COMMITTEE ON                        EPA 542-R-98-001C
THE CHALLENGES OF                         June 1998
MODERN SOCIETY                         www.clu-in.com
                                   www.nato.int/ccms
       NATO/CCMS Pilot Study
    Evaluation of Demonstrated and
     Emerging Technologies for the
      Treatment and Clean Up of
        Contaminated Land and
             Groundwater
               PHASE
           FINAL REPORT
      APPENDIX IV — PROJECT SUMMARIES
               Number 219
     NORTH ATLANTIC TREATY ORGANIZATION

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)



                                            CONTENTS


Table 1: Discussion of Projects by Chapters 	v

1 Trial of Air Sparging of a Petroleum-Contaminated Aquifer	   1

2 Bioremediation of Petrochemicals Following a Major Fire	   3

3 Bioclogging of Aquifers for Containment and Remediation of Organic Contaminants  	   5

4 Remediation of Methyl Ethyl Ketone Contaminated Soil and Groundwater  	   8

5 In S/Yw/On-Site Bioremediation of Industrial Soil Contaminated with Organic Pollutants: Elimination of Soil
    Toxicity with DARAMEND® 	  10

6 In Situ/On-Site Bioremediation of Industrial Soils Contaminated with Organic Pollutants: Elimination of Soil
    Toxicity with DARAMEND® 	  11

7 Demonstration of Thermal Gas-phase Reduction Process	  15

8 Biodegradation/Bioventing of Oil Contaminated Soils  	  21

9 Field Demonstration of an In Situ Process for Soil Remediation Using Well Points	  23

10 Integrated Treatment Technology  for the Recovery of Inorganic and  Organic Contaminants from Soil ...  24

11 On-site Biodegradation of PAHs in Soil at a Former Gasworks Site  	  26

12 Groundwater and Soil Remediation at a Former Manganese Sulfate Production Plant  	  36

13 Rehabilitation of a Site Contaminated by Tar Substances Using a New On-site Technique	  38

14 Ozone Treatment of Contaminated Groundwater	  42

15/45 Combined Chemical and Microbiological Treatment of Coking  Sites/Bioremediation of Soils from Coal
    and Petroleum Tar Distillation Plants  	  44

16 Combined Vacuum Extraction and In  Situ  Stripping of Chlorinated Vapors  	  45

17 Treatment of Polluted Soil in a Mobile Solvent Extraction Unit  	  47

18 Biological In Situ Remediation of Contaminated Gasworks   	  49

19 Cleaning of Mercury-Contaminated Soil Using a  Combined Washing and Distillation Process  	  52

20 Fluidized Bed Soil Treatment Process—BORAN	  59

21 Mobile Low-Temperature Thermal Treatment Process	  61

22 Environmental Evaluations of Former Soviet Military Bases in Hungary  	  67

23 Modeling and Optimization of In  situ  Remediation	  71

-------
NATO/CCMS Pilot Study, Phase II                                         Final Report (Appendix IV)


24 Combined Remediation Technique for Soil Containing Organic Contaminants: Fortec®  	  73

25 Slurry Reactor for Soil Treatment	  75

26 Treatment of Creosote-Contaminated Soil (Soil Washing and Slurry Phase Bioreactor)  	  77

27 Soil Washing and Chemical Dehalogenation of PCB-Contaminated Soil   	  79

28 Use of White-Rot Fungi for Bioremediation of Creosote-Contaminated Soil	  81

29 Sorption/Solidification of Selected Heavy Metals and Radionuclides onto Unconventional Solvents   ....  83

30 Using Separation Processes from the Mineral Processing Industry for Soil Treatment  	  86

31 Decontamination of Metalliferous  Mine Spoil	  88

32 CACITOX™ Soil Treatment Process  	  90

33 In-Pulp Decontamination of Soils, Sludges, and Sediments	  92

34 Chemical Fixation of Soils Contaminated with Organic Chemicals  	  95

35 In Situ  Soil  Vapor Extraction within Containment Cells  Combined with Ex Situ Bioremediation and
    Groundwater Treatment  	105

36 Enhancement  Techniques for Ex Situ Separation Processes Particularly with Regard to Fine Particles  ... 110

37 Bioventing of Hydrocarbon-Contaminated Soil in the Subarctic Environment	112

38 Demonstration of Peroxidation Systems, Inc. Perox-Pure™ Advanced  Oxidation Technology	114

39 Management of Soil Vapors at the Basket Creek Site  	117

40 An Evaluation of the Feasibility of Photocatalytic Oxidation and Phase Transfer Catalysis for Destruction
    of Contaminants from Water (In  Situ  Treatment of Chlorinated Solvents)	122

41 In Situ Microbial Filters	125

42 Integrated Pneumatic Fracturing and In Situ Bioremediation	127

43 Multi-Vendor Bioremediation Technology Demonstration Project 	129

44 Enhanced In Situ Removal of Coal Tar: Brodhead Creek Superfund Site	135

47 In Situ Electro-osmosis (Lasagna™ Project)  	139

49 Characterization of Residual Contaminants in Bioremediated Soil and  Reuse of Bioremediated Soil  .... 144

50 Integrated Rotary Steam Stripping and Enhanced Bioremediation for In Situ Treatment of VOC-Contaminated
    Soil (Cooperative approach to application of advanced environmental technologies)	149

51 Sobeslav, South Bohemia Wood Treatment Plant  	151

52 Permeable Treatment Beds	152
                                                  in

-------
NATO/CCMS Pilot Study, Phase II                                        Final Report (Appendix IV)







53 In Situ Bioremediation of Chloroethene-Contaminated Soil	154




54 Treatment of PAH- and PCP-Contaminated Soil in Slurry Phase Bioreactors  	156




55 Czechowice Oil Refinery Project	158




56 Spolchemie a.s.—Mercury-Contaminated Site	159
                                                IV

-------
               Final Report
         NATO/CCMS Pilot Study

     Appendix IV — Project Summaries
 Evaluation of Demonstrated and Emerging
Technologies for the Treatment and Clean Up
   of Contaminated Land and Groundwater
                (Phase II)
               June 1998

-------
                                        NOTICE
This report was prepared under the auspices of the North Atlantic Treaty Organization's Committee
on the Challenges of Modern Society (NATO/CCMS) as a service to the technical community by the
United States Environmental Protection Agency (USEPA). The document was funded  by USEPA's
Technology Innovation Office under the direction of Michael Kosakowski (Work Assignment Manager).
Michael A. Smith of Berkhamsted, U.K., was the principal editor for the Pilot Study reports. The Final
Report  was edited and produced by Environmental Management Support,  Inc.,  of Silver Spring,
Maryland, under USEPA contract  68-W6-0014. Mention of trade names or specific applications does
not imply endorsement  or acceptance by USEPA.

-------
NATO/CCMS Pilot Study, Phase II                                     Final Report (Appendix IV)
                                        PREFACE
The Phase II Final Report has been printed in two volumes. This volume contains Appendix IV —
Project Summaries. The other volume (EPA 542-R-98-001a) contains Chapters 1-12 and Appendices I,
II, III, and V.

The project summaries were prepared by a team comprising:
 Paul Bardos               R3 Environmental Technologies Ltd, U.K.
 Robert Booth              Water Technology International Corporation, Canada
 Diane Dopkin              Environmental Management Support, Inc., U.S.A.
 Mary Harris               Monitor Environmental Consultants Ltd, U.K.
 Mark Smith               U.S. Air Force
 Mikael Smith              M.A. Smith Environmental Consultancy, U.K.
 Kai Steffens               Probiotec  GmbH, Germany
 Neel Strobaek              Ramboll, Denmark
 Alison Thomas             U.S. Air Force
 Cathy Vogel               U.S. Air Force
While the authors have made their best to provide accurate summaries, limited material was available
for some of the projects. The summaries were edited and prepared for publication by Mike Smith and
Environmental Management Support, Inc.

The starting point for preparation was a set of one-page summaries prepared for the U.K. Department
of the Environment:

   Ian Martin and Paul Bardos. Summary Report on the NATO/CCMS Pilot Study on Research,
   Development and Evaluation  of Remedial Action  Technologies for Contaminated  Soil  and
   Groundwater: Technical Status May 1996, Centre  for Research into the  Built Environment,
   Nottingham Trent University, Nottingham, U.K., June 1996.

Table 1 is provided as a cross reference for the  chapters in which the projects are discussed.

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
                             Table 1: Discussion of Projects by Chapters
Project
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
o o
JJ
34
35
Chapter
4
/
/
/
/
5




6




7




8




9




10
/



project withdrawn
/


/


/


/
/

/




/











/




/



/


/

/




/

/
/


/
/
/
/


/

/


/



/








/
/
/

/


/



/

/





/





/
/
/





































/




/

















/
















/
/


/

/



/




/

/
/



/
/
/



-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
Project
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Chapter
4

/



/
/
/

5
/

/

/



/
6
/






/

7









8









9



/





10
/





/


replication of project 15
project withdrawn
/
/




/
project withdrawn






/









/
/



to be addressed in the Phase III report









/










/

/
/




                                           VI

-------
NATO/CCMS Pilot Study, Phase II
                           Final Report (Appendix IV)
    Project No. 1
                         Trial of Air Sparging of a Petroleum-Contaminated Aquifer
    Technical Contact:
    Ian Hosking
    Coffey Partners Int. Pty. Ltd.
    P.O.  Box 125
    North Ryde, NSW 2113
    Australia

    tel: +61/2-9888-7444
    fax: +61/2-9888-9977
    E-mail: sydney@coffey.com.au
Country:
Australia
Project Status:
Accepted into study    1993
Interim report           1994
Interim report           1996
1. INTRODUCTION

Leaking pipework under a petrol station located in Adelaide, South Australia, led to contamination of the soil and
groundwater with petroleum hydrocarbons (up to 2,100 mg/L), benzene (up to 1.5 mg/L), and xylene (up to 20
mg/L). Contamination occurred as dissolved, adsorbed, vapor, and minor free phases. The adsorbed phase occurred
as a relatively thin and widespread zone above the water table at a depth of 7.5 m below ground surface. Based on
initial investigations, it was believed that a remediation strategy involving soil vapor extraction with air sparging
below the water table would be appropriate. Air sparging involves injection of air beneath the water table.  As the
air bubbles rise, volatilization of dissolved contaminants occurs from the groundwater into the vapor phase. Airflow
carries the vapors to the soil surface. This vapor phase may then be extracted through soil vapor extraction wells.
This project presents the results of air sparging and soil vapor extraction trials undertaken between September 1993
and April 1994 (no  formal final report was provided for this project).

2. FIELD TRIALS
Three air sparging trials were conducted in total to determine the effectiveness of air sparging in combination with
vapor extraction. In the first trial, the three vapor extraction wells were arranged in a triangle with an injection well
at the center. Vacuum levels of 15-20 kPa resulted in an interpreted air permeability of 0.1 x 10"6 cm2. Injecting air
resulted in increased extraction of hydrocarbon vapor from the extraction wells.
In the second trial, vacuum extraction with and without air sparging was conducted. It was reported that the total
concentrations of hydrocarbons and BTEX compounds (benzene, toluene, ethylbenzene, and xylenes), as extracted
by the vacuum wells, were substantially increased as much as a factor of 10 by the injection of air into the sparging
well. Further, injection of air into one vapor extraction well while the other two wells were operating with vacuum
also resulted in a much greater rate of hydrocarbon extraction than when the wells operated without any air injection.
The system was estimated to be extracting between 1-3 kg of hydrocarbons per day.

A third trial, involving a longer operating time, was conducted to achieve a better indication of the effectiveness of
air  sparging and soil vapor extraction as a suitable combined remediation option. Detailed  sampling of stack
emissions and groundwater for hydrocarbons and oxygen was conducted throughout the remedial operation. Gaseous
oxygen levels were used as an indication of microbial activity. The third trial highlighted that while air sparging
substantially increased the amount of volatile hydrocarbons that can be removed by vapor extraction in the short
term, the rate slowed dramatically within just a few days. It was considered by the authors that this reduction in
extraction efficiency was due to the removal of hydrocarbons from the immediate vicinity of the  borehole and
through the more permeable soil layers being exhausted. The groundwater  concentration of hydrocarbons in the
sparging well was reduced to below  detection levels during the sparging trial. Concentrations of BTEX at wells up
to 3 m from the sparging well were also significantly reduced although petroleum hydrocarbon remained in the water.
It was concluded that the  zone of influence  of air sparging in terms of positive remedial effects on groundwater
chemistry appeared in this test to be restricted to a zone between 3 and 9 meters from the sparging well. Monitoring
of the soil oxygen  concentration during sparging indicated that levels of oxygen were  increased from 10.5% to as

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

much as 21% by sparging. The authors concluded that only a small percentage of this was being consumed by soil
bacteria.

3. CONCLUSIONS

The study found that air sparging substantially increases the amount of volatile hydrocarbons that can be removed
by vapor extraction from soil and groundwater. However, the zone of influence and rate of recovery of volatile
compounds are  site specific. The rate of extraction slowed dramatically after only a few days operation due to a
combination of low site permeability and the creation of preferential flow pathways. Air  sparging also produced
mounding of the water table around the sparge well which may have resulted in the migration of the contaminants
outside the influence of the extraction wells. Based on the results of the trials and given the interbedded sand and
clay  stratigraphy at the site clean-up to risk based standards were seen as difficult to achieve. The recommended
remedial approach at the site therefore included extraction of groundwater at the source of the plume to reduce and
contain the bulk of the contaminant mass, to employ soil vapor extraction with water table drawdown and in situ
bioremediation rather than air sparging.

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 2
                         Bioremediation of Petrochemicals Following a Major Fire
    Technical Contact:
    Stuart Rhodes
    Minenco (Sydney) Environmental Services
    Level 5, 77  Berry Street
    North Sydney 2060
    New South Wales
    Australia

    tel: +61/2-916-4923
    fax: +61/2-916-4930
Country:
Australia
Project Status:
Accepted by study  1994
Interim report       1996
1. INTRODUCTION

In August 1991, a major fire at the Coode Island petrochemical facility, near Melbourne, Australia, resulted in the
widespread contamination of soil and groundwater with phenol, benzene, and acrylonitrile from damaged storage
tanks. A site investigation in October 1991 found phenol concentrations as high as 24,000 mg/kg in near-surface soils
and 700 mg/L in the groundwater. The concentrations of volatile organic compounds (VOCs) found in soils and
groundwater included up to 200 mg/L of methyl ethyl ketone (MEK), 100 mg/L acrylonitrile, and 50 mg/L benzene.
In addition, considerable quantities of contaminated stormwater and water used to fight the original fire were stored
on-site.

The biodegradation of phenol has been well documented, and while it is possible to degrade phenolic compounds,
the inherent biotoxicity of the chemical is believed to significantly inhibit its practical remediation. This Pilot Study
project reported on the development of  an in situ bioremediation approach to remediate and manage the site.

2. LABORATORY AND FIELD TRIALS

A preliminary characterization of the contaminated sandy clay soil showed that it was very low in available nitrogen
and phosphorus and had a low microbial population especially phenol degraders. Bench-scale studies using 8 kg soil
samples showed that significant phenol toxicity to soil micro-organisms occurred at concentrations above 4000
mg/kg. Results  of a simple slurry bioreactor showed that rapid phenol degradation was  possible  with the addition
of a microbial inoculant and nutrient source. It was concluded by the authors that such a system would be impractical
for treating the  whole site. Therefore a model system was constructed that was  comprised of a percolation column
(containing 1 kg of soil) and a 1-L continuously-stirred tank bioreactor. This process was designed to take advantage
of the water solubility of phenol by transferring the contaminant into an aqueous solution that could be treated in
a conventional bioreactor. Results showed that soil phenol concentrations in the soil column could be reduced by up
to 90% in 14 days with up to 50% of the phenol degraded in situ. The flushing-biotreatment system was further
investigated at field-scale.

A field study using 100 tonnes of soil from selected hot-spots (containing up to 2,000 mg/kg phenol) was initiated
in March 1992.  The principal objectives of the study were to determine the effectiveness of the treatment system at
field-scale; to determine the  fate  of contaminants;  and to identify  significant factors  likely to limit  treatment
effectiveness. After four months of operation, the microbial population of the soil was found to have increased from
an initial 105 microbes per gram to more than 107 microbes  per gram, which included up to 50% phenol degraders.
Phenol concentrations declined steadily to a mean value of 120 mg/kg after 45 weeks.  A mass balance analysis
showed  that nearly  85% of the phenol had been  degraded in  situ with  only  10% flushed from  the soil  to the
bioreactor. This result had not been expected since the dissolved oxygen concentration of the irrigation water had
been relatively low. Mixing of the soil pile had occurred twice during the trial and results showed that degradation
rates were significantly increased after each mixing. The authors speculated that this may be due to the introduction
of air into the soil and  the creation of new macropores and fresh soil surfaces for microbial colonization.

-------
NATO/CCMS Pilot Study,  Phase II                                           Final Report (Appendix IV)

3. FULL-SCALE IMPLEMENTATION

The program of remediation for this site, developed in the aftermath of the fire, included disposal of contaminated
fire-, storm- and groundwaters to local treatment works for aerobic lagoon biotreatment; UV-peroxidation of firewater
and stormwater; and soil vapor extraction and sparging of VOCs; and in situ bioremediation of phenol-contaminated
soil using a soil flushing process. Hot spots were excavated and treated in a pond on site as a slurry. The Pilot Study
report only reviewed the soil bioremediation efforts.

Full-scale implementation of in  situ bioremediation was integrated with treatment of the large volume of ground-,
storm- and firewater stored on site. A damaged storage tank (60 m3) was refurbished to serve as a bioreactor to treat
the large volume of phenol-contaminated water generated by soil flushing on the site. Recovered process water was
treated in this bioreactor and used in turn as irrigation water for further soil flushing.

Treatment of 1,600 m2 of soil by soil flushing commenced in early May  1993 with addition of gypsum and 10 kg
controlled release nutrients, plowing to 0.5 m to break up compacted surface soil, followed by regular irrigation with
treated water. Gypsum was found to reduce the ped size of the massive clayey soil structure and therefore to increase
treatment accessibility. Phenol was  expected either to be mobilized from the  vadose zone into groundwater and
recovered for treatment above ground, or biodegraded in situ with the utilization of whatever electron acceptors were
available.

Phenol concentrations decreased rapidly in the top 0.6 m of soil after treatment  commenced. Monitoring of soil
microbiology during November  1993 showed high concentrations of phenol degraders (up to 5xl07 per gram) were
being maintained. Two months after in situ treatment began, phenol concentrations  in the groundwater rose to 1,000
mg/L, and then gradually declined to about 1 mg/L during early 1994. It is claimed that recent results show that
phenol hot spots (to a depth of 2.5 m) have smaller microbial populations than in other treatment areas. This supports
the findings of the treatability study, which predicted the biotoxicity of phenol towards the site microbiology.

Remediation of the remaining 3,000 m2 commenced in September 1993.

4. CONCLUSIONS

It was concluded that the in situ flushing-bioremediation process has substantially reduced the total soil contamination
burden at the site and as a consequence  greatly  reduced the  potential  migration of contamination off-site. No
economic data were reported to the  Pilot Study, nor was final disposition of the site remediation effort. However,
attainment of <1 mg/L of phenol in the groundwater throughout 1994 met the remediation criterion for this medium.

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 3

            Bioclogging of Aquifers for Containment and  Remediation of Organic Contaminants
   Technical Contact:
   Colin Johnston
   Division of Water
   CSIRO
   Private Bag
   Wembley
   Western Australia 6014

   E-mail: colin@per.dwr.csiro.au
Country:
Australia
Project Status:
Accepted by study  1994
Interim report       1996
1. INTRODUCTION

Recent studies have suggested a correlation between increased microbial biomass density in aquifer materials and
reduced saturated hydraulic conductivity of aquifers. This process may be termed "bioclogging." In the field, aquifer
clogging during in situ bioremediation has been recognized as a potential operational nuisance during treatment.
Bioclogging is believed to result from several different processes such as the production of low-solubility gaseous
end products, the excretion of extra-cellular polysaccharide, and the increase in bacterial cell numbers.

The aim of this Pilot Study project was to investigate the potential benefits of bioclogging to provide a temporary
partial subsurface containment of contaminated areas,  and to act as a site for enhanced biodegradation of organic
contaminants. The field site chosen for bioclogging tests is located at Largs North on the eastern side of LeFevre
Peninsula near Adelaide, Australia. Gasoline contamination detected in 1987 resulted from a fractured pipeline.
Contamination at the site consisted of petroleum hydrocarbons in soil and groundwater. The hydrocarbons (diesel
fuel and gasoline) may both have come from the fractured pipeline. Dissolved BTEX compounds (benzene, toluene,
ethylbenzene, and xylenes) existed to a  maximum of 2.5 meters below the water table.

