,1990,
     pp 425-432.

8.   G. L. Stacy and S. C. James, "The Superfund Innovative Technology Evaluation (Silt)
     Program," Control Hazardous Materials. 4(1): 23 (January/February, 1991).

9.   EPA Final Technology Evaluation Report: "SITE Program Demonstration of the
     Du Pont/Oberlin Microfiltration Technology," (July 1991).

10.  Oberlin Filter Co.'s Bulletin, "Oberlin Pressure Filter," (February, 1988).

11.  N. J. Sell, "Solidifiers for Hazardous Waste Disposal," Pollution Eng.. 20(8): 44 (August,
     1988).
                                         303

-------
12.  Federal Register, Vol. 51, No. 114 (June 13, 1986); and updates: Vol. 53, No. 159
     (August 17,1988) and Vol. 55, No. 61 (March 29,1990).

13.  J. Newton, "Understanding the Toxicity Characteristic Rule," Pollution Eng.. 22(9)- 90
     (September, 1990).                                                B    w

14.  J. Biedry, "Managing Waste to Meet Federal Land Ban Rules," Pollution Eng.. 22(10)-
     46 (October, 1990).                                                 '*"   U  }'

15.  J. Hauck and S. Masoomian, "Alternate Technologies for Wastewater Treatment,"
     Pollution Eng.. 22(5): 81 (May, 1990).

16.  EPA Final Applications Analysis Report: "Du Poni/Oberlin Microfiltration Technology "
     (May, 1991).
                                      304

-------
                          TABLE I
MEDIA
COST PER GALLON WASTE FILTERED
Actual Costs/Gallon Filtrate (cents/gal)*

Media
Microporous Membrane
Tyvek® T-980
PTFE Laminate
Melt-blown PP
Melt-blown PP
Nom.
Rating
(urn)
0.45
1
0.8
1
~5
Overall
Filtrate
Quality
Excellent
V. Good
V. Good
Good
Poor

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 H

        ACTUAL SAVANNAH RIVER  PLANT  OPERATION
  WITH  DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
. Aluminum (ppm)
Lead (ppm)
Zinc (ppm)
Copper (ppm)
Uranium (ppm)
Capacity (gpd)
Flux (gfd)
Actual State
Raw
Waste
110
687
127
1.6
0.5
2.0
2.3
35,000
•
discharge permit values.
T-980/
Oberlin
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
—
—

                            305

-------
                              TABLE Hi

     ACTUAL ELECTRONICS  MANUFACTURING PLANT WASTEWATER
        WITH  DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Lead (ppm)
Barium (ppm)
Cadmium (ppm)
Capacity (gpd)
Flux (gfd)
Cake % Solids
Cakes Pass TCLP?

Raw
Waste*
>1,000
1,000
20
25
2
2,500
—
—
No

Du Pont/
Oberlin
0.3
<1
0.04

-------
                               TABLE V
          CLARIFIER  UNDERFLOW HEAVY METALS  REMOVAL
       WITH  DU  PONT/OBERLIN MICROFILTRAT1ON TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Lead (ppm)
Zinc (ppm)
Copper (ppm)
Capacity (gpd)
Flux (gfd)
Cake % Solids
Cakes Pass TCLP?
Raw
Waste*
>1,000
17,000
40
410
1,050
6,800
—
—
No
Du Pont/
Oberlin
1.0
2.5
<0.01"
0.2
1.2
10,000
900
50
Yes
Discharge
Limits
NR
20
5.0
5.0
5.0
—
—
40
Must pass
                               TABLE VI
           DIRECT FILTRATION FOR HEAVY  METALS REMOVAL
               FROM A  BATTERY MANUFACTURING PLANT
      WITH  DU  PONT/OBERLIN MICROFILTRATION  TECHNOLOGY ***

Property
Turbidity (NTU)
TSS (ppm)
Nickel (ppm)
Cadmium (ppm)
Zinc (ppm)
Cobalt (ppm)
Capacity (gpd)
Flux (gfd)
Raw
Waste
175
1,600
30-50
15-40
0.2-1.0
0.4-1.5
40,000
—
Du Pont/
Oberlin
0.5
<1
<0.10
<0.05
<0.10
<0.05
60,000
2,400
Discharge
Limits
NR
20
2.27
0.4
1.68
0.22
— .
—
*    Existing plant clarifier underflow sludge was prevbusly hauled off-site to hazardous landfill.
*•    Detection limit
•**   Replaced clarifier and underflow press.
NR <= Not Regulated.
                                  307

-------
                     TABLE VII

HEAVY METALS REMOVAL  FROM CHEMICAL PLANT WASTEWATER
 WITH DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Copper (ppb)
Lead (ppb)
Zinc (ppb)
Nickel (ppb)
Chromium (ppb)
Capacity (gpd)
Flux (gfd)
Cakes Pass TCLP?
Raw
Waste
>100
100
2,000
100
3,500
400
1,000
75,000
—
No
Du Pont
Oberiin
0.06
<0.2
<5
<5*
<10*
<5*
<100*
130,000
3,600
Yes
Discharge
Limits"
<0.2
<1
15
60
250
130
150
...
._
Must Pass
                    TABLE VHI

           GROUNDWATER  LEAD  REMOVAL
WITH DU PONT/OBERUN  MICROFILTRATION  TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Lead (ppb)
Flux (gfd)
Cakes Pass TCLP?
* Detection limits
Based on EPA
NR m Not Regulated
Raw
Groundwater
<10
120
20,000
—
No
Chronic Toxicity tests
Du Pont/
Oberiin
0.1
<0.2
£10
1,400
Yes

Discharge
Limits "
NR
30
50
«.»
Must Pass

                       308

-------
                              TABLE IX

      CHEMICAL PLANT GROUNDWATER METALS  REMOVAL  WITH
          DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Fe (ppm)
Ba (ppb)
Cr (ppb)
Ni (ppb)
Zn (ppb)
Cake % Solids
Flux (gfd)
Cakes Pass TCLP?

Raw
Groundwater
100
80
40
900
45
50
300
—
—
No
TABLE
Du Pont/
Oberlin
0.17
0.3

-------
                  TABLE XI

     GROUNDWATER  METALS REMOVAL WITH
DU PONT/OBERLIN  MICROFILTRATION  TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Fe (ppb)
Cr (ppb)
Ni (ppb)
Zn (ppb)
Flux (gfd)
Cakes Pass TCLP?

Raw
Groundwater
160
180
50,000
40
300
14,000
—
. No
TABLE
Du Pont/
Oberlin
0.05
0.03
<100*
10
<5*
<10*
1750
Yes
XII
Discharge
Limits ••
NR
30
NR
150
130
300
...
Must Pass

MIXED WASTEWATER METALS REMOVAL WITH
DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Fe (ppm)
Cr (ppb)
Cu (ppb)
Zn (ppb)
Cake % Solids
Flux (gfd)
Cakes Pass TCLP?
* Detection limit
Based on EPA
NR = Not Regulated
Raw
Groundwater
<1000
1200
80
1800
1000
14,000
—
—
No
*
Chronic Toxicity Tests.
Du Pont/
Oberlin
0.2
0.4

-------
                 TABLE XMI
       WASTEWATER ZINC REMOVAL WITH
DU PONT/OBERLIN MICROFILTRATION TECHNOLOGY
Property
Turbidity (NTU)
TSS (ppm)
Zinc (ppm)
Cd (ppm)
Pb (ppm)
Cu (ppm)
Flux (gfd)
Cake Pass TCLP?

Raw
Wastewater
>1000
3000
1500
0.15
0.70
0.25
—
No
TABLE
Du Pont/
Oberlin
0.06
0.13
<0.01*

-------
                                 TABLE XV

           WASTEWATER Cu  REMOVAL WITH DU PONT/OBERLIN
                     MICROFILTRATION  TECHNOLOGY
       Property
 1. Chelated Case:
   Turbidity (NTU)
   TSS (ppm)
   Fe (ppb)
   Cu (ppb)
   Flux (gfd)
   Cakes pass TCLP?

2. Organics Case:
   Turbidity (NTU)
   TSS (ppm)
   Cu (ppm)
   Flux (gfd)
   Cakes pass TCLP?
   - with PROFIX
   -withPROFIXFOl
Organics Recovery
   (Vol %)	
   Raw
Wastewater
    20
    25
   500
  2,500

    No
  >1000
  30,000
   1600

   NO
  ~10%
    Du Pont/
     Oberiin
      <0.3
      <0.2
      20
      10
     4200
      Yes
       2
     <0.2
     300
     Yes
  (Metals Only)
(Metals & Organics)
    98.6%
   Discharge
     Limits
      50

   Must Pass



     0.2

   Must Pass


Cannot Discharge
                                TABLE XVI

 DU PONT/OBERLIN MICROFILTRATION  TECHNOLOGY  DEMONSTRATION
              BY EPA AT PALMERTON, PA SUPERFUND SITE
Property
Turbidity (NTU)
TSS (ppm)
Zinc (ppm)
Manganese (ppm)
Lead (ppm)
Cadmium (ppm)
Copper (ppm)
Selenium (ppm)
Cake % Solids
Cakes Pass EP Tox?
Cakes Pass TCLP?
* Detection Limit.
NPDES discharge
NRs Not Regulated.
Raw
Waste
>1,000
1,700
500
50
0.1
0.5
0.2
0.1
—
No
No

limits.

Du Pont/
Oberiin
0.1
0.8
0.25
40
Yes
Yes



Discharge
Limits"
NR
30
2.4
NR
0.7
0.2
NR
NR
NR
NR
Must Pass



                                    312

-------
                 U.S. ARMY CORPS OF ENGINEERS
             TOXIC AND HAZARDOUS MATERIALS AGENCY
     Composting of Explosives-Contaminated Soil

        at the U.S. Army UmatiSIa Depot Activity
Composting
    Composting  is  a  controlled  biological  process  by  which
biodegradable  materials   are   converted  by  microorganisms  to
innocuous,  stabilized by-products.  Inmost cases,  this is achieved
by the use of indigenous microorganisms.  Explosives -contaminated
soils are  excavated and mixed  with bulking agents such  as wood
chips, and organic amendments such as animal, fruit and vegetative
wastes.    Maximum  degradation   efficiency   is   controlled  by
maintaining moisture content, pH,  oxygenation,  temperature and the
carbon to nitrogen ratio.  There are three process designs used in
composting; aerated static  piles, windrowing, and mechanically
agitated in- vessel composting (Figure 1) .  This  technology requires
substantial space to conduct the composting operation and results
in approximately a two fold volumetric  increase in material due to
the addition of amendment  material.

    The first composting demonstration  for explosives -contaminated
soils  conducted  at  Louisiana  Army   Ammunition  Plant   (L&AP)
demonstrated  that  aerobic,  thermophilic  composting  is  able  to
reduce  the concentration  of explosives  (TNT,  RDX and  HMX)  and
associated toxicity to acceptable health  based clean-up levels.
However,   an   economic  analysis   determined   that  full  scale
implementation of composting of explosives -contaminated soils using
previously investigated design  parameters was not economically
competitive with incineration.  An optimization field demonstration
was initiated at a National Priority List (NPL) site at Umatilla
Depot  Activity,  Hermiston,  OR,  to investigate  several process
design  parameters  that  would  make  this  technology more cost
effective.  In  addition,   extensive chemical characterization and
toxicity studies were  conducted on the  final composted product.
                             313

-------
 General Site Information
     Umatilla Depot Activity (UMDA) ,  Henniston,  OR,  was  selected as
 the site for  the  compost  optimization field study.  Between 1950
 and  1965,   a  washout  facility  for  recovering explosives  from
 unserviceable munitions  was operated  at  UMDA.   Pressurized hot
 water was used  to melt out explosives  from munition bodies and
 upon cooling and  settling,  the explosives were recovered   Large
 quantities  of  water were  contaminated with explosives "in this
 operation.  It  was  standard practice,  at the time,  to discharge this
 water into unlined settling basins.  Explosives contained in these
 waters consisted  of TNT,  RDX and  HMX.   in 1987,  these  washout
 lagoons were  placed  on  Environmental Protection Agency's  EPA
 National   Priorities  List  (NPL)  because  of  the  presence  of
 explosives  in the water table aquifer  some 50  feet beneath the
 lagoons.   ^The  explosives   concentration  levels  in the  washout
 lagoons soils approach a maximum concentration of  10%  (100 000
 mg/kg} .  Explosives contaminated soils  from these  lagoons were hand
 excavated  for  this  composting  optimization  study.    The  mean
 concentrations  of TNT,  RDX and HMX were  13,380,  1,071, and 274
      ,  respectively.
 Optimization Field  Study OtHeetivea and  Deaicm


   The  primary objective  of  this composting optimization  field
 study  was  to  increase the  quantity   of  soil  processed  in  a
 composting treatment system per unit time.  Since soil  throughput
 is  dependent  on  the  rates of degradation and  the percent  soil
 loading,  the  key  variables   investigated  in  this  study  were
 amendment  mixture  composition  and  percent  contaminated  soil
 loading.   In  addition, two technologies were evaluated,  aerated
 static pile and mechanically agitated in-vessel composting systems.

   Amendment selection was based  on  adiabatic  testing using  a
 combination of fifteen readily available  agricultural wastes.  The
 amendments selected and their approximate cost  are provided in
 Table 1.  Percent  soil loading was investigated using seven 3 cubic
 yard  aerated  static  pile  systems  which  were  constructed  from
 fiberglass to model actual  static pile conditions  (Figure  2)  The
 tests were used to  investigate soil /amendment ratios using 6  7
 10, 20, 30 and 40 volume percent  of  explosives -contaminated  soil
and composted for 90  days.   Different soil amendment ratios and
amendment mixture compositions were investigated using a  special
seven- cubic yard pilot scale mechanically  agitated in-vessel (MAIV)
system which was constructed according to  rigorous explosive saf etv
standards by Fairfield Engineering Co.,  Marion,  Ohio (Figure 2)
                               314

-------
The mechanical pilot system uses rotating  augers attached to the
rotating  cover  to  mix  the  compost.     The  first  two  tests
investigated  two  different  amendment   compositions  using  10%
contaminated  soil  volume.   The  final   two  tests  were  used  to
investigate soil/amendment ratios using 25 and 40 percent volume of
explosives contaminated soil with the optimum amendment composition
selected from the previous tests.   The MA.IV  tests were composted
for 44 days.

    The static pile  systems and the  MAIV system were  housed in
greenhouses to protect them from the  environment and prevent the
spreading of contamination from explosives contaminated dust.   A
computer-based data  acquisition  and  control  system was  used to
monitor  and regulate  the  environment   in each  of the  compost
systems.  Temperatures were kept from exceeding 55°C using forced
aeration  and  the  moisture content maintained  between  45  -  50
percent.  Compost  samples were taken  at  various  time  intervals,
homogenized and split  into two fractions.   Chemical analysis for
the presence of TNT, RDX and HMX were conducted by High Performance
Liquid  Chromatagraphy  methods on  the first  fraction  while the
second fraction was subjected to toxicity testing.

    Since the implementation of this  technology will be based on
it's ability  to  meet health based  clean-up  criteria,  the second
objective  was   to  conduct  a  chemical   characterization  and
toxicological  evaluation  of  the  resultant   composted  material.
Leachates  from  the  EPA  Synthetic Precipitation  Leachate  Test
(referred  to  as  the  "Clean Closure  Leaching Test  or  CCLT) and
organic  solvent  leachates  were  subjected  to  Ames  bacterial
mutagenicity  tests,  acute and chronic  toxicity  tests  using the
aouatic crustaceans  Ceriodaphnia  dubia.  and  a  rat oral  toxicity
screening   of   finished  compost.     Characterization   included
determination of  the  explosives and TNT metabolites in the  composts
and organic solvent extracts.


Optimization Study Results


   The composting optimization study confirmed the LAAP composting
study results which indicated that composting can effectively  treat
TNT   RDX  and  HMX contaminated matrices.   The  optimization  study
indicated that both static pile and MAIV composting technological
approaches for implementing composting are effective in degrading
explosives.   The percent reduction of explosives observed in the
tests  are provided in. Table  2.    In  the  static pile tests, the
majority  of the  degradation occurred  in the  first 44  days of
composting while in the MAIV tests  the majority of the degradation
occurred  in the  first  10 days.   The results of this optimization
                               315

-------
 study indicated the amendment composition is an important parameter
 in achieving maximum reduction of RDX and HMX.  The maximum loading
 level for both  technologies  appears to be  30 volume percent and
 mixing is important in achieving rapid and extensive destruction of
 explosives.   Since mixing is an important process design parameter
 a composting windrow field demonstration has been initiated.    '

     Chemical characterization and toxicity  testing on the finished
 compost  sponsored  by  the  U.S.  Army  Biomedical   Research  and
 Development  Laboratory and  conducted by  the  Oak Ridge  National
 Laboratory  (ORNL) concluded that composting can effectively reduce
 the concentrations  of  explosives and bacterial. mutagenicity  in
 explosives-contaminated soil,  and can reduce the aquatic toxicity
 of leachable compounds.  .The  CCLT leachate toxicity to humans was
 estimated by comparing the concentrations of TNT, RDX and HMX with
 100 times their EPA Drinking Water Equivalent levels (a  100 fold
 dilution of  leachate in drinking water supplies was assumed,  as in
 RCRA) .   The  CCLT leachates  from the  composts met this  criteria
 indicating the  toxicity to humans  is not a  serious concern    The
 ORNL noced  that small levels  of  explosives  and  metabolites,
 bacterial mutagenicity,  and  leachable toxicity  do  remain  after
 composting;  however, they do not pose a serious health concern.  In
 fact,  no mortality  or toxic effects were  observed in a  rat oral
 toxicity screen.

    Chemical  analysis  of the finished  compost  indicated  that  the
 nitro groups in_TNT were reduced to amino groups, producing  first
 the monoamino-dinitrotoluenes and then diamino-mononitrotoluenes
 These metabolites have  been shown to  be  less  mutagenic  than  the
 parent compound in previous studies.   The  observed TNT metabolites
 did not quantitatively   account  for   the  decreases  in TNT
 concentrations.  The ultimate fate of the biotransformed explosives
 still remains unknown;  however,  extensive organic extraction and
 alkaline digestion of a  compost inoculated with radio-labelled TNT
 suggested that   a major portion  of  the  bio trans formed  TNT was
 chemically bound to the compost  and  not mineralized.   The nature
 and significance of the biotransformation product-soil  binding will
 continue to be investigated.
UMDA Washout Lagoon Feasibility Studv
    As the composting optimization study was concluding at UMDA in
1991, the Remedial Investigation (RI) of these washout  lagoons was
also nearing completion.  The washout lagoons soils were broken out
from the other sites in the RI and made the subject of  a separate
Risk Assessment and Feasibility Study (FS) .  An FS  was conducted in
accordance   with   the   Comprehensive   Environmental   Response,
                              316

-------
Compensation,  and Liability  Act  of  1980,   as  amended by  the
Superfund Amendments  and Reauthorization  Act of  1986.    The FS
summarized  the baseline  risk  assessment,  the  remedial  action
objectives,  screened potential remedial technologies and evaluated
in detail the most promising remedial  technologies.   Since there
are no federal cleanup standards  for explosives contaminated soils,
the cleanup  standards were developed based upon the remedial action
objectives  in  the Risk  Assessment and the evaluation of various
treatment alternatives.

    The remedial action objectives were used to determine a range
of  soil volumes  and  treatment  standards  to  be  considered  for
remediation.   Four  potential excavation  scenarios,  including a
cleanup  to  background,   were  considered  thus providing varying
degrees  of  contaminant  removal  and risk  reduction.   The state
cleanup  standard  requires  that soils  should be  cleaned  up to
background if possible,  or if  not, to a level which is protective
of human health and the environment.  Soil sample analysis within
the lagoons  determined that 90% of the explosives contaminants were
in the 5 feet of soil below  the  bottom of the lagoons.  Cleanup to
background  (i.e.,  excavation to  groundwater  at  47  feet deep)  was
evaluated, but the small amount  of additional protection provided
would  be  very expensive  ($14 million) .    It was determined that
excavation of the  soil to a depth of approximately 5 feet below the
lagoons to a cleanup standard of 30 parts per million for TNT and
RDX would reduce the excess lifetime cancer  risk to  7 x 10"6 in a
future  industrial land  use  scenario; well within  the acceptable
ranges specified by EPA (i.e., between 1 xlO"* and 1 x 10"6) .  This
represents a volume of soils of  approximately 6,800 tons.

    The remedial alternatives  evaluated in detail were no action,
excavation  followed by  incineration, and  excavation followed by
composting.     Incineration  is   a proven  method  of  destroying
explosives  contaminants  and could reduce  the concentrations and
associated toxicity by 99.99 percent.  Composting is an innovative
bioremediation  technology  for  which  extensive  site-specific
treatability studies have been conducted.  The optimization study
demonstrated  the  MAIV  composting  can  reduce  the  TNT  and  RDX
concentrations  by greater than  97  to  99  percent  and reduce the
toxicity by 90 to  98 percent.  Windrow composting is expected to be
comparably  effective.   Static pile composting was not considered
because of it's inability to meet the established cleanup criteria
in  the  treatability  studies.    The FS   evaluated  the  cost of
implementing MAIV composting,  windrow composting and  incineration
for several excavation scenarios.  Table 3 provides the  estimated
cost of implementing these remedial alternatives at UMDA.
                               317

-------
    The alternatives were evaluated for overall protection of human
 health and the environment;  compliance with applicable or relevant
 and  appropriate  requirements  (ARARs) ;  long-term effectiveness-
 reduction   in   toxicity,   mobility,   and   volume;   short-term
 effectiveness;   implementability;  and   cost.    The  no  action
 alternative failed to provide overall protection to human health
 and the environment and did not comply with ARARs.  The  FS showed
 that windrow  composting would be  expected to meet  the remedial
 action goals at a cost of approximately  $2 million, as compared to
 $4.1 million  for incineration.   The  FS indicated  that windrow
 composting may be the  most cost-effective application of  this
 technology due to the high capital costs of MAIV.  Composting was
 proposed to the public and regulatory  agencies and was selected as
 the final source control remedial action of  the washout lagoons in
 a Record of Decision on September 1992.  A composting windrow field
 demonstration is currently being conducted at UMDA to obtain design
 information for the full-scale remediation contract.  The U S  Army
 Corps  of  Engineers  -   Seattle  District   will   initiate*  the
 implementation of the remedial action in 1993.
 Conclusion
     Composting is  an  emerging,  innovative  technology  for  the
 remediation of explosives-contaminated soils.  Treatability studies
 have  demonstrated  that  composting   is   able  to  reduce   the
 concentration of  explosives  (TNT,  RDX and  HMX)  and  associated
 toxicity to acceptable  health  based  clean-up  levels.   The  UMDA
 Washout f Lagoon FS has  provided a cost analysis which  indicates
 composting  is a  cost  effective alternative  to incineration  for
 sites  with less  than  approximately 20,000  tons of  contaminated
 soxl.  The  implentation  of composting  to remediate  the  explosives
 contaminated washout lagoon soils at the U.S. Army  Umatilla Depot
 Activity will be the first application of bioremediation at an NPL
 site for explosives-contaminated soils.
    Army Technical
DRXTH-TE, Composting  of Explosives,  U.S.  Army Report, DRXTH-TE,
Juy 82 »


AMXTH-TE-CR-86077, Composting Explosive/Organic Contaminated Soils
Atlantic Research Corp., May 1986.
                              318

-------
AMXTH-IR-TE-88242, Field Demonstration-Composting of Explosives-
Contaminated  Sediments  at  the  Louisiana  Army  Ammunition Plant
(LAAP),  September 1988.


Evaluation of Composting Implementation:  A Literature Review, TCN-
89363, July 1990.

CETHA-TS-CR-89276, Proceedings for the  Workshop on Composting of
Explosives Contaminated Soils, September 1989.

CETHA-TE-CR-90027,  Composting  of   Explosive-Contaminated  Soil
Technology, October 1989.

Unnumbered Report, Evaluation  of Composting Implementation, August
1990.

CETHA-TE-CR-91053, Final Report  for the  Composting Optimization
Field Study at Umatilla Army Depot Activity (UMDA) , November 1991.

ORNL/TM-11573, Phase I,  Characterization of Explosives Processing
Waste Decomposition Due to  Composting, .January 1990.

ORNL/TM-12029, Phase II, Characterization of Explosives Processing
Waste Decomposition Due to  Composting,  November  1991.

CETHA-BC-CR-92014, Risk Assessment for  Explosive Washout Lagoons
(Site 4),  Umatilla Depot Activity Hermiston, Oregon, March 1992,

CETHA-BC-CR-92016, Explosives  Washout Lagoons Soils Operable Unit
Supplemental Investigation  Technical and Environmental Management
Support of Installation Restoration Technology Development Program
Umatilla Depot Activity,  Hermiston Oregon, April 1992.

CETHA-BC-CR-92017, Feasibility Study for  the  Explosives Washout
Lagoons  (Site 4)  Soils  Operable Unit  Umatilla  Depot  Activity,
Hermiston, Oregon, April 1992.
           Contacts
Captain Kevin Keehan or Mr.  Wayne Sisk
U.S. Army Toxic and Hazardous Materials Agency
ATTN: CETHA-TS-D
Aberdeen Proving Ground, MD,  21010-5401
(410) 671-2054 or DSN 584-2054
                              319

-------
Table 1; Amendment Mixture Composition and  Cost
Amendment
Sawdust
Apple pomace
Chicken manure
Chopped potato
Horse manure/straw
Buffalo manure
Alfalfa
Horse feed
Cow manure
COST PER TON
Mixture A
30 %
15 %
20 %
35 %





$15
Mixture B




50 %
10 %
32 %
8 %

$200
Mixture C
22 %
6 %

17 %


22 %

33 %
$11
                      320

-------
    Table 2:  Results of Umatilla Composting Optimization Tests
Test
% Soil
static
pile
0
7
10
10
20
30
40
MATV
10
10
25
40
Amend-
ment
Mix

A
A
A
C
A
A
A

A
B
C
C
Initial
cone.
TNT
mg/kg

0
1144
1008
3850
5716
7908
9858

3452
3126
5208
6950
Final
cone.
TNT
mg/kg

0 .
107
200
41
331
174
2086

90
5.6
14 '
209
% Reduction
TNT RDX HMX

N/A
91
96
99
94
98
79

97
99
99
gr

N/A
73
46
93
16
22
0

90
99
97
18

N/A
'39
21
80
5
11
2

29
95
68
0
half-
life
days

0
6.6
6.4
6.9
14.8
16.1
24.9

5.2
'5.1
6.4
14.9
Table 3: Estimated Cost per Ton to Implement Remedial Alternatives
Depth of Excavation
Tons of Soil
2 Feet / 3,700
5 Feet / 6,800
20 Feet / 30,000
47 Feet / 47,000
Windrow
$386/ton
$288/ton
$206/ton
NAa
MATV
$651/ton
$481/ton
$330/ton
NAa
INCINERATION
$740/ton
$660/ton
$280/ton
$300/tonb
Notes: "Composting was not considered a viable alternative ror tnis
large volume of soil due to the requirements for large quantities
amendments. blncludes the cost for special construction techniques
used for vertical sided excavation.
                              321

-------
                                     FIGURE 1:   COMPOSTING PROCESS DESIGNS
U)
ts>
to
                             STATIC PILE COMPOSTING
MECHANICALLY AGITATED IN-VESSEL

         COMPOSTING
                                                  WINDROW COMPOSTING

-------
                  FIGURE 2: COMPOSTING PROCESSES DEMONSTRATED DURING THE UMDA OPTIMIZATION STUDY
OJ
                 Deflector
                    \
On]
        To Blower
                                                Insulation
                                                    Wood Chips
                 AERATED STATIC PILE SCHEMATIC
                                                                        Totally enclosed
                                                                           reactor
                                               MECHANICAL IN^VESSEL  COMPOSTER

                                                         SCHEMATIC

-------
United States/German Bilateral  Agreement on Hazardous Waste
                     Site Clean-up Projects

       Stletzel, H.1,  Sannlng, D.2, Berberich, 6.3, Steffens,  K.3
 1  Federal  Ministry for Research and Technology, Refevat 523
   Heinemannstr 2,  5300 Bonn 2, Germany

 2  U.S. Environmental Protection Agency, RREL, 26 West Martin Luther
   King Drive,  Cincinnati, OH  45268,  USA

 3  Arbeitsgemeinchaft focon-PROBIOTEC   -  SchillingsstraBe 33 -
   D-5160 Duren 5, Germany
                                324

-------
      The problem of contamination to land and groundwater from improper
disposal of hazardous materials/waste are common to all industrialized
countries.  Research and the development of technologies to deal with these
environmental problems is very costly and requires the highest level of
creativity and ingenuity.  All countries resources and technical talents are
limited.  Good researchers do not operate in a vacuum.  Sharing technical
ideas and the expenses of developing new and innovative technologies for
dealing with contaminated land and groundwater appears to be a good idea based
on past experience.  Cooperation at an international level is an economically
cost effective way to access hard-to-get and very worthwhile information.  To
complicate the situation a wide variety of local and national circumstances/
factors lead to the development of varying standards and techniques used to
determine the effectiveness of various technologies to deal with these
environmental problems.  In order to reduce environmental  overload, close
international cooperation is indispensable, both in terms of measures
implemented and in the field of research and development,  including the
exchange of information and experiences.  Furthermore coordination of
environmental demands in different countries lessens the everpresent,
potential trade barriers and imbalances in competitiveness.

      Against this backdrop a United States/German Bilateral Agreement on
hazardous waste site cleanup projects has been developed.   The two lead
agencies are the United States Environmental Protection Agency (USEPA), Office
of Research and Development (ORD), and the Federal Republic of Germany (FRG),
Federal Ministry for Research and Technology (BMFT).  By leveraging existing
resources and programs within the bilateral agreement each agency will enhance
the potential impact of these activities to the overall hazardous waste site
problems within their respective countries.

The goals of this bilateral agreement are:

•     Facilitate understanding of each side's approach to the
      remediation of contaminated sites ("as if-approach":  as
      if the foreign project had taken place in their own country)
•     Demonstrate innovative remedial technologies
•     Compare quality assurance programs
•     Facilitate technology transfer

The technical approach to remediating a hazardous waste site is subject to a
different set of environmental and social factors in each country; therefore,
the usefulness of a particular technology can only be determined if it has
been evaluated in the light of these considerations within both countries.

Six sites on each side were selected for the cooperation.   Detailed executive
summary reports on each of the German sites are being prepared by the
contractors working for the BMFT.  These reports will be reviewed by
designated USEPA/RREL Technical Program Managers to determine what additional
analytical-type Quality Assurance (QA) and Quality Control (QC) measures
should be incorporated into the remedial action demonstration by the Germans
to meet the U.S. EPA Superfund Innovative Technology Evaluation (SITE))
criteria for determining the effectiveness of the individual technologies
being utilized at the sites.

The USEPA/RREL has prepared similar reports for the U.S. sites.  The BMFT will
                                      325

-------
 review these reports to determine what additional analytical QA/QC measures
 should be implemented into these remedial action demonstrations to meet German
 criteria for determining the effectiveness of the individual technoloqies
 being utilized at the EPA sites.

       In the Federal Republic of Germany support for environmental  related
 research and development has a central  function in the Federal  Government's
 environmental policy.  In its environmental  research program the Federal
 Ministry for Research and Technology defined priority areas  for the awarding
 of research grants in fields particularly interfering with the  environment.
 The present promotion concept for the area of waste  management  and  abandoned
 hazardous sites (1989 -  1994) presents  the future research  and development
 priorities in accordance with environmental  priorities.   In  1989 a  special
 program,  the "Model  remediation  of representative hazardous  waste sites"  was
 launched.   The aim of this  program is to  develop further  the state  of  the art
 and to  demonstrate new clean-up  technologies  as  well  as to assess their
 applicability to  actual  hazardous  waste sites.   For  the practical
 demonstration the remediation projects are funded where physical-chemical -
 thermal  and  biological techniques  or  combinations of these techniques  are
 applied.   Special  emphasis  is given to their  efficiency of clean-up, their
 ability to generate  a reusable soil,  their costs  and  their environmental
 impact.   Six of these R&D projects were selected  for  the .EPA/BMFT bilateral
 cooperation  on  hazardous waste site clean-up  projects.

      In order  to  obtain results within a  given time  frame the  schedule of  the
 remediation  projects  affects  the selection.  The  German project  sites, the
 remediation  technologies being applied, and the types of contamination being
remediated are  listed  in table 1.  (The technologies likely to start in 1993
and therefore that will be preferentially  studied in the bilateral agreement
are underlined)
                                      326

-------
 Table 1:  German Remediation Projects
German Sites
GASWERKE MUNCHEN
VARTA SUD,
Hannover
KERTESS, Hannover
HAYNAUERSTRASSE
58, Berlin
BURBACHER HUTTE,
Saabrucken
TNT STADTALLENDORF
Technologies
Soil-Washinq + Volume
Reduction of Residuals
Soil -Washing (+ Acid
Extraction)
Ground Water Pump +
Treat,
Soil Vapor Extraction
In-situ Soil-Washinq
Soil-Washinq.
Microbioloqy (ex situ).
Rotary Kiln Thermal
Treatment.
Fluidized Bed Inciner-
ation of Residuals
Thermal Treatment,
Soil -Washing,
Microbiology
Soil -Washing +
Incineration of
Residuals
Type of Contamination
PAH, Cyanides, Lead
Pb, Sb, As, Cd
CHC + Degr. Products,
CFC, HC, Honoaromatics
CHC, PHC,
Monoaromatics,
PAH, PCB, PCDD, PCDF
Sul fides, Cyanides, Pb,
Hg, Phenols, HC,
Honoaromatics
Munitions, TNT +
Degradation Products,
heavy metals
      A brief description of the activities at these sites is as follows:

Gaswerke Munchen

The remediation of a former coal-gasification and gas-distribution facility in
Munich, Germany, is coordinated by an interdisciplinary working group
comprised of the site-owner, the city and state authorities and various
technical consultants.  The entire site of the former gas-works is about 32.5
hectare (81.25 acres) in size and is located about 1 km northeast of the city
center.  Various site investigations since 1982 showed that both the soil and
the ground water are contaminated with PAHs and the top most layer of soil is
further contaminated by lead.  Slightly elevated concentrations of aliphatic
hydrocarbons and cyanides were found throughout the site, with some higher
peaks in limited areas.  One of the hot-spots of the contamination (the "C 1
area", approx. 1 acre) was found to pose a substantial  hazard to the ground
water and will be remediated in a first phase of the overall  site remediation
which is subsidized by the Bundesminster fur Forschung und Technologie (BMFT).

Soil washing was identified to be the appropriate remediation technology.
25,000 t of gravel  will be treated on-site using a soil washing process that
will be selected out of five optional technologies which are presently in an
off-site pilot scale testing-phase.  The five plants apply batch processes
with or without detergents.   Different systems are used for the separation of
                                      327

-------
 the clean soil  from the contaminant bearing fines.   To supplement the
 treatment process,  innovative treatment technologies like steam desorption  and
 vacuum distillation will  be applied in parallel  with soil-washing for the
 treatment of the soil-washing residues.

 Data on the  contamination and the cleanup  criteria  identified are listed  in
 table 2:

 Table 2

Pb
CN total
PAH
HC aliphatic
Soi 1 -Contami nati on
in mg/kg
maximum average
> 5,000 100 -1,000
> 1,000 > 50
> 7,000 < 500
> 10,000 < 500
Cleanup
soil
mg/kg
< 150
< 50
< 20
< 500
Criteria
eluate (pH7)
M9/1
-
< 50
< 0.2

VARTA-Sud, Hannover

The goal of this project is the remediation of the "Varta-Sud" area which was
contaminated by emissions from an adjacent battery factory between 1938 and
1989.  The remediation of the site owned by the City of Hanover is being
coordinated by a consultant with the support of the "Varta-Sud Assessment
Group" which consists of independent scientists and engineers.  The site is
located northwest of Hanover and covers an area of 4.5 hectare (approx. 11
acres).  From October 1989 to April 1991 the site investigation was performed
showing lead, antimony and cadmium as main contaminants in the soil.  Major
hot spots are deposits of sediments from a creek ("RoBbruchgraben") which was
used for waste water discharge by the battery factory.  Slags from smelting
and coal firing are distributed unhomogenously throughout the area and are of
concern, as well as, landfills of debris, slags and residues from the battery
production.  Besides the heavy metals, a PAH contamination of unknown origin
was detected in the south-eastern part of the site.  The groundwater is not
yet substantially contaminated with heavy metals. Some typical values are
presented in table 3.
                                       328

-------
 Table  3

Pb
Sb
Cd
2 PAH
landfill material
maximum concentration
[rag/kg]
38,000
11,000
1,300
1,000
slag/soil mixture
range of concentrations
[mg/kg]
120 - 5,000
3 - 380
0.1 - 14
0.05 - 10
At present  (November 1992), the results of investigations and trial runs of
different physical treatment technologies (water extraction, high pressure
soil washing, chemical extraction, or combinations) are being evaluated.
Concepts for the volume reduction of residues are designed.  Decisions
concerning  the further proceeding in the project are expected by winter 1992.

KERTESS, Hannover

The project involves the remediation of an industrial  area which was used for
the handling and storage of organic solvents and detergents, as well as
aromatic and halogenated hydrocarbons from 1946 until  1985.  The owner of the
site is the DB German Railroad (Deutsche Bundesbahn).   The area is about 1
hectare (2.5 acres) in size and is located in the southern part of the City of
Hannover.  Site investigations in 1975 detected groundwater contamination and
in 1976 the first remediation attempt (groundwater extraction and treatment)
started.  Due to the lack of efficiency of these measures the project
management was reorganized and a new remediation concept was designed.

Major contaminants are chlorinated and aromatic solvents found in the vadose
zone and on the bottom of the underlying aquifer (DNAPL).  Table 4 lists
average concentrations of selected contaminants as well  as cleanup criteria
identified by the State Secretary of Environment:
                                       329

-------
 Table 4

trichloroethene
perch! oroethene
chloroform
vinyl chloride
benzene / toulene
2 CHC / 2 VOC*
2 BTX
Cleanup Criteria
2 CHC
ground water [mg/m3]
4,160
6,050
65
410
110 / 310
18,300 / -
1,220
< 10 - 20 mg/m3
soil gas [mg/m3]
4,185
2,150
80
120
230 / 270
- / 14,935
200
< 1 mg/m3
*   chlorinated
Berlin:   Haynauer  Str.58

The  site  "Haynauer StraBe 58"  is situated  in the  south of Berlin,  in  the
district  of  Steglitz  and has an overall size of 3,100 m    The site was used
for  the reconditioning of solvents, chemical wastes and waste oils.   The
contamination  of the  site happened between  1952 -  1960.  In  1980,  the company
was  closed down.   Major contaminants are petroleum hydrocarbons  (PHC),  BTEX,
volatile  chlorinated  hydrocarbons (CHC), polychlorinated biphenyls  (PCB),
PCDD/F were  found  in  the debris of demolition works.

Clean-up  standards  for soil are fixed by the "Berlin List":
•     petroleum hydrocarbons (PHC)
      BTEX
•     total high-volatile chlorinated
      hydrocarbons  (CHC)
•     total PCBs
                                                150 mg/kg
                                                2.5 mg/kg
                                                2.5 mg/kg
                                                0.5 mg/kg.

For ground water the following clean-up standards are fixed:
      CHC
•     BTEX
.     PHC
*     PCBs
                                                25
                                                10
                                                250   /ig/1
                                                0.25  iig/1.
The remediation concept for this site is a combination of different individual
                                        330

-------
technologies:
•     PCDD/F-contaminated  debris are treated thermally  in  a rotary  kiln
•     full-scale,  off-site soil washing plant will  be used for the  main
      portion of the  contaminated soil (40,000 t)
•     residuals will  be  incinerated in a fluidized  bed  process or treated
      microbiologically
•     slightly contaminated  soil will be treated microbiologically
•     full scale soil  vapor  system is used for the  decontamination  of  soil
      vapor
•     ground water is  treated.

The soil of the site will  be excavated emission-free in three phases.
Remediation phase  1 was  started in fall 1991 and finished  in August  1992.   In
this step an overall contained third of the site (approx.  700 m ) was
excavated to a depth of  7  m  below ground level.  An amount of 7,100  m3 soil
was treated by soil washing, 600 t microbiologically.
reembedded on site.
The cleaned soil was
Soil vapor extraction has been performed since 1990.  In the first half of
1993 the thermal treatment process will be started for the residuals and
debris as well as the ground water treatment.

In fall 1992 the 2nd remediation phase was prepared.  About 1,600 m2 of the
site was covered in order to avoid emissions during excavation.  Work was
finished in October 1992.  Soil excavation was started in November 1992.
Again the contaminated soil is treated by soil washing and microbiologically.

Burbacher Hutte

The "Burbacher Hutte", founded in 1857 is located in the western part of
Saarbriicken.  In times of peak productivity the Burbacher Hutte reached its
maximum size with approx. 60 ha (600,000 m).

Contamination of the soil started when the steelworks was founded in 1857 and
continued until  the end of the 1980's when all production plants were
demolished. In the past, there were different individual  production plants,
e.g. coke factories, benzene works, gas generation plant etc. which caused
specific contamination at the Burbacher Hutte site.  The following main
contaminants were found in different parts of the site:

      BTEX
      phenols
      PAHs
      heavy metals (e.g. lead and mercury)
      sulfides and cyanides.

In addition, in some areas of the site ammonium has been  detected at elevated
concentrations.

In 1991 -  1992 full-scale pilot-tests were performed with different
technologies on-site and off-site:

•      soil  extraction
                                      331

-------
•     bioremediation
•     soil washing and
•     high temperature incineration.

At present the results of these pilot-tests are being evaluated.  The
remediation concept will be a combination of these different technologies for
the various, individual groups of contaminants.

A decision on a suitable combination of technologies is expected to be made in
late fall of 1993.  Start of the actual remediation will be in 1994 after
passing the permitting process.

Stadtallendorf

The former ammunition and explosives factory Stadtallendorf is located approx.
100 km north of the city of Frankfurt.  The production area of the former DAG-
site had a size of ca. 4.2 km  with  413 buildings  having an overall  base of
ca. 62,000 m .   In March 1941,  the TNT production  was started.   Within the
following 3 years the production raised from zero to 5,400 tons per month.
The overall production and processing amounted to ea. 126,000 tons of TNT and
27,000 tons of other explosives.  Production stopped on March 28, 1945.  After
World War II the U.S. Army used the  site as storage for ammunition,  gear and
tools.  Between 1946-1949 almost all production plants were demolished or
blasted.  Today, the site of Stadtallendorf is mainly characterized by mixed
use including residential areas, industrial sites and the former ammunition
and explosives factory.  About 3,000 inhabitants of Stadtallendorf are living
and working in the former bunkers or in the converted production buildings or
related facilities remaining on the  site.

Between 1938 and 1945 the site of the ammunition factory was contaminated
with:
•     explosives, e.g. TNT, H.N.D., RDX;
«     raw materials and pre-fabricates of the explosives production, e.g.,
      toluene, mono-and dinitrotoluenes;
•     degradation products, e.g. (di-)nitro-aminotoluenes;
•     miscellaneous pollutants like phenols,  heavy metals and others.

Contamination of soil and groundwater at Stadtallendorf was caused by  unsafe
production methods, uncontrolled waste disposal, insufficient waste water
treatment, accidents, leakages and others.

The envisaged remedial technology for contaminated soil is a combination of
soil washing with a high-temperature combustion of the contaminant-bearing
residues.  For soil decontamination the clean-up targets are defined as
1 mg/kg for explosive-related contaminants.  Soil washing and thermal treatment
is to be started in 1995.

At present additional investigation of a 6 ha hot spot area are being
conducted.  In a trianguarily shaped grid additional  borings are being carried
out at the former production plants (TNT Groups 1-3 ruling plants).  First
results indicate that the main contamination  with varying concentrations was
found at a depth of 0-lm below ground level.   The detailed results of this
investigation phase will be the foundation of the future remedial concept.
                                       332

-------
 Furthermore, a new analytical method for the analysis of TNT  by means  of HPLC
 and electro-chemical detection was performed by a scientific  expert  team
 because there was no analytical standard procedure for TNT  in Germany   This
 analytical method will be published in January 1993.

       The United States sites are summarized with their main  characteristics
 in table 5.

 Table 5.  U.S.  Remediation Projects
U.S. Sites
Waukegan Harbor Site
(Outboard Marine
Corp.), IL
Kelly Air Force Base
San Antonio, TX
Electro-Voice NPL
Site Buchanan, MI
Lawrence Livermore,
Site 300, GSA
Altmont Hill, CA
Indiana Wood
Preservers Site
Bloomington, IN
Pennsylvania Power &
Light (PP&L) Brodhead
Creek Site
Stroudsburg, PA
Technologies
Anaerobic Thermal
Treatment on Site
Radio Frequency
Induced Heating
Subsurface
Volatilization and
Ventilation System
(SVVS)
Advanced Oxidation,
Perox-Pure
Bioremediation
Contained Recovery of
Oily Waste (CROW)
Types of
Contamination
PCB, Oil & Grease
VOC'S, SVOC'S, TPH
BTEX, PCE, TCE, 1,
1-DCE
VOCs, TCE, PCE
Creosote/Coal Tar,
PAHs Naphthalene,
Anthracene
Organic Liquids, Coal
Tars, PAH, Phenols
      A brief description of these activities at these sites  is  as  follows:

Outboard Marine Corporation (OMC), Waukegan, Illinois

Outboard Marine Corporation (OMC) used PCBs as a hydraulic fluid at the
Waukegan Harbor facility.  As a result of leaks and spillage major parts of
Waukegan Harbor's 37 acres and parts of the manufacturing facility were
contaminated with PCBs up to 10,000 ppm (mg/kg).  The OMC site remediation
included dredging and excavating the PCB contaminated sediments and soils;
treating the high PCB concentration soils and sediments; and placing low
concentration PCB soils and sediments in one of several isolation cells.

Thermal desorption was the selected treatment technology for the high
concentration PCB soils and sediments.   SoilTech's Anaerobic Thermal Processor
(ATP) treated these solids after natural  dewatering occurred.  The ATP system
mixes and heats the solids in  the processing unit,  which consists of four
                                      333

-------
zones;  preheat,  retort, combustion and cooling.  Temperatures in the preheat
zone  (400  -  650  ) cause the water and volatile organic compounds (VOCs) in the
solids  to  vaporize.  The higher temperatures  in the retort zone (900 -
1,150 F) cause (1) heavy organics to vaporize and (2)  thermal cracking of
hydrocarbons creating coke and low molecular  weight gases.   In the combustion
zone the coked solids are combusted and either recycled to the retort zone or
passed  into  the  cooling zone.  Gases from the preheat and retort zones are
cooled  and condensed to recover the water and organics; while the
noncondensable organics are burned in the combustion zone.

In June 1992 a USEPA SITE Demonstration was performed on the Soil Tech ATP
system  operating at the OMC Waukegan facility.  Three replicate test runs were
conducted  at the typical operating conditions used during the OMC site
remediation.  Each test run consisted of 8.5  hours of solids and liquids
sampling and 8 hours of stack sampling.  A total of 224 tons of PCB
contaminated soil and sediment was treated during the Demonstration.
Extensive  process operating data were collected to document the ATP operating
conditions.   Key findings from the OMC/SoilTech SITE Demonstration included:

1)    PCB  concentrations were reduced from an average of 9,761 ppm in the
      untreated  solids to an average of 2 ppm in the treated solids.

2)    Approximately 0.12 milligrams of PCBs were discharged from the ATP stack
      per  kilogram of PCBs fed to the ATP (0.12 mg/kg or 99.9999% ORE).

3)    The  majority of PCBs removed from the solids during treatment were
      accumulated in the organic storage from the gas condensation.

4}    No dioxins, other than a low concentration (0.1 nanograms (ng) per dry
      standard cubic meter (dscm) of octachlorinated dioxin, were detected in
      the  stack  gas from the ATP system.   Tetrachlorinated furans were found
      in both the untreated (88 ng/g) and treated (5 ng/g) solids and the
      stack  gas  (0.07 ng/dscm).

5)    teachable  VOCs, semi volatile organic compounds (SVOCs) and metals in the
      treated solids were below RCRA toxicity characteristic standards.

6)    No operational problems affecting the ATP's ability to treat the
      contaminated solids were observed.

KELLY AIR  FORCE  BASE - San Antonio, Texas

      Site S-l,  located near the northern boundary of Kelly Ait Force Base was
used from  1960 to 1973 as an intermediate storage area for wastes awaiting
off-base reclamation.  Waste liquids consisting of mixed solvents,  carbon
cleaning compounds and petroleum oils and lubricants were hadled here.   Spills
and storage  tank flooding resulted in soil contamination from these chemicals.
At the  conclusion of waste handling operations in 1973,  the area was
backfilled and regraded.  This resulted in soil  contamination at depths of up
to 27 to 30  feet.  The vadose zone extends to this depth.

      Volatile and Semi-volatile organic contamination at the site occurs at
concentrations ranging from 8.5 to 104,000 ppb.   Total  Petroleum Hydrocarbon
(TPH) concentrations range from 20 to 17,000 ppm.   Specific contaminants found
                                        334

-------
 at  the  site  include a variety  of  short  chain  chlorinated hydrocarbons  such as
 1,2-Dichlorobenzene,  1,4-Dichlorobenzene  and  1,3-Dichlorobenzene,  aromatic
 hydrocarbons  such  as  benzene,  ethyl benzene, chlorobenzene and  styrene  and
 polycyclic aromatic hydrocarbons  such as  Dibenzo  Anthracene, Benzo Pyrene  and
 Benzo Fluoranthene.   Some  very low  concentrations of PCBs have also been
 detected.

      In-Situ Radio-Frequency  Heating will be  studied  at site  S-l  to determine
 its  efficacy  in  removing volatile and semi-volatile  organic contamination  from
 vadose  zone soils.  This technology was developed by the Illinois  Institute  of
 Technology Research  Institute  (IITRI) and will be demonstrated for the Air
 Force with the assistance  of Haliburton NUS.   The system consists  of a 120kW
 RFH  power source,  excitor  and  ground electrodes,  a vapor barrier and a vapor
 treatment system.

      For the test, the excitor and ground electrodes  will be  placed in the
 ground  at S-l  in a  predetermined array.  Thermocouples will be placed  in the
 ground  with the  electrodes to  monitor ground temperature.  Heat generated  by
 radio-frequency  radiation  will heat the soil to approximately  150 C.  The
 elevated temperature  combined  with steam distillation  resulting from the
 vaporization  of  soil  moisture  will sweep organic  contamination  from soil.  A
 vacuum  will be maintained  on the soil area under  treatment in  order to capture
 released contamination.  A vapor barrier installed atop  the soil treatment
 area will prevent the release  of contaminant laden gasses to the atmosphere.
 These gasses  will be.captured  by the vapor barrier and treated  by  both
 condensation  and carbon filtration.