The intention was to investigate both the benefits of containment and biodegradation through laboratory and field
studies of an aquifer contaminated with BTEX. An important aspect was identification of indigenous strains  of
polysaccharide-producing bacteria  at the field site and determination of the optimal conditions (nutrient and carbon
source requirements) for polysaccharide production. A final step was to be  determination of how these identified
conditions can be recreated under field conditions.

The interim report presented the results of the laboratory studies and preliminary characterization of the field site
up until February 1996. Field studies focused on collecting geotechnical, geochemical, and hydrogeological site data.
No bioclogging in the field had yet been attempted.  No additional information on the progress of the study had been
provided by the time this summary was prepared in November 1997.

This work was funded by  South  Australian Department  of Environment and Natural Resources Office of the
Environment Protection Authority, BP Australia Ltd., Minenco and CRA.

2. LABORATORY STUDIES

The objectives of the preliminary laboratory studies were:

    1)  to establish whether polysaccharide-producing bacteria were present at the field site;

    2)  to determine the hydrogeochemical conditions required for polysaccharide production; and
    3)  to determine the carbon and nutrient requirements for optimal polysaccharide production by indigenous
        bacteria.

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

Other researchers had reported that:

    1)  Bacteria produce polysaccharides in the presence of excess sources of carbon and energy, and a deficit of
        nutrients; and

    2)  The conversion of a carbon source to extra-cellular polysaccharides often requires a nitrogen source.

Initial  laboratory studies investigated  the  optimal C:N ratio  (using  glucose and  ammonium  additives)  for
polysaccharide-production for a known bacteriumXcmthomonas campestris in liquid culture. No significant interaction
between glucose and ammonium additives was observed. Glucose concentration alone had the most significant effect
on polysaccharide and biomass production.

Subsequent bench-scale studies examined the effect of aerobic and anaerobic (with nitrate as the electron acceptor)
conditions on polysaccharide-production by indigenous bacteria collected from the field site. Substantial differences
were observed in polysaccharide production in soil slurry cultures using aquifer material from different depths. For
the  aquifer material from 4.0-5.0 m below ground, aerobic cultures with atmospheric oxygen produced the highest
polysaccharide concentrations. Tests of the oxidation of aquifer materials by hydrogen peroxide (as a potential mode
of oxygen delivery) suggested that high concentrations of the reagent and dissolved oxygen were unlikely to persist
in the  aquifer due to matrix reactions (e.g., oxidation of the aquifer material).  Overall, samples with hydrogen
peroxide did not attain high polysaccharide concentrations and in the long term, performed only marginally better
than control  samples. In both aerobic and anaerobic cultures, there was an observed loss of polysaccharides after their
concentrations had reached a maximum. The study recommended that further investigation of these loss mechanisms
was warranted since overcoming them could greatly enhance the potential forbioclogging. Polysaccharide production
under anaerobic  conditions  for the aquifer material from 1.8 - 2.8 m below ground was much greater than for the
material from 4.0 - 5.0 m under both aerobic  and anaerobic  conditions. The  study  suggested  that a  further
investigation of polysaccharide production with aquifer material depth would be useful to identify the matrix and
ecological properties which affect polysaccharide production.

The effect of polysaccharide production on hydraulic conductivity was investigated using a test column of 0.5 m
length by 0.08 m diameter, packed with dry aquifer material  (from depth of 1.8 - 2.8 m) to a bulk density of 1.6
Mg/m3. The  test leachate (containing growth medium, glucose, and sodium nitrate) was passed through the column
at a constant volume flux density of 0.11 nrVday. Results showed that polysaccharide production can lead to a
significant (at least one order of magnitude) reduction in column hydraulic conductivity. The study also suggested
that after initial delivery of amendments (e.g., nutrients) the reduction in hydraulic conductivity appeared to be long-
lived. The most rapid change in Ksat occurred in the first 18 hours of the experiment, with further slow reductions
after 50 hours.

3. GROUNDWATER MODELING

Groundwater modeling activities were aimed at helping identify the critical hydrological  parameters in the Largs
North  aquifer  system; developing models and  data  sets  to describe how the unperturbed aquifer behaves; and
designing and  optimizing the system for delivering the bioclogging amendment solution.

A two-dimensional plan model investigated pumping strategies for delivering the amendment solution to promote
bioclogging. The model  considered two pumping wells,  one for injecting the  amendment solution  and  one for
extracting groundwater. For the computer simulation, the two wells were set 10 meters apart, and the pumping and
injection rates were equal. Desired pumping rates should be large enough to push water 10 meters through the aquifer
in no  more than 5 days, the time during which clogging occurs. Results from this modeling showed that pumping
over the 10 meter distance would not significantly disturb  the existing BTEX plumes at the site. Unfortunately, the
time taken to deliver the amendments over this distance was longer than 5  days. Uncertainty in the Ksat value used
in the  model leads to uncertainties in the predicted time to deliver the bioclogging amendment.  Obtaining better
estimates in the field for Ksat will improve on the  model's  ability to provide acceptable pumping  strategies.

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

4. FIELD TRIAL

The objectives in bioclogging at the field site were to produce a pilot-scale barrier of reduced hydraulic conductivity
10 m long by at least 2 m wide and extending to 4 m below the water table. The goals were to stimulate this barrier
by a once-off addition of the appropriate amendment solution and be able  to determine the effect on hydraulic
conductivity. Challenges in designing a cost-effective system to deliver amendments include: efficiently targeting
the specified volume of the aquifer; delivering the amendments before clogging takes place; overcoming effects of
injecting a dense fluid into the aquifer; and minimizing disturbance to existing contamination.

A 3-dimensional, density-coupled groundwater flow and transport  simulation package was used to design the
amendment emplacement strategy and to assess techniques for measuring changes in hydraulic  conductivity. Results
of modeling showed a linear, balanced injection/extraction scheme with five wells would deliver the amendment
solution to the desired volume of the aquifer most efficiently. A total pumping rate of 50 mVday would ensure
delivery between wells after  1-2 days,  after which a short reversal of the  scheme will  give a  near rectangular
distribution of amendment. Density-coupled simulations showed that it may be impossible to  limit the amendment
solution to the top half of the aquifer given the density of the amendment solution.  Relatively large dispersivities
were used to maintain stability in the simulations, which may lead to underestimations of the effect of density-driven
flows. On the other hand, anisotropy and stratification within the aquifer may limit density-driven vertical migration.

An amendment solution containing a glucose carbon source, NO3" electron acceptor, and NH4+ as a nitrogen source
was successfully added at the field site. Although the injection was designed to deliver the amendment to a maximum
of 4  m below the water table, there was evidence that the  greater density caused the amendment to move to the
bottom of the aquifer. This was in accord with model predictions. Microorganisms within the aquifer were stimulated
by the addition of the amendment. The most obvious manifestation was gas evolution within the aquifer for a period
of several days. Cores taken from the site also showed increases in the amount of polysaccharides in the aquifer and
groundwater samples showed utilization of the glucose carbon source. Pump tests specially designed to evaluate the
change in hydraulic conductivity in that part of the aquifer targeted by the amendment showed a reduction by an
order of magnitude. However, this reduction did not persist, and the reduction in hydraulic conductivity was  only
about a factor three a month after the addition of the amendment. Evaluation of results from the field trial are
continuing.

5. REFERENCE

Johnston, Colin D., et al., Studies of bioclogging for containment and remediation of organic contaminants, Centre
for Groundwater Studies, CSIRO Land and Water, Perth, Western Australia, November 1997.

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 4

                 Remediation of Methyl Ethyl Ketone Contaminated Soil and Groundwater
   Technical Contact:
   James Mantle
   Rust PAK.
   Level 6, 163 Eastern Road
   South Melbourne
   Victoria 3205
   Australia
Country:
Australia
Project Status:
Accepted by study  1994
Interim report       1996
1. INTRODUCTION

Significant soil and groundwater contamination has occurred on the site of an operational oil terminal in Melbourne,
Australia.  The petroleum storage center has operated as a distribution center for approximately 65 years with
subsurface contamination resulting from minor "operational spills" and specific instances of pipeline and storage tank
leaks. This Pilot Study project reports on the development and implementation of the site remedial strategy, which
incorporates in situ recovery of free product and contaminant vapors. Full-scale remediation started in mid 1994 and
was  still in  progress in February  1996 when the Interim Report was prepared.  No additional information  was
available in November  1997.

2. THE SITE

The  site geology consists of a thin layer of clay and reworked soil (up to 3 m deep) overlying basaltic lava flows
to a depth of 30 m. The basalt is an extremely heterogenous host aquifer (water table at depth of 12 m) comprising
discontinuous very low to moderately permeable layers with occasional interconnecting joints and fractures. Extensive
hydraulic  testing at the  site indicated a range of two orders of magnitude in local permeability. Contamination
resulted in development of three free product contaminant plumes  covering an area of over 7 ha. These plumes
included a lubrication oil plume up to 2 m thick, a methyl ethyl ketone (MEK) plume with concentrations over
10,000 mg/L, and a mixed plume containing BTEX compounds (benzene, toluene, ethylbenzene, and xylenes), petrol,
turpentine, and kerosene.

3. REMEDIATION SYSTEM DESIGN AND COMMISSIONING

The  initial remedial strategy consisted of a combination of pump-and-treat for free phase product, and soil vapor
extraction for volatile and adsorbed contamination.  Since contaminant distribution can be distinguished as three
separate plumes, the remediation plan addressed separate  recovery  and treatment systems for each of them. Over 50
boreholes  were sunk on site with specially designed "top loading" pumps to recover floating free product. These
modified displacement pumps use compressed air to pump the recovered fluid to the surface. Injection of "clean"
water around each plume was used to force free  product towards  the recovery wells. The pumping operations for
all three recovery schemes were controlled by a computer using a Programmable Logic Controller which enabled
constant monitoring and control of each borehole (e.g., turning pumps on/off and varying pumping rates). The project
managers  had concluded that due to the size of the project, standard air control systems  for each borehole would be
impractical. Fifteen vapor extraction boreholes were  nominated for the site and were integrated with the boreholes
constructed for free product recovery. An advantage of this approach is that vapor recovery takes  place almost
simultaneously with the creation of unsaturated ground from de-watering. The vacuum generated at each well in the
range of 10-20 kPa.

Recovered vapor,  free product and groundwater were treated at the surface in a "treatment compound" comprising
the following stages or  processes:

     1)  Oil water separation via three oil water separators;

    2)  Groundwater treatment via an heated air stripper (operating at 20-60°C). MEK recovery ranges from 40-
        90%, which is  lower than anticipated from its volatility (MEK partitions strongly into water);

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

    3)  Transfer to a treated wastewater storage tank that is drained or pumped to a municipal sewer (consent
        concentration of 300  mg/L MEK); and

    4)  A thermal destruction system (using a methane gas carrier to ensure full product combustion) to treat all
        hydrocarbon waste streams from the air stripper as well as the vacuum manifold system. When operated
        at 760°C and a retention time of 0.5 seconds, the system ensured total destruction of hydrocarbons.
        Emissions met the 10 mg hydrocarbon/Nm3 objective.

At the air stripper, influent groundwater contained an average of 3,000  mg/L MEK, but this value sometimes peaked
at 5,000 mg/L. In contrast, the stripped water often contained MEK at well below the discharge criterion of 300
mg/L. Off-gases  leaving the stripper had an average MEK concentration of 1,500 mg/Nm3. However, water leaving
the  air stripper contained around 10 ppm benzene. This residual benzene would normally  be removed by an air
stripper. The solubility of benzene in MEK, another contaminant in the extracted groundwater, may explain why
higher than expected levels of benzene remain in air-stripped groundwater.

1,2  dichloroethene was found in both recovered liquid hydrocarbon and in stack emissions. Concentrations in the
stack gases were well  below the threshold limit value  (TLV).

4. FULL-SCALE REMEDIATION

Full-scale  remediation began in August  1994 with the  initial remediation phase characterized by the recovery of a
large volume of free product with an associated high concentration of dissolved contamination. Remediation was on-
going through February  1996  with average rate of groundwater recovery in the order of 50,000 L/day. By the end
of 1995, over 25 million liters of groundwater had been recovered and  treated including 0.5% free product. Total
recovered  hydrocarbons could be categorized as free phase product (113,000 L), dissolved  phase (61,000 L), and
vapor phase (70,000 L).

Since operation commenced, the rate of hydrocarbon recovery declined  significantly over time. In situ bioremediation
was being considered in 1996  to treat the residual contamination within the basalt aquifer after pumping operations
were completed.

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
   Project No. 5

    In S/MOn-Site Bioremediation of Industrial Soil Contaminated with Organic Pollutants: Elimination
                                of Soil Toxicity with DARAMEND®
Technical Contact:
Country:
Austria
Project Status:
Project withdrawn
                                             10

-------
NATO/CCMS Pilot Study, Phase II
                        Final Report (Appendix IV)
  Project No. 6

    In S/MOn-Site Bioremediation of Industrial Soils Contaminated with Organic Pollutants: Elimination of Soil Toxicity with
                                              DARAMEND®
 Technical Contact:
 Dr. Alan G. Seech
 GRACE Bioremediation Technologies
 3451 Erindale Station Road
 P.O. Box 3060, Station A
 Mississauga, Ontario L5A  3T5

 tel: (905) 272-7427
 fax: (905) 272-7472
 E-mail: gracebio@fox.nstn.ca
Country:
Canada
Project Status:
Accepted by study
Interim Reports
Final Report
      1992
1993-1995
      1996
1. BACKGROUND

The key to consistent achievement of effective bioremediation is understanding the complex relationships that exist
among the soil matrix, contaminants, and the microorganisms degrading the contaminants. A fundamental point often
neglected in soil remediation is that all the biological processes take place in water held against gravity within soil
pores. Water-holding  capacities of contaminated and uncontaminated soil from the same sites are often dramatically
different—hydrocarbon contamination often coats part or all  of a soil agglomerate with oil, thereby rendering the
soil hydrophobic. The hydrocarbons close the soil pores and coat nutrient-providing soil organic matter, limiting the
supply of nutrients, oxygen, and biologically available water. As a result, even in soil containing microorganisms
with an enzymatic capacity to degrade the target compound, the rate of bioremediation may be too slow to be useful.
DARAMEND®  bioremediation was developed  to overcome these limitations naturally and cost-effectively.

The DARAMEND® process was demonstrated in an ex situ application on dredgings from Hamilton Harbour (see
also Project  10), and in situ at a fuel-contaminated military site in the far north at Alert, Northwest Territories. The
U.S. Environmental Protection Agency audited a full-scale demonstration at the Domtar Wood Preserving facility
in Trenton, Ontario, under the auspices  of the Superfund Innovative Technology Evaluation (SITE) program. At the
Domtar  site, initial chlorophenol concentrations of up  to 276 mg/kg and total polynuclear aromatic hydrocarbons
(PAHs)  from 577 to  2068  mg/kg were encountered. Both in  situ  and ex situ  applications of the technology were
demonstrated at Domtar.

2. TECHNICAL CONCEPT

DARAMEND®  bioremediation is based upon careful microbial  husbandry,  creating  optimal conditions for the
destruction   of  organic  pollutants  by  indigenous  microorganisms.  Organic amendments  are  homogeneously
incorporated into the contaminated soil to supply biologically available water, nitrogen, phosphorous, micronutrients,
and oxygen to support biodegradation of the pollutants. Concurrently, the amendments reduce the acute toxicity of
the  soil's aqueous phase by  transiently adsorbing pollutants, and providing surfaces for microbial adhesion and
development of biofilms. The composition of DARAMEND® organic amendments is soil-specific and based upon
the results of a thorough physical and chemical  characterization (e.g., texture, moisture retention, C:N ratio, nutrient
profile, and identity and concentrations of target compound) and treatability studies of the soil or waste to be treated.

3. PROCESS DESCRIPTION

The DARAMEND® process  is suitable for ex situ or  in situ remediation of  soils, and can also be used to treat
sediments ex situ following dewatering. Key features include:

  •  Addition of proprietary solid-phase organic amendments with soil-specific particle size, nutrient  profile, and
    nutrient release kinetics;

  •  Incorporation of  amendments and aeration of soil using specialized tillage equipment; and
                                                   11

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

  •  Strict control of soil moisture content.

Soil amendments are introduced at 1-5% by weight using conventional agricultural equipment (tractor, tiller, etc.),
followed by regular tilling and irrigation. For ex situ application, the soil or sediment to be treated is transported to
the treatment area and homogenized by tilling with a power take-off driven rotary tiller. Homogenization is necessary
both to distribute the organic  amendments uniformly and to minimize the small regions of very high contaminant
concentration, which may especially  inhibit biological detoxification. In situ, roto-tilling equipment is used for
homogenization. Soils and sediments  can be treated as deeply as effective tillage can be achieved.

During the remediation process, soil moisture content is maintained within a narrow range by drip irrigation. Excess
moisture can limit diffusion of oxygen through the soil matrix to microbially active microsites where biodegradation
occurs. Conversely, if soil moisture falls below the optimum range, biodegradation can be inhibited by an inadequate
supply of biologically available water. Optimal soil moisture contents are determined during the treatability studies.

Soil undergoing treatment is tilled approximately once every two weeks. Soil moisture is monitored gravimetrically
in the 0-200 mm zone and in the  200+ mm zone. The shallower depth is where most of the moisture flux occurs
due to microbial utilization, evaporation, and gravitative flow. Deeper samples indicate accumulation of excess water.
The two  moisture samples together are used to  determine irrigation requirements.

4. RESULTS

Data indicate that DARAMEND® bioremediation provides a significant improvement  in bioremediation rates, as
compared to standard bioremediation practice (i.e., inorganic nutrients, irrigation, and tillage). Conversion of added
14C-labeled phenanthrene to 14CO2 verifies that DARAMEND® bioremediation promotes  complete biodegradation of
the target compound. Impressive removals and toxicity reductions were achieved. In addition, soils previously thought
to be so highly contaminated  that bioremediation was impossible, were successfully treated.

Hamilton Harbour sediments were treated for 46 days, which resulted in an 84% reduction in total PAH, from 1,146
mg/kg to  187 mg/kg. Sediments which had been pretreated with CaNO3 showed a 72% reduction. A control plot
(tillage  only)  showed a  40% reduction.  Recalcitrant  PAHs  were effectively treated  with  DARAMEND®.
Benzo(a)anthracene reduction was 60%, 58%,  and 30%, respectively, for non-pretreated, pretreated, and  control
sediment.

At the Alert site, the most northern permanently-inhabited settlement in the world, the pilot in situ plots were covered
by temporary greenhouses to control soil moisture content and to increase soil temperatures through solar heating.
Three DARAMEND® formulations were tested,  and after 10 days, 89%, 62%, and 75% reductions in total petroleum
hydrocarbons (TPH) were observed. No TPH reduction was observed in the control plot.

After 254 days in the ex situ demonstration at the Domtar site, PAHs were shown to be reduced 94%, from 1,710
mg/kg to 98 mg/kg; chlorophenols 96%, from 352 mg/kg to  13.6 mg/kg; and TPH by 87%. The control area showed
a reduction of 41% in PAHs, but no reduction in either chlorinated phenols or TPH. Toxicity,  as measured by
earthworm mortality and seed germination, was eliminated or greatly reduced only in the treated soil.

Grace Bioremediation Technologies  has  provided a composite  results table (see Table 1) based on numerous
treatability studies and demonstrations.

5. COSTS

Detailed  economic  analysis  can be  found  in  the SITE report  (3), based upon the  experience  of the Domtar
demonstration. The cost estimates assume a contamination scenario similar to the Domtar case, and neglect extremes
of geology, geography, and climate. Estimates also do not include the cost of treatability studies.