      Development of  the demonstration plan is currently  under  way.  Electrode
 placement and  pre-test soil sampling should occur  in early 1993. System set  up
 and treatment  of 500  cubic yards of contaminated  soil  from S-l  will  take place
 in mid-1993.    Post test soil  sampling will be completed by the  end  of July
 1993.   Data analysis  and reporting will  be completed as soon as possible
 thereafter.

 SVVS Demonstration, Buchanan,  Michigan

 This SITE Program Demonstration will assess the effectiveness of the Soil
 Volatilization and Ventilation System™ (SVVSK) developed by Billings & Assoc.,
 Inc. of Albuquerque,  NM and offered  under license through Halliburton NUS.
 The technology is a combined  air-sparging/soil-vacuum-extraction technique
which moves air into  the water table through injection wells and draws air
 from the vadose zone  through  extraction  wells.  At some installations,  one
 injection and multiple vacuum wells  are  placed in the same soil boring,
 resulting in  a "well  nest."  In this installation, injection and vacuum wells
 are placed in  individual  borings.   The movement of air through the subsurface
 stimulates the activity of aerobic microorganisms resulting in a quicker
 remediation of organic contamination than by soil  vacuum extraction  (SVE)
 alone.   Vapors collected in the SVE  portion of the system can be treated in
Biological  Emissions Control™  (BEC™) units reducing or removing the need for
carbon  adsorption or other expensive emissions controls.

      The Demonstration site  chosen  is the Electro-Voice site,  a Superfund
 (National  Priority List [NPL])  site  in Buchanan,  Michigan.  This site in
southwestern  Michigan is  an operating factory  producing electrical  sound
                                      335

-------
 amplification  equipment.   Formerly,  liquid  painting  residues  were  discharged
 to  a  "drywell,"  a  lined  pit  backfilled with  gravel,  next  to the  factory
 facility.  This  practice was discontinued and  the  drywell  filled with  soil  and
 construction debris  in  1973.

      Sampling conducted during  the  NPL process  revealed  a sludge-like layer
 at  the bottom  of the former  drywell.  Initially, peak  soil concentrations were
 observed for alky!benzenes  (454,000  ppb), benzene  (350 ppb),  2-butanone  (4,900
 ppb), 1,1-dichloroethane (200 ppb),  l,2(cis)-dichloroethene (200 ppb)
 ethyl benzene (95,000 ppb), styrene (3,400 ppb),  tetrachloroethene  (14,000
 ppb), toluene  (330,000 ppb)  trichloroethene  (590 ppb),  and total xylenes
 (710,000 ppb).   During sampling  prior to the Demonstration, peak soil
 concentrations were  sought for toluene (96,000 ppb), ethyl benzene  (64,000
 ppb), m,p-xylene (190,000 ppb),  o-xylene (29,000 ppb),  benzene (210 ppb), 1,1-
 dichloroethene (130  ppb), trichloroethene (18,000  ppb), and tetrachloroethene
 (16,000 ppb).

      The SVVS™  technology is to be  evaluated  for  its  ability to meet a
 claimed 30% reduction in  seven combined contaminants (benzene, toluene,
 ethylbenzene,  xylene [BTEX],  tetrachloroethene [PCE],  trichloroethene [TCE],
 and 1,1-dichloroethene [DCE])  in the vadose zone within 12 months  of
 operations.  Further, data collected during the SITE Demonstration will
 investigate the  relative  contribution of the two primary mechanisms of
 removal:  bioremediation  and air-stripping.  Finally,  the  effectiveness of the
 BECW  unit at meeting State of Michigan air emissions standards will be
 evaluated.  Vinyl chloride levels in the soil will also be monitored.  In
 order to assess  the  claimed  30% reduction, twenty  soil borings will be located
within the 113 m area before the start  of operation, and twenty more soil
 borings will be  obtained  following 12 months of operation.  From each soil
 boring, seven  soil  samples will be obtained for Volatile Organic Contaminant
 (VOC) analysis with  special  quality  assurance/quality  control  measures
performed for  the seven  "critical analytes" and vinyl  chloride.  Analyses of
the gasses entering  and leaving the  BEC™ unit will be  conducted  periodically
throughout the 12 months.  Additionally,  soil gas measurements,  groundwater
 analyses,  soil  VOC levels from the saturated zone, and in-situ biological
respiration data will be obtained.

Lawrence Livermore National  Laboratory;  Site 300 Demonstration,
Altamont Hill,  California

      The perox-pure™ technology is designed to destroy dissolved organic
contaminants in groundwater  or wastewater through an advanced chemical
oxidation process using ultraviolet  (UV)  radiation and hydrogen  peroxide.
Hydrogen peroxide is added to  the contaminated water,  and the mixture is then
fed into the treatment system.  The treatment system contains  four or more
chambers in the reactor.  Each chamber contains one high-intensity UV lamp
mounted in a quartz  sleeve.  The contaminated water flows  in the space between
the chamber wall  and the quartz tube in  which each UV  lamp is  mounted.

      UV light catalyzes chemical oxidation or organic contaminants in water
by its combined effect upon  the organics  and reaction with hydrogen peroxide.
First, many organic  contaminants that absorb UV light may undergo a change in
their chemical  structure or may become more reactive with chemical  oxidants.
Second,  and more importantly,  UV light catalyzes the breakdown of hydrogen
                                      336

-------
 peroxide  to  produce  hydroxyl  radicals, which are chemical  oxidants.

 Site  Description:

       Site 300  is a  high-explosive test area, established  by  LLNL  in  1955.   It
 is  comprised of 11 square miles and is about 60 miles east of San  Francisco,
 and 15 miles east of Livermore, California.  Operations at Site 300 include
 hydrodynamic testing; charged particle beam research; physical, environmental,
 and dynamic  testing;  and high-explosive formulation and fabrication.  The
 demonstration took place in a section of Site 300 called General Services Area
 (GSA).  This area is  contaminated with VOCs, including trichloroethane  (TCE),
 tetrachlorethane (PCE), 1,1-dichloroethane (1»1DCE), 1,2-dichloroethane  (1,2-
 DCE),  and 1,1,1-trichloroethane (1,1,1-TCA).

 Additional monitoring measures included the following:

 •      VOC Purge and Trap GC-MS (method 8240), additionally to the  target
       compounds, TIC  (10 largest peaks) are identified applying NBS-search

 •      SVOC Method 8270  (GC-MSO, additionally to the target compounds, TIC (10
       argent peak?) are identified applying NBS-search.

 «      AOX/TOX During each run of the plant at the fed-line and at  the
       effluent-line on sample each is taken for TOX (method 9020).  "AOX"
       including the purging step was performed with 20% for the samples
       additionally.  Thus, a basis for a comparison between the two methods
       will be available, the risk of errors due to poor data  quality  is
       minimized.  A summary of the results will be supplied.


 Indiana Wood Preservers (IWP) Site, Bloomington, Indiana

       The IWP site was operated from 1976 to 1987.   A solution of
 creosote/coal tar was used as an agent for preservation of railroad ties.  IWP
 utilized the Boulton process which involves the application of an  initial
 vacuum, the pressure impregnation of the creosote/coal  tar solution into the
 railroad ties.   There is a total  of 35,000 gallons  of liquid  and 355 cubic
yards of sludge/contaminated soil  existing in two holding ponds, as well as
890 cubic yards of contaminated debris in a waste pile on site.

      Bioremediation was chosen for the treatment of the contaminated soils
removed from the holding ponds.   Because of the limited amount of  space
 available at the site, it was decided that the  biodegradation would take place
 in  constructed compost piles.   This treatment process would also add
additional organic sources to allow co-metabolism to assist in the degradation
as well as the direct mineralization of the polynuclear aromatic hydrocarbons
 (PAHs) present in the contaminated soil.   Preliminary studies were conducted
degraded.
      The compost piles are made of a mixture of 17 parts of soil, 4 parts of
horse manure, and one part of straw.   Each compost pile is 20 meters long, 10
                                       337

-------
 meters wide, and 2.5 meters high.   As the pile was being built, hollow plastic
 piping was placed from the outside of the pile leading into the heart of the
 pile.  This piping was added all  around the pile and at different heights in
 order to allow air into the pile  to supply oxygen for the microbial
 degradation of the PAHs.   Laboratory studies had shown an average oxygen
 consumption rate of 64.8 mg 02/kg  hour.

       The compost piles have been  built and are in operation.   The analytical
 test results have not yet been received.   Consideration is being giverv to a
 series of treatments consisting of biological  treatment,  followed by chemical
 treatment with Fenton's reagent (hydrogen peroxide and ferrous iron salt,
 which is a powerful  radical  oxidant),  followed by additional  biodegradation to
 attempt to bring the final  PAH concentration of the treated soil  to a level
 significantly below that  which could be achieved by biodegradation alone.

 Pennsylvania Power and Light (PP&L),  Brodhead  Creek Site,
 Stroudsburg, Pennsylvania

       The Brodhead Creek  Superfund site in  Stroudsburg,  Pennsylvania has  been
 selected for the demonstration of  the  contained recovery  of oil  wastes (CROW)
 process.  The primary contaminant  at the  site  is coal  tar produced as a
 byproduct of a coal  gasification plant.   Identifiable  contaminants include
 benzene, polynuclear aromatic  hydrocarbons  (PAHs),  and phenols.   A large
 amount of coal  tar has  migrated through the surficial  aquifer  and is now
 pooled in a depression  approximately 6 meters  wide  and 18 meters  long.  The
 aerial  extent of the pooled  coal tar will be better defined during a predesign
 sampling effort  tentatively  scheduled  for early December  1992.

       The CROW process  was selected  in a  record of  decision (ROD)  as an
 interim remedy to  reduce  the amount  of free coal  tar pooled at the base of the
 aquifer.  The CROW  process will use a system of injection  and recovery wells to
 remove the coal  tar.  Water  heated to a temperature  of approximately 90
 degrees Celsius  will  be injected at  the base of the  aquifer through  sic to
 eight  injection  wells on  all sides of the pooled  coal  tar.  Dual  completion
 recovery wells  located  in the  center of the coal  tar pool  will be  used to
 recover ground water  and  coal  tar.  The lower  completion  recovery  well is
 designed to  recover  the injected water and  coal  tar; the upper completion
 recovery well  is intended to maintain a copy of cooler ground water  over  the
 treatment area.  The maintenance of a cool water  cap over  the treatment area
 is intended  to prevent contamination from migrating  into the unsaturated  zone.
 The fluids produced from the recovery wells will  be  sent to an oil-water
 separator for treatment.  Most of the outflow water  from the oil-water
 separator will be reinjected by the CROW process, while the remainder of  the
water will be treated in a bioreactor and discharged to Brodhead Creek.   If
 the water does not meet the requirements of a National Pollutant Discharge
 Elimination  System (NPDES) permit,  treatment in the bioreactor will  be
 followed  by  carbon adsorption.  The U.S. Environmental Protection Agency  (EPA)
Region  3  has set the cleanup criteria at 60 percent removal of the free coal
tar measurable in monitoring wells.  Remediation Technologies, Inc.  (ReTech),
the prime contractor supervising the remedial effort for the potentially
responsible  parties, has indicated that the CROW process will be operated
until the amount of recovered contamination has stabilized.
                                          338

-------
       The Superfund  Innovative  Technology Evaluation  (SITE)  program will
 monitor all  phases of CROW process  installation,  operation,  and post-operation
 sampling.  The  steps  that  will  be undertaken  to design  and  build the CROW
 process include predesign  sampling,  process design, construction,  and
 operation.   If  the current schedule  is  followed,  the  predesign  sampling will
 occur in December 1992,  the design will  be completed  by May  1993,  construction
 will  be completed by  the end of July 1993, and the  system will  be  operated
 from  August  1993 through October and possibly into  December  1993.   During the
 predesign sampling, the  SITE program will  collect samples of the site soil  and
 pure  coal  tar.   The samples will be  used  to refine  and  validate the analytical
 methods.   The SITE program demonstration  will  include soil and  ground-water
 sampling and monitoring  of operational  parameters.  Soil samples will  be
 collected before and  after the  CROW  process operation to determine the
 reduction of subsurface  contaminants.   The ground water, injected  water,  and
 recovered water will  be  sampled  to determine  the  amount of contamination
 recovered.  During CROW  process  operation, the SITE program  will monitor
 pumping  rates'  injected, recovered,  and ground-water  temperatures;  and water
 levels.   The data collected will be  used  to evaluate  the effectiveness of the
 CROW  process at the Brodhead  Creek Superfund  site and to determine the factors
 that  will  affect the  CROW  process at other sites.
Acknowledgement:  The authors would like to express their appreciation to the
following Project Managers from the Risk Reduction Engineering Laboratory for
their technical contributions:

            Kim Kreiton
            Paul dePercin
            Ron Lewis
            Norma Lewis
            Laurel Staley
                                          339

-------
       Groundwater Remediation: Extraction and Removal of TCE & Cr6* at an NPL Site
                                  Subijoy Dutta, P.E.,
   '  U.S. EPA, OSW/CAPB/PSPD/ATS, OS-341;  401  M St. SW, Washington D.C. 20460.
                        Phone: 202-260-1371; FAX #202-260-0096

      This case study focuses on  the engineering analysis of a treatment process used in a
National Priority Listed (NPL) site in  a midwestern State within the continental United States. The
primary contaminants intended to remove from the groundwater at this site are Trichloroethylene
(TCE) and hexavalent chromium (Cr6+)

      The groundwater treatment plant, for this site belonging to an aircraft repair and paint
company, is designed for a flow rate of 150,000 gal./day (567 m3/day). The cleanup period is
estimated to be 30 years. The treated water is planned for industrial reuse at the same facility.

      The waste characteristics of this NPL site are very common to many aerospace industries
and other industrial sites. The cleanup method used at this site for TCE and Cr6+ removal could
be used in similar other sites since it renders a high (99.99%) contaminant removal efficiency and
offers an environmentally clean method with very minimal sludge/waste generation.

      Based upon the result of a study, involving a complete engineering analysis of several
different processes that are available for removal of organics, mainly trichloroethylene-(TCE), and
metals, primarily hexavalent chromium (Cr6*), the following treatment train was selected for this
NPL site.
            + Organic Removal Process
            * Metals Reduction Process
            + Metals Precipitation process
            * Final polishing process
below:
      The treatment processes for the groundwater treatment plant (GWTP)  are summarized
      1 .     The Organic Removal process consists of the Aquadetox system, developed by Dow
Chemical and patented for removal of high boiling organic compounds. This process has been
found to be effective in removal of most of the organic compounds, which are listed as hazardous
by the U.S. Environmental Protection  Agency (EPA). The stripping technology, whether an air
stripper or  a steam stripper, can provide 99.9999% removal efficiency. The effluent from this
process is  expected to have non-detectable  concentration (less than 1  ppb) of the  organic
contaminants. The Aquadetox process is also accepted under the EPA's Superfund Innovative
Technology Evaluation (SITE) program. This technology, working at over 10 locations, eliminates
any carbon polishing of the effluent water. The conventional air stripping provides only about 90-
95% removal of volatile organic compounds.
                                       340

-------
      2.     The Metals Removal Process consists of the following:

   "   The incoming groundwater contaminated with hexavalent Chromium (Cr6+) is treated by
a reduction process first. This reduces the Cr6* to the trivalent form,  Cr3+. The metal is then
removed  as precipitate from the groundwater.

      In the first stage of treatment, the pH is lowered to 2.5 by the addition of 98% Sulfuric acid
(H2SO4); a pH probe in the tank controls this addition. The stream is made acidic to facilitate the
reduction reaction. The reducing agent  to  be used in the process  is  sodium metabisulfite
                                                                                   .3 +
         . This chemical reduces the hexavalent chromium, Cr  , to trivalent chromium, Cr

      The automatic addition of sodium metabisulfite is controlled by an Oxidation Reduction
Potential (ORP) measuring instrument. Proper reduction of chromium can be seen readily by a
color change from yellow to blue in the treatment tank.

      After being reduced, the stream from the first stage flows to the second stage tank for
primary precipitation treatment. The pH is raised to 9.0 to precipitate the trivalent chromium as a
chromium hydroxide. This is  accomplished by adding a 50% sodium hydroxide solution;  the
amount of addition is controlled by a pH probe. Chromium is an amphoteric metal which means
that it is highly soluble  at both a low pH as well as at  a high pH.  For  this reason the pH is
maintained at the optimum value of 9.0. Practically, this value can not be precisely controlled due
to the addition of the sodium hydroxide in large quantity. If the pH differs greatly from this value,
excess trivalent chromium remains in the solution. At a pH of 9.0 the trivalent chromium and other
heavy metals precipitates as metal hydroxides.  The sludge generated by this process is expected
to be 75% less than that generated by the Ferrous Sulfate process, which is commonly used in
old metal processing/plating industries,

      A secondary precipitation of these metals occurs in the pH adjustment tank, where the pH
is maintained at 9.0 (for further precipitation) by adding sodium hydroxide  or sulfuric acid as
required. Thus if the pH goes to a value of 10.0, it is lowered to 9.0 by adding sulfuric acid.
Likewise, if the pH falls to 8.0, sodium hydroxide is added to raise the pH to 9.0.  This addition
is controlled by a pH probe located within the tank. The precipitated metal hydroxides  exists as
light density particles.

      The treated water is then transferred to the flash mix tank of the clarifier, using a centrifugal
pump.  The pumping is controlled automatically by a level monitoring probe which cycles  the
pump on or off. To promote flocculation, a polyelectrolyte is added to the stream.

      The effluent from the precipitation process has a chromium concentration of about 20-50
parts per billion (ppb).  The effluent then passes through the final polishing process to further
reduce the concentration of chromium and other heavy metals to 10-15 ppb.

       3.    The Fine Filtration Process involves a Polymer Enhanced Cross Flow Sand Filtration.
                                         341

-------
This sand filtration system.Dynasand™ is a patented process. It has some unique features in sand
filtration. It is a continuous, backwash, upflow, deep and granular media filter. The filter media is
continuously cleaned by recycling the sand internally through an airlift pipe and sand washer. The
regenerated sand is redistributed on top of the sand bed allowing for a continuous uninterrupted
flow of filtrate and reject water.

      Feed is introduced into the bottom of the filter, then flows upward through a series of riser
tubes and is evenly distributed into the sand bed through the open bottom of an inlet distribution
hood.  The influent flows upward through the downward moving  sand bed with solids being
removed. The clean filtrate exits from the sand bed, overflows a weir, and is discharged from the
filter.  Simultaneously, the sand bed, along with the accumulated solids, is drawn downward into
the suction of an airlift pipe which is positioned in the center of the filter.  A small volume of
compressed  air is introduced into  the bottom of the  airlift.  The sand,  dirt, and water are
transported upward through the pipe at a rate of about 200 gpm/ft2.  The impurities are scoured
louse from the sand during this turbulent upward flow.  Upon reaching the top of the airlift, the
dirty slurry spills over into the central reject compartment. The sand is returned to the sand bed
through the gravity washer/separator, which allows the fast settling sand to penetrate, but not the
dirty liquid. The washer/separator is placed concentrically around the upper part of the airlift and
consists of several stages to prevent any short circuiting.  By setting the filtrate weir above the
reject weir a steady stream of clean filtrate flows upward, countercurrent to the sand, through this
washer  section and acts as a liquid barrier that carries away the dirt and reject water. Since the
sand has a higher settling velocity than the dirt particles, it is not carried out of the filter. The sand
is redistributed by means of a sand distribution cone. The sand bed is continuously cleaned while
both a continuous filtrate and a continuous reject stream are produced.

      For an  effluent of 10 ppb to 20 ppb at a flow rate of 65 to 70 gpm, a Filter with a Polyblend
emulsion polymer feed system is recommended for the polishing process. This proven process
has been successfully used at an AT&T plant in Mesquite, Texas and in a Groundwater treatment
plant of Stauffer Chemical Co. in Martinas, California. The effluent chromium level of 10 ppb was
attained by using just the sand filters at that treatment facility.   A simplified flow chart of the
proposed groundwater treatment process is shown on Rgure 1.
                                        342

-------
                                       A








Carbor
Beds







^


Clear
Emiss
*
. I
T Ste^m Cond.
1 1
                                     Stripping Tower
            u
                                                                            Solvent
                                                                            Storage
                             Metals
                            Removal
                             Process
Sand
Filtration
Unit
                                                  Filter Press
 Clean
Affluent
 to
 Industrial
 Reuse
FIGURE l  Flow chart of the proposed groundwater treatment plant

-------
      The annual operating cost of this system is very low compared to the reverse osmosis or
ion exchange systems, since the Dynasand™ system gets continuously regenerated.  The only
operational cost involved in this process is the electric utility cost and the polymer cost. Labor cost
is minimal in this operation.

      The automated control mechanism routes the water through the optional polishing process
if, and only if, the concentration of chromium in the effluent is over 50 ppb, or any other desired
value. The polishing process involves fine filtration. The whole process is completely automated
which cuts down a significant  cost of this long-term cleanup operation.

REFERENCES

1.     Aldrich, James R., "Effects  of pH and proportioning of Ferrous and Sulfide Reduction
      Chemicals on Electroplating Waste Treatment Sludge Production", Proceedings of the 39th
      Annual      Purdue Industrial Waste Conference, West Lafayette, IN, May 1984.

2.     Christensen, Erik, R., and Delwiche, John, T. "Removal of Heavy Metals From Electroplating
      Rinsewater by Precipitation, Flocculation, and Ultrafiltration", Water Research, Vol. 16 (1982)
      p. 730-744.

3.     Collie,  M.J. "Industrial Water Treatment Chemicals and Processes Developments since
      1978",  NDC Publications, NJ 1983.  Chemical Technology Review No.  217,  Pollution
      Technology Review No. 98,  1983.

4.     Cushnie, George C., Jr.  "Electroplating Wastewater Pollution Control Technology", Pollution
      Control Review No. 115, Noyes Publications, 1985.

5.     DeFilippi, Richard P., "Ultrafiltration", a part of "Filtration, Principles and Practices, Part I",
      Mercel Dekker Inc., New York and Basel, 1980, p. 475.

6.     Eilbeck, WJ. and Mattock,  G.  "Chemical  Processes in  Waste Water Treatment" Ellis
      Norwood  Limited, England, 1987.

7.     Faust,  Samuel  D., and Aly,  Osman, M. "Chemistry of Water Treatment",  Butterworth
      Publishers, 1983.

8.     Fisco, R.,  "Plating and Industrial Waste Treatment at the Expanded Plant of Fisher Body,
      Elyria,  OH", Proceedings of the  25th  Purdue  Industrial Waste Conference, Purdue
      University,
      May 1970.

9.     Fresenius, W. and Schneider, W. "Waste Water Technology",  Springer-Verlag Berlin
                                       344

-------
      Heidelberg, 1989.

10."   Germain, J. , Vath, C., and Griffin, C., "Solving Complex Waste Disposal Problems in the
      Metal Finishing Industry", Presented at the Georgia Water Pollution Control Association
      Meeting, Sept.  1968.

11.    Glynn, W.,  Baker, C., LoRe, A., and Quaglieri, A. "Mobile Waste Processing Systems and
      Treatment  Technologies", Pollution Technology Review No. 147, 1987.

12.    Gould,  J.  "The Kinetics  of Hexavalent Chromium Reduction by Metallic  Iron", Water
      Research,  Vol.  16 (1982), p. 871-877.

13.    Henold, Kenneth L, and Walmsley, Frank "Chemical Principles, Properties, and Reactions",
      Addison-Wesley Publishing Company, 1984.

14.    Middlebrooks,  E. Joe "Water Reuse", Ann Arbor Science Publishers Inc., 1982.

15.    Naval  Facilities Engineering Command (NFEC),  Technical Memorandum no. 71-85-01,
      "Manufacturer/user study of selected recovery-related plating technologies", Port Hueneme,
      California,  October 1984.

16.    Patterson,  James, W., "Industrial Wastewater Treatment Technology", second edition,
      Butterworth Publishers, Boston, 1985.


17.    USEPA, "Treatability Studies for the  Inorganic Chemicals Manufacturing Point Source
      Category", Publication no. 440/1-80/103, July  1980.

18.    USEPA, "Development Document for Effluent Limitations Guidelines  and New Source
      Performance Standards for the Fabricated and Reclaimed Rubber Segment of The Rubber
      Processing Point Source Category", USEPA Publication No.440/1-74/030a, Group I, Phase
      II, December 1974.

19.    USEPA, "An Investigation of Techniques for  Removal of Chromium  for Electroplating
      Wastes", Publication no. 12010 EIE 13/71, 1971.
                                        345

-------
                          THE USE OF HORIZONTAL WELLS IN
                            REMEDIATING AND CONTAINING
                                   A JET FUEL PLUME -
                                PRELIMINARY FINDINGS

                        by  Daniel Oakley, Science & Technology, Inc.;
              Mark Thacker and Karen Singer, CDM Federal Programs Corporation;
                      Jack Koelsch. Martin Marietta Energy Systems, Inc.;
                            and Bill Mabson, Dynamac Corporation
                                        ABSTRACT

A Pilot Study/Demonstration Study (PS/DS) is being performed at Williams Air Force Base in
Chandler, Arizona, to demonstrate the ability of horizontal wells to remediate and contain a jet fuel
plume. The jet fuel plume is floating on the water table at approximately 220 ft below ground
surface.  The associated dissolved plume flows east-southeast in the shallow aquifer that is
approximately 25 ft thick.

One horizontal well is installed in an east-west direction with the screen located approximately 15 ft
below the floating jet fuel. This well is parallel to the long axis of the plume.  During the PS/DS,
this well  will be evaluated for the ability to capture the most contaminated section of the plume.
Based on testing of the first horizontal well, a second horizontal well may be installed in a north-south
direction near the downgradient edge of the dissolved plume.  This well will be perpendicular to the
long axis of the plume.  During the PS/DS, this well will be evaluated for the ability to act as a
barrier to plume migration.

A short-term pumping test has  been performed on the horizontal well.  Short- and long-term pumping
tests have been performed on vertical extraction wells also  installed at the site. Preliminary findings
of these pump tests are compared and discussed.
                                    1.  INTRODUCTION
The U.S. Air Force Air Training Command is conducting a Pilot Study/Demonstration Study (PS/DS)
at Williams Air Force Base in Chandler, Arizona, which is listed on the Environmental Protection
Agency National Priorities List. The purpose of this PS/DS is to evaluate the ability of horizontal vs
vertical extraction wells for the remediation and containment of a jet fuel plume at the liquid fuels
storage area (LFSA).  The PS/DS is performed under an interagency agreement between  the
Department of Defense and the Department of Energy that is being administered by Martin Marietta
Energy Systems, Inc.'s, Hazardous Waste Remedial Actions Program  (HAZWRAP).  CDM Federal
Programs Corporation has been contracted by HAZWRAP to conduct the PS/DS.

The LFSA covers approximately 5 acres and has been used since 1942.  The fuel storage system was
composed of numerous underground storage tanks and several thousand feet of 4-in.- and 6-in.-diam
delivery pipes. Most of this fuel storage system has been removed. Leaking tanks and fuel lines
                                         346

-------
have resulted in a JP-4 jet fuel plume in the uppermost aquifer.  This plume consists of a free product
plume (650,000-1,400,000 gal) (IT Corporation 1992) and an associated dissolved contaminant plume.

Beneath the site, an unsaturated zone of intermixed clay, silt, sand, and several gravel zones and thin
caliche layers extends to a depth of 220 ft below land surface (bis). These deposits are representative
of channel, floodplain, terrace, and alluvial fan deposition (Laney and Hahn 1986).  The shallow
aquifer is approximately 25 ft thick and is composed of intermixed clay, silt, and sand.  Underlying
the shallow aquifer is a clay aquitard approximately 20 ft thick.  Below the aquitard is a semiconfined
aquifer that is several hundred feet thick and is used as an irrigation and drinking water supply.

Investigations between 1984 and 1992 resulted in the installation of 31 shallow and 5 deep monitoring
wells (Fig. 1).  Potentiometric data from the site indicate that groundwater flow in both aquifers is to
me east-southeast. The eastern extent of the aquitard is currently unknown, although it has been
identified in diree downgradient compliance wells more than 1000 ft from the LFSA.  Apparent
product thickness measurements from wells within the free product plume vary from <0.1 ft to > 10
ft.  Groundwater sampling results indicate a large dissolved contaminant plume in the shallow
unconfmed aquifer associated with die free product plume.
                             2.  PROJECT PURPOSE AND SCOPE
The purpose of die PS/DS is to compare the effectiveness of horizontal and vertical extraction wells
in containing and remediating contaminated groundwater from the shallow unconfined aquifer. Before
contracting installation of the first horizontal weil, a groundwater model was used to help evaluate
potential effectiveness of horizontal wells at this site.  The model, which will be discussed in more
detail later,  indicated that horizontal wells would be an effective means of capturing die contaminant
plume.

Based on die plume geometry, hydrogeological characteristics, and modeling results, it was
determined mat die optimum placement of the first horizontal well would have an east-west
orientation with a screen 500 ft in length located approximately 15 ft below the floating jet fuel (235
ft bis).  This weil, parallel to die long axis of die plume, is the longest and deepest horizontal well
ever installed for environmental cleanup applications. During die PS/DS, this well will be evaluated
for the ability to capture die most contaminated portion of die plume. If testing of mis horizontal well
indicates efficient plume recovery, a second horizontal well with a screen 500 ft in lengdi will be
installed in a north-south orientation near the downgradient edge of the dissolved plume. This weil
will be perpendicular to die long  axis of the plume.  During the PS/DS, this well will be evaluated for
the ability to act as a barrier to plume migration.

Two vertical extraction wells were installed near die upgradient edge of the free product plume to
compare die efficiency of horizontal and vertical wells at this site .  Both a 4-in.- and a lO-in.-diam
extraction well were installed to evaluate the  increase in efficiency widi well diameter.  Two
additional monitoring wells  were installed for use as observation wells during aquifer testing.

Upgradient of the plume, four 4-in.-diam injection wells were installed in the unsaturated zone.
Groundwater will  be extracted during long-term tests from the horizontal and vertical extraction wells,
treated in a wastewater treatment plant, and reinjected into the shallow unconfined aquifer.  This
                                           347

-------
00
Approximate Extent     SL  Surface Location
of Free Product       ....
                 EH  Exit Hole
Fuels Wastewater
Treatment System
(FWWTS)
                                                                                                             200  10QO  JQO 200  300

                                                                                                                  SCALE IN FEET
            5 ppb Benzene
                                             Figure 1.   WELL LOCATION MAP.

-------
should prevent dewatering of this relatively thin, low-yielding aquifer and accelerate the rate of
cleanup by creating a groundwater mound upgradient of the plume.
                               3. PRELIMINARY MODELING
A three-dimensional model (DYNFLOW) using a horizontal finite element grid of 725 nodes and
1344 elements was used to simulate the potential effectiveness of horizontal wells at this site.  Based
on previous aquifer test results, a horizontal hydraulic conductivity (KJ of 1 ft/day was used.
Vertical hydraulic conductivity (KJ/Kh was varied between 1/10 and 1/100.  Specific yield was
assumed at 0.20.  The elevation of the horizontal wells was evaluated at 5 and  15 ft below the
groundwater surface.  Pumping rates  were varied from 13 to 40 gpm and test duration from 10 to 360
days.  Figure 2 illustrates me predicted groundwater surface after 360 days of pumping at 15 gpm
from two horizontal wells wim screens located 15 ft below the groundwater surface.  The leading
edge of the contaminant plume is engulfed within the zone of capture.  Therefore, the preliminary
results indicated that two horizontal wells should effectively contain the contaminant  plume.
                            4.  VERTICAL WELL INSTALLATION
The two vertical extraction wells were installed to a depth of approximately, 250 ft with a mud rotary
drill rig.  The saturated zone was cored, and screen slot size was determined based on sieve analysis.
Forty feet of 0.010-in. slotted, wire-wrapped, stainless steel screen was placed to extend  10 ft above
the static water table.  A carbon steel riser was installed above the screen to ground surface. The
annuius was backfilled with 20/40-size quartz sand to 3 ft above the screen.  A 2-ft-thick bentonite
seal was installed above the sandpack. The annuius was then backfilled to ground surface with
cement bentonite grout.  Monitoring well construction was similar except for the use of 0.020-in.
slotted screen and  PVC riser material.

The four injection wells were installed to a depth of 200 ft with a mud rotary drill rig. Construction
specifications were 100-180 ft of 0.020-in. slotted, inverted V-wire wrapped,  stainless steel screen
with stainless steel riser to the ground surface. The annuius was backfilled in the same manner as the
extraction wells.
                          5. HORIZONTAL WELL INSTALLATION
 The horizontal well was installed using river crossing technology to a depth of 235 ft. Two boreholes
 were abandoned because of drilling or installation difficulties before successful installation of the first
 horizontal well. The custom-made rig used for installation included a hydrauiically operated, gear-
 driven feed frame trailer, a control trailer, and a mud tank/mixer trailer.  The gear-driven feed frame
 trailer is raised at one end, and the odier end is placed in a mud pit to set the entry angle. For this
 well, the entry angle  was 20° to the ground surface.
                                          349

-------
Ul
o
               •230-

         BUCHANNON  ST.
      LEGEND
            Reinfection Well
      f*   '\  Approximate Extent
            of Free Product
Fuels Wastewaler
Treatment System
(FWWTS)

5 ppb Benzene
-240—Water Level Contour

 •HHB Well Screen

 SL  Surface Location

 EH  Exit Hole
                        Figure 2.   MODELING OF  EXTRACTION AND REINJECTION.
                                    Extraction/Reinjection  Rate - 15  GPM/Well
                                          360 Day Water  Level Contours
                                                                                                 200  100 _JL JOO 200  300

                                                                                                     SCALE IN FEET

-------
Figure 3 illustrates the well installation technique.  A pilot hole (step i) is drilled at a 20° angle and
becomes horizontal at 860 ft measured length (235 ft true vertical depth).  The pilot hole is then
steered horizontally for 500 ft, and the angle is then built to 16° toward an exit location 1250 ft
beyond the end of the horizontal section.  The pilot hole is drilled with 6-in.- diam drill string and a
9-in.-diam tricone bit.  The drill string is steered using a bent subassembly and a drill bit with one
jet.  Increasing the borehole angle up or down, left or right, is achieved by orienting the bent
subassembiy in the desired direction and jetting while advancing the drill string without rotation.
Drilling a straight hole without changing the angle is accomplished by rotating  the drill string while
jetting.

A downhole magnetic guidance system, located in  the probe directly behind the drill bit, is connected
to the surface  by wire line and  determines the real-time inclination and azimuth of the bent
subassembly.   Based on this information and the pipe length, trigonometry is used to determine the
borehole location. Secondary confirmation of the borehole location is obtained by halting drilling
operations every 30 ft (one joint length) and applying an electrical current to a  wire loop on the
surface, thereby creating a magnetic field  of known geometry and intensity over the padi of the
borehole. This  magnetic field is measured at the probe and  modeled.  The location of the probe  is
then determined relative to the proposed borehole path.

The downhole location equipment is a proven technology for river crossing boreholes, which are
usually no deeper than 100 ft bis.  Drilling of the three boreholes during this project showed that the
secondary confirmation tools lost depth accuracy below 150 ft.  The secondary confirmation tool
azimuth readings generally remained accurate throughout the three boreholes.  The primary downhole
magnetic  guidance system remained accurate for depth and azimuth throughout the three boreholes.
A downhole gyroscopic tool was used to confirm the borehole location for the first two boreholes.
This tool  confirmed that the boreholes were within 2.5 ft (plus or minus) of the proposed depth.

After the pilot hole is successfully installed, a 16-in.-diam reamer bit is attached to the drill string at
the exit location and pulled back to the entrance location, thereby enlarging the borehole (step 2).
Drill string  is  shuttled from the entrance to the exit location and added behind the reamer bit so that a
complete  drill string remains in the borehole.   After completing the first ream, an 18-in.-diam
reamer bit is added at the exit location and pulled back by the same method,  further enlarging the
borehole (step 3). After completion of the second ream, the 16-in.-diam reamer bit is again attached
to the drill string at the exit location. Immediately behind the reamer bit is a swivel assembly.
Behind this  swivel assembly are the well completion materials.  The swivel assembly allows rotation
of the drill string and reamer bit without rotation of the well completion materials. This minimizes
stress on the well completion materials during installation. All well completion materials are welded
together and placed on rollers on the ground surface before  installation. The well materials are then
pulled into place from the exit to  the entrance location for well  completion (steps 4 and 5).

Figure 4 illustrates well construction specifications including 500 ft of prepacked, 0.008-in. slotted,
wire-wrapped, stainless steel screen with 40/60 sandpack. The inside diameter of the screen is 6.0 in.
and the outside diameter is  10.75 in.  The prepack sandpack is  1.7 in. thick. Epoxy-coated carbon
steel riser material  11 in. in diameter and 860 ft long is installed from the screen to ground surface.

Well development was accomplished by first attaching a fire hose to the well and flushing  water down
the well,  out the screen,  and up the annulus, exiting at both the entrance and exit locations.  The well
was flushed for 6 hours at 150 gpm until returns cleared significantly. The exit borehole was then
                                           351

-------
                        Ground Surface
                    Horizontal
                                 able
                               1250-ft —
	600-1	+— 500-1 -H

 Directional drill 9-in. pilot hole


     •Boo-t—Jt— soo-fl.

       2)  16-in. putt ream
eOO-fl.	J-	500-ft.	-I1	1250-ft.-

 3)  18-in. pull ream
N)
                               Rfinghead
                                                      Borehole
                      Dri Slam
                                            Ground Surfact
                              0.220-
                                         Horizontal
                                -800-fl-
                                         -500-fl-
-izsan.
                               Pulling well completion materials
                                       into place
                                                                              Scnen
                                     End Cap •
                                      Ground Surface
                                                                                      Horizontal
                          -BOO-flr
                                                                                                         Abandoned
                                                                                                         BordKle
                        5)  Completed horizontal well
                                                                                                                Not to Scale
                 Figure  3.  DRILLING AND  HORIZONTAL WELL INSTALLATION PROCEDURE.
                                                   Williams  Air Force  Base

-------
       Entry
                -Cament/grout
                 seal to surface
                                        Cement/grout—,
                                        seal to surface \
                                                                                             Exit
 235 ft.
11-ln. epoxy-eoated
steel riser
          Centralized—
          submersible  /
            pump    /
                                         Native materials
                                        Miiih i,!• i  111,
                                        •••—••mTTTTTI
                                     IUHHin P i"1' in ''I
                                          V-40.75-IR. OJ}., 6JWn. I.D.
                                            stainless steel
                                            prepack screen

                                               500ft. 	
Not to Scale
           Figure  4.    HORIZONTAL  WELL CONSTRUCTION DIAGRAM.
  plugged and a jetting tool inserted in the screened section of the well.  An environmentally safe
  spotting fluid designed to break up mudcake on the borehole walls was pumped into the well screen at
  three locations. The borehole was then allowed to sit for 24 hours before another high volume flush
  was performed. The entrance borehole was then plugged and a pump placed in the well.  Pumping
  was performed with the pump located in different sections of the well screen until the water became
  clear of suspended solids.
                                      6.  AQUIFER TESTS



  Aquifer tests of the vertical and horizontal wells were performed to provide data for development of a
  more realistic interpretation of the aquifer underlying the site.



                            7. VERTICAL WELL AQUIFER TESTS



  Step-drawdown and constant rate  72-hour aquifer tests were performed on the 4-in.- and lO-in.-diam
  vertical extraction wells, EX-01 and EX-02.  Additionally, EX-02 was pumped for 21 days. Water
  levels were recorded using data loggers and transducers placed in monitoring wells MW-30 and
  MW-31 and extraction wells EX-01 and EX-02.  All wells are fully penetrating and screened in the
  same interval.
                                           353

-------
Before conducting the first aquifer test, background water levels were monitored. Barometric pressure
and precipitation measurements from the Sky Harbor Airport in Phoenix were obtained during the
testing periods to evaluate their effect on water levels.  Eight-hour step-drawdown tests initially were
performed on the vertical extraction wells to determine the specific capacity of the wells and to
choose the optimum pumping rate for die long-term tests. Constant rate pumping tests were
performed on the two vertical extraction wells.  After completion of each constant rate test, recovery
of the water table was monitored.
                              8. VERTICAL WELL RESULTS
Vertical extraction well EX-01 was pumped at a constant rate of 3.8 gpm for 72 hours.  Well EX-Q1
is 254 ft deep and screened between 212 and 252 ft bis.   Before the start of pumping, the static water
level was 215.4 ft bis.  Transducers were placed in the pumping well (EX-01) and three observation
wells fMW-31, MW-30, and EX-02) to record drawdowns.  Wells MW-31,  MW-30, and EX-02 are
27, 102, and 127 ft, respectively, from the pumping well, EX-01.

After 72 hours, a total of  14.6 ft of drawdown was measured in the pumping well.  Drawdowns in
the observation wells ranged from 0.61 ft in MW-30 to 0.42 ft in EX-02.  In general die data
indicate steep cones of depression surrounding the pumping well because of low transmissivity of the
formation.

Distance drawdown plots indicated the largest drawdown was recorded in observation well MW-30,
more than 100 ft from the pumping well. This suggests an increased hydraulic connection between
EX-01 and MW-30 relative to the other observation wells. This is possible considering the site
stratigraphy of interbedded sands, silts, and clays.  Well MW-31 recorded the next largest drawdown
and is widiin 27 ft of the pumping well.

Time vs drawdown plots from the pumping and observation wells resemble delayed gravity drainage
as described in Fetter (1988). The corrected drawdowns in  EX-01, MW-31, and MW-30 were used
to determine hydraulic conductivity and specific yield values using the Neuman Method (1975) for
delayed gravity drainage.  Horizontal  hydraulic conductivities calculated from observation well data
ranged from 9.62 ft/day to 11.73 ft/day.  Calculated vertical hydraulic conductivities ranged from.
0.07 ft/day in MW-30 to 0.08 ft/day in MW-31. These correspond to a k/k,, ratio of approximately
 1/140.  Specific  yields (Sy) ranged between 0.023 (MW-30) and 0.286 (MW-31). The time
drawdown plot for MW-30 may not have reached  the late curve, resulting in the lower values of Sy
Time drawdown data for EX-01 resulted in a hydraulic conductivity of 3.0 ft/day.  Recovery data
from EX-01 was used to calculate a hydraulic conductivity of 2.58 ft/day.

Extraction well EX-02 was pumped for 72 hours at approximately 2 gpm. Drawdown in the
observation wells increased until 1100 minutes into the test.  At this time, the drawdown declined
steadily until the last day of the 3-day test when it began to rise  in a curve similar to the initial
drawdown.  Total rainfall for the month before this test was three times the monthly average. The
sudden drop in drawdown may be partially caused by delayed gravity drainage from low permeability
clays. Because of the later anomalous data, drawdown data from MW-30 for the first 1000 minutes
of the test was used to calculate a hydraulic conductivity of  1.8 ft/day.
                                          354

-------
Sustained pumping rates from the two vertical extraction wells varied considerably.  Also distance
drawdown plots did not decrease linearly with increased distance from the pumping well.  These
findings indicate heterogeneous, nonisotropic conditions. Increasing well diameter did not increase
recovery.  Recovery rates are a function of formation properties and well efficiency.
                          9.  HORIZONTAL WELL AQUIFER TEST
A 3-day constant rate aquifer test was performed on the horizontal well.  The pump was placed in the
middle of the screened section, and a transducer was installed inside a drop pipe placed at the
beginning of the horizontal section.  Transducers were installed in monitoring wells LI-06, W-01,  W-
03, W-05, and W-08 surrounding the horizontal screened section.

During the test, hand water level and product-water interface measurements were taken to confirm
transducer readings and to determine the effect that pumping has on the product-water interface.
                           10.  HORIZONTAL WELL RESULTS
The pumping test on the horizontal well consisted of pumping the well at 9.8 gpm for 3 days.  Wells
within 30 ft of the horizontal well recorded drawdowns of less than 0.40 ft indicating a steep cone of
depression similar to that observed in the vertical well aquifer tests.  During the test, product
thicknesses rose steadily in the observation wells within the free product plume.  The water table
decline in the cone of influence was minimal. The apparent product thickening is likely the result of
removing the skimming system from these wells 48 hours before initiating the test.

The value of hydraulic conductivity from the horizontal well test data was calculated using the
following analytical equation for flow to a trench (Powers 1981).
              _Q     =     KfH2 - h2
              2X              2880 L
       where:
              Q      = discharge (gpm)
              X      = length of trench (screen) (ft)
              K      — hydraulic conductivity (gpd/ft2)
              H      = saturated thickness at distance L from trench (ft)
              h      = saturated thickness above screen at end pumping (ft)
              L      = distance between trench and point where H
                         is measured (ft).
                                         355

-------
Calculated hydraulic conductivities ranged from 0.14 to 0.33 ft/day. These values are much lower
than those calculated from the vertical well aquifer test data.  This is not unexpected because of the
adverse impact to the formation from the three attempts at well installation. Upon completion of an
on-site treatment system, the well will be developed  further and pumped to obtain additional data.
                             11.  FUTURE MODELING EFFORTS
Currently the  finite element numerical model (DYNFLOW) developed by CDM Federal is being
adapted to include site geologic conditions as well as calculated hydraulic properties at the site.
Results of aquifer tests from the horizontal well will be used for transient calibration of the model.
This will enhance the model's capability to predict aquifer response to pumping in the horizontal well.
                            12.  SUMMARY AND CONCLUSIONS
The PS/DS being performed to evaluate the ability of horizontal vs vertical extraction wells to
remediate and contain a JP-4 plume will require additional data before definitive conclusions can be
reached.  Four-in.- and lO-in.-diam vertical extraction wells and a 6-in.-diam horizontal extraction
well have been successfully installed.  Only preliminary aquifer testing of the horizontal well has been
performed.

Aquifer test results from vertical extraction wells show low transmissivity and vertical hydraulic
conductivity and a very steep cone of depression.  Therefore,  a tight vertical well spacing (25-50 ft)
would be required to remediate the site with vertical wells. These tests also  indicate that increasing
well diameter does not increase the recovery rate in this heterogeneous formation.

A short-term aquifer test performed on the horizontal well also shows a very steep cone of
depression.  Based on the initial results, a horizontal well may perform more effectively as a barrier
than for plume recovery at this site.   After further testing, a decision will be made as to whether a
second horizontal well will be installed at this site.

Future testing of the horizontal well will include pumping at different locations in the horizontal well
screen, long-term (30 day) and short-term tests, low and high pumping rate tests, an evaluation of
whether product recovery from the skimming pumps is increased during water table depression, and
pumping from two pumps simultaneously in the horizontal well.
                                            356

-------
                                      REFERENCES
Fetter. C. W. 1988.  Applied Hydro geology, Second Edition. Merrill Publishing Company,
       Columbus, Ohio.

IT Corporation 1992. U.S. Air Force Remedial Investigation/Feasibility Study, Williams Air Force
       Base, Arizona, Remedial Investigation Report, Liquid Fuels Storage Area - Operable Unit 2
       Vol. 1.

Laney, R. L. and Hahn, M.E.  1986.  "Hydrogeoiogy of the Eastern Part of the Salt River Valley
       Area, Maricopa and Pinard Counties, Arizona,"  U.S. Geological Survey, Water-Resources
       Investigations Report 86-4147.

Neuman, S.  P. 1975. "Analysis of Pumping Test Data From Anisotropic Unconfmed Aquifers
       Considering Delayed Gravity Response," Water Resources Research, Vol. 11, No. 2.

Powers, J.P. 1981. Construction Dewatering, a Guide to Theory and Practice, John Wiley & Sons.
                                        357

-------
EVALUATION OF THERMAL E)CrRACTION TECHNOLOGIES FOR TREATMENT OF
   SOILS CONTAMINATED WITH COAL TARS AND WOOD PRESERVATIVES
             AT THE PACIFIC PLACE SITE, VANCOUVER, B.C.
             Sandra Whiting, Gregory Helland, John Kinsella
                 SCS Engineers, Bellevue, Washington
                                358

-------
                                      INTRODUCTION
 The British Columbia Ministry of Environment, in cooperation with Environment Canada's
 Demonstration of Site Remediation Technologies (DESRT) Program, conducted treatability studies of
 several treatment technologies to evaluate their effectiveness for the treatment of soil from a
 contaminated site.

 The site is a 185 acre area of industrial land located in Vancouver, B.C. Historically, a wide variety
 of industrial activities operated on the site, including two manufactured gas plants, sawmills, boat
 building, metal plating, wood preservation, carpet cleaning, railway yards, and fuel storage. In
 particular, coal tars and metal oxide wastes from the coal gasification plants and wood preservatives
 were mixed with fill material.  Contaminants detected at elevated concentrations on the site included
 polycyclic aromatic hydrocarbons (PAHs), cyanide, lead, sulfur, total petroleum (extractable)
 hydrocarbons, and chlorophenols.

 The technologies  evaluated included incineration, thermal extraction, bioremediation, and
 stabilization/solidification. This paper focuses on the testing of two innovative thermal extraction
 technologies:  the Chemical Waste Management XTRAX process; and the UMATAC Industrial
 Processes AOSTRA Taciuk Processor. This paper includes a brief description of each technology,
 procedures  utilized in each study, results, and estimated costs for full-scale treatment.
                           CHARACTERISTICS OF SOILS TESTED
 Soil samples were collected from the site by excavating pits to the desired depths; the targeted soils
 were then mixed in the excavation using the bucket of the excavator.  After mixing, samples were
 coarsely screened and placed into shipping containers. All samples were shipped to vendors in
 accordance with the Canadian Transportation of Dangerous Goods Regulations and the U.S.
 Department of Transportation Regulations.

 Four different sample types, with varying concentrations of PAHs, chlorophenols, oil and grease, and
 metals, were provided for testing in the thermal extraction systems. As would be expected from
 such a site, the fill was heterogenous and contaminant concentrations exhibited a wide variability,
 even between split samples.  In general, all the soil samples were poorly sorted (well graded) and
 were characterized as consisting primarily of fine to coarse sand with  approximately  17 to 34% silt
 and clay.  Samples also contained debris such as brick fragments, metal fragments, and wood
 chunks.  Sample moisture content ranged from 10 to 45% and averaged 25%.  The pH of each
 sample was near neutral (6.9 to 7.2) except for Sample 1, which had a pH ranging from 4.2 to 5.0.
 The chemical characteristics of each of the four samples are summarized below.

 Sample 1.  This sample location is characterized by the presence of coal tars; the average
 concentration of total PAHs in Sample 1 was 3,375 mg/kg. Oil and grease concentrations ranged
from 29,000 to 200,000 mg/kg with average concentrations of  112.0OO mg/kg. Total cyanide
 concentrations in Sample 1 were between 2,900 and 5,000 mg/kg. Sample 1 also was notably high
 in total sulfur (ranging from 1 to 17 percent).