The unit cost of using DARAMEND® technology was estimated to be U.S.$92/m3 ($70/yd3) for in situ application
and $140/m3 ($108/yd3) for the ex situ  case. This estimate includes the cost of disposal of residual oversized particles
screened out during pretreatment and deemed to be  hazardous.  Without this cost, unit costs decreased to  $46/m3
($35/yd3) for in situ application and $96/m3 ($73/yd3) for ex situ application. Either way, in situ application is more


                                                   12

-------
NATO/CCMS Pilot Study, Phase II
                                          Final Report (Appendix IV)
          Table 1: Initial and intermediate concentrations of a variety of contaminants in soil and/or
                             sediment treated using DARAMEND® technology.
             Parameter
Initial Cone.1
   (ing/kg)
Final Cone.12
   (ing/kg)
 Destruction
Efficiency (%)
 Treatment
Time (days)
     Hydrocarbons
     PAHs
     PAHs
     PAHs
     Heavy Oil
     Heavy Oil
     Hydraulic Oil
     Fuel Oil
     TPH
   16,886
   2,029
    1,442
   100,367
   42,033
   35,000
   41,000
   8,700
    1,546
     100
     36
    4,260
    1,875
    5,400
    2,700
     35
      91
      95
      98
      96
      95
      85
      93
     >99
    241
    211
    240
    293
    343
    204
    225
    182
     Phthalates
     Total Phthalates
     OOP3
     DNOP4
    4,350
    2,700
    1,640
     26
     19
     7.1
      99
      99
     >99
    130
    130
    130
     Nitroaromatics
     TNT
    7,200
     19
     >99
    125
     Pesticides/Herbicides
     Metolachlor™
     Total CPs
     PCP
     PCP
     PCP
     Total OCPs
     Total OCPs
     Total OCPs
     p,p-DDT
     p,p-DDT
     p,p-DDT
     Dieldrin™
     Toxaphene™
     Toxaphene™
     139
     180
     11
     155
     702
     249
    2,125
     354
     66
     684
     124
     128
    1,050
     40
     4.0
     1.5
     1.5
     0.8
     4.0
     87
     404
     164
     1.1
     1.9
     14
     38
     244
     1.9
      97
      99
      86
      99
     >99
      65
      81
      54
      98
     >99
      89
      70
      77
      97
    187
    109
    61
    109
    345
    13
    147
    28
    13
    147
    28
    147
    147
    142
     1. Rounded to the nearest  0.1 for concentrations >0 and <10
                         1 for concentrations >10 and <1,000
                         10 for concentrations >1,000
     2. These values do not necessarily represent final residual levels;
     3. bis(2-ethylhexyl)phthalate
     4. di-n-octylphthalate
                  however, they provide an indication of remediation rates.
                                                  13

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

economical,  although  ex situ  treatment is  more robust to  higher toxicity or contaminant levels.  The Domtar
demonstration produced no leachate, so effluent treatment or disposal was not a cost factor.

Higher ex situ cost was due to increased materials required to construct the pad, sand buffer, and greenhouse and
to the additional labor required to build the initial ex situ plot and to clear treated soils and spread fresh soils for the
batchwise operation.

6. SAFETY

The DARAMEND® process poses no inherent dangers to workers or the community. Hazardous material safety
procedures appropriate to the target contaminants apply, as well as appropriate safety precautions for the machinery
used.

7. CONCLUSIONS

DARAMEND® is an effective treatment technology for remediating soil or sediment highly contaminated with a
variety of organic materials. Total contamination and toxicity can be reduced to target levels in a reasonable amount
of time, and the treated soil regains a high degree of usefulness as fill or topsoil.

8. REFERENCES

1.  Bucens,  P.,  A.  Seech, and  I.  Marvan, 1996.  Pilot-scale Demonstration  of DARAMEND®  enhanced
    bioremediation of  sediment contaminated with polycyclic aromatic hydrocarbons in Hamilton Harbour. Water
    Qual. Res. J. Canada. 31:433-451.

2.  Seech, A.G., I. J. Marvan, and J.T. Trevors, 1994. On-Site/Ex-Situ bioremediation of soils containing chlorinated
    phenols  and polycyclic aromatic hydrocarbons. In: Hinchee, R.E., A. Leeson, L.  Semprini, and S.K. Ong, (Eds.)
    Bioremediation of Chlorinated and Polycyclic Aromatic Hydrocarbon Compounds, pp. 451-455. CRC Press,
    Boca Raton.

3.  Grace  Bioremediation  Technologies DARAMEND®  Bioremediation  Technology-Innovative  Technology
    Evaluation Report—Superfund Innovative Technology Evaluation (SITE). U.S. Environmental Protection
    Agency, Office of Research and Development.  Washington, DC,  20460, EPA/540/R-95/536. July 1996.

4.  Performance Evaluation of the DARAMEND® Bioremediation Process at  the Domtar Wood Preserving Site in
    Trenton, Ontario—Final Report. An Independent Evaluation Prepared for Environment Canada and the Ontario
    Ministry of  Environment  and Energy  under the Development and Demonstration  of Site  Remediation
    Technologies  (DESRT)  Program.   September  1995.  Unpublished Report. Environmental  Technology
    Advancement Directorate, Environmental Protection Service, Environment Canada, Ottawa, Ontario, K1A OH3.

9. FURTHER READING

1.  Seech, A.G., J.E.  Cairns, and I.J. Marvan, 1997. Composition and method for degradation of nitroaromatic
    compounds. U.S. Patent 5,618,427.

2.  Seech A.G., J.E. Cairns, and I.J. Marvan, 1996. Composition for dehalogenation and degradation ofhalogenated
    organic  contaminants.  U.S. Patent 5,480,579.

3.  Seech A.G., J.E. Cairns, and I.J. Marvan, 1995. Method for the dehalogenation and degradation ofhalogenated
    organic  contaminants.  U.S. Patent 5,411,664.

4.  U.S. Environmental Protection Agency, Grace Bioremediation  Technologies,  Daramend™ Bioremediation
    Technology:  Innovative Technology Evaluation Report.  USEPA (Washington, DC, 1996), EPA/540/R-95/536.
                                                  14

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
   Project No. 7
                             Demonstration of Thermal Gas-phase Reduction Process
   Technical Contact:
   Doug Hallett
   ELI ECO LOGIC International, Inc.
   143 Dennis St.
   Rockwood, Ontario NOB 2KO

   tel: +1/519-856-9591
   fax: +1/519-856-9235
Country:
Canada
Project Status:
Accepted by study  1992
Final Report        1995
1. INTRODUCTION

The ECO-LOGIC thermo-chemical process employs hydrogen to chemically reduce organic compounds to mineral
components at temperatures of about 900°C. Soils are treated with the aid of a novel thermal desorber involving use
of a molten metal bath. A technology demonstration at the Middleground Landfill in Bay City, Michigan, USA was
conducted in 1992 under the U.S. Environmental Protection Agency's (USEPA) Superfund Innovative Technology
Evaluation (SITE) Program with partial funding from the  Canadian and Ontario Governments(l,2). The wastes
processed were oily-PCB contaminated water, high-strength PCB oils, and PCB-contaminated soil.

2. BACKGROUND

The Middleground Landfill accepted municipal wastes for approximately 40 years. A 1991 investigation indicated
elevated levels in groundwater of trichloroethene, polychlorinated biphenyls (PCBs), 1,2-dichloroethene, methylene
chloride, toluene and ethylbenzene. The groundwater also contained lesser concentrations of benzidine, benzene, vinyl
chloride,  chlorobenzene,  polycyclic aromatic hydrocarbons  (PAHs),  lindane, dieldrin, chlordane,  and DDT
metabolites.

3. TECHNICAL CONCEPT

The ECO-LOGIC thermo-chemical process employs hydrogen to chemically reduce organic compounds to mineral
components at temperatures of 850°C or above at near ambient pressure. Chlorinated hydrocarbons,  such as PCBs
and polychlorinated dibenzo-p-dioxins, are converted to methane and  hydrogen chloride, while non-chlorinated
organic compounds, such as PAHs, are reduced to methane and ethene. (Incomplete reduction and combustion in
the system may result in benzene formation; as noted below residual benzene was a problem in the demonstration
project). Ethene formed by contaminant breakdown may undergo additional conversion to methane. Methane reacts
with water vapor to form hydrogen and carbon monoxide. The hydrogen chloride is scrubbed out in a caustic soda
scrubber downstream  of the processor. The reformed process gas can then be recirculated in the system or used as
fuel in various stages of the process. The absence of free oxygen in the reactor inhibits dioxin (PCCD) and furan
(PCDF) formation. The process employs automatic monitoring to maintain optimum operating conditions.

The reactor feed is dependent upon the soil or waste being treated.  Aqueous streams, such as  groundwater, are
preheated in a vaporizer using steam from a boiler and fired by propane and/or process gas, before injection. When
soils or sediments are to be processed, they are first  pretreated in a linked thermal desorption unit (TDU). In this
novel unit, the soil or sediment is floated on a bath of molten tin. Volatilized contaminants are sent to the reactor
through a separate port. Contaminated solids, such as transformer parts, can be handled in sequencing batch vaporizer
chambers. The process is shown schematically in Figure 1.

4. THE DEMONSTRATION REACTOR

The demonstration-scale reactor (Figure 2) was 2 m in diameter and 3  m tall, mounted on a 15 m drop-deck trailer.
The trailer carried a scrubber system, a recirculation gas system, and an electrical control, center. A second trailer
held a propane boiler, a waste preheating vessel, and a waste storage tank. ECOLOGIC designed the process to treat
                                                  15

-------
NATO/CCMS Pilot Study, Phase II

Figure 1: Reactor and Thermal Desorption Unit Schematic Diagram
                                                     Final Report (Appendix IV)
                   N2-
                   H2-
                                                                             Recirculating Gas
Propane  Exhaust     900°C.
Air  I    Fl ^35      I     ''
        l^vXX/VV^
        I Quench  I
        I  TankI
          Thermal
       Desoiptlon Unit
3.6 tonnes/daya of waste oil, 9.1 tonnes/day of wastewater, and 22.7 tonnes/day of soil, depending on the nature of
the contaminants, their degree of chlorination, and their water content.

For the  demonstration, a heat exchanger  evaporated contaminated  aqueous feedstock  to  form steam and  a
concentrated liquor. Atomizing nozzles sprayed the heated liquor with associated  particulates into the  reactor. The
PCB-rich oil was injected directly into the reactor through a separate set of atomizing nozzles. Compressed hydrogen-
rich recirculation gas passed through a gas-fired heat exchanger and entered the top of the reactor tangentially. The
tangential entry swirled the fluids to provide effective mixing. As indicated in Figure 2, the swirling mixture traveled
downwards  in the annulus formed by the reactor wall and the central ceramic-coated steel tube, past electrically
heated silicon carbide elements. These elements heated the mixture to 900°C. At the bottom of the reactor, the
mixture entered the central tube and flowed upwards to the reactor outlet. The reduction reactions occurred as the
gases traveled from the reactor inlets to the  scrubber inlet.

Heavy particulates dropped  out of the gas stream and collected at the base of the reactor in a grit box.  The gas
leaving the reactor was scrubbed using a caustic alkaline wet scrubber (pH 9), which removes steam,  particulates,
and gases such as hydrogen chloride. Ninety-five percent of the exhaust gases (reheated to 500°C) were recycled,
with an additional 5% being used as a supplementary fuel for the propane-fired boiler. The boiler produced steam
used in the heat exchanger and burned the reformed gas. The exhaust from this boiler was the only source of air
emissions from the process.

5. THERMAL DESORPTION UNIT (TDU)

The TDU desorbs organics at 500-600°C into a hydrogen-rich carrier gas from soil supported on a molten tin bath.
Hydrogen and tin are used because they do not react. Tin offers favorable properties such as high density, low vapor
pressure, high surface tension  (which  means it does not "wet" the soil and  enter into pores), high thermal
        The SITE reports uses U.S. units (e.g., 1 ton = 2,000 Ibs). These have been converted to metric units on the basis that 1 Ib = 0.45359
        kg. 1,000 kg = 1 (metric) tonne. Thus, 1 U.S. ton = 0.907 tonne.  1 U.S. gallon = 3.785 liters (1000 liters = 1 m3).
                                                    16

-------
NATO/CCMS Pilot Study,  Phase II

Figure 2: The ECO-LOGIC Reactor
Final Report (Appendix IV)
                                                                         To Scrubber
                                                                 Waste Injection Ports
                                                                 Reactor Steel Wall

                                                                 Fibreboard Insulation
                                                                 Refactory Lining

                                                                 Electric Heating Elements

                                                                 Ceramic-coated Central
                                                                 Steel Tube
                       To Grit Box
conductivity, and good solvent properties for heavy metals such as lead, cadmium, and arsenic.

Some of each volatile metal present in the soil passes to the reactor; some dissolves in the molten tin bath, and the
remainder stays in the soil. Non-volatile  metals remain in the treated soil. Quench water cools the soil before
disposal.

During the demonstration,  a  hopper with a screw feed dropped waste soil onto the tin bath. The screw feeder
provided a gas seal between the hydrogen and the outside air. Once inside the TDU, the soil floated on the molten
tin. A paddle wheel removed treated soil from the end of the tin bath and fed it to the quench tank.

6. MATERIALS PROCESSED

At the Middleground Landfill site, the wastes processed were oily-PCB contaminated water, high-strength PCB oils,
and PCB-contaminated soil extracted directly from the landfill. The tests on the two liquids provided information
on reactor performance. The tests on the soil provided information on the functioning of the complementary thermal
desorption unit.

A liquid pool of waste within the landfill provided feedstock for the tests. Perchloroethene (PCE) was added to the
feedstocks to  serve  as a tracer to determine  DEsb. The reactor  program treated approximately  2.6  tonnes  of
wastewater contaminated with 3,757 mg/L PCBs and 3,209 mg/L PCE at a rate of 1.73 kg/min, and 0.2 tonnes  of
waste oil containing 25.4% PCBs and 6,203 mg/kg PCE was treated at a rate of 0.385 kg/min. Additional feedstock
contaminants included fluoranthene, naphthalene, phenanthrene, other PAHs, chlorobenzene, chlorophenol, methyl
chloride, tetrachloroethene, toluene, and various metals.
        DE is a measure of the system's ability to achieve destruction of organic compounds as measured around the system and all output
        streams. DE(%) = {(l-MassOIltpJ/Massmput}*100
                                                    17

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

Two runs were  carried out with the  TDU on PCB-contaminated  soil.  The two  runs  treated  1 tonne of soil
contaminated with 627  mg/kg PCB and 14,693  mg/kg hexachlorobenzene  (HCB).

7. RESULTS

The runs employing liquid feeds confirmed the feasibility of the gas-phase  reduction process for treating PCBs and
other chlorinated organic compounds, producing a fuel gas from contaminated liquids, and providing environmentally
acceptable air emissions. In general, the reactor system effectively destroyed  PCBs, reducing them to lighter
hydrocarbons.

Theoretically, the overall effectiveness of the destruction process could depend on the functioning of both the reactor
system's gas phase reduction process (which produced the reformed gas), and the propane/reformed gas-fired boiler.
Destruction and removal efficiencies (DREs)° for PCBs in the scrubbed reformed gas were essentially equal to the
DREs at the boiler stack. This shows that combustion of the reformed gas  in the boiler is not required to complete
PCB destruction.

Stack emissions generally met stringent regulatory standards. However, average benzene concentrations in the stack
gas (corrected to 7% oxygen)  and scrubber liquor required close monitoring. Benzene emissions of 73-113 micro-
grams per dry standard cubic  meter  (ug/dscm) exceeded the regulatory  limit. The scrubber liquor required either
disposal as a RCRA waste or  recycling through the system for additional treatment.

Demonstrated DREs for PCBs ranged from 99.9999  to 99.99999%.  The demonstrated DE for PCE was  99.99%.
There was no net formation of PCDD or PCDF.

The TDU did not operate to specification. The most important finding was inefficient desorption from soil during
one of the runs. The DE for HCB ranged from 72.13 to 99.99%.

8. AIR EMISSIONS

The  USEPA evaluated emissions of criteria air pollutants  and HC1, as well as compliance with the Michigan
Department of Natural Resources (MDNR) permit. During the tests, continuous emission monitors (CEMs) measured
the  concentrations of the criteria air pollutants at the stack: nitrogen oxides (NOX),  sulfur dioxide (SO2), total
hydrocarbons (THC), and carbon monoxide (CO). Each of these pollutant  concentrations was well under the level
established in the permit.

The demonstration boiler operated between high and low fire, depending on  the system's steam requirements. The
test analysis  showed out-of-range spike concentrations of THC and CO (indicators of combustion efficiency) during
low-fire operation, most notably during treatment of the wastewater when cycling between high- and low-fire
conditions.

HC1 emissions were well below the MDNR permitted level of 4 Ibs/hr (1.81 kg/hr), or 99% removal). Average stack
concentrations ranged from 0.66 mg/dscsm at  109 mg/hr to 0.81 mg/dscm at 198 mg/hr. Removal efficiencies
reached 99.98%.

9. PROCESS RESIDUALS

Process residuals included reactor grit, scrubber sludge, scrubber decant, scrubber liquor, compressed tank condensate
and stack gas—and when the TDU was used—treated soil. The SITE reports  contain information on the production
rates of these residuals and their composition. The reactor grit contained PCBs, PAHS, and a number of other organic
compounds.  However, ECO LOGIC intends to recirculate the grit through the reactor in a full-scale plant. Scrubber
residuals also contained metals and a variety of organic compounds, but again recycling through the plant should
be possible.
   = DREs compare the mass flow rate of selected feedstock compounds, in this case PCBs, to their mass flow rate in the boiler stack gas.
    DRE(%) = {(l-Massstmt)/MasslrlpJ}*100

                                                   18

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

10. COSTS

The 12 cost categories established for the  SITE Program* formed the basis for the cost analysis for the treatment
of liquid wastes. Costs were estimated for the reactor system, processing an average 2.2 kg/min, as operated at the
Middleground Landfill site. For this estimate, it was assumed that 378 m3 (100,000 U.S. gallons) of wastewater and
114 m3  (30,000  U.S. gallons) of waste oil were  stockpiled for treatment. Based on the economic analysis, the
estimated cost in 1994 U.S. dollars for treating liquid wastes  similar to those at the Bay City site ranged from
U.S.$2,205/tonne (60% utilization factor) to U.S.$1,840 (80% utilization factor). The  most important  element
affecting cost is labor (52%), followed by site preparation (15%), supplies (12%), and start up/mobilization (12%).

Site preparation  costs for the  demonstration were U.S.$127,400. Capital costs (for the  reactor and immediately
associated equipment, but excluding the TDU) for a commercial operation are estimated to be U.S.$585,000. The
costs of treating contaminated soils were estimated at U.S.$695/tonne (at a 60% utilization factor) and U.S.$550/tonne
(80% utilization  factor). Important elements of these costs are fuel (67%), equipment (11%), and labor (9%).

11. SAFETY

The principal safety considerations concern the physical hazards common on construction and remediation sites,
chemical use, equipment integrity, and process control. The chemical hazards arise from the use of propane,  liquified
nitrogen/oxygen, hydrogen, industrial chemicals, and hazardous feed materials. In addition, the process generates
methane. Standardized industrial procedures provide adequate guidance for storing, transporting, and handling these
materials.

There should be no undue concern associated with  hydrogen usage in the process. Well-established  and proven
procedures are available for safe hydrogen  storage and use. Hydrogen is no more dangerous than  gasoline or
methane, but it must be handled with regard to its unique properties. Verification of system integrity is essential for
process  safety. Hydrogen is more difficult to  contain because of its small molecular size. Therefore, interfaces of
equipment, instruments, and piping must be leak-free. To provide additional safeguards, the process is operated at
a slight positive  pressure to prevent oxygen ingress; internal oxygen concentrations are monitored; and gas  feeds
(propane and hydrogen) are maintained at low pressure to minimize the likelihood of pipeline breaks. The plant is
fitted with an automatic safety system, which initiates plant shutdown in the event of a hazardous situation.

13. CONCLUSIONS

The SITE Program concluded that the ECO LOGIC Process efficiently treated liquid wastes containing oily PCBs
and other organic compounds,  and water containing PCBs, other organic compounds, and metals. Stack emissions
met stringent regulatory levels. The scrubber effluent (the principal residual stream) concentrated metals and some
organics (benzene, PCBs, and PAHs), indicating that additional treatment (e.g., recycling through the process) might
be required prior to disposal. However, throughput reliability was only 20-55% of design and system availability—a
reliability of 24% indicates that  system reliability  needs  improvement. A number of technical problems were
encountered during  the operation of the plant and are recorded in the SITE program report. The boiler should be
operated at firing rates and air/fuel ratios that prevent the spikes of THC and CO observed during the  trial.

The reactor system  is best suited for processing liquids and TDU off-gases and water vapor.  The waste's organic
content limits the demonstration-scale system's feed rate because of the reformed gas generation. Currently, ECO
LOGIC plans  to  improve throughput by storing  excess reformed gas after compressing it.  Future users should
consider the implications, logistics, and costs  of this approach.