Sample 2.  Total PAHs in this sample ranged from  42 to 348 mg/kg (averaging 117 mg/kg). Oil and
grease concentrations ranged from 4,200 to 15,600 mg/kg. Lead concentrations were between  116
and 5,600 mg/kg in this sample.  Zinc was also high, averaging 2,347 mg/kg.
                                           359

-------
Samp/e 3.  This sample was collected from an area historically used as a dip tank for wood treating.
Pentachlorophenol concentrations ranged from 9 to 309 mg/kg (averaging 136 mg/kg) and
tetrachlorophenol ranged from 125 to 500 mg/kg (averaging 242 mg/kg). PAHs were also detected
in this sample, averaging 260 mg/kg total PAHs. Oil and grease ranged from 8,300 to 23,200
mg/kg.

Samp/o 4.  Sample 4 was characterized by high concentrations of lead, ranging from 8,000 to
12,500 mg/kg. Total PAHs averaged 151 mg/kg; oil and grease concentrations averaged 700
mg/kg.
                             TREATABILJTY STUDY PROCESS
The vendors were required to perform detailed chemical and physical analyses of the soil samples
using specified analytical protocols, prior to treatment and at the end of the treatment period.  In
addition, detailed analysis of treatment by-products or residuals was also required.

The major objectives of the treatability studies were to:

              Determine or validate technology effectiveness.
              Develop pricing data for full-scale operation.
              Evaluate the likelihood of delisting treated waste under B.C. procedures.
              Demonstrate compliance with emissions and effluent standards.
              Provide information potentially useful for other contaminated sites.

Cleanup goals were established for the treatability study program in order to evaluate technology
performance. The "British Columbia Standards for Managing Contamination at the Pacific Place
Site', developed by the British Columbia Ministry of Environment, formed the basis for the target
clean-up levels for organic contaminants. The clean-up goals are included in Tables 1 and 2.

A brief description of each thermal extraction technology, the study procedures used in testing, the
testing results and estimated costs are discussed in the following sections.
                                         X*TRAX
Technology Description
The X*TRAX system, developed by Chemical Waste Management, Inc., is a low temperature thermal
separation process that vaporizes or steam strips organic compounds from soils or sludges. It
consists of an externally fired rotary dryer and gas treatment system. A simplified flow diagram of
the X*TRAX system is provided in Figure 1.

Soils or sludges are fed into a rotary dryer and heated.  Treated solid residues exit the dryer and are
collected in a hopper. Nitrogen is used as a carrier gas for the volatilized organics which are
transported from the rotary dryer to a gas treatment system.

The gas treatment system includes a spray tower/scrubber, primary and secondary shell-and-tube
heat exchangers for condensing the gases. In the full-scale X*TRAX system, most of the nitrogen
gas is recycled, except for about 10  percent that is vented to the atmosphere. The bench-scale
system is equipped with two activated carbon filters for treating residual organics in the gas stream.
                                            360

-------
  Liquids and particulates (and less volatile organic compounds) from the spray tower are passed
  through a gravity phase separator and are then filtered to separate out solids. Condensed liquids
  from the heat exchangers are composited, filtered, and phase separated for further treatment as
  necessary.

  Review of the flow diagram indicates that the X*TRAX system produces several treatment residuals:

               Treated solids from the dryer
               Phase separator water (this is continuously recycled to the spray tower)
               Phase separator filtrate (particulates and some organics)
               Phase separator oil
               Filtrate from condensate
               Condensed water
               Condensed oil (not always produced in lab-scale tests)
               Emissions

 The full-scale X*TRAX system is a transportable unit capable of treating up to  125 tons per day of
 soil with a moisture content of 20%. CWM also operates a pilot-scale unit that is mobile and can
 treat about 5 tons per day of material containing 30% moisture.

 Testing Program

 Bench-scale treatability testing was conducted using  a  laboratory-scale X*TRAX unit which uses the
 same components as the full-scale X*TRAX system.  Samples 1, 2, and 3 were individually tested at
 two operating temperatures:  482°C and 371 °C.  In all cases, residence time was 85 minutes  Prior
 to processing, the samples were screened through a 1/4 inch sieve and then homogenized   The
 feed was sampled once each hour, beginning and ending at steady state conditions.  The feed
 samples were composited for each temperature condition and later submitted  for laboratory
 analyses. An average of about 7 kilograms of each soil sample was fed to the system. Feed rate
 ranged from 8 to 13 grams per minute, depending on the sample type.

 Treated solids and  condensates were collected and weighed every 15 minutes. Solids were
 composited for each steady state condition and then  were subsampled for analytical testing  The
 condensates and phase separator liquid were filtered through 25 micron filter paper  The filtered
 condensate was poured into a separatory funnel from which samples for analysis were taken  The
 filter cakes and phase separator water products were subjected to chemical and physical analyses.

 Because of the small scale, no organic liquid phase products were generated in these studies nor
 were emissions collected and analyzed.

 Results

 The analytical results indicate that the X*TRAX process was effective at removing organic
 compounds from contaminated soils and meeting the treatment goals for the treatability study
 Table 1  provides a summary of results for the 482°C operating condition.

 In Sample 1, the highly contaminated sample, total PAHs (16) were reduced by 99.93 percent at the
 482°C operating condition.  The concentrations of PAHs (as well as other organic constituents) in
 the treated soil were well below the established treatment goals.  Similar PAH and other organic
 compound removal efficiencies were achieved in Samples 2 and 3.

 Pentachlorophenol was reduced by an average of approximately 98 percent, and
2,3,4,6-tetrachlorophenol by an average of approximately 92 percent for all three samples tested.
                                           361

-------
                                  Figure   1.  X*TRAX Flow Diagram
ON
to
                               LIQUID NITROGEN
              GAS
                               HEATER
                      DRYER
                 r
                     TREATED
                     SOLIDS
                     HOPPER
                  GAS
         GRAVITY
         PHASE
         SEPARATOR
                     T
                            LIQUID
                             SOLIDS
                             FEED

                                 GAS
                            SPRAY
                            TOWER/
                            SCRUBBER
                                                       RECYCLED NITROGEN
                                                       FULL/PILOT-SCALE
                                           1
                                  CONDENSER
                                  (PRIMARY)
GAS
                                             LIQUID
                                       CONDENSATE
                                               L_
                                                                                TO
                                                                            ATMOSPHERE
                 CONDENSER
                I(SECONDARY)
GAS
                                                        •M-
              LIQUID!
                   M
                CONDENSATE
             	I
                                 CARBON
                                 FILTER
                                 (LAB-SCALE
                                 ONLY)
                  FILTRATION

                      \
            WATER     OIL    FILTER CAKE
                                                         T
                                                      FILTRATION
  Source:
Modified from "Pacific Place Project Treatability Technical Study for British Columbia Hazardous Waste Management Corporation", Chemical
Waste Management of Canada, Inc.

-------
 Concentrations of chlorophenols in some of the treated soils appeared to be slightly above the
 treatment goals.

 The vendor reported that concentrations of dioxins and furans in the Sample 1 feed soil (3.7 ppt
 TCDD equivalency) decreased after processing through the X*TRAX system (to 1.6 ppt TCDD
 equivalency).

 Total metals concentrations in the samples did not change significantly as a result of the X*TRAX
 treatment process, nor did the solubility of the metals appear to be affected by treatment, based on
 extraction testing.  Total cyanide in the treated soils from Sample 1 was reduced from 2,500 mg/kg
 to 6 mg/kg in the lower temperature run and to less than 0.5 mg/kg in the 482°C run.

 Organics concentrations in the condensed water and phase separator water produced during the
 treatability study were generally  very low or insignificant except for oil and grease and total
 phenolics. Oil and grease concentrations ranged from less than 5 to 24 ppm.  Total phenolics were
 detected at high levels in both condensed and phase separator water (54 and 58 ppm, respectively)
 from Sample 3 soils and in condensed water from Sample 1 soils (5.51 ppm). These concentrations
 exceed the B.C. effluent standard for phenols.

 There were insufficient products to allow for the analysis of liquid-phase organics. Analysis of
 emissions during the treatability  study was not performed due to the small scale of the equipment.

 Examination of the analytical testing performed on treatment residuals from this study shows that a
 significant proportion of the PAHs from all sample types and chlorophenols in Sample Type 3 were
 removed in the spray tower and  were deposited on the phase separator filter cake. A small amount
 of the total PAHs and chlorophenols were also removed from the piping between the dryer exit and
 the spray tower ("carryover holdup"). Some of the cyanide was also removed on the phase
 separator filter cake and from the piping between the rotary dryer and the spray tower.

 Estimated Treatment Costs

 The vendor estimated a unit cost for soils treatment and disposal of condensed liquids and filtrates
 at about $600(US) per ton.  The  cost assumed treatment of about 27,000 tons of soil, and included
 mobilization, labor, health and safety, sampling and analysis, and ambient air monitoring.

 Discussion

 Based on the data provided in the treatability study report, the X*TRAX process appears to be an
 effective technology for removing organic contaminants from the tested soils.  However, several
 treatment residuals are produced that require proper management/disposal (i.e., filtrates, condensed
 oil, condensed water, phase separation water).

 During full-scale operation of the X*TRAX, condensed water and phase separator water produced
 during the treatment are typically applied to the treated soil to control dusting. As noted above, the
 concentrations of some organic contaminants in the two wastewater streams were of concern,
 making the use of the water for dust control a questionable practice.  However, CWM reports that in
full-scale treatment the water is tested first to ensure that the quality of the water is acceptable (for
 disposal of the treated soil) before it is applied to the treated soil.

The treatability study could not provide data on condensed organic-phase liquids, and full-scale data
 pertaining to this waste stream were not available at the time of the testing program.  However, if
free-phase organic liquids were produced during full-scale operation, it is likely that they would
require incineration at an approved hazardous waste incinerator.
                                           363

-------
TABLE 1. Selected Organic Contaminant Concentrations, X*TRAX TreatabllHy Study
Contaminant
PAHs
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(g,h.i)perylene
Pyrene
Benzo(a)pyrene
lndeno(1 ,2,3-cd)pyrene
Acenaphthene
Acenapthylene
Anthracene
Fluoranthene
Fluorene
Napthalene
Phenanthrene
Total PAHs
Chlorophenols
2,4,6-Trfchlorophenol
2.4,5-Trichlorophenol
2,3,4,6-Tetrachlorophenol
PsntachloroDhenol
Sample 1
Run Feed
(ppm)
<0.014
65
60
27
190
220
400
61
<0.005
45
100
530
110
1,300
320
3,428
<0.42
<0.66
<0.66
<1 8
482 'C
Treated Soil
(ppm)
0.18
0.20
<0.001
0.050
<0.0047
0.35
<0.010
0.033
<0.005
<0.008
0.057
0.68
<0.0006
0.41
0.29
2.25
<0.42
<0.66
<0.66
<1.8
Sompte Type 2
Run Feed
(ppm)
3.2
3.5
4.3
5.1
<0.0047
4.4
9.5
3.7
<0.005
<0.005
8.8
6.3
1.8
<0.005
3.5
54.1
<0.07
1.6
2.4
9.4
462* C
Treated
Soil (ppm)
<0.014
<0.001
<0.001
<0.0004
<0.0047
<0.0025
<0.010
<0.001
<0.005
<0.008
<0.0007
<0.0007
<0.0006
<0.005
<0.005
0
<0.07
<0.11
0.58
<0.30
Sampto Typ« 3
Run Feed
(ppm)
12
11.0
9.1
6.8
8.9
36.0
18.0
6.8
<0.008
<0.008
5.3
43.0
4.3
<0.005
27.0
192.6
<0.07
<0.11
<0.30
<0.035
482' C
Treated
Soil (ppm)

<0.001
<0.001
<0.004
<0.0047
<0.0025
<0.010
<0.001
<0.005
<0.008
<0.0007
<0.0007
<0.0006
<0.005
0.18
0.18
<0.07
<0.11
<0.11
<0.3
TreatmBntGoate
(Pacific Place Level
B Standard*
(ppm)
1
1
1
1
1
10
1
1
10
10
10
10
10
5
5
20
0.5
0.5
0.5
0.5

-------
                              AOSTRA TACIUK PROCESSOR
Technology Description

The AOSTRA Taciuk Processor (ATP) is owned and operated by UMATAC Industrial Processes in
Calgary, Alberta (the technology is also known as the SoilTech ATP Plant in the U.S.). The ATP is a
thermal extraction technology that separates organics from solids by pyrolysis and distillation. A
flow diagram of the ATP is provided in Figure 2.  For the treatability study testing, UMATAC was a
subcontractor to Newalta Corporation of Calgary.

The processor consists of a rotary kiln with a preheat zone, operating at about 260°C; a reaction
zone, operating at between 317°C and 649°C; and a combustion zone, operating at 538°C to 815°C.
Water vapor and the more volatile organics are extracted in the preheat zone. Oils and heavier
organics are removed in the reaction zone. Both the preheat and the reaction zones are maintained
in anaerobic conditions to prevent oxidation of the organics. The combustion zone bums coke from
thermal cracking that forms on the solids in the reaction zone. Oxygen is added to the system in
the combustion zone.

The flue gas from the combustion zone is passed through a baghouse and a wet scrubber. Carbon
absorption is sometimes added for final gas scrubbing in full-scale treatment.  Treated solids are
discharged from the kiln as tailings. The steam and organics produced in the preheat zone and the
organics separated out in the reaction zone are withdrawn under vacuum and are condensed in
separate vapor trains.  The  condensed water generally requires treatment before disposal. The oily
waste must either be incinerated or used as fuel  for the processor.

As can be determined from examination of the process flow diagram, this treatment process
produces several treatment by-products:

        •      Preheat Zone

                     Vapor stream:  water, oil, gas (recycled to combustion unit)
                     Solids stream:  heated, dried solids (to reaction zone)

       •      Reaction Zone

                     Vapor stream:  cyclone solids, bottom oils (removed at 150°C - 210°C),
                     heavy oils (removed at 100°C - 150°C), light oils (removed in heat
                     exchanger), sour water, off gases (flared or recycled to combustion zone)

                     Solids stream:  coked solids (to  combustion zone)

       •      Combustion Zone

                     Flue gas stream: cyclone fines, baghouse fines, scrubber water, scrubber
                     solids, cleaned gas

                     Solids stream:  treated tailings
                                           365

-------
Oi
o\
            Figure   2. Aostra Taciuk Process (ATP) Flow Diagram
      -n STACK
       WATER
   OIL      FUEL
                 FLUE GAS
                TREATMENT
        FEED
                                                t
                   VAPOR
                 TREATMENT
                                PREHEAT
                                VAPOR
    ATP
PROCESSOR
                          VENT
                          GAS
                                                     OFF GAS
OIL
                                            VAPOR
                                             SOUR
                                             WATER
                                                                   T
                                                                             WATER
   OIL
RECOVERY
PRODUCT OIL
                     TAILINGS
  Source:    "Treatability Study Report, Pacific Place Soils Remediation Project", Newalta Corporation, June 30,1992.

-------
Testing Program

Both bench- and pilot-scale studies were conducted using the ATP.  In addition, a bench-scale
wastewater treatment study was performed to evaluate the effectiveness of a combination of
standard treatment technologies for treating preheat water and sour water. Only the pilot-scale
study is presented here.

Based on the success of the batch-scale testing program, pilot-scale testing in UMATAC's 5 tonne
per hour prototype unit was undertaken. A total of four tests were performed using the 5 tonne per
hour ATP. Samples 1 and 3 were processed separately and a blend of Samples 1, 2, and
4 and a blend of Samples 2 and 4 were also tested.  Because of the high  sulfur content in Sample 1
and the concern that the sulfur dioxide emissions would be exceeded, UMATAC added sand in a
four-to-one ratio to the sample material before processing. The vendor also added sand to the third
and fourth test runs to reduce moisture content.

Each test lasted between 3.4 and  4.4 hours; feed rates ranged from 2.7 to 4.1 tonnes per hour,
depending on the sample type. Exit temperatures in the reaction zone averaged 526°C.  Exit
temperatures in the combustion zone averaged 648°C.

During each test run, samples of the feed and samples of the treated tailings (treated soils,
baghouse ash, and cyclone ash) were collected every 30  minutes. Composite samples of each of
these were submitted for analytical testing. Condensates, both water samples and oil samples, were
also collected every 30 minutes and analyzed from each run.

Monitoring of stack emissions was also performed during  three test runs to demonstrate compliance
with emissions standards and for calculating destruction and removal efficiencies (ORE) and mass
balances.

Results

Comparisons of the concentrations of selected organic compounds in the feed material and the
treated soils and ash are provided in Table 2.

In general, the ATP achieved nearly 100 percent removal of organic contaminants from the soil
samples. The concentrations of PAHs in the tailings were all less than detection limits (practical
quantitation limits of 0.33 - 0.53 ppm).  It should be noted that concentrations of organics in the feed
were low due to dilution from the addition of sand before processing. For  example, after dilution
with sand, the concentration of total PAHs in the Sample 1 run averaged about 46 mg/kg. In the run
combining Sample Types 1, 2, and 4, total PAH concentrations averaged 309 mg/kg.  These
concentrations are considerably lower than those found in undiluted soil samples.

Analyses of dioxins and furans were performed for the second run (Sample 3 soils).  Concentrations
in the feed, treated  soils, and ash  streams were all less than 1  ppm.  The data provided by the
vendor is suspect, however, due to poor surrogate recoveries reported  by the vendor's laboratory.

Useful information about reduction of chlorophenolic compounds was not generated in this testing
program because the concentrations of chlorophenols in the soils received by UMATAC were
non-detectable (6.6 ppm detection limit).

Some volatile organic compounds were detected at high levels (as high as 300 ppm) in baghouse
ash, most notably toluene and xylene.  No explanation of these levels was  offered by the vendor.

Cyanide was reduced to less than detection limits in the treated soils containing soil from Sample 1.
Generally, low concentrations of cyanide were detected in cyclone dust and baghouse ash.
                                           367

-------
                           TABLE 2. Selected Organic Contaminant Concentrations, ATP Treatablllty Study
Contaminant
TEH"'
Oil & Grease
Phenols
Total
Chlorophenols1'1
Dloxlns/Furans'3'
Total PAHs m
BTEX B
.
Ftttd
Soil
(ppm)
145
118
<0.01
ND
NT
46.2
ND
Sampt
Treated
. Soil
(PPm) .
ND
26
<0.01
ND
NT
ND
ND
• 1 +Sand
Cycfon*
Duct
(PPm)
ND
22
<0.01
ND
NT
<
ND

BaghouM
Ash
(ppm)
ND
13
0.53
ND
NT
3.3
ND
Sampled
Feed
Sol!
(ppm)
ND
1180
0.13
ND
0.566
2.17
ND
Treated
Soil
(ppm)
ND
21
0.01
<
0.005
ND
ND
Cyclone
Duct
(PPm)
ND
65
0.02
ND
0.553
<
30
BaghouM
A*h
(ppm)
ND
74
0.05
ND
. 0.596
16.5
133
Sample 1, 2, 4 + Sand
F««d
Soft
(PPm)
ND
159
0.05
ND
NT
309.2
<
Treated
8ofl
(ppm)
ND
<5
<0.01
ND
NT
<
ND
Cyctott*
Duct
(PPm)
ND
7
<0.01
ND
NT
<
ND
Bftghout*
Ach
(PPm)
ND
9
<0.01
ND
NT
3.2
130
Treatment Goalt
PucWIcPlac* :
Lov.lB
Standards (ppm)
150
1000
1.0
1.0
NA
20
5.0, 30. 50. 50
Notes: (1) Total Extractable Hydrocarbons (Modified EPA Method 8015), average of duplicate columns. Detection limit In ash streams ranged from 9 to 48 ppm.
(2) Detection limit for pentachlorophenol in treated soils and ash was 1.65 ppm.
(3) Data suspect due to poor surrogate recoveries.
(4) Detection limits for PAHs ranged from 0.33 to 0.462 mg/kg.
(5) Detection limits for BTEX ranged from 20 to 70 mg/kg.
< = Below detection limit
NT = not tested
ND « not detected
Data from: Treatabllity Study Report, Pacific Race Soils Remediation Project," Newalta Corporation, June 30, 1992.
o\
GO

-------
In general, the concentrations of metals were reduced somewhat in the treated soils from what was
reported in the feed, and were higher in the cyclone dust and baghouse ash (particularly arsenic,
copper, lead, tin, and zinc).

The testing program produced about 2.5 tons of oil (1.6 tons of which was from make-up oil for
the hydrocarbon vapor stripper). In general, the oils contained very low levels of inorganics (except
for calcium and sulfur) and low levels of benzene, toluene, ethylbenzene, and xylene (BTEX). The
vendor did not analyze the oils for PAHs or chlorophenols. However, data from the third party
laboratory analyses of the oils, indicated significant concentrations of PAHs in the oils, as would be
expected. Dioxins and furans were also present in significant concentrations (ranging from 0.5 to
130 ppb TCDD toxicrty equivalence) in the bottoms oil from processing Samples  1 and 3 (oil from
the blended run was not analyzed for dioxins or furans). This is based on data from the third party
laboratory. The data from the vendor's laboratory, which also reported high levels of TCDD, are
suspect due to poor surrogate recoveries.

Stack Emissions

During the testing programs,  sulfur dioxide limits were exceeded in both tests using Sample 1
material. This, according to UMATAC, was due to the scrubber on the prototype equipment which
can only achieve 50 to 60 percent sulfur dioxide removal.

Particulate limits were exceeded in the second and third test runs (emissions were not monitored in
the fourth  test).  According to the vendor,  this was  due  to damaged bags in the baghouse that were
not discovered until testing was completed.

Total oxides of nitrogen  exceeded the emissions criteria for NO2 in one test run.  Since  NO2 was not
directly measured, the vendor has stated that it is possible that the NO2 limit was not exceeded.
Carbon monoxide limits were also exceeded. This was explained by the vendor as attributable to
low combustion efficiencies attained during the testing program.

All metals discharges were well below emissions standards.  Dioxins and furans were not measured
in emissions samples as planned, because of insufficient sample volume.

The vendor calculated DREs for PAHs  using both the B.C. Special Waste Regulation method, and
Alberta Environment's method (which takes into account more of the waste streams). Using the B.C
method for calculation, the vendor reported DREs ranging from 98.41 to 100.00%. DREs of 99.99%
were not achieved for anthracene, fluoranthene, naphthalene, phenanthrene, and pyrene in one or
more test  runs.

In general, all of the water products from the processor (preheat, sour, and  scrubber water) were
found to be contaminated with various compounds that would require treatment before discharge.
Contaminants present in significant concentrations in one or more of the water samples included oil
and grease,  phenols, ammonia, hexavalent chromium, aluminum, lead, manganese, and cyanide.
Naphthalenes were detected in the sour water and preheat water in the parts  per billion range. The
scrubber water also proved to contain  several contaminants in excess of effluent  limits including oil
and grease,  aluminum, arsenic, zinc, manganese, total  suspended solids, cyanide, and  ammonia.

Estimated Treatment Costs

The vendor estimated a unit cost for soils  treatment and disposal of condensed liquids and filtrates
at about $200 (US)  per ton.  The cost assumed treatment of about 27,000 tons of soil, and included
mobilization and demobilization, materials preparation, labor, health and safety, sampling and
analysis, quality assurance/quality control, site preparation, and closure.
                                           369

-------
Discussion

In general, the ATP demonstrated that it was effective at removing organic compounds and cyanide
from the contaminated soils and at meeting the treatability study goals. However, it is not certain
how the pilot-scale processor would have performed had the feed soils not been diluted with sand
and the concentrations of contaminants been higher in the feed.

Emissions standards for metals were met in the study. However, meeting stack emissions for
conventional parameters (SO2, CO, NO2, and particulates) proved problematic for the ATP in this
treatability study.  The vendor has concluded that this is due to the use of old equipment that was
not properly functioning. It is expected that some of these problems could  be corrected by
improvements in the system  (for example, by using properly functioning baghouse filters and a more
efficient scrubber). However, sulfur levels in Sample 1, if they are as high as those found in the
study soils, may require dilution of the soils and/or reduced throughput to meet SO2 emissions
standards, thus increasing treatment costs. Carbon monoxide emissions are indicative of poor
combustion  efficiency. The vendor has indicated that CO emissions could be controlled by adding
an afterburner to the unit. However, it is uncertain whether such a modification will be effective, and
it may increase treatment costs.

ORE calculations were based on measurements of PAHs in the soils and the stack emissions. Due
to the variability of contaminant concentrations in the feed soils, the DREs may be somewhat
uncertain. However, the vendor's ORE calculations show that the ORE performance standard of
99.99% was not met for several of the PAH compounds. This may be partly due to the low
concentrations of PAHs in the feed, which make it more difficult to meet DREs.

Based on data from both the vendor's laboratory and the third party laboratory analyses of oils,
significant concentrations of dioxins and furans were found in the oils.  Since  only very low levels of
dioxins and furans were detected in feed samples (from this and other studies), this may mean that
the ATP is creating dioxins.  However, a mass balance has not been calculated to verify this
hypothesis.  No information was obtained for evaluating the concentrations  of these compounds in
stack emissions.
                              SUMMARY AND CONCLUSIONS
Bench-scale testing of the Chemical Waste Management XTRAX process and the AOSTRA Taciuk
Processor were performed using soils contaminated with PAHs, chlorophenols, lead, cyanide, and
varying levels of oil and grease.  Pilot-scale studies were also performed using the ATP. Based on
the results of the treatability studies, both of .these technologies were effective at removing organic
contaminants from the treatability study soils.  Organics  in the treated soils ranged from
non-detectable levels to the low parts per billion range and were virtually all less than the treatment
goals.