The TDU did not perform to design specifications.  The USEPA categorized the TDU test data as a system proof-of-
concept, rather than as a comprehensive evaluation of a fully-developed unit. The TDU only achieved acceptable
desorption efficiencies at the  expense of throughput.  In addition, ECO  LOGIC experienced material handling
problems with the TDU feed. The  combination of feed problems and inadequate organics desorption showed a need
        Site preparation, permitting and regulatory, capital equipment, mobilization and start-up, operations labor, supplies, utilities, effluents,
        residuals, analytical, repair and maintenance, demobilization.

                                                    19

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

for further development.  Nevertheless, the  demonstration did show that the TDU can desorb PCBs,  and that
satisfactory treatment in the reactor system was possible.

Cold-weather operations may inhibit efficient destruction because of the incremental amount of energy required to
heat the reactor. In addition, feedstock liquids would require melting prior to treatment, and liquid residuals could
freeze in the unheated storage tanks. Winterization (including heat tracing) is necessary to provide adequate feedstock
and to ensure uninterrupted processing.

14. REFERENCES

1.   U.S. Environmental Protection Agency, Eco Logic International Gas-Phase Chemical Reduction Process -The
    Reactor System: Applications Analysis Report, Washington, DC, 1994, EPA/540/AR-93/522.

2.   U.S. Environmental Protection Agency, Eco Logic International Gas-Phase Chemical Reduction Process -The
    Thermal Desorption Unit: Applications Analysis Report, Washington, DC, 1994, EPA/540/AR-94/504.
                                                  20

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 8
                           Biodegradation/Bioventing of Oil Contaminated Soils
   Technical Contact:
   Yvan Pouliot
   Biogenie,  Inc.
   Pare technologique du Quebec
   metropolitan
   350 rue Franquet, entree 10
   Sainte-Foy, Quebec G1P 4P3

   tel: +1/418-653-4422
   fax: +1/418-653-3583
   E-mail: biogenie@ic.net.net
Country:
Canada
Project Status:
Accepted by study  1992
Interim report       1993
Interim report       1994
Final report         1996
1. INTRODUCTION

Biogenie, Inc., under the auspices of the DESRT technology demonstration program, has developed a biopile-based
approach to soil treatment. This is an ex situ bioremediation technique whereby soil is heaped in contained areas,
(e.g., on an asphalt pad with drainage collection) and supplied with oxygen, moisture, and nutrients via an integrated
irrigation and air blower system to enhance natural degradation processes. Biopiles are reportedly suitable for a wide
range of organic compounds and can incorporate addition of surfactants,  co-substrates, or acclimated bacteria, when
needed. This Pilot Study project reviewed the bench-, pilot-, and full-scale development of this technology including
reports on specific aspects of process development—such as the exhaust air treatment system (Oxford meeting, 1994).

2. THE FIELD TRIALS

Under the DESRT program, two full-scale trials were conducted at sites selected for their contrasting contamination
and soil characteristics. The first trial was  conducted at Boucherville, near Montreal. Boucherville was a former
electrical distribution station with a heavy clay soil contaminated by up to 14,000 mg/kg of mineral oil and grease.
The second site, located in Jonquiere, north of Quebec City, was contaminated by up to 54,000 mg/kg mineral oil
and grease, which resulted from spills of locomotive fuel oil. The interim results of these two studies included bench-
scale treatability studies (optimizing aeration, tilling, type and amount of soil amendments, and nutrient additions)
and field tests (Quebec City meeting, 1993).

The field tests used 500 m3 of contaminated soil from each site which was transported offsite to prepared treatment
facilities. At the Boucherville site, remediation was continued during the winter by installing a 15-kW heater to the
pile during November. After 34 weeks of treatment, 72% of the hydrocarbons  present were  degraded (final
concentration 3,800  mg/kg) despite over half the test period taking place during the winter. At the Jonquiere site,
after 28 weeks of treatment, up to 68% of the hydrocarbons present had been treated (final concentration 17,100
mg/kg). This  trial did not continue over the winter.

It was concluded that biopiles were effective for soils with high levels of contamination and for difficult soil types
such as heavy clay. However, the Jonquiere soil did not meet criteria for  industrial soil, and neither cleaned soil met
the criteria for residential redevelopment. The bench-scale treatability studies showed a degradation efficiency of
about 80-90% after 20 weeks  which was not reproduced by the field tests.

Application costs  predicted from these  studies were around 90 Canadian dollars" (Cdn$90) per m3  for the whole
treatment.
   * It is assumed that costs were reported in Canadian dollars, rather than U.S. dollars.
                                                   21

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

3. TREATMENT OF EMISSIONS

The Biogenie, Inc. approach reportedly uses an innovative filtration system for treating off-gases (containing several
hundred parts per billion volatile organic compounds (VOCs) from the exhaust of the air blower system. Whereas
traditional technologies include carbon adsorption, thermal or catalytic oxidation, and wet scrubbing, the Biogenie
Inc. process uses a biofilter. The biofilter technology was field tested at a petrol station near Quebec City where soil
and groundwater had been contaminated with petrol leaks from an underground storage tank. A combination of in
situ bioventing and air sparging was used to clean the site with the biofilter employed to treat the off-gases produced.

The biofilter technology consists of an air/water separator followed by  a trickling filter and a biofilter in series. The
separator was used to eliminate excess water in the inlet gas. A mixed culture of microbial hydrocarbon degraders
(sustained on VOCs) was housed in a trickling filter, which was supported by nutrients added to irrigating water.
Contaminated air was  bubbled through  the  filter countercurrent to  the irrigation water. After the trickling filter,
contaminated air and water were passed through a biofilter consisting of a column filled with composting materials.
Contaminants were sorbed to the filter medium and degraded by indigenous microbes (from site soil) and transported
microbes (from trickle filter). No nutrient supplementation was needed  since the compost provided sufficient levels.
During the demonstration, the system operated under winter conditions with results after 127 days showing a removal
of 90% of BTEX compounds (benzene, toluene, ethylbenzene, and xylenes) with a filter residence time of less than
2 minutes.

4. OTHER STUDIES

By February 1996, a further 17 field studies were conducted using the biopile and biofilter technologies.  In total,
over 390,000 tonnesb of contaminated soil have been treated, with  average removal efficiencies of 75% for mineral
oils, 99% for BTEX, 95% for  pentachlorophenols, and 90% for total PAHs (>99% for naphthalene).

5. COSTS

Remedial costs were suggested to vary according to soil quantity and site location but, in general were reported to
be between Cdn$45-90 per ton for soil quantities up to 5,000 tons, and between Cdn$15-30 per ton for soil quantities
greater than 25,000 tons.
        The reports do not make it clear whether the term "tons" refers to metric ton(nes), short (U.S.) tons (2,000 Ibs), or imperial tons (2,240
        Ibs).
                                                    22

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 9

       Field Demonstration of an In Situ Process for Soil Remediation Using Well Points
   Technical Contact:
   Arnold Ross
   Serrener Consultation Inc.
   855,  rue Pepin, Bureau 200
   Sherbrooke, Quebec J1L 2P8

   tel: +1/818-829-0101
   fax: +1/819-829-2717
Country:
Canada
Project Status:
Accepted by study
1993
1. INTRODUCTION

An in situ remediation process combining  soil flushing using surfactants and  bioremediation is being field
demonstrated  at  an  aliphatic-  and  hydrocarbon-contaminated  site.  This  demonstration  is  sponsored  by
Serrener/Varisco Consortium, the National Research Council of Canada, Environment Canada, and the Ministry of
Environment of Quebec.

2. TECHNICAL CONCEPT

The process is based on a wellpoint system  developed by VARISCO SPA for the control of the water table in
construction projects. The treatment process consists of a recirculation system with injection and extraction wells.
A surfactant/co-surfactant solution is injected into the wellpoints to immobilize the contaminants. The extracted
contaminated washings are sent to an effluent treatment plant. The  remaining hydrocarbon residual located in the
subsurface is biodegraded by injecting nutrients and air or hydrogen peroxide into the wellpoints.
The project scope includes: laboratory-scale selection of  surfactants; in situ soil washing tests using the selected
surfactant; in situ biodegradation testing of residual hydrocarbons and the contaminated washing solution; and
subsequent monitoring to verify contaminant removal.

Over 50 types of surfactants and co-surfactants were tested to establish the pairings and concentrations required to
extract over 95% of the hydrocarbons from the contaminated soil.  Tests on the toxicity of the surfactants on the
microbial populations and effects on degradation are being conducted.

BTEX (benzene, toluene, ethylbenzene, and  xylenes) and aliphatic hydrocarbon degraders have been identified.
Degradation of the aliphatic hydrocarbons present in the  soil appears to be inhibited, perhaps by the presence of
heavy metals, and will be studied further.
                                                  23

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 10
               Integrated Treatment Technology for the Recovery of Inorganic and Organic
                                          Contaminants from Soil
   Technical Contact:
   Bruce Holbein
   Tallon Metal Technologies Inc.
   67 Watson Road South, Unit 4
   Guelph, Ontario N1H6H8
Country:
Canada
Project Status:
Accepted by study  1993
Final report         1994
1. INTRODUCTION

Two sites in Canada exceeded guidelines for industrial reuse  due to  soil concentrations  of heavy metals  and
polycyclic aromatic hydrocarbons (PAHs). The Ataratiri site covers 32 hectares in Toronto and was contaminated
with PAHs and heavy metals such as lead, copper, and zinc. It was estimated that a minimum of 500,000 tonnes of
soil on the site did not meet the commercial/industrial usage  criteria.  At the Longue Pointe site  in Montreal,
approximately 155,000 tonnes of soil was contaminated with lead smelter fly ash and by lead battery recycling. In
places, site soils had a high clay content or very high lead content. Bulk soil samples (up to 35 tonnes) representative
of distinct areas of both sites were collected for pilot scale evaluation of the treatment technology. The technology
was also  evaluated at the bench scale on sediments collected from Hamilton Harbour in Ontario. These sediments
were enriched in organic and metal contaminants and had a total organic carbon load exceeding 10%. Some of the
metals (e.g., Cd, Cr, Fe, Mn, Ni, Pb, and Zn) exceeded the "severe effect level" based on Ontario's sediment quality
guidelines.

Demonstration and evaluation of the technology was sponsored by the Ontario Ministry of Environment and Energy,
Environment  Canada through its Development  and Demonstration of Site Remediation Technologies  (DESRT)
program and the Contaminated Sediment Treatment Technologies Program, and National Defence Canada.

2. THE PROCESS

The evaluated remediation technology incorporated physical-chemical treatments to recover both metal and organic
contaminants. Pretreatment included the screening of large site debris, such as bricks, and a wet scrubbing to break
down softer agglomerates  such as clay clods. The slurried soil was then classified by particle size, density, or
magnetic susceptibility to produce contaminant concentrates which were reported to be recycled offsite, and a "clean"
fraction, which required further treatment through the organic and metal washing processes.

Organic contaminants were washed from the soil using a solvent extraction process. The organic-depleted soil slurry
was subsequently treated in a hydrometallurgical circuit where fine metal contamination was selectively leached from
solution and recovered using metal selective adsorbents. It was suggested that the filter cake produced by this process
was suited to off-site metal recycling. The depleted soil slurry was then washed, dewatered, and recombined with
the other soil products.

3. RESULTS

3.1 Ataratiri

The pretreatment of the Ataratiri soils produced metal-contaminated concentrates containing up to 55% by weight
of iron for potential recycling. Overall,  the "clean" soil fractions, which represented up to 95% of the original soil
weight, appeared to comply with residential criteria for some contaminants, and industrial criteria for others, such
as certain PAHs. For  example, benzo(b)fluoranthene in the soil was reduced from 14 mg/kg to 4 mg/kg after treat-
ment, but this value still exceeded residential guideline values. Reduction in metal concentrations ranged from 4,026
mg/kg to 360 mg/kg for zinc, while copper levels (100 mg/kg) were unaffected by treatment.

A report highlighted the different outcomes of Ataratiri soils treated by coarse metal recovery and organic recovery
only, and with the added hydrometallurgical recovery step. After treatments to recover coarse metal and organic
                                                  24

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

contaminants, two soils that originally exceeded industrial guidelines for various heavy metals and PAHs resulted
in 95% of the original material that were within industrial guidelines for certain PAHs. Other soils that had the
additional hydrometallurgical recovery step produced 86% of the  material as clean soil suitable for reuse for
residential purposes.

3.2 Longue Pointe Soils

At Longue Pointe, treatment tests on the soils revealed that most of the lead contamination occurred in fine soil
particles and could be recovered by the hydrometallurgical extraction and Vitrokele™ adsorption. After this treatment,
soil contained as little as 7% of its initial lead content, and this residual concentration met the regulatory limits of
either 1000 mg/kg for industrial use or 500 mg/kg for residential use.

It was reported that the Longue Pointe  site is now undergoing full-scale remediation. Overall, 115,000 tonnes of lead-
contaminated soil were to be treated through the plant at a rate of 600 tonnes per day. Excavation of the soil from
the Longue Pointe site was to be secured in a temporary on-site holding cell by late autumn 1994 with treatment
scheduled to begin by summer 1995.

3.3 Hamilton Harbour Sediments

Bench-scale tests with a 20-kg test sample collected from Hamilton Harbour included flotation, magnetic separation,
and gravity separation techniques. When compared to the original feed, clean tailings contained only 4.1% of the
total oil and grease, 4.9% of the PAHs, less than 2% of the lead, and 11% of the zinc. In contrast, metals like Cd,
Ni, and Fe were reduced but remained between 24% and 30% of the original load. The recovered organic product
contained 27% of the total mass and captured 62% and 74% of the original oil and grease and PAHs, respectively.
It was also enriched in Pb, Zn, Cd, Ni, and Fe at 81%, 60%,  54%,  and 32%, respectively, of the original content.

4.  COSTS

It was indicated that for Ataratiri and  Longue Pointe, treatment produced a soil which was of a suitable quality for
recycling. An estimate of the remediation cost of the Ataratiri site, involving a total of 500,000 tonnes of soil, was
presented as U.S.$75 per tonne.

Based on bench-scale tests, the treatment process for Hamilton Harbour sediments would consist of screening,
magnetic separation, concentration of organics by flotation, and a hydrometallurgical extraction if required to remove
residual Zn, Cu,  Ni, and Pb from feeds to the  leach circuit. For this treatment, the  contractor  estimated costs of
U.S.$60-75 per tonne assuming at least 20,000 tonnes require treatment.

5.  RESIDUALS

Residuals consist of inorganic concentrates for disposal or recovery of metals, organic concentrates, waste water from
sizing and classifying material during treatment, as well as treated material for use as fill on-shore or off-shore.
                                                   25

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)
    Project No. 11

                    On-site Biodegradation of PAHs in Soil at a Former Gasworks Site
   Technical Contact:
   J0rn Bo Larsen
   Ramb0ll Consultants
   Teknikerbyen 31
   DK-2830 Virum
   tel: +45/45 98 8300
   fax: +45/45 98 8520
Country:
Denmark
Project Status:
Accepted by study  1993
Interim report       1994
Final report         1997
This project reports on the experimental studies at the Frederiksberg  Gasworks  site  located in the middle  of
Copenhagen, Denmark. The gasworks operated from 1895 to 1964 for the production of coal gas which gave rise
to several wastes, such as naptha, benzol, ammonia, cyanide,  sulfides,  and sulfates. Investigation of the site  in
connection with the demolition of the site buildings identified  a large area of soil contaminated  with polycyclic
aromatic hydrocarbons (PAHs). Analysis of the  site  soil indicated that concentrations of individual PAHs were in
the range 0.8-23 mg/kg, and the average total PAH concentration was 400 mg/kg.

The objective of the project was to define the most suitable method for the biological treatment of the tar compounds
in the soil. Laboratory tests were used to determine the population of tar degrading micro-organisms present within
the site soil. Identified degraders were  subsequently evaluated for their interaction with temperature, humidity,
nutrients, and composting materials such as bark and wood chips. The effect of adding proprietary detergents to the
system was also evaluated.

Among the more important conclusions of the study, was that the condition of the soil at the time of sampling and
analysis (e.g., consistency and soil moisture content) may affect the measured PAH concentrations.

1. INTRODUCTION

The pilot project at Frederiksberg gasworks site was one  of six pilot projects on the  remediation of municipal
gasworks sites financed by the Danish government and the municipalities  involved. The pilot project involved three
phases, including both bench-scale and field tests:

  •  Phase 1: The population of microorganisms and the population of tar-degrading microorganisms in the soil from
    the site was investigated. In addition, 10 detergents were assessed for their ability to mobilize PAHs and make
    them accessible to biodegradation.

  •  Phase 2: Microcosm tests were conducted to monitor the degradation of PAHs  under different test conditions
    over a period  of  325 days.  Column tests  were conducted over a period of  126 days. Subsequently, the
    degradation of PAHs was examined in a slurry test in which soil samples from the microcosm test were
    suspended in distilled water and put on a shaking table in the dark for 69 days.

  •  Phase 3: The degradation of PAHs was monitored for 496 days in  10 windrows: five in the  open air on the
    former gas tank foundation and five in the former covered tar basin. The affect of the following test conditions
    on PAH degradation was examined: addition of wood chips, addition of compost, addition of detergents, aeration
    by windrow turning, continuous  aeration,  windrow location (inside vs. outside).

2. BENCH-SCALE TESTS

2.1 Assessment of Microbial Populations

In Phase 1, the number of natural PAH-degrading bacteria was  measured by adding 14C-marked PAH compounds
to soil samples and determining the quantity of radiolabeled carbon dioxide that was released. Results showed that
                                                  26

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
a population of 107 bacteria per gram of soil were present on the site and that they showed potential for degrading
phenanthrene, anthracene, and pyrene.

2.2 Examination of the Microbial Degradation over a Period of 325 Days

During the Phase 2 bench-scale tests (microcosm tests), the degradation and removal of PAHs from site soil by three
treatments (radiation, temperature, and detergent) was monitored over a period of 325 days. The soil was thoroughly
homogenized and particles greater than 4 mm were removed prior to the test.

PAH degradation was insignificant during the first four months (120 days), but after four months, the concentration
of the PAHs examined (phenanthrene, anthracene, fluoranthene, pyrene, andbenzo[a]pyrene) decreased significantly.
The number of phenanthrene-degrading bacteria decreased in the initial period, but generally increased during the
last 198 days of the test.  The total number of cultivable bacteria did not  show similar fluctuations. The PAH
degrading bacteria were, therefore, probably affected by the initial test conditions, but adapted during the first four
months. The reason for this temporary inhibition is believed primarily to be the high pH of in the soil after addition
of nutrient salts.

During the first 127 days of the test, the concentrations of PAHs were generally lower in the samples treated by
detergents than in the corresponding radiated samples and background samples. The detergent mix applied probably
had an initial effect on the degradation of PAH that diminished and disappeared in time, probably due to total or
partial microbial conversion of the detergent. The addition of detergent to the  soil did not enhance the growth of the
phenanthrene-degrading bacteria.

An increased incubation temperature of 25°C compared with 15°C did not generally affect the removal of PAHs.

During the  last 198 days of the test period, 66%, 74%, 66%, 64%, and 71%  of phenanthrene, anthracene, fluoran-
thene, pyrene, and benzo(a)pyrene, respectively, was removed independently of the treatment given (Table 1).
       Table 1: Percentage of PAHs removed in the microcosm test from day 127 to day 325 (198 days)
                              calculated using estimated regression equations
PAH
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)pyrene
mg PAH/kg soil DM
day 1271
9.8
6.2
17.4
12.8
6.6
day 325
3.4
1.6
5.9
4.6
1.9
%PAH removed
65.7
73.5
66.4
64.3
71.2
tl/2
128
103
126
133
110
1 No degradation or removal of PAHs occurred from day 0 to day 127. Note: t!/2=half-life
2.3 Examination of the Microbial Degradation in Slurry Tests over a Period of 69 Days

Soil samples from the microcosm test were suspended in distilled water and put on a shaking table at 100 rpm. The
samples were shaken in the dark for 69 days (slurry test) with the temperature maintained at 22°C. The PAH content
of the soil and water phases were analyzed and it was concluded that:

  •  Incubation of the suspended soil generally did not lead to any reduction of PAHs. Exceptions were anthracene
    and fluoranthene, the concentrations of which tended to decrease after the 69-day test.
                                                  27

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
  •  There was  no difference in degradation rates between the different treatments (radiation, temperature, and
    detergent) of the test soil.

  •  The content of benzo(a)pyrene tended to increase.

The reason for the lack of reduction and degradation of PAHs may be the age of the soil contamination (i.e., a very
strong bond of the PAHs to the soil particles and possibly to tar residues) and the low initial concentrations of PAHs.
The benzo(a)pyrene content significantly increased at the conclusion of the test. This may have been due to either
analytical variation or to mobilization of benzo(a)pyrene during the test. If the latter is the case, both mobilization
of PAHs and partial mineralization of the mobilized PAHs may have occurred at the same time.