Stack emissions in the ATP were acceptable for metals but were unacceptable for several
constituents (SO2, CO, NO2, and particulates). The use of equipment in poor repair, limited the
extent to which the emissions results from the testing program could be evaluated and used to
predict performance in full-scale operation.

Emissions in the X*TRAX system were not measured in the study, due to the small quantity of soils
processed. Therefore, additional information from full- or pilot-scale operation of the unit would be
needed to assess compliance with any applicable emissions standards.
                                            370

-------
Of concern in thermal extraction systems is the potential for the production of dioxins and furans
when precursors (particularly chlorophenols) are present.  In the ATP test results, concentrations of
these compounds in feed soils and treated soils and ash were insignificant. However, significant
concentrations of dioxins and furans were observed in the condensed oils. In the X*TRAX system,
dioxin and furan concentrations in the feed and treated soil were insignificant and well below
regulatory levels.  No dioxin testing was performed on the condensate phases or the off-gases
because these treatment residuals were not produced in sufficient quantities for analytical testing.
Therefore, the question of whether dioxin and furan generation may occur in the X*TRAX system
was not answered in this treatabiiity study.

In conclusion, based on the results of these treatabiiity studies, both the X*TRAX and the ATP
systems appear promising for the treatment of soils contaminated with manufactured gas plant
wastes and wood preservatives.  However, additional research or data on full-scale operation of
these systems is needed to address the potential production of dioxins and furans, before either
technology is selected for treatment of soils containing precursors.
                                             371

-------
         CF Systems
Solvent Extraction Technology
     For Site Remediation
A Post "Site Demonstration Update"
              372

-------
CF SYSTEMS' SOLVENT EXTRACTION TECHNOLOGY
                     FOR SITE  REMEDIATION
        A POST "SITE DEMONSTRATION UPDATE"
INTRODUCTION

In September 1988, CF Systems' Solvent
Extraction Technology, on a pilot-scale,
was tested and evaluated under EPA's
Superfund Innovative Technology Evalua-
tion (SITE) Program. The final Applica-
tions Analysis Report (EPA/540/A5-90/002)
was issued in August 1990.  This current
paper contains a synopsis of that demon-
stration, an update on recent develop-
ments of the technology,  and a review of
the commercial scale application of the
technology.

CF Systems'  proprietary Solvent Extrac-
tion Technology uses a liquefied gas, such
as propane, butane or carbon dioxide, as
the solvent to extract organics from solids,
sludges and wastewaters. The  solvent,
containing extracted organics, is separated
from the treated product and the solvent
is recovered by distillation in the Solvent
Recovery System, condensed and recycled
to the extraction system.  The extracted
organics are discharged from the system
and typically sent either to a disposal or
recycling facility.

The unique physical properties of a lique-
fied gas, such as low viscosity,  density
and surface tension, result in a solvent
which has a significantly higher rate of
extraction in  comparison to conventional
solvents. Additionally, these enhanced
physical properties accelerate the rate of
settling of the soil/solvent mixture follow-
ing extraction which broadens die applica-
bility of the process to include the treat-
ment of  fine days and silty sediments.
Due to high volatility, liquefied gas sol-
vents are easily and economically recov-
ered from die treated waste matrix by
distillation.  This minimizes the potential
for solvent residues in the treated soils
and aqueous streams as well as in the
extracted organics.

Materials that can be successfully and
economically extracted include:

•  Petroleum-derived hydrocarbons
•  Chlorinated hydrocarbons,  including
   PCBs, dioxins, and pesticides
•  Phenols, alcohols, and organic acids

Areas of applications of this Technology
include:

•  Remediation of contaminated sludges,
   soils, sediment, and fill at Superfund,
   production and disposal sites thus:
   -  eliminating the need for  off-site
     disposal
   -  meeting mandated clean-up
     standards
   -  recovering or separating  organ-
     ics for recycle or disposal

•-  Integration with refining, petrochemi-
   cal, and industrial processes to:
   -  minimize or eliminate hazardous
     waste production
   -  minimize treatment costs
   -  recover product or by-product for
     recycle or resale

•  Removal of organic contaminants from
   wastewater streams  to:
   -  meet regulatory standards for waste-
     water discharge
   -  recover product for recycle
   -  minimize treatment and  disposal
     costs
                                       373

-------
•  On-site treatment of hazardous wastes
   at refineries or petrochemical units to:
   - meet existing and projected stan-
     dards for land disposal
   - minimize waste volumes and dispos-
     al costs
   - minimize liabilities of transportation
     and disposal of hazardous waste

CF Systems is a subsidiary of Morrison
Knudsen Corporation, a $2 billion diversi-
fied services company with primary em-
phasis in the engineering, construction,
and environmental sectors.  CF Systems
supplies engineered solvent extraction
systems and remediation services to the
refining, process, and environmental in-
dustries. Several commercial-scale sys-
tems employing the proprietary technolo-
gy have been built and the technology has
been specified for the remediation of a
Superfund sites.

CF Systems successfully completed its first
commercial Best Demonstrated Available
Technology (BDAT) treatment operation at
the Star Enterprise Refinery in Port Ar-
thur, Texas.  The propane-based solvent
extraction unit processed listed refinery
wastes (F and K) and produced treated
solids that met land-disposal treatment
levels.  The unit was operated continuous-
ly from March, 1991 to March, 1992 and
had an on-line availability in excess of
90%.

The first commercial wastewater treatment
unit was put into service in October, 1991
at the Clean Harbors waste treatment
facility located in Baltimore, Maryland.
This unit is a COo-based extraction unit
designed to treat large volumes of waste-
water containing hydrocarbon contami-
nants.

CF Systems has been selected by the Tex-
as Water Commission (TWC) and the EPA
as the sole-source supplier of a solvent
extraction system for remediation of the
United Creosoting Superfund site in
Conroe, Texas. This unit will employ a
batch process to treat non-pumpable sol-
ids containing poly aromatic hydrocarbons
(PAHs) and dioxins.  The unit capacity
will be in excess of 200 tons per day
(TPD).

SITE DEMONSTRATION
PROGRAM

The SITE program demonstration of the
CF Systems solvent extraction technology
was conducted to obtain specific operating
and cost information that could be used to
evaluate the potential applicability of the
technology to Superfund sites. The dem-
onstration was conducted concurrently
with dredging studies managed by the
U.S. Army Corps of Engineers at the New
Bedford Harbor Superfund site in Massa-
chusetts.

Contaminated sediments were treated by
CF Systems' Mobil Demonstration Unit
(MDU), which had a rating of 20 barrels
per day of sludge feed, on a  once through
basis. The MDU used a liquefied propane
and butane solvent mixture to extract
organics from contaminated sediments.
During the tests, treated sediments were
recycled through the unit to simulate the
design and operation of a full-scale, four-
stage unit.  Recent MDU operation has
employed a batch processing technology,
similar to that now proposed for full-scale
remediation.

The MDU was tested on sediments ob-
tained from New Bedford Harbor, Massa-
chusetts which contained PCB concentra-
tions of 350 and 2,575 parts per million
(ppm). Table 2.1 summarizes the feed
and treated solids PCB concentrations for
two of die tests.

The objectives of the SITE testing included
an evaluation of:

(1)   the unit's performance,
(2)   system operation,
(3)   health and safety considerations,
                                         374

-------
(4)   equipment and system materials
     handling problems,
(5)   projected system economics.

The conclusions drawn from the test re-
sults are given in Table 2.2, with an evalu-
ation of the findings versus current devel-
opments in table 2.3. The primary conclu-
sions were:

   Polychlorinated biphenyl (PCB)
   extraction efficiencies of 90 percent
   were achieved for New Bedford
   Harbor sediments containing PCBs
   ranging from 350 to 2,575 ppm.
   Concentrations of PCBs in the clean
   sediment were as low as 8 ppm.
Some operating problems occurred
during the SITE tests; such as, inter-
mittent retention of solids in system
hardware and foaming in the treated
sediment collection tanks.

Operation of the MDU at New Bed-
ford did not present any threats to
the health and safety of the opera-
tors or local  community.

The projected cost of applying the
technology to a full-scale cleanup at
New Bedford Harbor ranged from
$148 to $447 per ton. The predicted
onstream factor for the full-scale
commercial unit is the variable that
introduces the greatest uncertainty
to the cost estimates.
Table 2.1
Summary of "Site Test Data"
Stage
Number
Feed
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
Stage 8
Stage 9
Stage 10
Test 2 PCB
Cone.
(ppm)
350
77
52
20
66
59
41
36
29
8
40
Cumulative
PCB
Removal %

78
85
94
81
83
88
90
92
98
89
Test 4 PCB
Cone, (ppm)
2575
1000
990
670
325
240
200




Cumulative PCB
Removal %

61
62
74
87
91
92




                                         375

-------
Oi
>J
ON

Performance Test Objective
Removal Efficiency
Foaming
Flow Control
Solids Retention
Extract Quality
Health & Safety
On-Stream Factor
Propane Loses
PreTreatment /PostTreatment
Remediation Costs
Table 2.2
New Bedford Harbor SITE Demonstration Findings
Findings
The technology can separate organics from harbor sediments, sludges, and soils. PCB
extraction efficiencies greater than 90 percents were achieved for New Bedford Harbor
sediments and treatment levels as low as 8 ppm PCBs were attained.
Foaming of treated sediments occurred.
Solvent flow fluctuated widely and caused the solvent to feed ratio to fall below specifications.
Solids were retained in process hardware.
Solids were observed in organic extracts.
No significant releases of pollutants to the atmosphere or surrounding area soils occurred.
Results of the demonstrations tests show that the CF Systems technology is capable of
operating without risk to personnel or the surrounding community.
For the purposes of the economic evaluation an 85 % availability was assumed. This level of
availability was considered optimistic in the Applications report.
Propane losses were not quantified in the evaluation; however, they were reported to be
significant.
Pretreatment (removal of coarse solids) was required to maintain feed sediment particle sizes
below 1/8 inch. Post-treatment such as thermal destruction of the concentrated extract may be
necessary.
The Applications Analysis Report estimated total treatment costs for New Bedford Harbor
(Pile to Pile) to be $450/ton for Hot Spots and $150/ton for the Base Case. These were
considered to be optimistic in view of the required availability to achieve these costs.

-------
Table 2.3
Historical Performance Comparison SITE Program to Present Day
Test Objective Performance/ New Bedford
Design Criteria || Demonstration
Removal Efficiency
Foaming
Flow Control
Solids Retention
Extract Quality
Health & Safety
On-Stream Factor
Propane Loses
Pretreatment / Post-Treatment
Remediation Costs
Test 2: 92 % PCB
Test 4: 97.5% PCB
Caused Poor
Performance
Caused Poor
Performance
Cross Contamina-
tion
Disposal Costs
No Problem -
Design Criteria
85% Estimate
10% Of Feed
Sieving 
-------
 BENCH AND  PILOT SCALE
 WORK

 CF Systems has been operating a bench-
 scale testing facility since 1980 and its
 experience in performing laboratory-based
 feasibility studies for both commercial
 clients and governmental agencies is con-
 siderable. The composition of samples
 sent to CF Systems for testing has ranged
 from dry soils/sludges to sediments and
 wastewaters. These samples have origi-
 nated from a variety of sources including,
 Superfund and RCRA Closure sites, on-
 going waste streams generated from in-
 dustrial operations, and planned remedia-
 tion sites. The results obtained from these
 bench-scale  extractions have been used to
 develop a broad experimental data base
 for the extraction  of organics from hazard-
 ous wastes using  a variety of liquefied
 gases as extractive solvents. These data
 are the foundation for the different sol-
 vent extraction units CF Systems has de-
 signed and built.  Typical bench-scale data
 is given in Table 3.1.

 The MDU, used for the New Bedford
 Harbor study, has been modified to allow
 for operation in a Batch treatment mode
 for non-pumpable wastes and sludges.
 This, and other changes in the system
 design and operation, has resulted in
 significant enhancement of the technology
 for remediation service.

 The following is an outline of some of the
 studies carried out at bench scale and pilot
 scale.

 Massachusetts Superfund Site

 CF Systems  was contracted to perform a
bench scale treatability study to determine
 the feasibility of using liquefied propane
 as a solvent  to remediate PCB-contaminat-
 ed soils from a Superfund site in Massa-
chusetts. This study was done as part  of
 the remedial design phase of work at this
Superfund site, and was performed at CF
 Systems testing facility under a TSCA
 R&D permit granted from EPA Region I.

 The results from this study (Table 3.2),
 indicate that the composite sample (i.e.,
 697 ppm PCBs) was reduced to 0.36 ppm
 after six stages of extraction. The second
 "hot spot" sample, which contained 13,800
 ppm PCBs in the feed, was reduced to 1.5
 ppm after six stages of extraction. The
 percent reduction of PCBs from these two
 soil samples was calculated to be 99.95%
 and 99.99% respectively, and are an indi-
 cation of the high PCB removal rates that
 can be achieved using liquid propane as a
 solvent.

 Star Enterprises Refinery, Port Arthur,
 Texas.

 The pilot-scale Mobile Demonstration Unit
 (MDU) had its initial start-up at Star En-
 terprise's Port Arthur refinery in  Septem-
 ber, 1987. Waste materials processed
 through the unit included  contaminated
 soils from a clay pit, ditch  skimmer
 sludge, and tank bottoms.  The resulting
 treated solids product streams were ana-
 lyzed by Star and the concentration of
 individual components, including PAHs
 and volatiles, were found to be well below
 the regulatory levels established for the
 treated solids.

 PCB-contaminated NPL site in Augusta,
 Maine

 This bench-scale treatability study was
 carried out to determine die feasibility of
 using CF Systems' solvent extraction pro-
 cess to treat five soil and sediment sam-
 ples collected from a PCB-contaminated
 Superfund site in Augusta, Maine. A
 variety of soil types (i.e., clay, fill, and
 sediment) were processed and it was dem-
 onstrated that liquefied propane extraction
 was capable of reducing the PCB  levels to
below 1 ppm.  The overall PCB removal
 efficiency for  this study exceeded 99%,
with PCB concentrations in the treated
 solid as low as 30 ppb.  Results from this
                                         378

-------
bench-scale study resulted in a contract
for CF Systems to conduct a Pilot Scale
Remedial-Design Treatability Study at the
site.

PCB-contaminated Naval Station site in
Adak, Alaska.

The purpose of this bench-scale treatabili-
ty study was to determine the feasibility
of using CF Systems' solvent extraction
process for the remediation of PCB-con-
taminated soils at a Superfund site in
Alaska. Samples were collected from the
site and tested. The results of this study
indicated that solvent extraction utilizing
liquid propane as the extractive solvent
was an effective treatment process for the
remediation of PCB-contaminated soils.
Analytical results showed that an overall
PCB reduction of 99% was achieved in the
study.

Waterways Experiment Station,
Vicksburg, Mississippi.

In April of 1989, CF Systems conducted
a pilot-scale technology demonstration
for the EPA at the Waterways Experi-
mental Station in Vicksburg, Mississippi.
CF Systems utilized its Mobile Demon-
stration Unit (MDU) for this study.  The
purpose of this demonstration was to
develop treatment standards for the
KO48-KO52 listed refinery wastes, and
CF Systems' solvent extraction process
was in fact the
only solvent extraction technology to be
field-tested for that purpose. The results
of this treatability study were used by the
EPA to develop BOAT treatment stan-
dards for five listed refinery wastes.

United Creosoting Superfund Site,
Con roe, Texas

CF Systems conducted a pilot-scale treat-
ability study for EPA Region VI and the
Texas Water Commission at the United
Creosoting Superfund site in Conroe,
Texas. The objective of this treatability
study was to evaluate the effectiveness of
CF Systems' solvent extraction process
for the remediation of soils  containing
creosote-derived organic contaminants,
such as PAHs, PCPs, and Dioxins.
Treatment data from this field demon-
stration showed that the overall concen-
tration of PAHs were reduced by 95.7%
in the two soil samples, and subsequent
full-scale treatment data indicates that
PAH reductions of > 99% are possible
through the optimization of extraction
operating conditions.  Based on the
results of this pilot-scale study, CF Sys-
tems' critical fluid  solvent extraction
process was selected for full-scale remed-
iation at the United Creosoting Super-
fund site.
                                         379

-------
OJ
00
o
TABLE 3.1
SELECTED CF SYSTEMS TREATABILITY STUDIES
Concentration (PPM)
Client
Texas Water Com-
mission
Metcalf & Eddy
Occidental
Chemical
Central Maine
Power
U.S. Navy
NY Dept. of Envir
Conservation
Metcalf & Eddy
Kerr McGee
Site
United Creosoting
Superfund Site
Pine St. Canal
Superfund Site
No. Tonawanda
O'Connor Superfund
Site
Adak Naval Air Station
Booth Oil Site
Norwood Superfund
site
Wood Treating Site,
OK
PCOC's
PAHs
PCPs
Dioxins
PAHs
Dioxins
PCBs
PCBs
TPHs
PCBs
TPHs
PCBs
Dioxins
PAHs
PCPs
Feed
2324.0
33.9
312.0
240.
23. ppb
113.
305.
7850.
12.4
40.3
13,800.
7.068 ppb
In progress
After Removal
Treatment | Efficiency
51.7
5.59
14.8
16.
0.6 ppb
3.4
2.2
34.
<0.24
0.24
1.4
0.09924 ppb
In progress
97.7%
83.5%
95.3%
93.4%
97.4%
97.0%
99.3%
99.6%
>98.%
99.4%
>99.9%
>99.9%


-------
Table 3.2
PCB Contaminated Superfund Site
Solvent Extraction Results
Sample
PCB in Feed
PCB After
Extraction Stage
No.
1
2
3
4
5
6
Composite
Sample
(ppm)
697.0

14.0
1.4
2.9
0.2
0.07
0.3
Cumulative
% Removal


97.99
99.80
99.58
99.97
99.99
99.96
Hot Spot
Sample
(ppm)
13800

178
26.8
10.3
5.3
2.6
1.4
Cumulative
% Removal


98.71
99.81
99.93
99.96
99.98
99.99
COMMERCIAL-SCALE
PROCESS

CF Systems' proprietary Solvent
Extraction soil treatment process involves
the use of liquefied gases, such as
propane, as the solvent to extract and
separate organic contaminants from soils
and sludges.  The unique physical
properties of a liquefied gas, such as low
viscosity, density, and surface tension,
result in a solvent with significantly
higher rates of extraction in comparison to
conventional solvents. Additionally, the
enhanced physical properties accelerate
the rate of gravity settling of the
soil/solvent mixture following extraction
which broadens  the applicability of the
process to include the treatment of fine
clays and silty sediments. Due to high
volatility, liquefied gas solvents are easily
recovered from the waste matrix which
minimizes the potential for solvent
residues in the treated soils.

CF Systems has successfully completed
the first commercial on-site BOAT
treatment operation at the Star Enterprise
Refinery in Port Arthur, Texas.  The
propane-based solvent extraction unit
processed listed refinery K- and F-wastes
and produced treated solids that met land-
ban requirements. The unit operated
continuously from March 1991 to March
1992, with an on-line availability in excess
of 90%. Following fixation for heavy
metals, the treated solids were disposed of
in a Class I landfill. During operation,
100% of the feed material was treated to
meet the land-ban specifications.  Multiple
feeds,  including API separator solids, slop
oil emulsion solids, slop oils, and
contaminated soils were treated. Typical
                                          381

-------
 treatment results for a refinery waste are
 given in Table 4.1.

 The first commercial wastewater treatment
 unit was put into service in October 1991,
 at the Clean Harbors waste treatment
 facility in Baltimore, Maryland. This unit
 is a CO2-based extraction unit designed to
 treat large volumes of wastewater
 containing hydrocarbon contaminates.
 Recovered hydrocarbons will be disposed
 of in an incinerator while treated
 wastewater  will be discharged to  the local
 POTW.

 CF Systems has been selected by the
 Texas Water Commission (TWC) and the
 EPA as the sole-source supplier of a
 solvent extraction system for remediation
 of tlie United Creosoting Superfund site
 in Conroe, Texas.  This unit will employ a
 batch process to treat non-pumpable
 solids containing poly aromatic
 hydrocarbons (PAH's) and dioxins. Unit
 capacity will be in excess of 200 tons per
 day.

 Process Description

 The CF Systems solvent extraction process
 for remediation of organic-contaminated
 soils and sediments is comprised  of the
 following systems:
   A Feed Delivery System
   An Extraction/Gravity Settling System
   A Treated Solids Filtration System
   A Solvent Recovery System
   A Vent Gas Recovery System
Feed Delivery

Soil delivered to the process battery limits
is screened to less than 1/4 inch to remove
oversized material.  Oversize materials are
segregated and reprocessed to an accept-
able size through the use of mechanical
size reduction equipment.  Oversized
material can also be washed to remove
attached "fines' and surface
contamination.  The screened soil is sent
 to the extractor(s) via an enclosed
 conveyor/screw auger system.  In most
 cases, die process does not require other
 forms of soil pretreatment, such as
 dewatering or the addition of chemical
 reagents.

 Extraction System

 The extraction system is comprised of one
 or more agitated extraction vessels where
 contaminated soil is contacted widi
 liquefied propane.  The number of
 extraction vessels and die size of die
 vessels and agitators determine the rate of
 throughput and die degree of organics
 removal.  Exact configuration of the
 extraction system is determined during
 the bench test phase of die project.

 Contaminated soil is fed to die extractors
 where it is contacted with liquid solvent
 pumped from die Solvent Recovery
 System. Following die extraction step,
 the agitators are stopped and settling of
 the soil from the solvent occurs.  Settling
 of soil from the solvent occurs rapidly in
 the extraction vessel due to die enhanced
 physical properties of the solvents.
 Following settling, die solvent/organics
 phase is drained to the Solvent Recovery
 System where the solvent is recovered.

 The extraction-settling-draining process is
 repeated in the extraction vessel until die
 extraction is complete. The final step
 involves injecting water into the extraction
 vessel in order to displace residual liquid
 propane, which is insoluble in water, and
 is displaced from die top of die extractor.
 This final water displacement step in die
 extractor vessel forms a treated soil/water
 slurry that is fed to a belt filter press
 operation.

Filtration System

This system includes the soil/water slurry
day tank, belt filter press, and all required
drums and pumps for routine filter press
operation.  Treated filter cake from the
                                          382

-------
filter press has high compressive strength,
and is suitable for land disposal. Treated
filter cake also has the appropriate
moisture content for the addition of
solidification/stabilization reagents for the
fixation of metals, if required.

Solvent Recovery System

This system contains an extract surge
vessel, the main solvent recovery still, a
solvent condenser and a solvent recycle
pump.  The solvent/organics mixture
(extract) flows from the extract surge drum
to the main solvent recovery still,  where
the solvent is vaporized using steam, hot
oil or the heat from a vapor recompression
cycle.

Solvent vapor from the still is condensed
and flows to the solvent storage drum.
The solvent recycle pump takes a gravity
suction from this drum and pumps liquid
solvent, on demand, to the extraction
system.

From the main solvent recovery still
reboiler, organic-rich extract flows to a low
pressure organics recovery drum,  where
residual solvent is removed from the
extracted organics and sent to the Vent
Gas Recovery System.  Recovered
extracted organics are sent to storage and
disposal.

Vent Gas Recovery System

A low pressure compressor recovers low
pressure solvent vapor from the organics
recovery drum, recovered organics storage
and from low pressure vents from the
treated soils system. The compressed
recovered solvent is returned to the main
still for recycling.

CF Systems' Solvent Extraction: Benefits

The primary benefit of using CF Systems'
solvent extraction process at large scale
organic-contaminated sites is low
treatment costs.  The low operating (i.e.,
utility) requirements for this process
results in treatment costs in the range of
$100 to $350 per ton at sites containing
greater than 10,000 cubic yards of soil.

The process has been commercially proven
at Star Enterprises where it operated for
over one year, 24 hours a day, seven days
per week, with an on-line availability
exceeding 90%.

The systems is totally enclosed, with no
process vents or emissions, thereby
reducing the level of complexity associated
with obtaining required permits.

No chemical additives or reagents are
required  for this process, and the organics
are extracted and recovered unchanged
and unreacted. For most applications, the
pretreatment requirements are minimal,
consisting of size reduction to less than 1
inch particle size diameter.

The process is operated at ambient or
near-ambient conditions, thereby
eliminating die potential for the formation
of toxic compounds and their subsequent
release to the environment.
                                         383

-------
Table 4.1
TYPICAL TREATED SOLIDS - REFINERY WASTE
Compound
Volatiles
Benzene
Ethylbenzene
Toluene
Xylene
Semi-Volatiles (PAH)
Naphthalene
Phenanthrene
2-Methyl Phenol
Anthracene
Benzo(A)-
Antliracene
Pyrene
Chrysene
Benzo(A)-
Pyrene
Phenol
4-Methylphenol
Bis(2-E.H.)
Phthalate
Di-n-Butyl
Phthalate
Typical Feed
(mg/kg)

65
130
250
480

350
450
10
37
5
35
19
5
10
10
10
ND
Product
(mg/kg)

<1.0
<1.0
<1.0
1.5

1.8
<1.0
ND
<1.0
ND
<1.0
ND
ND
<1.8
ND
ND
. ND
Percent
Removal

>98
>99
>99.4
99.7

99.5
99.8
N/A
>97
N/A
>97
N/A
N/A
>90
N/A
N/A
N/A
BOAT
(mg/kg)

14
14
14
22

42
34
6.2
28
20
36
15
12
3.6
6.2
7.3
3.6
384

-------
COMMERCIALIZATION OF INNOVATIVE TECHNOLOGIES -
           CHALLENGES AND SOLUTIONS

          (A Technology Vendor's Perspective)
Frederic A. Eidsness, Jr., Director of Government Relations
      Michael J. Taylor, P.E., Senior Vice President
           Alistair H. Montgomery, Director -
       Corporate Waste Treatment Technologies

         Canonie Environmental Services Corp.
         6300 South Syracuse Way, Suite 300
            Englewood, Colorado  80111
                  (303)  290-8336
                        385

-------
            COMMERCIALIZATION OF INNOVATIVE TECHNOLOGIES-
                        CHALLENGES AND SOLUTIONS
                      A Technology Vendor's Perspective
 INTRODUCTION

 Commercializing innovative technologies for the remediation of hazardous waste is a
 task fraught with obstacles and challenges.  Experience in the field indicates that
 many of the obstacles and challenges can be overcome when the industry, including
 owners, regulators and  purveyors of innovative technologies,  work  together.
 However, should the  various parties continue to operate from their own isolated
 spheres without considering the others' needs, goals and constraints, then the use of
 innovative technologies will be stifled and new cost-effective solutions will not be
 applied to the hazardous waste problems of today and the future.