3. FIELD TESTS

Field tests were initiated at Frederiksberg gasworks site in June 1994 and were concluded in October 1995 (a total
of 496  days). The soil used in the field tests was excavated in and around the former tar basin in 1990. During this
time, the soil was processed to remove wood, bricks, concrete, and heavily tar-contaminated soil and then stored in
two windrows at the gasworks site.

The test soil was a sandy soil with a total PAH concentration of around 400 mg/kg. Prior to placing the soil into the
test windrows, the soil was mixed and homogenized, and particles larger than 80 mm were removed. The soil was
then split into five piles,  and water, nutrient salts in the form of NPK fertilizer, wood chips, compost, and detergent
solution were added. Then the soil was placed in 10 windrows of which five were located in the open on the former
gas tank foundation (windrows 1G-5G) and five were  located in the covered former tar basin (windrows 1B-5B).

Table 2 shows the treatment given to each test windrow and the treatment carried out in the test period.  Table 3
shows the results of chemical analyses and measurements carried out on the test soil.
                                 Table 2. Treatment of each test windrow
Windrow
1G (control)
IB (control)
2G
2B
3G
3B
4G
4B
5G
5B
Soil Volume
(m3)
33
21
19
19
20
20
32
32
33
36
Aeration Method2
-
-
Turning
Aeration
Turning
Aeration
Turning
Turning
Turning
Turning
Treatment at
Initiation of Test
-
-
+ 1 1 m3 wood chips
+ 1 1 m3 wood chips
+ 10 m3 compost
+ 10 m3 compost
+ detergent solution3
+ detergent solution3
-
-
Treatment During the
Test
Addition, as required, of:
- NPK fertilizer
- calcium nitrate
- water

4. RESULTS OF THE FIELD TEST

4.1  Initial Processing of Results: Correlation Analysis

A correlation analysis of the results of all 946 samples analyzed was carried out to determine whether the results
for  total PAHs correlated with the individual PAHs in the samples. From the results of the correlation analysis, it
                                                   28

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
generally appears that there is poor correlation between total PAH and individual PAH concentrations. In further
processing of the results, only the individual PAH results were used.
                    Table 3. Time for sampling of chemical analyses and measurements
Chemical analyses and
measurements on
samples1
Tar (PAH)3
Tar (PAH)3
Ignition loss
Dry substance
pH
Inorganic material4
Microbiology3
Measurements on site
Humidity
Oxygen (O2), carbon
dioxide (CO2) and
methane (CH4)
Time in days2
0
+
-
+
+
+
+
-
55
+
-
+
+
+
-
-
103
+
134
+
173
+
289
+
320
+
372 406 44
+ + +
1 468 496
+ +
Samples frozen and after the conclusion of the test freeze-dried
and analyzed for their content of tar
+
+
+
-
-
+
+
+
+
-
Time
+
+
+
-
-
+
+
+
-
+

Measurements carried out continuously
Measurements carried out on the dates
during short periods of the test period
+
+
+
+
-

+ + +
+ + +
+ + +
+ +
+ - +

+ +
+ +
+ +
+
+

throughout the test period
mentioned in the table
and every day
1 It not otherwise indicated, all windrows were subjected to analyses and measurements.
2 Time 0 is 94.05.20 and time 488 is 95.10.02
3 12 PAH single substances by GC/MS/SIM, and total PAH by GC/FID
4 Total carbon, total nitrogen, total phosphorous, total potassium, nitrate number, potassium number, and
phosphorous number
5 Determination of the population of total cultivable bacteria and phenanthrene decomposers by plate
count and determination of PAH degradation potential for phenanthrene and pyrene decomposers
(the latter was conducted on day 441 and only in windrows 1G-5G).
4.2 PAH Content in Ordinary Samples and in Freeze-dried Samples

Soil samples from sampling rounds t2 through tu were collected from each windrow and frozen for analysis the
conclusion of the test. At this time, all of the frozen samples were freeze-dried and analyzed for PAHs. The purpose
of freezing and freeze-drying samples from each round for concurrent analysis was:

  •to examine whether when the samples were analyzed might have affected their results. This may be indicated by
the results from analysis rounds VV Hence, results from the same sampling round, in general, tended to be similar
in that when a  high concentration was measured in one sample, high contents were typically measured in the other
samples.

  •to assess whether the water content (the soil consistency) of the samples affect their results. (Very different soil
consistencies were observed between the windrows located inside and outside).

Generally, the effect on the PAH content of ordinary samples (samples analyzed immediately after sampling rounds
Vtn)  and those samples that were frozen and freeze-dried for analysis at the conclusion of the test), respectively,
is considered significant (a = 5%). The freeze-dried samples from  the windrows located inside (1B-5B) had an
average PAH content that was about 70% of the PAH content of ordinary samples. The ordinary samples from the
windrows located outside (2G-5G), on the other hand, had an average PAH content that was about 60% of the PAH
content of the  freeze-dried samples. The exception was samples taken from windrow  1G where the freeze-dried
samples had a  PAH content of around 70% of the content found in the ordinary  samples.
                                                  29

-------
NATO/CCMS Pilot Study, Phase II
                                                              Final Report (Appendix IV)
Comparison of the results for ordinary and freeze-dried samples showed that there was a significant difference
between the PAH content of ordinary samples from windrows 1B-5B and 1G and the PAH content of windrows 2G-
5G. In contrast, the content of PAHs in frozen and freeze-dried samples was almost identical (Figure  1).

Figure 1. Average content of PAHs in group  1-5B and 1G and group 2-5G, respectively
                          Average content in groups 1-5B-HG and 2-5G, respectively
20-
           I
           •S   15
               O -
                                                                    • 1 -5B-H <3 Ordinary
                                                                    • 1 -SB-i-1 G Freeze-
                                                                    n 2-5G Freeze-dried
                                                                    d 2-5G Ordinary
The only general difference between the ordinary and the frozen freeze-dried samples was that the water content in
the freeze-dried samples was markedly lower than in the ordinary samples. The water content could possibly have
a negative or a positive impact on the efficiency of the extraction of PAHs from the samples; however, a clear-cut
connection was not established between the water content of the samples and the PAH concentrations in the ordinary
and freeze-dried samples.

It was not immediately possible to explain the difference in PAH content of the ordinary samples measured in test
windrows 1B-5B & 1G and test windrows 2G-5G. As mentioned previously, the ordinary samples from windrows
2G-5G generally had a lower dry matter content than the ordinary samples from 1B-5B and 1G. A marked difference
in the soil consistency of the above two groups of samples was also registered during sampling and by the analysis
laboratory. Thus, the  samples from test windrows 2G-5G had a muddier consistency than the samples from test
windrows 1B-5B (located inside) and windrow 1G (located outside, but not turned during the test period).

In summary, it was concluded that freezing and freeze-drying of samples created similar soil conditions (including
soil consistency and water content). This,  in turn, resulted  in similar PAH contents among test groups. In contrast,
the difference in soil conditions of ordinary samples affected their PAH content. Thus, it was concluded that the soil
consistency and water content affect PAH content.

4.3 Change  in PAH  Content over a Period of Time

The reduction in concentration of the six PAHs analyzed during the test, under the test conditions applied, was
examined regression  analyses.  Soil samples were collected from both the inside and outside test windrows. Soil
samples were collected during each of the 12 sampling rounds (t0-tu) over a period of 496 days. In the regression
analyses, the most probable linear reduction in the PAH content of the samples with time was estimated. It was also
examined whether the slope of the decay curve could be  assumed  to differ from 0, which would then indicate a
significant (a = 5%)  reduction in the PAH content of the samples over the test period.
                                                  30

-------
NATO/CCMS Pilot Study, Phase II

4.3.1 Effect of different treatments
Final Report (Appendix IV)
During the test period of 496 days, a significant reduction in the PAH content in test windrows IB, 1G, 4B, 4G, and
5B was registered. From the results, it appears that there was no significant difference in the half-lives (tYz) of these
windrows. Hence, the addition of detergent solution did not lead to an increased removal of PAHs, as compared with
the removal of PAHs in test windrows IB and 1G (see Table 4).

In none of the windrows with a significant reduction of the PAH content, could the  effect be unambiguously
attributed to the treatment the windrows received. Neither addition  of wood chips nor compost had any positive
impact on the removal of PAHs over time. On the other hand, the results achieved for test windrows 1G and IB
could be interpreted to mean that addition of wood chips and compost inhibited the removal of PAHs.

           Table 4. Reduction of PAHs in test windrows located inside (1B-5B) and test windrows
                        located outside (1G-5G) during the test period of 496 days.

IB
2B
3B
4B
5B
1G
2G
3G
4G
5G
Phenanthrene
Effect
Effect
None
Effect
Effect
None
Effect
Effect
Effect
None
Effect
tl,2
573

491
608

751
583
466

602
Anthracene
Effect
None
None
None
None
None
None
None
None
None
Effect
tl,2









809
Chrysene
Effect
Effect
None
None
Effect
None
Effect
None
None
Effect
Effect
tl,2
610


626

636


412
575
Fluoranthene
Effect
Effect
None
None
Effect
None
Effect
None
None
Effect
Effect
tl,2
720


742

760


623
600
Pyrene
Effect
Effect
None
None
Effect
None
Effect
None
None
Effect
Effect
tl,2
634


701

807


576
565
Benzo(a)pyrene
Effect
Effect
None
None
Effect
None
Effect
None
None
Effect
Effect
tl,2
565


578

717


415
474
ti;2 = half-life, in days, following t0
Effect= indicates significant decrease of substance content over time
None= indicates no significant decrease of substance content over time
In general, no significant removal anthracene was observed in any of the test windrows. This may be due to the fact
that the content  of anthracene,  as compared with the  content of other PAHs, was generally low. Hence, the
anthracene may not have been accessible to microbial degradation, or the removal of anthracene might have been
overshadowed by the relatively great standard deviations of the results.

4.3.2 Effect of location of test windrows: inside/outside

No difference was found in the development of PAH content between test windrows located inside and test windrows
located outside. An exception was test windrow  5G in which PAHs were significantly reduced, whereas there was
no significant reduction in windrow 5B. The reason for the difference in the results from windrow 5B (located inside)
and windrow  5G is not readily explainable. No difference  in the development of PAH content between windrows
aerated by turning (2G, 3G) and windrows aerated continuously (2B, 3B) was found.

From the results  of the ordinary  samples, it appears that within the test period a relatively great difference in the
PAH content  was observed during period t4-t8, when the content in test  windrows 2G-5G was lower than in test
windrows  1B-5B and 1G. The fact that  no mineralization of the PAHs occurred in windrows 2G-5G during this
period appears from the results from time period t9-tn, in which the PAH content is generally larger than the content
of PAHs in period t4-t8.

A lower PAH content was observed in windrows 2G-5G than in 1G, but these windrows were all located outside
so the difference  in the results cannot be explained their location. The consistency of the soil differed greatly for the
two groups of windrows  in the period t4-t8. Hence, the soil in windrows 2G-5G had an almost muddy consistency.
Windrow 1G, which was not turned, did not have this muddy consistency. It should be noted that this  great
difference  in soil consistency was not clearly reflected by data on dry matter content.
                                                  31

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

The manner in which the difference in soil consistency affected the accessibility of the PAHs was not conclusively
determined. However, possible reasons for the low content of PAHs in windrows 2G-5G during t4-t8 include:

  •  The PAHs were leached to the bottom of the windrow. This scenario is not very likely because soil samples
    were taken for analysis while turning the windrows; therefore, soil was taken from the entire  soil profile.

  •  Elution of PAHs to leachate. This is also not very likely because the content of PAHs was high during time
    period VV Moreover, a number of elution tests (bench-scale tests) established that the elution of PAHs from
    the soil was insignificant.

  •  The change of the soil consistency resulted in changes in bonding and adsorption, etc., of PAHs to the soil
    particles so that the accessibility of PAHs in the soil from windrows 2G-5G was  reduced. The ability  of the
    extraction agent dichloromethane to extract PAHs is affected by the soil structure. Considering the results of the
    freeze-dried samples in which the content of PAHs in this period was the same for windrows located outside
    as for those located inside, this explanation is believed to be possible.

4.3.3 Estimate of the Amount of Degraded PAH

In test windrows IB, 1G, 4B, 4G, and 5G, a 23.5-45.5% reduction in the content of the selected PAHs was estimated
in the test period of 496 days (see Figure 2). An exception was anthracene. This corresponded to half-lives between
529 and 967 days  for each PAH (see Table  5). In the group of test windrows comprising 2B, 2G, 3B, 3G, and 5B,
there was a significant reduction of phenanthrene (39.3%, corresponding  to a half live of 618 days), fluoranthene
(18%, corresponding to a  half live of 1,428 days) and  pyrene (15.6%, corresponding to a half life of 1,494  days).
No significant reduction was observed in anthracene, chrysene, or benzo(a)pyrene  (see Table 5).
                                                  32

-------
NATO/CCMS Pilot Study, Phase II
           Final Report (Appendix IV)
Figure 2.Reduction rates for PAHs and standard deviations for windrows IB, 1G, 4B, 4G, and 5G, and
windrows 2B, 3B, 3G, and 5B, respectively.
                                            Reduction rates
                           -0.030-
                                                                      D Phenanthrene
                                                                      • Fluoranthene
                                                                      • Pyrene
                                                                      01 Chrysene
                                                                      D Benzo(a)pyren|e
                                                                      D Anthracene
                            0.000
                                         1B-4B-1G-
                                           4G-5G
2B-3B-5B-
  2G-3G
Generally, it can be expected that due  to the chemical structure and physical and chemical properties (water
solubility, log Kow, etc.) the biodegradation of phenanthrene and anthracene will be greater than other PAHs. From
Table 5, it appears that the conversion of anthracene is significantly smaller than the conversion of the other PAHs,
which may be attributed to the low output concentrations or the great standard deviation of results, which may have
overshadowed the effect.

Addition of  14C-marked PAH substances  to the test soil showed a relatively quick mineralization of phenanthrene,
anthracene, and pyrene with a t!/2 of greater than 100 days. In the microcosm test, a reduction in the PAH content
with a half-life of 110-133 days for each PAH was measured. After conclusion of the microcosm test, slurry tests
(soil suspended in water on shaking table) were carried out for 69 days. The lowest and highest concentrations of
the selected  PAHs were 1-2 mg/kg DM anthracene and around 5 mg/kg DM fluoranthene, respectively. After the
69 days, no significant reduction in the content of each PAH could be  registered. This indicates that a further
reduction of the PAH content was not possible even if the microcosm test had been longer.

As mentioned above, a reduction of the PAH content with a half life of 529-967 days was observed in the field test.
At the conclusion of the test on the 496th day, the content of each PAH was around 3-4 times higher than in the
microcosm test (which was the initial concentration in the slurry test). It may well be that the lower concentrations
for degradation of the PAH substances  in the test  soil  had been reached in the field test. Thus,  the achievable
minimum concentrations are higher in the field test than in the bench-scale test, which can be explained by the fact
that the soil in the field test is much more homogeneous and was manipulated less than the soil applied for the
bench-scale test. This becomes obvious when looking at the significantly greater standard deviations observed in the
analytical results from the field test, as compared with those of the bench-scale test. It should be noted that soil
particles greater than 4 mm were removed from the bench-scale test soil, whereas particles greater than 80 mm were
removed from the field test soil.
                                                   33

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
 Table 5. Percentage of PAH removed from windrows IB, 1G, 4B, 4G and 5G and windrows 2B, 2G, 3B, 3G,
                  and 5B in the test period (496 days), calculated using regression equations.
Test Windrows
IB, 1G, 4B,
4G, and 5G
2B, 2G, 3G,
3B, and 5B
PAH
Phenanthrene
Anthracene
Chrysene
Fluoranthene
Pyrene
Benzo(a)pyrene
Phenanthrene
Anthracene
Chrysene
Fluoranthene
Pyrene
Benzo(a)pyrene
mg PAH/kg DM
t = 0 days
19.444
8.436
13.156
28.842
21.527
12.005
16.425
-
-
21.683
15.922
-
t = 496 days
12.996
6.452
7.204
18.426
13.095
6.549
9.977
-
-
17.715
13.442
-
%PAH
Removed
33.2
23.5
45.2
36,1
39.2
45.5
39.3
-
-
18.3
15.6
-
tl/2
(days)
627
967
556
674
636
529
618
-
-
1428
1494
-
Note: t1/2= half-life
On the basis of the test, it is not deemed realistic to attempt a complete degradation/reduction of the selected PAH
in the soil from the Frederiksberg Gasworks site through on-site biodegradation. This should be seen in the light of
the results achieved in the present tests and  results reported in literature. Whether  the same is  true for PAH-
contaminated soil from other gasworks sites, and from other types of tar contamination,  will to a great extent depend
on the following factors:

  •  age of the contamination;
  •  "form" and distribution of the contamination in the soil; and
  •  soil composition. (Is the soil sandy  or clayey with a high or low content of organic matter? Which qualitative
    properties does the organic fraction  have? At old gasworks sites a relatively high content of coal and coke dust
    particles to which the PAHs can adsorb can be expected.

It was also concluded that it is doubtful whether the biodegradation of PAHs in contaminated soil can meet the limit
values of 5 mg total PAH/kg soil DM and 0.1  mg benzo(a)pyrene/kg  soil DM that currently exist in Denmark. At
any rate, it is a precondition that the extraction in the chemical analysis is conducted with dichloromethane.

The report emphasizes that the part of the PAH reduction that can be attributed to biological processes (total and
partial mineralization) has not been fully determined. Consequently, total mineralization of PAHs is often deemed
less important than adsorption processes or than partial mineralization combined with adsorption processes. It has
not been determined whether PAHs adsorbed to soil particles (which are not accessible to the chemical analysis) can
be mobilized  again at a later time  and thereby become accessible to microbial degradation.

Toxicity investigations of PAH-contaminated soil have shown a lower toxicity in soils with a high content of organic
matter, because PAHs are adsorbed to the organic fraction in the soil. In these tests, the PAH substances were not
bound more than they could be extracted in the chemical analysis. The fact that soil with the same content of PAHs
can have widely different toxicity  is interesting and requires further examination.

5. CONCLUSIONS
                                                   34

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

The study showed shown that on-site biodegradation of the 3- to 5-ringed PAHs is a slow process. Moreover, none
of the different treatments of the test soil (addition of wood chips, compost, detergent etc.) resulted in a significant
reduction in treatment time. At many gasworks sites there will be large standard deviations in the results, even with
a large number of chemical analyses conducted at the same time. This is believed to be due to the distribution of
PAHs in the soil and the form of the PAH substances (e.g., large or small tar residues etc.).

Analyses of the  same soil at different times is deemed problematic as changes in the soil conditions (e.g., soil
consistency, including water content, etc.) affect the ability of the extration agent to extract PAHs and, thus, affect
the analytical  results.  The field test, which was  based on a large number of analyses (a total of 960 chemical
analyses), has  shown that more detailed standardized methods for determining the PAH content in soil are required.
The standardization should include:

  •  pre-treatment of PAH-contaminated soil for chemical  analysis;
  •  method of analysis, including choice of extraction agent and method;
  •  selection of PAHs representing the contamination; and
  •  the result  should be compared with a standardized description of the soil.

Moreover, execution of intercalibrations between laboratories is required. Such intercalibrations could be part of an
accreditation scheme for the laboratories.

As far as PAH-contaminated soil is  concerned, an assessment of remedial  action (e.g., biodegradation) should not
be based on measurement of total PAH concentration  alone,  but instead on individual PAH  concentrations.
Furthermore,  due to the large standard deviations, it is insufficient to base remedial actions on a few chemical
analyses or measurements in an area. In addition to a greater number of chemical analyses and measurements it is
important to conduct a more detailed assessment of the physical and chemical properties of the soil.  As a final basis
for a decision  on remedial actions, it may prove relevant to assess the toxicity of the soil.

6. REFERENCE

    J. O. Larsen. Experience of on-site biodegradation of tar at Frederiksberg gasworks site, Paper to ATV Meeting
    on Contaminated  Gasworks Sites, Schaeffergarden, October  1996.

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 12

           Groundwater and Soil Remediation at a Former Manganese Sulfate Production Plant
   Technical Contact:
   Karin Christiansen
   Carl Bro Consultants
   Graskoven 8
   Glostrup 2600
   tel: +45/43-48-6979
   fax: +45/43-48-4414
Country:
Denmark
Project Status:
Accepted by study 1993
Final report 1994
1. INTRODUCTION

The former manganese sulfate (MnSO4) plant in the Municipality of Tinglev, Denmark was found to contain severe
sulfate (SO4) and cyanide (CN) contamination of the soil and groundwater resulting from on-site disposal of about
45,000 m3 of hazardous waste. The plant operated from 1932-1970, and industrial operations included the use of coal
gasification wastes as a raw materia; the disposal of production residues and gasworks waste is the source of most
of the site contamination. The production residues contaminated groundwater with manganese and sulfate (sulfate
concentrations up to 1,600 mg/L), while the gasworks waste caused cyanide contamination.

Off-site incineration and off-gas  treatment  of the buried wastes was estimated to  cost 8-13 million ECU. If this
traditional remediation approach  were  taken at each of the estimated 125 gasification sites in Denmark, the total
clean-up costs could be economically prohibitive. To reduce these costs a pilot-scale demonstration was initiated to
evaluate alternative remedial designs.

2. EXPERIMENTAL STUDIES

2.1 Treatment Concept

The materials requiring treatment were the disposed solid wastes, leachate from the solid wastes that has infiltrated
the unsaturated zone, and contaminated groundwater. Conceptually, the leachate and most contaminated groundwater
could be collected for treatment by insertion of drains beneath  the deposits. Treatment of the wastes could be
achieved by accelerated leaching through installation of infiltration drains at the  top of the wastes to introduce
suitable treatment agents. It was estimated that if leaching at a rate of 2-3 L/S annually were achieved (40-60 times
the natural rate), it would take 7-10 years for contaminants to be leached from the wastes. A total of 160,000-240,000
m3 would have to be collected and cleaned each year. Ex situ batch processing of the wastes was also investigated.

2.2 Leachability of Wastes

Initial laboratory investigations focused on the teachability  of the solid wastes present on the site. Results indicated
that CN and SO4 leaching was highest above pH 10 and at a liquid/solid ratio larger than or equal to 20. Pilot-scale
leaching tests conducted on heaps of the production residue and the gasworks waste showed that sodium hydroxide
effectively leached sulfate  and cyanide from the production waste. However, fine particles in the gasworks  waste
resulted in clogging when the leachant was added, so the leaching tests with this waste were unsuccessful.

2.3 Treatment of Leachate  and Groundwater

Five possible methods for cleaning the leachate and contaminated groundwater were evaluated: biodegradation,
chemical precipitation, chemical  oxidation, evaporation, and reverse osmosis.  An evaluation of the economics  of
evaporation and reverse osmosis  suggested that these approaches were too expensive. Furthermore, both methods
have the drawback of producing small volumes of highly concentrated liquors.  As a result, they were not evaluated
experimentally.
                                                   36

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

Pilot-scale tests of biological treatment of the leachate required continuous supervision and were laborious, and the
biomass was insensitive to fluctuations in pH. The treatment was not evaluated further.

Laboratory experiments examined the potential of chemical oxidation and precipitation to treat specific components
of a contaminated leachate or contaminated groundwater. Experiments showed that ultraviolet (UV) light destroyed
complex cyanides in the concentration range of 1-25 ug/kg within 10 minutes.

Precipitation of excess sulphate in the leachate using barium chloride suggested this approach would not be practical
at a large scale. The procedure removed all sulphate by forming a barium sulphate precipitate. However, high levels
of barium remaining in treated liquids prevented these  liquids from being recirculated  into the  subsurface by
infiltration drains because  of the likelihood of precipitation. Also, the liquids could not be used downstream without
further treatment. In addition to this technical disadvantage, the process was estimated to be too expensive.

2.4 Treatment of Solid Wastes

Treatment of the sulfate-rich solid waste was proposed as an ex situ batch or in situ remediation process. In  the ex
situ batch process, solid waste would be mixed with water in a "ponded" area to increase chemical solubility through
intimate contact between  solution and waste.  Retention times were  as high as 120 hours. An improved design
consisted  of a fluidized bed batch treatment which eliminated the need for stirring, and water rich in sulfate was
constantly removed from the top of the reaction reservoir.

In situ treatment would involve accelerating the leaching process by using pH-neutral groundwater or an aggressive
leachant (containing sodium hydroxide) on the waste, and collecting leachate in collection  drains installed beneath
the waste. It was suggested that at a liquid/solid ratio of 20, the leachate derived from pH-neutral water may have
a concentration acceptable for being returned to the groundwater. A pilot-scale treatment facility was constructed
onsite. Heaped waste was leached using a sodium hydroxide solution. Objections to this technology are linked to the
use of chemicals during treatment. It was concluded that batch treatments are unlikely to be feasible due to high costs
and long treatment times.  Furthermore, disposal of effluent  is costly. In situ may  be more viable, but it will  take a
long time. Costs for disposing effluent may also  make this  approach unfavorable.

Treatment of the cyanide-contaminated waste was evaluated using accelerated in situ leaching and an ex situ batch
process. In situ treatment  was evaluated at pilot-scale, but was found to be unsuccessful since  fine-sized particles
present in the wastes rapidly reduced permeability  of the  heap to leachant additions. Ex situ batch processing was
evaluated at bench-scale using water at pH 7 and aqueous sodium hydroxide. Ex situ batch processing was shown
to be unworkable because subsequent solid/liquid separation proved extremely difficult. The use of water was not
recommended by the authors, since the total leaching of the cyanide was not achieved.  It was therefore concluded
that treatment of the cyanide wastes was not feasible, at present, due to lengthy treatment  times and high costs.

3. CONCLUSIONS

In conclusion, the sulfate-contaminated wastes could be treated by an in situ method, although the treatment times
could be lengthy and disposal of effluent costly.  A viable option may be to leave the material onsite and cover it
with clean soil from other  parts of the site. Growing plants on the surface will minimize infiltration of precipitation.
A  membrane cover may be another option. Regardless of the approach used, leachate and groundwater should be
monitored. The cyanide-contaminated wastes are not suitable for on-site treatment  and should be  excavated and
removed for incineration or treatment.

Future work at the site will involve modeling of the site hydrogeology to evaluate the groundwater threat.
                                                    37

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 13

         Rehabilitation of a Site Contaminated by Tar Substances Using a New On-site Technique
   Technical Contact:
   Inge-Marie Skovgaard
   Environmental Protection Agency
   29 Strandgade
   DK-1401  Copenhagen K

   tel: +44/3-266-0100
   fax: +44/3-296-1656
   E-mail: ims@mst.dk
Country:
Denmark
Project Status:
Accepted by study  1994
Final report         1996
1. INTRODUCTION

This on-site remedial demonstration project combining excavation of tar contaminated soil followed by on-site ex
situ thermal desorption was conducted at an old gasworks site in a densely populated area of Copenhagen. Excavation
was conducted inside a ventilated tent. The project was regarded as a demonstration of a new on-site technique.

2. BACKGROUND

The Valby  Gasworks was one of the largest in Copenhagen with an operating capacity of up to 300,000 m3 of coal
gas per day. Site investigation showed a wide distribution of contaminants in the soil and groundwater, including
coal tars, phenols, ammonium  compounds, cyanides, and  heavy  metals. One area of the site  contained two tar
reservoirs that had been partially filled with demolition debris after site closure. A large amount of tar was left at
the base of the reservoirs. The upper part was filled with rainwater mixed with tar substances. The total amount of
contaminated material in the two reservoirs was about 12,000 tonnes.

The objective of the demonstration project was to remediate the heavily-contaminated soil and debris within and
around the pits using an approach that minimized off-site migration (for example, as air emissions) since the site was
located in an urban area.

The project was funded by the European Union LIFE-program (44%),the Danish Agency of Environmental Protection
(22%), the  City of Copenhagen (30%), a consulting firm, and several private contractors. The estimated total cost
was 18M DKK (U.S.$2.6 M). Seventy percent of this cost was attributed directly to soil and groundwater treatment.

3. TECHNICAL CONCEPT

Free-phase  tar and heavily-contaminated materials were excavated within the tent, which covered an area of 40 by
50 m (with a height of up to 10 m). The air within the tent was cleaned using three powerful (extraction) ventilators
connected to a two-stage filter system consisting of particle and active carbon filters. About 60% of ventilation air
was blown in actively as fresh air at the end of the tent away from the extraction fans; the remaining 40% came from
general leakages.

Although on-site  thermal desorption  is claimed to be  suitable for treating tar contaminants,  it was considered
uneconomical to treat all the excavated material in this manner. A soil washing system was therefore used to provide
a volume reduction step by producing clean fractions in the particle size ranges greater than 50  mm and 2-50 mm
through screening and high-pressure spray washing. All solid fractions were dewatered before further treatment or
re-use with the contaminated process effluent, recycled, and subsequently treated before discharge. It was observed
that tar removal efficiencies were increased through use of recycled water. This was believed to  be due to elevated
levels of ammonia in the recycled water (up to 20,000 mg/L) which assisted tar solvation. The contaminated fraction
(<2 mm)  was treated using the thermal desorption process.

The thermal treatment plant consisted  of two separate indirect heat treatment units, the first operating at 250-300°C
while the second operated at  800-900°C.
                                                  38

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

In the first stage, water and volatile substances were evaporated from contaminated materials The off-gas was treated
using a paniculate dust trap and an air/oil/water separator/condenser. The second stage, operating at the higher
temperatures, was used to volatilize the heavier tar substances, which were recovered in an air/tar condenser.  Each
treatment unit was transported in a standard 12  x 9 m container and had a maximum operating capacity of 2.5
tonnes/hour. However, during operation, its practical operating capacity was only 1 tonne/hour, the limiting factor
being the rate at which energy could be transferred  from the reactor walls to the material. Treated material had a
residence time of approximately 3 hours inside the  reactors. After treatment, the decontaminated soil was cooled
using a closed single-axle screw conveyor with a box cover to prevent dust emissions.

The  first stage desorber was a specially designed indirectly-heated rotary  unit operated in a  strictly controlled
atmosphere that was slightly overpressurized and secured by inert gas. The processor was heated by a closed loop
thermal fluid (boiler). The slight overpressure led the gaseous hydrocarbons and steam to the controlled condensation
stage. Non-condensable gases (nitrogen, argon, methane, hydrogen, etc.) were removed from the condenser and fed
to the oil-fired burner of the stage-2 processor. Stage 2 employed a indirect rotary heat exchanger with an external
shell temperature of 1,100°C produced by  oil burners.

4. ANALYTICAL APPROACH

Soil samples were extracted for 60 minutes with dichloromethane, after addition of water. The extract was analyzed
on a gas chromatograph with a flame ionization detector. Identification and quantification was done by comparison
with known standards—in this case coal tar. Water samples were extracted with pentane and analyzed in a similar
way. Identification and quantification was done by comparison with standards for polycyclic aromatic hydrocarbons
(PAHs) and BTEX compounds  (benzene, toluene, ethylbenzene, and xylenes). For analysis of phenol, pentafluor-
benzoylchloride  was added to the sample prior  to  extraction with heptane.  The extract was analyzed on a gas
chromatograph with an electron capture device (BCD).

In Denmark, analysis for "total  tar substances" is carried out by comparison with a "natural" coal tar as standard,
rather than through determination of individual PAHs (e.g., the U.S. Environmental Protection Agency's list  of 16
PAHs).

5. RESULTS

Analysis of the "clean"  products from the  soil washing process proved difficult, due to the variable nature of the
coarse material (comprising pieces  of brick, concrete, and stones). Only limited chemical  analysis of the coarse
washed fraction (2-50 mm) was carried out. Reliance was largely placed on visual inspection. Tar concentrations in
the wash water increased due to its recirculation, and washing efficiency decreased because at high tar concentrations,
the water lost its ability to emulsify the tar.

Concentrations of total tars in the fine concentrate (<2 mm) passed to the thermal treatment plant were found to
range between 1,500 mg/kg and 83,000 mg/kg. The  material was regarded to be of optimal size and homogeneity
for the thermal treatment.

Because of the variable nature of the feed material and the three-hour plug-flow operation, it was difficult to match
input and output samples. The characteristics were therefore expressed statistically—the distribution of values for
both was approximately log normal.

Treatment efficiency was reported to be on the order of 99.8%, with total tar concentrations reduced to 22 mg/kg
from a median input of 11,000 mg/kg. Comparable performance was achieved for  individual PAHs, including
benzo(a)pyrene, which was reduced from a median value of 100 mg/kg to less than 0.1  mg/kg.

6. HEALTH AND  SAFETY

During excavation,  the  release of  volatile  substances was  very high. Although ventilation managed three air
exchanges per hour, it was still necessary for workers within the tent to use personal protection. Two principal levels
of protection were provided:
                                                   39

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

(1) chemically-resistant suits with fresh-air supplied through an air tube for manual work in the tent involving direct
    contact with heavily-contaminated material; and

(2) fabric coveralls and with a fresh-air mask with multifilter for all other work operations.

Excavator operators either used the latter level of protection or were supplied with a pressurized cabin with filtered
air supply.

The effectiveness of the chemically resistant suits was checked. Volatile tar substances such as BTEX compounds,
phenols, and naphthalene, could not be detected inside the suits.

Both of the desorber units were protected by two independent systems:

(1) Jets spraying water on the hot internal surface thereby producing steam, which has three functions:

        as a start-up procedure to remove air/oxygen from internal hot parts before feeding materials;

        to keep the system overpressurized, relative to  atmospheric pressure during operation; and

        as back-up in shutdown situations, when steam production from wet material decreased.

(2) Nitrogen was automatically added to the processor  if the internal pressure dropped below atmospheric.

7. ENVIRONMENTAL IMPACTS

One of the key elements of the project was to avoid emissions to atmosphere by using a ventilated tent. One of the
project's key observations concerned the noise associated with the remedial process. Air emissions from the thermal
treatment process were destroyed by feeding them to the stage-2  oil burner.

It was found that with a measured level of  about 60 dB(A) at the working site, it was difficult but practical to meet
the maximum  permitted noise level of 50 dB(A)  at the  site boundary.

Water from the tar reservoirs was used in the soil washing process and/or sent to a treatment plant (1 mVhr capacity)
that consisted  of an oil skimmer, cellulose filter for particulates, and an activated carbon filter.

8. COSTS

The total project cost was about 18M DKK (about U.S.$2.6M), of which 15M DKK (about 70%) was for contracted
clean-up costs. The cost breakdown was as follows:

  •  Personnel  costs (consultant and contractor), 13 %
  •  Operation and maintenance,  7 %
  •  Establishment and rental of the tent, 7%
  •  Soil, air, and water cleaning, 70%
  •  Analytical costs, 1%
  •  Miscellaneous, 2%

9. CONCLUSIONS

  •  The tent was necessary during remediation of the tar reservoirs to protect surrounding areas against volatile tar
    substances.

  •  On-site remediation of soil is difficult in densely-populated areas, mainly because of possible noise nuisance.

  •  Soil washing of the coarse fraction was successful, providing efficient cleaning of coarse materials (>2 mm)
    contaminated with tar substances
                                                   40

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

  •  Water in the tar reservoirs was beneficial in the soil washing process, probably due to a high content of
    ammonia.

  •  The indirectly heated thermal treatment plant was capable of treating tar polluted soil with a cleaning effect
    greater than 99.8%.

  •  Protective masks and suits are essential and effective when working on old gasworks sites.

10. REFERENCES

1.  Skaarup J., and M.  Pedersen. Rehabilitation of a site contaminated by tar substances using a  new on-site
    technique. Agency of Environmental Protection, City of Copenhagen, 1996.

2.  Erik K. Jergensen, AS Consulting Engineers. Rehabilitation of a site contaminated by tar substances using a
    new on-site technique,  Report for City of Copenhagen,  Danish Environmental Protection Agency and the
    European Commission,  Life93/DK/A14/DK/2510.
                                                  41

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 14

                             Ozone Treatment of Contaminated Groundwater
   Technical Contact:
   Pascale Antonus
   SAFEGE Ingeniers Conseils
   Pare de I'ile - BP 727
   92007 Nanterre Cedex
   tel: +33/1-46-14-7260
   fax: +33/1-46-14-7253
Country:
France
Project Status:
Accepted by study 1992
Interim report 1993
Final report 1994
1. INTRODUCTION

In 1991, ADEME, the French Agency for Energy and Environment, received responsibility for the remediation of
an abandoned quarry located in Vaucelles, France. The quarry had been used from 1963 to 1972 for the disposal of
various chemical wastes, including chlorinated solvents. By 1980, the first signs of pollution—including foul-smelling
odors and dying vegetation—had appeared in the surrounding area.

This Pilot Study project showed that combining biological pretreatment with ozone/ultraviolet (UV) oxidation to
groundwater containing organic contaminants achieves levels lower than regulatory limits set for drinking water or
for discharge to a river. Results of the pilot-scale study will be used to design a full-scale application by 1996.

2. REMEDIATION PROGRAM

The remediation program consisted of three phases:

    Phase 1: Site investigation.
    Phase 2: Study of treatment feasibility and treatment selection.
    Phase 3: Installation of a pilot-plant to test an ozone based treatment process.

Phase 1, started in June 1993, identified soil and water contamination.  Over 20 toxic chemicals were identified in
the groundwater, including chlorinated and non-chlorinated solvents, alcohols, petroleum hydrocarbons,  aliphatics
and aromatics, and volatile organic compounds (VOCs).  Ethanol, phenols, chlorinated solvents, BTEX, acetone, and
several aliphatic hydrocarbons were noted in concentrations between 0.8 and 360 milligrams per liter (mg/L).
Chemical oxygen demand was 250 mg/L; total organic  carbon was 60 mg/L; and AOX was 5.5 mg/L.

Contaminated groundwater was found to be emerging at the base of a limestone hill downgradient from the  quarry,
which is near the Autumn River Valley. Rainwater mobilized contaminants from wastes deposited in the quarry, into
infiltrated site bedrock. Contaminants  infiltrated the limestone strata is a sandstone drinking water aquifer that
underlies it. The contaminated groundwater flowed from the aquifer directly into the Autumn River.

During Phase 2 (begun in September 1993), a treatment process was selected for reducing contaminant concentrations
in the groundwater to acceptable levels.  A key selection criterion was that the process should not form any hazardous
by-products. The selected process consisted of two elements: (1) a biological pretreatment; and (2) a combined ozone/
UV oxidation system. The biological pretreatment involved the degradation of organic compounds within the ground-
water by microorganisms, in a process similar to that used in conventional sludge treatment. However, the activated
sludge was separated from the purified water using a membrane filter which in principle reduced sludge production
by a factor of five over the conventional process.  In the second stage of treatment, a UV/ozone process was used
to treat the waste water from the biological process since it was hoped that pretreatment would degrade contaminants
into a form more amenable to photochemical oxidation.

Phase 3 of the study involved a pilot-scale assessment of this remedial scheme. First, a long and deep drainage trench
was constructed at the base of the limestone hill to  intercept the groundwater flow. The  downstream wall of the
trench was lined with an high-density  polyethylene (HDPE) membrane to restrict groundwater flow beyond the

                                                   42

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

trench. Contaminated water was pumped from the trench into an aeration reactor where microbes degraded organic
compounds.  As noted earlier, using a membrane filter  produced one-fifth of the activated sludge  produced in a
conventional process. UV/ozone oxidation was then performed on the pretreated wastewater. The objectives of the
pilot-test were to optimize treatment cost and performance and to determine dimensions of a full-scale plant. Tested
treatment combinations were as follows:

  •  ozone, and ozone/UV without biological treatment;  and

  •  ozone, ozone/UV, and ozone  H2O2 after biological treatment.

3. RESULTS

The main results of the study were:

  •  Combined biological/chemical treatment was more effective than chemical oxidation alone;

  •  Ozone consumption was higher in the ozone/H2O2 system than in the ozone/UV system, but the treatment results
    were the same.

  •  Biological pretreatment combined with ozone/UV and having a flow rate of 30 L/hour reduced chemical oxygen
    demand by 90-95%,  total organic carbon by 80%, and  VOCs by  100%.

4. COSTS

Estimated costs for reducing the VOC concentration in the groundwater to acceptable drinking water limits were
reported as FF15 million (capital  cost) plus FF23.7 million annually (1992).