 Even with more harmonious institutional relations, the current mind-set of the federally
 driven "command and control" regulatory program for cleaning up contaminated sites
 thwarts the commercial development and application of innovative technologies. That
 mind-set is the false assumption that complex sites  with complex wastes can be
 adequately understood based on "studies" that justify a selected design of a total site
 remediation system that will work. Consequently,  hazardous waste generators are
 being asked to sign on up-front for implementation of a  "total solution" with imperfect
 knowledge, unproven technology and astronomical costs. These risk-posing factors
 flow downhill from site owners to remediation contractors and vendors in the form of
 contracts, hence, the deterrent to owners, contractors and vendors to invest in the
 development and use of innovative waste treatment technologies.
                               i
This paper discusses several key obstacles which emerge from this mind-set that must
 be overcome to foster innovative technologies in the market place.  Solutions to
                                     386

-------
overcoming these obstacles are also presented. These obstacles and solutions are
based  on  reaMife experiences of bringing,  or attempting  to bring,  innovative
technologies to the commercial level in the hazardous waste industry of today.

With a co-operative spirit, a more workable approach to remediating sites and a few
changes  in the  contractual  methods  of handling these endeavors,  innovative
technologies can be developed and commercialized to achieve more cost effective
solutions to our hazardous waste problems.
                                     387

-------
 OBSTACLES AND APPROACHES TO SOLUTIONS

 Predicting Performance (or Lack Thereof)

 Comprehensive long-term performance of innovative technologies cannot be predicted
 with any degree of certainty for the complex waste  and subsurface conditions that
 exist at hazardous waste sites. Yet innovative technologies often may represent the
 most cost-effective method of cleaning all or portions of the site. However, in today's
 regulatory approach, agencies are asked to chose and owners are asked to accept a
 technology  based  on  site investigatory information  and  to  guarantee  that the
 technology  will work to clean the  site to  often unrealistically  low  levels  of
 concentration.  Without the certain predictability, the innovative technologies often
 do not make the first cut. Even  if the technology is selected, the  purveyors of the
 technology cannot predict long-term performance to the  unrealistic low levels for the
 ill-defined complex conditions  based on the limited  investigatory  site  information.
 They therefore do not bid the work, or have to add such high contingencies that the
 technology does look financially attractive.   Only if they are allowed to apply their
 technologies in a step-wise sequential  remediation approach with near-term realistic
 goals can they successfully apply and  commercialize their technologies.

 Reducing Microscopic Regulation

The  current  ad-hoc system for  determining residual waste stream performance
standards during construction impedes innovation. The  characteristics and volumes
of residual wastes generated by the application of treatment technologies during the
construction  phase are not  predictable  at bench-scale or even pilot-scale testing
because of the varying  site conditions and waste characteristics encountered as site
remediation proceeds.  Yet, engineering designs attempt to define in absolute terms
the performance standards and emission rates that must  be met during construction.
                                      388

-------
 When remediation  is undertaken under the Superfund program where  Potentially
 Responsible  Parties (PRP) have assumed the lead role,  residual  waste  stream
 performance standards and rates are typically incorporated into Remedial Action Plans
 which themselves are considered a part of Consent Decrees and Records of Decision.
 Therefore, any change in performance standards resulting from site conditions during
 full-scale remediation may require agreement of all the parties to the decree including
 the EPA,  Department  of  Justice, State and  PRPs. When  Superfund projects are
 financed by  the Fund  and construction administered by the U.S.  Army Corps of
 Engineers (COE) under an  agreement  with EPA, contractors and technology vendors
 are faced  with convincing the COE.  For historic, cultural reasons, the COE tends to
 treat  remediation projects as standard civil engineering public works project where
 engineering designs tend to be viewed as sacrosanct.

 Within this ad hoc decision framework, risks and uncertainties abound. Vendors are
 expected  to  warrant that they can  meet these prescribed  emissions  performance
 standards and limitations. Failure to  meet them usually results in shutdowns and
 protracted negotiations. Enforcement actions and penalties are a possibility.

The market  responds to these risks  and  uncertainties  in the only way  it can, by
increasing contingencies and bid prices, taking the chance the problem will not arise
and fighting it out with the owners and regulators (the change order route) if it does,
or declining  to bid  the technology.  None of these alternatives  is a good  option
financially or in terms of developing  and  maintaining good vendor and technology
reputation in  the industry.

A decision protocol is needed that will allow performance standards to be established
based on  full-scale proof-of-process testing on  site wastes that will assure the
technology will operate to its best capability and assure that the off-site public is
protected and the on-site condition does not worsen. Such a decision protocol should
                                    389

-------
allow technologies to be operated uninterrupted from  proof-of-process through to
completion, provided that the off-site public is protected.

Risk Sharing

To realize savings through the use of cost-effective, innovative technologies, the
regulators, the owners of the site and the vendors of the technologies must share the
risk as well as the benefits. Each of these entities needs to understand that it cannot
regulate, contract and/or receive payments as would when more developed, mature
technologies are used.  The risk sharing and the rewards must be fair and equitable
to all concerned. Regulators must accept the premise that as long as the short-term
treatment activities do not pose an unacceptable risk to  the off-site public or worsen
the on-site condition, the technology should be allowed to operate.

Institutional Attitudes and Mechanisms

In our industry today there is a tendency to think that the key to any problem, in this
case the commercialization of innovative technologies, has at its core an "engineering
solution." In reality, the problems are institutional and, once resolved, the engineering
component, that which most of us are trained to concoct, can be brought to the table.

The main barriers to commercialization of innovative technologies are, at their core,
institutional  problems. In summary, they include: regulator's "command-and-control"
and enforcement mentality; regulators, owners and contract administrators penchant
to view the use of innovative technologies as  no different than standard civil
engineering  projects;  tendency  to pass  on  risk to remediation  contractors and
technology vendors; and, a mind-set "letting the perfect be the enemy of the good"
in terms of designing solutions to complex hazardous waste sites.
                                      390

-------
 Canonie Environmental Services  Corp.  (Canonic)  has had the  most success  in
 commercializing technologies where the risks have been shared and the institutional
 mechanisms for the project have been well developed and mutually understood by the
 parties involved. Case studies discussed herein include such projects as the McKin
 Super-fund site  where the Low Temperature Thermal Aeration (LTTA®) Technology
 was originally developed and commercialized; the Gould Superfund site where the
 patented Waste Battery Treatment System is being commercialized; and the San Jose
 Superfund site where the Accelerated Vacuum Extraction Process was developed and
 applied on a commercial scale.

 Canonie also has commercialized innovative technologies on projects where the risk
 sharing and the  institutional mechanisms were less than desired and the results have
 not been satisfactory.  Even when the technology was technically successful, the
 financial  burden on the vendor and the schedule  of  performance have been so
 adversely effected  that participation on  future projects of the  type become very
 unattractive and commercialization of the technology is severely curtailed. Such was
the case on the Wide Beach Superfund Site and the Waukegan Harbor Superfund Site
 where the SoilTech Anaerobic Thermal Processor (ATP) Technology was applied in
two different modes for the first commercial applications of this technology.
                                    391

-------
COMMERCIALIZATION EXPERIENCES - CASE STUDIES

Case Study #1 - McKin Superfund Site
  The  Mckin  Superfund site  provides  a good example  of how  innovative
  technologies can be applied at a commercial level while still being developed, if
  the parties are cooperative and the risks are shared^ This project was done during
  the early days of the Superfund program when such innovation and risk sharing
  were common. Low Temperature Thermal Aeration fLTTA^J was the innovative
  technology developed and used in the successful cleanup.  At that time,  no
  commercial thermal aeration systems existed.

  The McKin Superfund site was formerly a liquid waste storage, treatment and
  disposal facility for volatile organic solvents, chemicals and heavy oils. As a
  result of improper operation, the soils and ground water were impacted by VOCs
  and oils.  The site was also subject to a high degree of public exposure because
  it was within 300 feet of several residences. This site was ranked No. 32 on the
  National Priorities list at the time of the Record of Decision (ROD) in  1985.

  The ROD had called for the soi/s to be excavated and disked on the surface to
  remove the VOCs prior to being backfilled.  However, the VOC concentrations
  were to be kept below a certain level at the site boundary.  Disking to remove
  VOCs would certainly have caused exceedances at the boundary and would have
  taken an unacceptably long time to complete.

  Canonie,  while working with one  of the main PRPs, realized that ex-situ
  processing was the only way to accomplish the intent of the ROD without
  causing unacceptable emissions.  However, to design, build and permit a system
  would have caused the project to go  far beyond the specified completion date,
  resulting in substantial costs, and still would not result in a proven commercial-
  level treatment technology.

  Canonie was aware of an existing asphalt aggregate dryer that had an air permit
  to work in the area  of the McKin Site. A technical review indicated that the heat
  and mixing action in the dryer would vaporize the organics in the soil and clean
  the soil to meet the specified criteria. The vapors could not, however, be released
  into the air.  Therefore, to capture the vapors and assure that the provisions of
  the dryer air permit were not violated,  Canonie engineered an extensive and
  innovative air treatment train using primarily existing air treatment components
                                     392

-------
   from other projects or commercially available, including vapor phase carbon and
   slurry scrubber. This combination of existing and engineered systems was the
   initial formulation of the patented Canonie L TTA * system.

   At the  time the LTTA® treatment train was proposed to  the owners and :th^
   agencies on the McKin project, there was no guarantee that the system would
   clean the soils,  would not violate emission standards at the site boundary or
   would not cause microscopic emission excursions of the internal components of
   the system: Yet the system offered the owners the potential for substantial cost
   savings and the ability to meet the project schedule. All that was needed was an
   agreement to share some of the  risk of applying the system at full production
   level and a cooperative spirit in entering the project.

   No threat to the residents living near the site could be tolerated,  and they had to
   be part of the  overall plan.  Consequently, a  door-to-door campaign  was
   implemented to discuss what was to occur, what precautions were going to be
   taken and what the risks and rewards (i.e., the cleanup of the site)  would be.
   The public supported the total program.

   The system was mobilized and test runs were conducted.  The system  was
   adjusted to improve performance. Full production was implemented under the
   monitoring of emissions and treated soils by the regulatory agencies which was
   conducted at the boundaries of the site and not at the microscopic emission
  points at,  or internal to, the equipment.  Equipment performance criteria for less
   innovative types of technology were not imposed on the unit (i.e., the agencies
   did not try to apply incineration performance criteria merely because this was a
   new type  of thermal device).  The owners, in cooperation  with the equipment
   developer/vendor, monitored the costs.

   The project was  the first NPL site cleanup to be successfully completed in  EPA
  Region L The remediation was used by the EPA as an example of success of the
  Superfund program.  The owner received a clean site at less than the original
  estimated  cost and the contractor was adequately reimbursed for his efforts.
  Most important, the public was the all-around winner.

Case Study #1 (Continued) - McKin  Superfund Site
                                    393

-------
Case Study #2 • Wide Beach Superfund Site
  The application of the innovative SoilTech ATP Technology to remediate PCB-
  contaminated soiis at the Wide Beach Superfund Site represents an case wherein
  technical success can be achieved in spite of non-cooperative behavior on the
  part of some of  the parties, provided, however, that the developer/vendor is
  willing to take a large portion of the financial and performance risk.

  The Wide Beach Superfund site  was a residential area on the shores of Lake Erie
  near Buffalo, New York, where PCB-contaminated oils had been used for dust
  suppression on community roads. The PCBs existed throughout the roadways as
  well as in driveways and lawns.  The ROD called for the PCB-contaminated soils
  to  be excavated and  treated  with a dehalogenation process, a  new and
  developing technology.  The original estimated quantity to be treated was nearly
  20,000 tons. The project was managed by the COEas the Contracts Manager for
  the EPA.

  Even though dehalogenation had never been accomplished at a commercial level
  anywhere in the world, the COE wrote a standard performance specification and
  solicited unit-price  hard bids for the work.   The successful bidder for the
  treatment portion of the contract was a technology developer/vendor of a batch
  process for dehalogenation which had been doing some test  work on the site.
  The technology was, however, not developed much beyond the bench-scale, or
  at  most the pilot-scale level,  and application of this process to  full-scale
  remediation was speculative at best.  The technology developer was supported
  in  the bidding and contracting efforts by the prime contractor, who was
  responsible for the excavation and other miscellaneous work.

  When the prime contractor began to realize the excessive risks he had assumed,
  he  began to seek alternate solutions.  He discovered that Canonie, through its
  subsidiary company, SoilTech, had a technology developed to commercial level
  that would desorb the PCBs from the soil. The addition of the dehalogenation
  capability looked feasible, but was not proven. Desorption was a developing
  innovative  technology already,  and adding the dehalogenation process made
  matters even more innovative and speculative.

  Since the contractual and bidding mechanism for this project was not geared
  toward risk sharing and application of innovative approaches, the prime contractor
  could only apply the SoilTech ATP Technology with dehalogenation systems
  added by submitting a Value Engineering Proposal with fixed unit prices. SoilTech
  was asked to supply these prices and to take all the risk that
                                     394

-------
  performance would be successful. There was no mechanism for nor willingness
  by the COE to allow any risk sharing or to allow adjustments for changing site
  conditions,  equipment performance  on this innovative application,  or other
  adjustments typically required to respond to the  needs of applications of
  innovative technologies at the commercial level on a first-time basis.  The COE's
  attitude was (and continues to be) that the developer and vendor should take all
  the performance and financial risk.

  On the Wide Beach project, Canonie and SoilTech  elected to take the risk.
  Canonie/SoilTech had to agree that no compensation would be received for the
  mobilization and testing if the processing was not successful in this first-time
  commercial level application. This was a gamble of over $ 1 million, not something
  a small company with a new innovative technology would be likely, or able, to
  take.

  The gamble proved to be a technical winner. The SoilTech A TP Unit successfully
  removed and dechlorinated the PCBs in the soil and provided soils with non-
  detectable levels of PCBs after treatment.  The gamble, however, turned out to
  be  a financial disaster due to the rigid contractual position of the COE and its
  inability and unwillingness to share in any of the risk.

  As with most Superfund work, the site characterization was valid to define the
  extent of the contamination,  but was  inadequate in defining the parameters
  necessary to apply and adequately price the application of the  treatment
  technology,  especially an innovative treatment technology.

  When excavation began, it was discovered that the majority of the  soils to be
  treated were very  clayey in  nature, more so than  expected  from  the site
  characterization. Even though the equipment had to be adjusted significantly and
  the production rate had to be slowed (with attendant cost impacts) to accomplish
  the technical success {not an unusual occurrence when applying  innovative
  technology at the commercial level for the first time), the COE would not consider
  an adjustment in the quoted unit price. The developer/vender of the technology
  (Canonie/ SoilTech) was told  to keep  working  regardless of the  costs and
  difficulty.
Case Study #2 (Continued) - Wide Beach Superfund Site
                                    395

-------
  The project continued, and the quantities of treatment doubled; again, not an
  unusual occurrence at a CERCLA site.  The innovative application of the SoilTech
  A TP Technology with dechlorination capabilities continued to perform well and:
  cleaned over 42,000 tons of soils to non-detectable levels. This was the first
  commercial application at this scale world-wide and was hailed by EPA and the
  scientific community as a major success.  {The technology received a national
  award in the Association of Bay Area Governments for the best innovative tech-
  nology.) However, the COEhas denied all claims and continues to administer the
  contract as if the work were standard ordinary construction work,

Case Study #2 (Continued) - Wide Beach Superfund Site
                                    396

-------
Case Study #3 - Gould Battery Superfund Site
      Gould Battery Superfund site in Portland, Oregon, is another example of how
  innovative treatment technology can be developed and applied through joint
  cooperation among the owners, the regulators and the technology vendors. This
  project is a first-of-a-kind and has significant potential for the remediation of
  similar sites nation-wide.

  Automobile batteries were recycled at the site over a period of 32 years.  Lead,
  copper and  other  metals  were  extracted from the batteries  by crushing,
  separation and smelting methods.   The waste materials, principally battery
  casings and furnace products (matte, slag, etc.) were used as fill material both
  on and around the  property.  The site was placed on the NPL in 1983  and a
  Remedial Investigation/Feasibility Study (RI/FS)  was completed in 1988.  The
  ROD issued in 1988 specified excavation, treatment and recycling of fill material
  as the preferred remediation  method.    Specifically, the  selected  remedy
  contemplated recycling plastics, hard rubber (ebonite), metallic lead and lead
  oxide/lead sulfate products from the process.  The feasibility study, however,
  failed to demonstrate that the requirements of the ROD could be achieved with
  any known  treatment,  technology,  including processes then  being  used
  commercially for automobile battery recycling.

  At this point in the project, the owner and the EPA entered into a Consent Decree
  to conduct predesign studies to collect information on the  treatment, recycling,
  and health and safety aspects of the project. Canonie was retained to perform
  this work, which resulted in the successful development of innovative technology
  for the treatment of the wastes on the Gould Superfund site.

  In electing to pursue the predesign studies route, the EPA  avoided the pitfall of
  specifying a remediation approach  which could not be met  with existing
  technology and which  would have resulted in a costly  failure for both the
  regulators and the owners.  Realizing there was no precedent for remediating
  abandoned battery wrecking sites, Canonie sought to develop a  treatment
  approach which would minimize the risk associated with technology innovation.
  Canonie accomplished this by characterizing the site from a treatment rather than
  contamination standpoint. Existing unit processes, used previously in the mining
  industry, were selected for the conceptual process flowsheet, but configured in
  such a way as to meet the site regulatory requirements for the remediation. The
  process developed by  Canonie hot  only  produced  the  required recyclable
  products, but addressed water treatment and off-site disposal issues byreducing
                                     391

-------
  or eliminating toxic chemicals from the process. Public health exposures to dust
  and air borne lead emissions were mitigated by developing and incorporating a
  wetting agent/dust control system which is totally compatible with the process.

  This approach resulted in a significant reduction of the remediation cost over
  projections made during the R//FS phase*   Most importantly, the treatment
  technology was developed through the stages of bench-scale, pilot-scale and on-
  site demonstration.                  •    •

  In the Gould project the risks were shared in some proportion by the participants.
  The owner paid for the technology development, but may receive revenue from
  its application on other sites.   The EPA  took the risk that a process and/or
  approach  could be  developed to meet the  requirements of  the ROD.   The
  technology developer (Canonie) took the risk that the remediation project  could
  be completed at a cost substantially below the amount projected in the RI/FS.
  By performing the work on a target value basis, Canonie  had the incentive to
  complete the project at even  lower costs if possible.

  The confidence built among the participants during the predesign studies phase
  was sufficient to persuade the owners to negotiate the remediation contract with
  Canonie on a sole-source basis.  This relationship  continues to strengthen during
  the construction phrase of the project through the participation of the owners,
  regulators, and  the contractor in  resolving problems  as  the  remediation
  progresses.

Case Study #3 (Continued) - Gould Battery Superfund Site
                                     398

-------
Case Study #4 - Waukegan Harbor Superfund Site
  The SoilTech ATP Technology applied at the Waukegan Harbor Superfund site
  represents an example of a technical success for an innovative technology, but
  where the variable nature of the site conditions and where micro-monitoring of
  the equipment operation by the agencies caused unwarranted financial burden on
  the developer/vendor of the  technology.
  The sediments of the Waukegan Harbor and an adjacent stream were heavily
  impacted with PCBs with concentrations in excess of 20,000 parts per million
  (ppm).   This site ranked No. 86 on the National Priorities List.  Remediation
  alternatives were limited because the harbor was located adjacent to a public
  beach and was used for both recreational and commercial boating activities.

  The selected remedy was to excavate or dredge the contaminated soils, then
  remove 97 percent of the PCBs and dispose of the resultant soils in an on-site
  vault. Innovative technologies were required to remove the PCBs from the soils
  since, at the  time of the ROD, incineration was the only method used  for
  treatment of PCB contaminated soils. Solvent extraction and thermal desorption
  were considered. The SoilTech A TP Technology thermal desorber was selected.
  Even though  SoilTech  had not been applied commercially  and, in  fact, a
  commercial transportable unit did not exist, the technology was well advanced
  with operating pilot units of similar size to the planned transportable unit.     ;

  Canonie agreed to a lump-sum contract to perform the excavation and dredging
  and to treat the estimated 19,000  tons of material by thermal desorption using
  the SoilTech A TP Technology. On the basis of this contract, the transportable
  unit was manufactured at a cost of more than $5 million, to be immediately sent
  to the Waukegan site to process the soils.

  A provision was included in the work plan at the insistence of the agencies that
  stack emissions had to meet "six-nines" Destruction Removal Efficiency (DRE) for
  PCBs, a requirement that was not applicable to  this technqlogy {but one the
  agencies felt comfortable with) because the SoilTech ATP Unit did not destroy
  PCBs and the  emissions volume was orders of magnitude below that of an
  incinerator. However, the agencies were adamant and the developer/vender of
  the technology had little choice but to accept this stipulation in  the work plan if
  the project was going to progress. There were some indications from bench-scale
                                     399

-------
   work that this  could be achieved and, if it was close, the developer/vender
   assumed that the agencies would be reasonable, given the difference between
   this  technology and the incineration  technology for which  the  criterion was
   established.         •••••-  '••'-:/•"»'^h-^^:-*:-••;,-'•• •K^y^^--^"^- ••••: -,   ..i^xA-,^^..,.,
   Within the first few days  of excavation  at the s/te; unanticipated PAH-
   contaminated soils were discovered and the  whole project was stopped so the
   agencies could determine what to do. The resultant delays lasted for well over
   two years.  Meanwhile, the technology developer/vendor was sitting with a $5
   million piece  of equipment having to pay debt service   with  no income.
   Notification of delays were given to the owner and delay claims are  being
   discussed, but this did not help the technology developer with  the difficult
   financial conditions caused by the delay.  Eventually, when the extent of the
   delay was finally clear, SoitTech was able to find another interim project, but not
   without unanticipated costs and interruptions in the overall commercialization plan
   for the technology.  Without strong financial backing, this  initial perturbation
   could have resulted in the end of this technology's commercialization.

   Eventually the Waukegan project was once again under way. The SoilTech ATP
   Unit  was mobilized to the site, test  work was done  and production level
   processing began.  The soil types and water content were ideal for the unit and
   production progressed at rates near the name plate capacity.  The soils clean-up
   criterion was near 500 ppm PCB.  The cleaned soils had levels of less than 4 ppm
   maximum and often were non-detectable.  The project was going to  be a
   technical, schedule and financial success,  when the agencies again brought the
   whole process to a halt for failure of the A TP Unit to demonstrate the requisite
   ORE of 99.9999 percent after repeated tests (the ATP was demonstrating
   99.9996 percent ORE).

   Risk assessments conducted by EPA indicated that the health risk at the boundary
   of the site was less that Iff9, a level two orders of magnitude below requirements
   cited in the Site Health and Safety P/an. The owner's experts determined that the
   amount of natural volatization ofPCBs from stockpiled waste awaiting processing
  in the ATP would exceed the  mass of emissions  that would result  from
  completing the project with a demonstrated ORE of 99.9996 percent The site's
  citizens advisory  committee  was  appraised of  the  situation and  voted to
  recommend completion of the SoilTech portion of the project since,  at the time
  of shutdown, more than 50 percent of the  contaminated material had  been
  treated.                                         '  "    '  -
Case Study #4 (Continued) - Waukegan Harbor Superfund Site
                                    400

-------
  Faced with these arguments, EPA allowed the project to restart and proceed to
  completion. However, it choose the lengthy administrative remedy of modifying
  the Consent Decree to relax the emissions standard to 99.999 percent DRE. TJhis
  required agreement on the part of the owner, the State\,ot[MicJifgan and the
  Department of Justice.     '" """ '' S:f:^;S::i::<; S^^fe^SK'^^i!:^?^'^!-^^:

  The decision to relax the emissions standard was a difficult one for the EPA
  Region as this Region has a commendable track record of successful pollution
  control by imposing strict "technology-based" standards and, in the 1970s
  established the  knowledge base within EPA that allowed EPA to promulgate
  nationallyenforceable Best Available TechnologyEconomically'Achievable effluent
  guidelines for industrial discharges  under the National Pollutant Discharge
  Elimination System  (NPDES) program  in the early 1980s.   However,  the
  administrative remedy of modifying  the Consent Decree proved to be more
  complicated and time consuming than anticipated.

  During the negotiation process, however, the equipment and labor were sitting
  idle. Frantic efforts were underway by the on-site crews supported by technical
  staff in an attempt to make the equipment meet the criterion so operations could
  continue.  Carbon beds were added, dehalogenation agents were tried and other
  near-research-level attempts were incorporated. None of these efforts improved
  the performance of the unit  to  clean  the soils,  added to public health or
  environmental protection or in any way helped the remediation of the project.