5. CONCLUSIONS

The project showed that applying a combined biological pretreatment with  ozone/UV oxidation to groundwater
containing organic contaminants achieves levels lower than regulatory  limits set for drinking water or for discharge
to a river.  Results of the  pilot-scale study will be used to design a full-scale application by 1996.
                                                  43

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 15/45
                    Combined Chemical and Microbiological Treatment of Coking Sites/
                  Bioremediation of Soils from Coal and Petroleum Tar Distillation Plants
   Technical Contact:
   Pascale Antonus
   SAFEGE Ingeniuers Conseils
   Pare de I'ile - BP 727
   92007 Nanterre CEDEX

   tel: +33/1-46-14-72-60
   fax: +33/1-46-14-72-53
Country:
France
Project Status:
Accepted by study 1992
Incomplete
In France,   former disposal sites of petroleum refinery wastes, including coal tars, are often characterized by
concentrations of total hydrocarbons of up to 2,200 mg/kg in soil and 4,800 mg/kg in settling ponds; 3-10 mg/kg
phenols; 850-1,500 mg/kg polycyclic aromatic hydrocarbons (PAHs), and 10-300 mg/kg cyanides. During this Pilot
Study project, PAH-degrading bacteria were identified and the practical bioremediation of PAHs in the contaminated
soils evaluated. Furthermore, an oxidation-based pretreatment to enhance subsequent bioremediation was evaluated.

Pilot-scale  assessments were conducted with five different  fungi and  nine bacteria  isolates in a recirculating
bioreactor.  Biopiles were constantly mixed and aerated with amendments of straw,  saw dust,  and uncontaminated
soil. Inorganic nutrients were added to the system  using a proprietary time-release nutrient that the manufacturers
claimed was specially formulated to speed up degradation rates. Three oxidizing pretreatments  (hydrogen peroxide,
sodium hypochlorite, and  ozone) were assessed by mixing them as amendments  into the biopiles during their
construction.

The results after two months of the trial showed that only four of the bacterial isolates and none of the fungi showed
significant PAH degradation.  A practical problem encountered during the pilot-study was that the high concentrations
of tar present  in the soil made it difficult to turn and  mix the  material, which  may  have significantly  reduced
contaminant accessibility. Furthermore, it was speculated that a similar problem may occur with heavy clay-rich soils.
Despite these difficulties, degradation rates of 75% for PAHs (800 to 200 mg/kg after 12 months), 75% of phenols
(after 7 weeks), and 50% of cyanides (after 2-3 months) were reported. Unfortunately no information was provided
on the relative success of each degrader organism.

Little information was provided on the success of the oxidizing pretreatments other than that considerable additions
were required to observe any increased degradation of PAH concentrations. The addition of an oxidizing catalyst,
such as ferrous sulfide, was found to reduce the amount of nutrients necessary for the system. In 19 weeks, 1,000
mg/kg PAH was reduced to 50 mg/kg with pretreatment, hydrogen peroxide, and  a  ferrous sulfide catalyzer.

No detailed written information has been provided for this project.
                                                   44

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 16

                 Combined Vacuum Extraction and In Situ Stripping of Chlorinated Vapors
   Technical Contact:
   Yves Guelorget
   ICF Kaiser Environnement
   14-30,  rue Alexandre - Batiment C
   92635  Gennevilliers CEDEX

   tel: +33/1-468-89900
   fax: +33/1-468-89911
Country:
France
Project Status:
Accepted by study  1994
Interim report       1996
1. INTRODUCTION

The NOVOC™, an in situ volatile organic compound (VOC) removal system,  is based on an air-lift pumping
technology that use air injection. A remediation system installed at a former pigment manufacturing facility located
in Aubagne, France,  consisted of a series of NOVOC™ wells and soil vapor extraction wells and blowers. ICF
Environnement, ICF's Kaiser International's French subsidiary designed, constructed and operated the system for
a period of 22 months, beginning in June 1994.

2. BACKGROUND

The facility  specialized in the manufacturing of paint and pigments for paints, inks, plastics, and textiles. The
pigment manufacturing process involved the use of metals-based colorants and solvents. The facility operated for
approximately 50 years and closed in 1986. Contaminants present included perchloroethene (PCE) and heavy metals.
Initial PCE concentrations ranged from 0.085 to 3.7 mg/L in groundwater and from 10 to greater than 5,000 mg/kg
in soil. The negotiated cleanup levels for PCE were 1 mg/L in groundwater and 50 mg/kg in soil.

3. TECHNICAL CONCEPT

The NOVOC™ system employs a special well design that consists of a well within a well. The outer-fully screened
well extends from ground surface through the unsaturated  zone and terminates at the water table.  The inner well
extends from ground surface through the saturated zone and is screened in the zone of contamination. An air injection
line is placed in the inner well where it releases  bubbles through a diffuser above the zone of contamination.
Contaminated water enters into the bottom well screen rising with the bubbles. As the bubbles rise, contaminants
are transferred from the liquid phase to  the gas phase (in situ air stripping). The VOCs are separated from the water
at the top of the well and are collected by a vacuum system. The treated groundwater flows back to the aquifer
through the top screened outer casing.  The VOCs are treated above ground in a carbon adsorption unit. Figure 1
shows the system schematic.

Numerical modeling  of the gas and water flow in the inner well provided estimates of flow patterns,  pressure
gradients, gas and water velocities, and mass transfer characteristics within the treatment scheme. The remediation
system consisted of nine air extraction wells and three NOVOC™ wells. A total of four NOVOC™ wells (NV1, NV2,
NV3, and NV4) were installed, however, only three wells were located in a productive aquifer. NV1, NV2, and NV3
were installed at depths of 20 m, 36 m, and 20  m,  respectively. Intercepting highly productive aquifers proved to
be a limitation during operation. Pits were constructed around NV1, NV2, and NV3 (60 m3, 120 m3, and 120 m3,
respectively). The pits were filled with washed  gravel. A 650-mm diameter, 5 m long,  screened PVC casing was
installed at the top of each of the pits to the depth of the saturated zone to facilitate the reinjection of the pumped
water. Six monitoring wells and two pumping wells were installed during the construction of the  pits for system
monitoring. The pits were lined at the surface with plastic to prevent vapors from escaping.

The system operated for a period of 22 months. System monitoring included measuring water levels daily, recording
well depression values, and collecting and measuring the volume of PCE recovered from the decanter, stripper, and
carbon unit  condenser. NOVOC™ system efficiency was controlled quarterly and consisted of collecting water
samples from monitoring wells located adjacent to the remediation wells. Static samples were performed quarterly
                                                  45

-------
NATO/CCMS Pilot Study, Phase II
                                                    Final Report (Appendix IV)
           |F*re»B4jre I
           •MlQf>t»Qt-»  I
                                                              ZONt
                                                                       NOT TO SCALE
                                              Upper Screened Interval tor outflow
                                                                 Outer and Inner Wells.
.«• I «>.           I  Upper Screened inter*
•,!    A          I Screen Exists in both O
•j-v    \          ~ Infiltration gallery i* nen
                                 GROUNDWA TEFt
mr
»C«EV
WATER »
VOC VAP
VOC
o«
                                                           I-ower Screened Interval for inflow
Figure 1: NOVOC™ system schematic
by turning off the soil vapor extraction wells, allowing them to equilibrate for two weeks, then sampling the soil gas
from the well head. Results determined if the desired cleanup levels had been achieved and whether operation of the
wells  should be continued.

NV1 operated at 125 gallons per minute  (gpm) and NV2 at 60 gpm. The zone of circulation was estimated to be 40
m radially from both NV1 and NV2. The circulation zone distance was determined by measuring dissolved oxygen
(DO)  levels in the groundwater.  DO background levels were 1 to 2 mg/L, while DO levels in the circulation zone
were measured at 5 to 8 mg/L.

4. RESULTS AND CONCLUSIONS

Only NV1 and NV2 wells were operated, due to  the inability to pump a significant amount of water from NV3. The
volume of PCE removed confirmed system efficiency. By March  1996, over 2,070  liters (3.4 tonnes) of PCE  had
been recovered. Results  from the final soil and  groundwater sampling conducted in March 1996 showed that the
target concentrations of 1 mg/L in groundwater were achieved. PCE concentrations measured in the NOVOC™ wells
after the system was turned off for a 1-month period were between 200 and 565 ug/L.

It has been suggested that advantages of  this system over conventional pump and treat are a reduction in cost
(between  10-15%  cheaper than pump and treat),  minimal maintenance  requirements, and consolidation of
contamination  into a single treatable phase. Reported disadvantages include  the  dependence on favorable  site
conditions, potential  cross-contamination  of a  any shallower aquifer  and the impact of this process on local
groundwater levels.
                                                  46

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 17

                       Treatment of Polluted Soil in a Mobile Solvent Extraction Unit
   Technical Contact:
   Georges Pottecher
   Generale de Rehabilitation de Sites
   13 Villa de la Croix Nivert
   75015  Paris
   tel: +33/1-53-69-6180
   fax: +33/1-47-34-6855
Country:
France
Project Status:
Accepted by study  1994
Interim report           1996
1. INTRODUCTION

The mobile ORG-X solvent extraction system has been developed as part of the EUREKA RESCOPP program to
remediate a wide range of organic contaminants and soil types. It has been specifically developed to remediate high
molecular weight compounds (for example, polycyclic  aromatic hydrocarbons [PAHs] and tars), especially where
contamination is associated with the silt and clay soil fractions. Under this Pilot Study project, the ORG-X process
was combined with other technologies to treat PAHs, polychlorinated biphenyls (PCBs), petroleum hydrocarbons and
heavy metals in a variety of soil types.

2. THE TREATMENT SYSTEM

The ORG-X treatment system is a continuous process using a combination of standard and patented equipment. It
is based on several treatment steps. First, excavated soil and crushed debris are screened to remove oversized clean
material (#4 mesh). Undersized material from the screens is passed to a solvent mixing tank where the soil and debris
are mixed with solvent at ambient temperature. This provides the first solvent extraction stage. From the mixing tank,
the slurry is passed to an inclined auger screw, which provides a second extraction stage in addition to solid-solvent
separation.  Coarse dewatered material is transferred from the top of the auger to a dryer which evaporates residual
solvent, condenses it, and  sends the condensate to a vaporizing recovery unit. The treated  dry soil is then  either
stabilized, if contaminants  such as heavy metals are present, or reused. Solvent exiting the base of the screw  auger
is further treated in a decanter to remove suspended soil fines, which are returned to the primary mixing tank.  Spent
solvent is treated in a vaporizer recovery unit to produce an organic contaminant concentrate and clean solvent. The
recovered solvent is recycled to the initial mixing tank, while the concentrated contaminant oil requires off-site
treatment. The composition of the proprietary solvent has not been reported, but it is claimed to be non-chlorinated,
non-toxic, biodegradable, and oxygenated, and to  have  a low boiling point.

The feed rate of the ORG-X system depends on the soil type and ranges  from 2 to 5 tonnes per hour. The system
is transported in modular components on five separate trailers.

3. PERFORMANCE

Table 1 contains limited case study information presented for this technology. The ORG-X mobile solvent extraction
unit operates best under optimum moisture content and grain size. In one case study, the soil moisture content of
20% made the soil  difficult and more costly to treat. The process achieves higher efficiency on sandy material, but
is being improved for fine-grained soils.

4. COSTS

Costs depend on the amount of material requiring further treatment, ranging from U.S.$ 200 per tonne of soil for
quantities less than 2,000 tonnes to U.S.$ 90 per tonne for quantities of material over 100,000 tonnes (1994).
                                                   47

-------
NATO/CCMS Pilot Study, Phase II                                      Final Report (Appendix IV)




                 Table 1: Case Study Summary Information for the ORG-X Technology
Material Type
Silty sand, 200 t
Loam soil, 12 t
Silty sand, 80 t
Principal Contaminants
2-10 g/kg coal tar
250-2,000 mg/kg PAH
200-500 mg/kg
2-4 g/kg BTEX
2,000 mg/kg heavy organics
Residual Contaminant Levels
coal tar, not reported
6-30 mg/kg PAH
1-6 mg/kg PCB
BTEX, not reported
20 mg/kg heavy organics
                                             48

-------
NATO/CCMS Pilot Study, Phase II
                        Final Report (Appendix IV)
  Project No. 18
                       Biological In Situ Remediation of Contaminated Gasworks
 Technical Contact:
 Hilke Wurdemann
 Institut fur Bodenmechanik und Felsmechanik
 Universtat (TH) Fridericiana Karlsruhe
 Postfach 6980
 Richard-Willstatter-Allee
 76128  Karlsruhe

 tel: +49/721-608-3290
 fax: +49/721-696-096
Country:
Germany
Status Project:
Accepted by study         1992
Final Report               1994
1. BACKGROUND

This Pilot Study project followed the progress of a three-year in situ bioremediation project of a former gasworks
site in the German town of Karlsruhe. The specific test area was located beneath the location of a tar-ammonia
separating sump,  where spills and  leaks  had resulted in contamination  of the underlying ground. Maximum
contamination was located at a depth between 5-7 m. The <2 mm grain size soil fraction recovered from this zone
contained a maximum of 55,000 mg/kg of extractable lipophilic organics and 14,000 mg/kg of polycyclic aromatic
hydrocarbons (PAHs). The intention of this technology was to enhance natural degradation processes on site by
providing an optimized environment through controlled addition of oxygen and nutrients.

2. TECHNICAL  CONCEPT

The test area was  isolated from its surroundings by sealing walls keyed into an aquiclude of impervious clay at 17
m depth, forming a test cell (Figure 1). The water level in the test cell was lowered below the level of contamination
to enhance the air flow and promote aerobic degradation of hydrocarbons. A network of lances was sunk into the
ground to serve as injection and extraction points for oxygen, to try to attain homogenous, horizontal, subsurface air
flow. Above ground, an irrigation system was installed to allow vertical seepage of inorganic nutrients and moisture
into the soil.

An objective of the study was to follow the changing concentration of leachable contaminants during the course of
the experiment. This was achieved by large-scale flushing of the soil body at large intervals during the test period.
Flushing was achieved by using a dedicated infiltration and an extraction well.  This process  was later used to
introduce nutrients into the ground and to raise soil temperature by preheating the infiltration water.

3. RESULTS

Prior  to  field testing,  considerable work was  carried out by  researchers to quantify  the type  and  level of
contamination, both horizontally and vertically, in the soil. Three cores were collected from the site and sampled at
1-m vertical intervals  for chemical analysis. Principle  methods included extractable lipophilic  organics, infrared
spectroscopy of hydrocarbons (IR-HC), and gas chromatography  of the  16  PAHs  contained  on the U.S.
Environmental Protection Agency list and the two methylnaphthalenes. Considerable disparity resulted from estimates
of total contamination using the various assays. Detailed investigations (not elaborated for the Pilot Study) showed
the non-specific parameter, lipophilic organics, to be the most suitable measurement.

After one year of bioremediation, a second set of cores was extracted from the site. The second cores were taken
near the initial cores to minimize the influence of the inherent heterogeneity of contaminant distribution. Nonetheless,
some  paired samples  showed decreases  in contamination,  while  some actually showed increases. For technical
reasons, the cores could not be closer than 0.5 to 1 m, which proved to be too far for comparison. In order to account
for the very  heterogeneous contaminant distribution observed in the soil, a new method was developed that gave
more detailed  information about the degree and distribution of contamination. Freezing the ground allowed the
researchers to extract undisturbed soil cores as long as  12 m from the test area and also from the untreated (but
                                                   49

-------
NATO/CCMS Pilot Study, Phase II
                                       Final Report (Appendix IV)

AIR-
COMPRESSOR
t

j— <

i
>_ ,,
^ ,-J


FILTER
.1 J
SPRINK
i
u- ^

.INC
_<
>S
•*•
YSTEM <
t
WATER
SEPARATING PLANT

                                                  - r8_^,
       GRAVEL,
        SAND
 17m
    BORDER OF 2 '  *
    CONTAMINATION
         CLAY
                            :a
GW
                                    •.  • • i:
                                   AIR LANCES
                            SUCTION WELL
                                                                ^__£
                                                                        GW
                                            AIR FLOW

                                            FLOW OF
                                            PERCOLATING
                                            WATER

                          SEALING WALL
                                   (9in x  15m)
Figure 1: Illustration of principal surface and subsurface installations

uncontaminated) ground around the test site. Initial observation indicated that the tar composition between treated
and untreated soils had changed during treatment. The viscosity of the tar in the treated soil was markedly higher
than that of the untreated soil, suggesting a reduction in the lighter weight components. As an alternative measure
of biodegradation, the authors suggested using the method of Douglas et ai, (1993). Assuming that in similar source
material, the PAH composition remains relatively  constant, and that during soil bioremediation, lower molecular
weight compounds that are relatively easier to degrade than higher weight compounds will be destroyed first, then
the ratio of higher to lower molecular weight compounds is an indicator of biodegradation. The authors used the 4-
ring PAHs as an indication of initial concentration. Using this technique, the researchers reported that 54% of total
PAHs present in the soil were degraded after 2.5 years of remediation.

Monitoring of CO2 and O2 over three years yielded an estimate of organic material decomposed, as well as certain
conclusions about the factors which most influence degradation rates. Assuming that the hydrocarbons consumed
were predominantly aromatic, the O2 consumed and CO2 produced account for 2,400  kg of organic  material
mineralized.  Maximal degradation rates were seen when the soil was warmed either by flushing with warm air or
water, or by natural warming during the summer months. Rapid degradation was also supported by supplying oxygen
by air flushing, rather than as dissolved oxygen in water. This latter observation  reinforced the desirability of
lowering the water table below the level of contamination to facilitate aeration of the contaminant zone.
                                                  50

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

4. RESIDUALS

Apart from biomonitoring, determining contaminant emissions is very important for assessing the course and progress
of the remediation process. It provides an indication of the hazard potential that the contaminated site still possesses.

Water was flushed through the contaminant zone approximately every three months to extract elutable contaminants.
Chemical oxygen demand (COD) and dissolved organic carbon (DOC) in water samples obtained during these large-
scale elution experiments are of particular interest. Because the water used for flushing was repeatedly passed through
the  body of contaminated soil (10-20 fold flushing of mobile soil pore water volume), it became enriched with
contaminants. During the 3 years of bioremediation, the COD peaks thus obtained were reduced by about 83%, while
the  DOC  peaks dropped about 76%.  During the same  period, there  was a parallel decrease of 97%  in  the
concentration of PAHs.

5. COST DATA

No  cost data are available for this study.

6. CONCLUSIONS

Exactly assessing the efficiency of bioremediation  in the field was difficult due to the extreme heterogeneity of the
soil. Analysis of contaminant composition suggested that the more available components of the tar oil were  readily
biodegraded. The main limitation to biodegradation appeared to be the bioavailability of contaminants.

Bulk removal of 54% of PAHs were mirrored by large reductions in elutable components. Large-scale elution tests
showed a reduction of 89% for DOC and COD, and 97% for PAH. Although in some places the soil still contains
relatively high concentrations of PAH, the hazard potential of the site has been reduced considerably. Determining
the  extent of contaminant reduction for soil not receiving special treatment will be an important subject of  further
studies.

7. REFERENCE

    Douglas,  G.S., R.C. Prince, E.L. Butler, and W.G.  Steinhauer (1993). "The  use of internal chemical indicators
    in petroleum and refined products to evaluate the  extent o/biodegradation," In Hinchee, R.E.,  B.C. Alleman,
    R.E. Hoeppel, and R.N. Miller (Eds.), Hydrocarbon Bioremediation, Lewis Publishers, Ann Arbor, MI.
                                                  51

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 19

       Cleaning of Mercury-Contaminated Soil Using a Combined Washing and Distillation Process
   Technical Contact:
   Winfried Grb'schel
   Fa. Harbauer & Co.
   Berlin

   Franz Defregger
   Bavarian State Ministry for State
   Development and Environmental Affairs
   Rosenkavalierplatz 2
   81925 Munich

   tel: +49/89-92-14-2259
   fax: +49/89-92-14-2152
Country:
Germany
Project Status:
Accepted by study 1993
                    Interim report
                    Interim report
                    Final report
                1994
                1996
                1997
1. ABSTRACT

About 57,000 tonnes (metric tons) of soil and debris were successfully treated in a combined soil washing/vacuum
distillation process. Soil and debris were contaminated with mercury at concentrations between 300 and 5,000 mg/kg
(dry weight), with an average concentration of about 500 mg. Although in this specific project the treatment objective
was to treat the soil and debris to meet the treatment criterion of 50 mg/kg to allow for landfilling, trial runs and
plant optimization resulted in residual contaminant concentrations of well below 20 mg/kg. The treatment cost was
estimated to be 480 Deutsche Marks (U.S.$320) per  metric ton. The system is  a full-scale, containerized, and
transportable plant equipped with treatment systems for process water and off-gas. The daily throughput capacity
amounts to 150 metric tons. Process residuals to be disposed or recycled offsite are condensed mercury, spent ion
exchange resin, and a fine-grained precipitation sludge from the water treatment unit.