  Finally, approximately two months after the processing was halted and before the
  criterion change process could be culminated, the developer demonstrated a DRE
  of 99.99999 percent, accomplished in cooperation with the Region and EPA's
  Risk Reduction Laboratory in Cincinnati, Ohio, by additions and adjustments to
  the equipment  and plugging a hole in  the  carbon bed that had remained
  undetected during previous attempts at improving air emission controls.

  The cost to the developer/vender was, however, over $500,000, the remediation
  of the  site gained nothing, and the developer/vendor missed a window of
  opportunity for another project.

Case Study #4 (Continued) - Waukegan Harbor Superfund Site
                                    401

-------
Case Study #5 - San Jose Superfund Site
  The development and application of a commercial-level Accelerated''Vacuum'.
  Extraction System at the San Jose Superfund Site is a good example of the way
  an innovative technology can be effectively commercialized by risk sharing among
  all parties involved.       •

  The San Jose Superfund site resulted from leakage of an underground tank
  containing waste so/vents.  The so/vents reached the underlying drinking water
  aquifers and caused significant concern.  The remediation was accomplished in
  a  classic  example of the Sequential Risk  Mitigation approach advocated by
  Canonie including pump and treat, source removal, containment with a slurry wall
  and contaminated soils remediation.  The use of the Accelerated Vacuum
  Extraction System was part of the on-site soils remediation.

 At the time of the soils remediation requirement, the owner of the property had
 accomplished much of the remediation through the sequential application of
 several different innovative technologies and approaches.  The remaining
 contamination was concentrated in the vadosezone of the on-site soils. The site
 was sufficiently remediated that prospective  buyers were interested in the
 property.  However, the property could not be sold with the soils contamination
 remaining at levels above  those considered acceptable to the agencies.

 Canonie assessed the situation and presented two options to the owner.

 1.     Regular Vacuum Extraction using many wells pumped at low rates with
       small pumps.   This was  the system most were  using at the time to
       accomplish soils aeration.  The capital cost was low; however, the time to
       remediation could be long and somewhat unpredictable, and the operation
       and maintenance costs could be high.
             .' •-.••:.'•..                                        f <•    V^
             :  '     '    "                                             4
 2.     Accelerated Vacuum Extraction consisting  of an innovative system
       designed by Canonie to accomplish acceptable levels in the soils in a time
       frame consistent with the impending sale and development of the property.
       The capital costs were higher, but if the system performed as predicted,
       the  overall costs would  be lower  diie to  the tower  operation and
       maintenance costs.  The system was, however, unproven at  commercial
       levels.
                                   402

-------
  Canonie worked with the agencies to establish turn-off criteria for both systems
  based on realistic anticipation of the performance of the respective technologies.
  The accelerated system was more difficult to discuss since the agencies had not
  seen such a system perform.
  The owner elected to go with the accelerated system and worked with Canonie
  on its development and application.  The owner shared in the financial risk and
  in the risk that the system might not perform as predicted. The agencies shared
  in the risk by not imposing unrealistic operating performance criteria and agreeing
  to reasonable shut-off criteria.                                           •

  The system was built and installed.  The performance was near that projected.
  The chemicals were removed from the ground on a schedule almost identical to
  that predicted. The site was remediated and ready for sale at the time indicated.

  The cooperation of all parties for this application of an innovative technology was
  excellent. The developer was not financially burdened nor did he receive windfall
  profits. The owner received remediation at a lower overall cost than with a more
  conventional  system. Since  the  owner also paid for the  equipment, it was
  available to do another project for the owner at no additional capital.  The
  agencies were able to show a site remediated in a timely manner.  The project
  was a success to all concerned because of the cooperative risk sharing of all
  parties involved.

Case Study #5 (Continued) -  San Jose Superfund Site
                                    403

-------
 PROPOSAL TO EPA

 Canonie's experience at commercializing innovative  hazardous waste treatment
 technologies as depicted above represents considerable technical success, but not
 always financial success. Ultimately, the viability and credibility of the hazardous
 waste remediation industry (including regulators and regulated alike) will depend upon
 two interrelated issues, financial success of those whom we rely upon to develop and
 commercialize technologies and a regulatory approach which will foster cleanup of
 hazardous waste sites quicker, cheaper and in a manner that will minimize risks and
 liability for generators and site owners.

 In response to this need, Canonie has made an unsolicited written proposal to EPA
 entitled, "Accelerating Superfund from the Remediation Contractors'  Perspective."
The  precepts of this proposal  as they relate to  commercialization of innovative
technologies are:
      1.
The selection of hazardous waste treatment technologies should  be
made on the basis of reducing risk on a sequential basis by tackling the
parts of the problem that pose the greatest overall risk to the public first.
In doing this, a site should be viewed in a holistic sense and the arbitrary
distinction made by the current regulatory structure of "operable units"
be eliminated.
      2.
Hazardous waste treatment technologies should  be  selected  only
according to a realistic expectation that they can remediate a waste to
a certain level of clean-up.
     3.    Once selected, the technology should be allowed to operate without
           microscopic regulation, provided the off-site public is protected and the
           on-site conditions do not worsen.
                                     404

-------
      4.    The marketplace will foster the development of innovative technologies
            whose performance capability continues to improve simply because of
            the desire of site owners and contractors to limit liability and risks.

The following summarizes Canonie's recommendations to EPA.

Sequential Risk Mitigation

Sequential Risk Mitigation as depicted on Figures 1  and 2 is a comprehensive and
structured approach to investigating, engineering and  remediating complex hazardous
waste sites  utilizing  the  "observational approach"  to construction. Selection of
remedial actions and technologies is made in a sequential fashion (see example shown
in Figure 3) according  to  five levels of  probable range  of risk  or status of
contamination that are estimated for the site (described in Figure 2).

Employing a repetitive approach  of  "investigate," "select,"  "implement,"  and
"observe," sites are remediated in a sequential fashion without much of the ritual (and
attendant "transaction" costs) of the RI/FS and remedial design phases of the current
Superfund program (or their RCRA equivalents). Knowledge obtained by observing the
effects of previous levels of remediation activity serves as the basis to select the next
remedial actions and technologies appropriate to the  next desired level of cleanup.

The role of risk assessment  in the Sequential Risk Mitigation model is significantly
different than that currently promulgated by EPA. We recommend that traditional risk
assessment methodologies to determine long term cleanup goals be deferred until the
site reaches  Level IV. Canonie's proposed  use of risk assessment is much more
truncated as it is designed for a different purpose, to determine the probable "range
of risk" that a site represents or would represent without intervention. This "range of
                                     405

-------
                             90P
                           HEALTH RISK
                               DECREASING
3]

Q

3)
rn

-------
S
     LEVEL
HEALTH  RISK
    RANGES
DESCRIPTION
                             ;iXifWEK;FBlEfPHASEp1[EMl<
                                IN DRINKIN
                    WITH CLEARLY DEFINABLE
                         MOVEMENT
                                    SUBSURFACE BURIED
                                    CONCENTRATIONS, DISSOLVED, CONSTITUENTS IN ,
                                    GROUND WATER WITH SlOW
                                    DEFINED PATHWAYS *
                                    CHEMICALS Or? LOW CONCENTMUdNS, IN
                                   ;GROUND WATER AT SIGNIFICANT DEPTHS BEIOW
                                    GROUND SURFACE, POORLY DEFINED PATHWAYS
        V
                    MINUTE CONCENTRATIONS IN SOIL AND <3ROUND WATEfc
                    WITH NO REALISTIC OR AT MOST REMOTELY POSSIBLE
                    PATHWAYS
                                                                FIGURE 2

-------
O
oo

REMEDIAL ACTIONS
1. EXCAVATE LARGER AREAS Of LOW
CONCENTRATION MATERIALS AND;
» IRLAI ON- Sill Ot.
MICH CAPACITY MOBILE
THERMAL OESORPTION
HIGH OPACITY MOBILE
INCINERATION SYSTEMS
WCM CAPACITY SOU WASHING
HIGH CAPACITY aOREACTIONS
PRODUCTION LEVEL IHNOVATrvE
KCHNOLOCCAl. SYSTEMS
B. TRANSPORT OFF-SHE TO;
MCINERATION
RECYCLE/TREATMENT CENTER
2. W-SIIU TREAT BY;
SUBfACE MUTING BT MUMC
VACUUM EXTRACTION
BIORCUCOUTION
STABILIZATION
PROOUC1ION LEVEL INNOVATIVE
TECHNOLOGICAL SYSTEMS
LEVEL III - DESIGNED AND/OR STRATEGIC ACTION REQUIRED TOWARD LEVEL IV

t


HYC
IAS
FUE


KKX
xm
i a


Mt
t.

a
CAHI
tvE
TO


cs
HIS



w
c
t


S1UVOLAME
ICMOCS (SUCH AS
HtOKuUCKZEtKS.
»tlCHANlHI!tHES.
WU PMIHAUItS)
} Jo Jo|




a
o
a
a
a
o






1


CHIORWAIED
PESliOOES
SUCn AS DOT.
OUDRW AND)
KEI1ML)
2 Jjoljol^D




O
a
(i
o
o





G
MNCHt DAMMED
PESTKOCS
SUCtl AS PARAIMON
ClUTARAlDEmDE.
AND UtCASAIL)






o
a
o
o
0





H 	
units






•
•
0
a
o
a
•
a





IUMUCIMS
(KADWU. PlUIOMUU
UOANH1U. Me)






•
•
a
o
•
a





COAL TAR






•
a
o
a
a
o
a
a
•
•
a




                                       LEGEND;
                                                 CONTAMINANT
                                          1    OPEN AND EXPOSED AT SURFACE

                                          2    DfiUUMED OR CONTAINED

                                         J"    SUBSURFACE IN SOIL - FREE PHASE

                                         *>    SUBSURFACE IN SOIL - PARTIALLY SATURATED

                                         .    SUBSURFACE IN GROUND WATER -
                                              FREE PHASE IN DRINKING WATER AQUIFER

                                         ,    SUBSURFACE IN GROUND WATER -
                                              DISSOLVED  PHASE IN UNUSED AQUIFER
     APPLICABILITY OF TECHNOLOGY;

VERY APPLICABLE

APPLICABLE UNDER CERTAIN CIRCUMSTANCES

NOT VERY  APPLICABLE BUT SHOULD BE CONSIDERED

NOT APPLICABLE
   METHOD FOR INITIAL SCREENING OF
    REMEDIAL ACTiONSAECHNOLOGIES
      FOR VARIOUS CONTAMINANTS
        AND  SITE CONDITIONS -
         LEVEL 111  TO LEVEL IV
              PHCI'AHED FOR
          UNITED  STATES
ENVIRONMENTAL PROTECTION AGENCY
                                                                                                                                           FIGURE 3

-------
risk" serves only as a bench-mark to determine the appropriate remediation strategy
and to  measure progress as the site  passes through various  phases (levels)  of
remediation.

Because sites are approached holistically, and interim goals for remediation are the
subject  of remedial actions (not 100 percent solutions based on studies of operable
units), innovative approaches will be fostered.

Emissions Control During Construction

To address the  central question of health protection during the operation of on-site
technologies, each  site  should be viewed as an  "envirosphere" as  depicted on
Figure 4.  This  figure shows various sources of emissions that can be predicted  to
occur at the site during  the remediation phase.  EPA would establish a "Boundary
Condition" for the contaminants of concern,  based on an acceptable level of the
temporary risk involved.  The "Boundary Condition concept" is illustrated in Figure 5.

The site owner and/or remediation contractor would develop a "Site Emissions Control
Plan"  (SECP). Methodologies for estimating and modeling emissions would be EPA-
approved.

Project construction would be allowed to proceed without interruption, provided site
boundary conditions are not exceeded. Treatment technologies could be operated in
a flexible manner.

In addition to  the  boundary conditions monitored for health protection  during
construction  activities, the policy framework would adopt a "technology-based"
component which would assure that the selected separation/treatment remediation
technologies perform to their maximum efficiency under site conditions. The
                                     409

-------
                                                     RURAL AREA
URBAN AREA
                                                       FIGURE 4

-------
BOUNDARY
                                              WASTE WATER
                                            TREATMENT PLANT
                                                              DISPOSAL OF
                                                             TREATED SQJL
                                                                                      FIGURE  5

-------
 determination of "Best Technical Performance" would be made during the "proof-of-
 process"  phase  based  upon  a  "Proof-of-Process  Plan"  prepared  by  the
 owner/contractor during the RD phase.

 Together, the Boundary Condition and Best Technical Performance components form
 a policy framework for establishing emissions performance standards which can be
 characterized as the "Best Technical Performance That is Health Protective."

 We view the combining of both health- and technology-based considerations as critical
 to address concerns of the public regarding health protectiveness. The EPA is also
 assured that the primary treatment/separation technologies achieve a greater degree
 of emissions control efficiency than would otherwise be allowed on  the basis of
 human  health.   If the proof-of-process phase is extended  to determine the  best
 technical performance on the contaminated medium as well as residual air emissions,
 remediation to levels below the action limit established in the ROD represents a factor
 of safety against future liability.

 Summary

 In summary,  the  following rules should  apply to the selection and operation of
 innovative technologies and approaches toward more cost-effective site remediation:

Selection of Technologies for Remediation
     1.
Investigation of sites should be limited to gathering that only information
needed to select an RA to reduce the health risk from one level to the
next.
                                     412

-------
     2.      An RA technology should be selected based on the health risk level of
            the site and the ability of that technology to reduce the health risk by at
            least one level.

     3.      Observation should be made during the implementation of an RA to allow
            realistic, pertinent, cost-effective selection of an RA to take the site to
            the next lower level  of health risk.

Operation Constraints

     1.      A health- and risk-based boundary condition, which is determined  by the
            EPA, should be established for the site.

     2.      Exceedance of the boundary condition would result in the shut-down of
            operation of the separation/treatment technology.

     3.      On-site proof-of-process testing should be the basis for determining the
            best technical performance of the technology on site wastes. Operating
            parameters and emissions limits will be established at this time to assure
            the technology will  perform to its capability, even when the resulting
            emissions are  more protective  than allowed  under the health-based
            boundary condition.
     5.
Continuous operation of the separation/treatment technology should be
allowed, provided proof-of-process testing demonstrates that boundary
conditions are not exceeded.

In the event that  emissions limits  or specified operating conditions are
exceeded following proof-of-process testing, but the boundary conditions
are not exceeded, the owner/contractor should be provided a reasonable
                                     413

-------
       response time to return the operation of the technology to a state of
       compliance without interrupting the operation. The EPA-approved SECP
       would define contingency actions and time frames.

6.     EPA should not take enforcement action against the owner for failure to
       meet emissions limits or operating conditions provided that the failure did
       not arise from knowing endangerment or willful misconduct.
                               414

-------
CONTRACTUAL MECHANISMS TO SHARE RISK

Contractual  methods of risk sharing are imperative  and feasible for application of
innovative technologies. The totally successful projects discussed above utilized such
types of contracting methods.

Unlike standard well-defined construction projects, lump sum or fixed unit prices are
often not good  contracting methods for applying innovative technologies.  Such
contracts assign financial risk to the innovative technology contractor.  Many such
contractors  cannot define  the magnitude  of that financial risk and  consequently
cannot bid the work or must add  such levels of contingency that their technology
looks unattractive.

Contract methods more suited  to innovative technologies include target-valve-
incentive, daily-rate-cost reimbursement or time-and-materials cost reimbursement
contracts  with  not-to-exceed limits.  These  contract types give incentives to  the
contractors  to make their  technology  perform, yet limit their financial exposure.
Contractors  should not, however, be given "blank checks."  Checks and balances are
required.  A technology can receive a bad reputation and not be selected on future
jobs if significant cost overruns occur on cost-reimbursable type contracts. This  can
be as detrimental to the technology's future as significant financial distress to  the
innovative technology contractor on fixed price contracts.

Cooperation in developing an appropriate contract should be part of this plan, but all
parties must be good business people.  A "win-win" situation can be established
whereby technical success  is coupled with  financial success for both the contractor
and the owner.
                                     415

-------
 CONCLUSION

 Cost-effective development and commercialization of innovative waste treatment
 technologies to remediate hazardous wastes can be fostered by a more cooperative
 spirit on the part of regulators, owners and vendors with each making a concerted
 effort to understand the goals and constraints of the other and a willingness to share
 the risk, and a change in mind-set in planning, designing and physically cleaning up
 contaminated sites. The regulatory decision process needs to  be redefined to allow
 innovative technologies to  be  selected on the basis  of a realistic expectation of
 performance to a specified level of cleanup.  Sequential Risk Mitigation incorporates
 the observational approach to construction wherein remedial actions are selected to
 address the most serious health risk first, with subsequent levels of cleanup defined
 on the basis  of observations and experience obtained in  the previous  stages of
 construction.

 Once construction  is under way, the contractors and  vendors must be allowed to
 modify operations  to adjust to changing site and waste  conditions  encountered
 without microscopic regulation, interruption of operations of  the technology and
 without fear of enforcement provided that the adjacent public health is protected.

 Canonie's proposals for  "Sequential Risk Mitigation"  and  the decision model for
 residual waste emissions control during construction can be implemented within the
 current Superfund statutory and policy framework. If adopted, transactional costs will
 be  reduced  and the 80/20 rule  applied - eighty percent of the problem  solved at
twenty  percent of the cost.

Without a comprehensive  revamping  of the  regulatory approach along the lines
suggested herein and in Canonie's report to EPA entitled "Accelerating Superfund
from the  Remediation Contractors Perspective,"  more cost-effective  innovative
technologies will not be developed and applied on a wide scale.
                                     416

-------
                         PREPRINT EXTENDED ABSTRACT
                  PRESENTED AT THE U.S. EPA FOURTH FORUM
        ON INNOVATIVE HAZARDOUS WASTE TREATMENT TECHNOLOGIES-
                          DOMESTIC & INTERNATIONAL
               SAN FRANCISCO, CALIFORNIA, NOVEMBER 17-19, 1992
           •                                          .       i.   •            '
            BIOREMEDIATION OF CHROMIUM (VI) CONTAMINATED
            SOLID RESIDUES USING SULFATE REDUCING BACTERIA

Louis J. DeFilippi, PhD, Allied-Signal Research and Technology.  50 E. Algonquin Rd.. Des
Plaines IL.  60017-5016 Tel (708) 391-3251: FAX (708) 391-3776

      Cr(VI)-containing residues -are of concern when present in the environment.
Chromate [hexavalent chromium, Cr(VI)] is a strong oxidant and has toxic potential.
Dissolved Cr(VI) has a propensity to pass through mammalian cell membranes and, due to its
water solubility, also has the potential for groundwater migration.  This material often arises
from the chromium roasting process, where chromium in iron-containing ore is oxidized to
chromates to enable separation of the water-soluble chrbmates from insoluble ferric oxide.
The residues from the aforementioned process contain Cr(VI) usually in a highly alkaline
environment arising from the presence of rather high  levels of lime (CaO), which is used in
large quantities in the chromium ore roasting process.  For example, water in contact with
solid waste  arising from lime extraction of roasted chromium bearing ores possess a pH of
between 10 and 11. An extreme pH necessitates pH adjustment prior to microbiological
treatment, but this often leads to  saline conditions that are not conducive to bacterial
proliferation.
      The  naturally occurring reduction  of Cr(VI) to Cr(in) by hydrogen sulfide produced
by sedimentary sulfate reducing bacteria has been noted by R. H.  Smillie, et al. (1).
However, until recently it was believed that sulfate-reducing bacteria were considered to be
unsuitable for treating chromium-containing industrial waste waters because of the inherent
toxicity  of chromium to microorganisms.
      We previously described a bioremediation system (2,3), based upon the generation of
H2S by sulfate reducing bacteria, to effectively reduce Cr(VI) present in saline aqueous
fractions to the low toxicity, low solubility state, Cr(ffl).  We have no™ extended this
approach to include bioremediation of solid residues in contact with a liquid aqueous
fraction, especially when salinity is high.  We feel the use of a sulfide intermediate to effect
bioremediation is superior to direct bacterial reduction for a number of reasons,  especially
that H2S may "seek out" Cr(VI) by diffusing to microenvironments poorly accessible or
inhospitable to microorganisms or their nutrients.  For lime-containing waste the process
involves the addition of a suitable acid, such as HC1, to effect a pH between 6.0 and 9.5,
followed by treatment with an acclimated, anaerobic sulfate reducing bacterial inoculum.
For example, 5.1  g tailings containing 950 ppm Cr(VI) and 16,000 ppm total Cr were
loosely packed into a 0.7 X 17 cm glass column and then treated  (down-flow) with 1.5 mL 1
M HC1, then water, then the inoculum. A black layer (indicative of metal sulfides) 1  cm
thick formed at the surface while below an off white  precipitate [indicative of Cr(OH)3] was
dispersed.  Treatment continued  until < 0.006 ppm Cr(VI) was detectable in the effluent.  In
another experiment untreated solid residue exhibited 103 ppm TCLP extractable Cr(VI).
                                          417

-------
  After bioremediation this value dropped to < 0.0002.
        We have field-tested a similar approach on a 2,000 gal/wk scale.  In this particular
  instance the pH of the groundwater was in the neutral range so that no pH adjustment of the
  groundwater was necessary in situ.  The heart of the bioremediation system  was a 1,000 gal
  fermenter. In it was fermented a solution of sodium sulfate,  ammonium sulfate, super
  phosphate and molasses dissolved in estuarine (brackish) water.  The function of the
  fermenter was to effect an acidogenic phase of the fermentation in a reactor  where the pH
  could be readily controlled.  In it the molasses carbon source was fermented, and the initial
 production of anaerobic, sulfate reducing bacteria and H2S-containing fermented broth
  occurred. The balance of the metabolic process occurred in situ. During  the acidogenic
  stag* of the fermentation the pH of the broth dropped due to the formation of VFAs  (volatile
 fatty acids).  The pH was maintained at about 8  tnrough the addition of soda ash (sodium
 carbonate).  Each batch of fermentation fluid was deemed to be ready for  injection when the
 pH of the broth ceased to change.
        The fermented broth and bacteria were applied to two experimental plots, a
 percolation site, which was  designed for remediation of the Vadose zone through flooding of
 that zone, and an injection site,  designed for remediation of the water saturated  zone.  The
 percolation site consisted of an open-bottomed box fitted with a lid and a Calgon H2S
 scrubber that effected retention of fermentation broth over a specified area.  It is one half of
 a percolation  and leachate recirculation  option that is a potential system for contaminated soil
 and groundwater remediation. The injection site consisted of a single injection well.  Both
 sites were surrounded by a number of piezometer (pipes) for groundwater  monitoring.
       We found that the bioremediation of Cr(VI) on site was successful within certain
 limitations. Possibly the most critical factor in success is the geology of the  site.  Where
 remediation was most complete, the conductivity properties of the soil or residue were
 consistent with the presence of high conductivity and, therefore, preferential  paths of  flow.
 Where there was lack of maintenance of anaerobic conditions  of Vadose zone areas due to
 rapid draining of fermented  broth there  was less than optimal reduction of  Cr(VI).
       We have performed cost estimates for the main raw materials used  for the
 bioremediation of lime-laden sites.  By titration of existing surface residue  we determined
 that it would require about 4.3 mEquiv/g to attain a pH of 8.  For a site containing
 1,000,000 tons of residue this translates to 3.9 x 109 equivalents, and the same number of
 moles of the monoprotic acid HC1.  This is equivalent to 16,700 tons pure  HC1 or 46,400
 tons of 22° Be" (12 M HC1) material. At $68/ton this is $3.2 x 106.  Waste acid  is often
 available for a considerably lower price.
       The concentration of Cr(VI) is typically 3.5 g/kg,  wet weight. Again, for a site
 containing 1,000,000  tons of residue this translates to 6.1 x 107mol of Cr(VI). The
 stoichiometry  of the reduction of Cr(VI) [as CrO4=]  to Cr(m)  [as Cr(OH)3] is as follows:

2CrO<" + 3H2S + 2H2O — > 3S + 2Cr(OH)3 + 4OH'

thus requiring 9.2 X 107 mol H2S. Since H2S is generated from SO4= on a one to one molar
basis, an equivalent number of moles SO4" is needed. At $90/ton this is $1.3 x  106 plus
shipping. A portion of this cost is defrayed by the use of sulfate naturally present in sea or
                                          418

-------
estuarine water.
       The cost of the carbon source is quite variable. Waste molasses is possibly obtainable
at nearly shipping costs.  Items such as "sugar water" are available free in the Industrial
Waste Exchange Program.  When these are not available, com syrup, 43 Be", is available for
$11.22/100 Ibs.  Although one mole of glucose is required for the production of three moles
sulfide, taking into account metabolic needs there is a requirement for 3.8 x 107 mol glucose.
At $0.20/lb this comes to $3 x 106.  Capital costs at well as ongoing pumping and analytical
needs would be an additional factor.
(1) Smillie, R.H., Hunter, K. and Loutit, M., "Reduction of Chromium (VI) by BacteriaUy-  ;
Produced Hydrogen Sulfide in a Marine Environment," Water Research, 15_,  1351  (1981).

(2) DeFilippi, L. J. and Lupton, F. S., Proceedings ofR&D92 National Research &
Development Conference on the Control of Hazardous Materials,  San Francisco CA, pp.  138-
141, February 4-6, 1992.

(3) Lupton, F. S., DeFilippi, L. J., and Goodman, J. R.,  U.S. Patent 5,062,956 (1991).
*U.S.GOVERNMENT PRINTING OFFICE:! 99 3 -750 -002/60163
                                           419

-------

-------