2. CHEMISCHE FABRIK MARKTREDWITZ SITE  BACKGROUND

2.1  The Site

The Chemische Fabrik Marktredwitz (CFM) site occupies 0.5 square kilometers and was previously operated as a
chemical production facility. It is located in the city center of Marktredwitz, Bavaria, Germany. Founded in 1788,
CFM was one of the oldest chemical manufacturing facilities in the world. The facility was closed in 1985 because
the  subsurface soil and groundwater was severely contaminated. Mercury was processed at the CFM site for the
production  of pesticides, herbicides, and  other mercury-containing products. There were  accidental spills of used
solvents, chemical wastes, and treatment  residuals that were stored onsite. The primary contaminant of concern at
the  site is mercury in the concrete and brick-structures of the buildings and in the subsurface soil; concentrations
between 300 and 5,000 mg/kg were detected. In 1988,  the state of Bavaria decided to fund the remedial action on
the  site. The County of Wunsiedel, a co-founder of the project,  was charged with the management of the remedial
action project.

2.2  Site Remediation

In 1988, the development of a concept  for comprehensive remediation of the CFM site was initiated with the
objective of allowing the site  to be  developed as a housing and shopping area.  The remedial concept consists of
applying the innovative Harbauer technology to clean up the soil and debris to an extent that allows landfilling of
the  treated  solids. The remedial approach incorporates  the following elements:

  •  Protection of the nearby creek, "Kosseine," by installation of a vertical groundwater barrier and a groundwater
    pump-and-treat system;

  •  Demolition of technical facilities and buildings;
                                                  52

-------
NATO/CCMS Pilot Study, Phase II
                                        Final Report (Appendix IV)
 •  Soil excavation and backfilling (The soil on the site had to be excavated to an average depth of 4 m below the
    original ground surface. The excavation pit was backfilled with clean soil);

 •  Soil and debris treatment (A total mass of 57,000 metric tons of excavated soil and debris contaminated with
    greater than 50 mg/kg mercury was treated in the off-site Harbauer treatment facility); and

 •  Landfilling of treated soil near the soil treatment plant (Excavated soil from the site containing less than 50
    mg/kg mercury was landfilled directly).

3. HARBAUER TREATMENT TECHNOLOGY

3.1 Plant  Description and  Treatment Principle

The Harbauer treatment system consists of two main processes: soil washing and vacuum distillation. The principle
of this combined treatment process is to concentrate the mercury in the fine-grained portion of the soil by soil
washing and to clean the resulting highly-contaminated, fine particle fraction using vacuum distillation (a thermal
evaporation process) as a downstream treatment. Figure  1 shows the flow of material, in principle. The feed stream
enters the  facility on the top-most point and passes the process units downward without pumping.
Figure 1. Material flow
                              FLOW OF  IX/IATERIALS
                                                  MATERIAL FEED
                  SPENT ION
                  EXCHANGE
                     RESIN
                                                 DRY SEPARATION
                                                SIEVING / CRUSHING
                                                            : 5O mm
       OFF-SITE
      RECYCLING
                    PRECIPI-
                     TATION
                    SLUDGE
      SUBSURFACE
      DEPOSITION
    WET MECHANICAL SEPARATION
WASHING / SORTING / CLASSIFICATION
        RINSING / DEWATERING
          WATER TREATMENT
            AIR TREATMENT
                             LOADED  LIGHT-WEIGHT
                            ACTIVATED   MATERIAL
                             CARBON
       OFF-SITE
      RECYCLING
                   MERCURY
                                   VACUUM DISTILLATION
                                   DRYING / DISTILLATION
                                COOLING / CONDENSATION
                                          CLEANED SOIL AND RUMBLE
The fine-grained, highly-contaminated fraction of soil that is separated in the soil washing process is transferred into
the vacuum-distillation process. This technology involves heating the soil to a temperature high enough to volatilize
mercury. The Harbauer vacuum-distillation process heats the soil to temperatures between 350-450 °C at a pressure
of 50-150 hPa. These low-pressure conditions reduce the boiling points of the contaminants, resulting in lower energy
consumption of the entire system and a small flow of process gas amounting to only 3-5% of the gas flow, which
is usual for incineration plants.  Thus, process  gas treatment systems with much smaller capacities are  needed,
resulting in significantly lower investment and operational costs. Furthermore, vacuum distillation is a low-oxygen
process that is assumed to  avoid the generation of dioxins  or  other unwanted oxidation products. The relatively
moderate heating temperature is considered to cause no severe  changes in the mineral structure of treated soils.
                                                   53

-------
NATO/CCMS Pilot Study, Phase II
    Final Report (Appendix IV)
The Harbauer soil treatment plant is a full-scale, commercial, transportable plant configured as shown in Figures 2
and 3. The plant is a modular system, and the process units are preassembled in about 60 containers (3-m wide, 3-m
high, 10-14 m long). The plant is gas-sealed; internal air is kept and treated; noise-reduced equipment is used; and
the subsoil of the operation site is protected by a bottom-sealing system.
                      Figure 2
         Contaminated soil <1QOurn
                                        > Vacuum
                        cleaned soil < 100pm  <=*  Lock
                  * f
                      cleaned soil 100um < X < 50mm
                  r
                     Cleaned Soil
Figure 3
                                                                                    50 mm 0
 In addition to soil washing and vacuum distillation, the Harbauer treatment system incorporates water treatment and
 process air treatment. All processes are described in the following sections

 3.2 Technology Description

 3.2.1 Soil Treatment

 Simplified flow  charts of the soil washing and vacuum distillation stages are  shown in Figures 2 and 3.  The
 contaminated material is transported from CFM to the soil treatment plant in closed containers. Here, it is crushed
 and screened at 50 mm and then conveyed to the soil washing unit.

 Blade washers and an attrition scrubber dislodge the mercury and other contaminants from the  soil particles.  The
 mobilized contaminants are separated by means of classification, sorting, rinsing, and dewatering. After the washing
 process, the  sand and debris fraction is discharged as cleaned material. The fine-particle fraction is concentrated in
 thickeners, dewatered in chamber filter presses, and then transported continuously to the vacuum distillation unit
 using belt-conveyors and a bucket conveyor.
                                                     54

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

The vacuum distillation process consists of the following primary units: a dryer, vacuum distillation unit, and cooler.
The contaminated solids are fed to a rotating drum dryer that is heated indirectly to approximately 100 °C by steam
produced in a non-contact heat exchanger. Soil exiting the dryer has a residual moisture content of less than 1% by
weight. The dried soil is fed through a vacuum lock to the vacuum distillation unit, which consists of a rotating drum
heated indirectly by propane burners. Under low pressure conditions of 50-150 hPa and an average temperature of
350 to 400°C, the mercury and other compounds are volatilized from the solids. After passing the vacuum distillation
unit, the soil is transported to a cooler, a water-cooled rotation drum. The cooled soil (temperatures less than 50°C)
passes a vacuum lock and  is transported to a mixer where it is mixed with washed coarse-grained material. The
treated soil was disposed in the CFM landfill located near the treatment facility.

3.2.2 Water Treatment

The process water runs in circuits and is reused after treatment. Process water from the vacuum distillation unit and
from the soil washing stage is first pumped into thickeners, where small soil particles are separated. The thickener
overflow is pumped into a buffer tank. This tank ensures a constant flow rate in the water treatment unit and reserves
a buffer volume  in case of technical problems. From the buffer tank, the contaminated process water is pumped to
the flocculation/precipitation stage, which consists of five stirrer vessels. In the first two vessels, Ca(OH)2 or HC1
is added to adjust the pH. In the next precipitation vessel, precipitation agents like Epofloc, FeQSO4, and FeCl3 are
added. If required, HC1 and Ca(OH)2 are also added to adjust pH. In the next step, the process water is pumped to
a plate separator.

Before entering the plate separator, flocculants are added to the process water in a flocculator to form larger particles
that can be separated in the  plate separator. From there, the process water is pumped to a sand filter followed by two
activated carbon filters. The sand filter is used to separate the remaining very fine particles,  and the activated carbon
filters used to adsorb any organic contaminants in the process water. The sand filter and the  activated carbon filter
need to be backflushed from time to  time, so a washwater vessel is  filled with cleaned water coming from the
activated carbon filters is used. After backflushing, the water is pumped into the washwater vessel to be treated again
in the whole water treatment stage. The precipitation sludge separated by the plate separator is  mechanically
dewatered using filter presses. Flocculants are added to achieve good dewatering results.

The final treatment step for the process water is an ion exchanger. This requires the pH of the process water to be
lowered in an acidification  stage. Screen filters are used to protect the ion exchanger against plugging. The cleaned
water may be discharged to the sewer with the pH adjusted to meet local requirements.

Typically, all the cleaned process water is reused in the soil treatment process. For this purpose, it is pumped into
the process water tank. The desalting water (which is regarded as clean) of the cooling towers is also collected in
the tank. From the process water tank, the cleaned water is pumped to  the mixer and is used for the batching tanks
of chemicals.

3.2.3 Exhaust Air Treatment

A dustfilter and  activated carbon filter are the main units of the exhaust air treatment unit. Approximately, 25,000
mVh of air are extracted from the receiving hall. The dust is separated from the air in the dustfilter and is conveyed
to the material feed of the vacuum distillation. The stream (20,000 m3/h) of exhaust air from the vacuum distillation
process is combined with the de-dusted stream from the dust filter, and the whole stream of 45,000 m3/h is treated
in the activated carbon filter stage. Impregnated activated carbon is used to separate  the  mercury. To avoid
condensation in the filters,  the contaminated air is heated in a heat exchanger.

After the activated carbon filter station, the cleaned air is released to the atmosphere via a stack where the mercury
concentration is  monitored. The height of the stack depends on local conditions.
                                                    55

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

4. TREATMENT RESULTS

4.1 Results of Routine Plant Monitoring

Thorough monitoring of the plant has shown that all requirements were met during operation of the system.
Emissions were found to be well below the specified criteria. Long-term plant monitoring has shown that residual
concentrations are well below 50 mg/kg, even if peaks of very high input concentrations amount to more than 5,000
mg/kg mercury (Table  1).

Besides the routine monitoring, a technology demonstration according to the U.S. Environmental Protection Agency's
(USEPA) Superfund Innovative Technology Evaluation (SITE) Program protocols has been done. The demonstration
results are summarized in Section 4.2.

                        Table 1. Mercury Concentrations in Treated Waste Streams
Unit
soil washing plus
vacuum distillation
vacuum distillation
water treatment
stack gas
Period
Sept. 93 - Oct. 94
Nov. 94 - Oct. 95
Nov. 95 - Feb. 96
Oct. 95 - Apr. 96
1993-96
1993-96
Avg. Cone, of
Mercury in Feed
Stream (mg/kg)
average = 500 peaks
up to 5,000
-
-
-
Avg. Cone, of
Mercury in Treated
Stream (mg/kg)
23
19
7.9
3.6
(see Section 3.2)
(see Section 3.2)
Treatment
Criteria
50 mg/kg
-
10 ug/L
50 ug/dscm
About 57,000 metric tons of soil were successfully treated between 1993 and 1996; the daily average throughput
amounted to about 150 metric tons.

4.2 Results of a Technology Demonstration under USEPA's SITE Program

The Harbauer Treatment System has undergone a technology demonstration in a joint project of the German Federal
Ministry of Science, Education, Research and Technology and the USEPA. The technology demonstration was done
according to the SITE protocols during routine operation of the plant. Based on the SITE demonstration results, the
following conclusions were drawn about the Harbauer soil washing and vacuum distillation soil treatment technology:

  •  Average total mercury concentrations in the treated sandy loam and loam soils were reduced from 875 mg/kg
    to less than 18 mg/kg (95% confidence  level).

  •  Average total mercury removal efficiencies for the soils ranged from 98-99%.

  •  Average Toxicity Characteristic Leaching Procedure (TCLP) mercury concentrations in the soil leachates were
    reduced from 82 ug/L to less than 6 ug/L.

  •  The average mercury concentration in treated process water discharged to the municipal sewer was 5 ug/L.

  •  The average mercury concentration in  the treated stack gas discharged to the atmosphere was 2.92  ug per
    normal cubic  meter (these measurements were not made during the same test runs as the demonstration soil
    samples that were collected).

  •  The estimated treatment cost is 480 Deutsche Marks (DM) per metric ton, which is approximately U.S.$320 per
    metric ton (assuming a 1.5 DM to 1 U.S. dollar exchange rate).
                                                  56

-------
NATO/CCMS Pilot Study, Phase II

4.3 Process Residuals
Final Report (Appendix IV)
  •  Lightweight particles separated in the soil washing process and spent activated carbon from water and off-gas
    polishing are collected in containers and treated in the vacuum distillation unit in batches.

  •  Precipitation sludge from the water treatment system is disposed to subsurface hazardous waste storage facilities.

  •  Spent ion exchange resin is reactivated offsite for recycling.

  •  Condensed mercury is disposed offsite for reuse or proper disposal. During plant operation, nearly 30 metric tons
    of mercury were recondensed and disposed.

  •  Treated wastewater and treated off gas were released with the mercury concentrations listed in Table  1. Both
    streams were subject to constant on-line monitoring.

5. SCOPE OF  APPLICATION OF THE  TECHNOLOGY

5.1  General

The modular design of the plant allows adaptation of the system to a wide range  of applications that might be very
different from the conditions at the CFM site. Although the soil washing and the vacuum distillation can also be
operated separately, the combination of these two processes is  considered to be the main advantage. In many cases,
clean-up criteria cannot be  met economically by using only one type of technology.

For the CFM site, a soil clean-up criterion of 50  mg/kg mercury (dry  weight) was specified by the authorities
because the soil was designated to be landfilled in a monofill. The soil was not used for backfilling at the CFM site.
In the future, it can be expected that in other remedial projects, reuse of the treated soil will be required. For these
cases, a residual mercury concentration of 50 mg/kg can be assumed to be too high. Therefore, optimization of the
plant operation during test  runs was  performed. The results show that the combined Harbauer process allows for
substantially lower residual contaminant concentrations (well below  10 mg/kg) for the combined coarse- and fine-
grained stream exiting the plant (see Table 1). The soil stream treated by vacuum distillation from October 1995 to
April 1996 was 3.6 mg/kg dry weight in average. Long-term monitoring of the plant has shown that the soil treated
by vacuum distillation had residual concentrations well below  10 mg/kg dry weight. Trial runs done with mercury-
contaminated soil from another  site  showed that a clean-up  criterion of 2 mg/kg can be reliably met using an
appropriate plant configuration.

5.2  Applicability to Contaminants

The technology  is suitable for volatile contaminants with boiling points between 350-450°C at pressure ranges
between 50-150 hPa. The results of trial runs with a small-scale  pilot plant,  serving as a model for the full-scale
systems, are listed in Table 2.
                        Table 2. Results of Trial Runs Using a Small-Scale Pilot Plant
Contaminant
PAHs (sum of 16)
Total Hydrocarbons
Cyanide
Mercury
Lindane
TNT
Feed Soil Concentration
(mg/kg)
32,000
102,000
1,728
8,000
3,400
100,000
Treated Soil Concentration (mg/
kg)
< 1
< 10
< 10
0.5 -4
0.002
0.4
                                                    57

-------
NATO/CCMS Pilot Study, Phase II                                             Final Report (Appendix IV)

5.3 Applicability to Soil Types

The soil washing separation point between the fine-particle fraction to be treated in the vacuum distillation unit
and the coarser fraction, which is discharged as cleaned material, can be adjusted according to the contaminant
load of the different grain size fractions. Adjustments can be made in the range between 100-um and 8-mm
particle sizes.

Three years of routine operation of the plant have shown that the vacuum distillation process is very insensitive
to variations in soil type. Gravel, sandy soils, fine-grained soils (up to 90% silt and clay), slags, concrete, bricks,
and other solids with mechanical characteristics similar to soil can be treated effectively and efficiently with this
technology.
                                                    58

-------
NATO/CCMS Pilot Study,  Phase II                                           Final Report (Appendix IV)
    Project No. 20

                              Fluidized Bed Soil Treatment Process—BORAN
   Technical Contact:
   Arnd Martin
   Bodenreinigung GmbH & Co Westhafen
   Alt-Mobait
   D-10559
   Berlin
Country:
Germany
Project Status:
Accepted by study  1994
To be continued in Phase III
1. INTRODUCTION

The BORAN thermal fluidized bed soil treatment process has been designed to treat slurry residues from soil washing
plants. The system is designed to treat contaminants such as polycyclic aromatic hydrocarbons  (PAHs) and poly-
chlorinated biphenyls (PCBs).  The plant is operated by the designers  and manufacturers of the plant, and while
intended to be a commercial operation, is also intended to be used as a  full-scale test bed for design modifications.
The basic technology is already in use for waste treatment and a number of other applications. Due to difficulties
with the feed mechanism, progress in becoming fully  operational was delayed. The project will consequently be
continued in the Phase  III Pilot Study.

2. PROCESS DESCRIPTION

Before contaminated soil is fed into the furnace, it is screened in a vibrating bar sizer with a cut-point at 20 mm.
Material greater than 20 mm in diameter is transferred elsewhere for alternative treatment and safe disposal. Material
less than 20 mm is fed into the furnace.

The furnace is fitted with two overbed screw feeders for soils and four in-bed feed points for fine residues. At the
design conditions, fuel  oil is supplied to the bed at the rate of 725 L/hr to maintain the bed temperature at 900°C,
which reportedly ensures that the organic content of the soil is fully oxidized and destroyed. The combustion chamber
is rectangular with one  side wall inwardly inclined to act as a deflector plate and limit bed expansion. Preheated air
is injected into the chamber to ensure that feedstock (contaminated soil), combustion gases, and bed materials are
circulated rapidly  in an elliptical path within the combustion chamber.  It is claimed that the controlled circulation
produces lateral mixing and turbulence, which enhance combustion efficiency.

Immediately downstream of the reactor is  a hot gas cyclone capable of removing up to 85% of the soil and fines
expected to be  carried by  the fast moving off-gas  stream. The cyclone is designed to  remove up to 7.5 tonnes of
particles per hour. This material is cooled using water-cooled screw feeders from around 900°C  to 180°C before it
is transferred to three storage silos as a clean product. The majority of the treated soil is collected at this stage,
although some of the coarser particles sink to the base  of the fluidized  bed within the furnace and are collected as
an heavy ash product.

After the hot gas cyclone, the off-gas stream passes to  an afterburner where it is heated to 1,200°C by an oil-fired
heater. The flue gas enters the treatment chamber at the top and passes down the center of the chamber to the exit
at the base. The afterburner gases are directed tangentially at high velocity to the flue gas stream  to mix turbulently
with the flue gas at the top of the chamber. At the base of the chamber, a water bath with an immersed screw feeder
extracts any soft entrained material.  When the afterburner is in use, the  following quench chamber reduces the
temperature of the flue gas back to 900°C. A heat exchanger system uses the  flue gases to preheat air about to be
injected into the main furnace.

The flue gas cleaning plant consists of a primary and a secondary system. Primary absorption of inorganic gases such
as hydrogen chloride and sulfur dioxide is  achieved by mixing finely-powdered limestone into the flue gas stream.
The limestone is subsequently collected in a bag house. In the secondary absorption system, any residual organics,
including  dioxins  and furans, are collected by activated brown coal coke filters. Fouled coke is  incinerated in the
furnace. No performance data for this technology are available at present (July 1997).
                                                   59

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

The plant was not in operation during a visit by some members of the Pilot Study in September 1996 because it was
undergoing modification to improve the feeding mechanism—the original version not being well-suited to handling
clayey soils. The plant is located on a very restricted site, and the main limiting feature appears to be the very limited
storage capacity for material awaiting treatment.
                                                  60

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 21

                           Mobile Low-Temperature Thermal Treatment Process
    Technical Contact:
    Jtirgen Fortmann
    Ruhrkohle Umwelttechnik GmbH
    Gleiwitzer Platz 3
    D-46263 Bottrop

    tel: +49/2041-166-610
    fax: +49/2041-166-612
Country:
Germany
Project Status:
Accepted by study  1994
Interim report       1996
Final report         1997
Many commercially available thermal treatment plants for contaminated soil are based on rotary kilns operating at
temperatures above 500°C. However, some hydrocarbons are volatile at temperatures below 300°C offering the
possibility of significantly reducing the  environmental and  economic costs of treating some soils. This project
concerned the design and evaluation of a mobile low-temperature thermal treatment technology. The results reported
were obtained from a pilot-scale plant. Although a full-scale plant has been designed, one has yet to be built.

1. BACKGROUND

Ruhrkohle Umwelttechnik GmbH (RUT) has operated thermal treatment facilities for contaminated soil since 1986.
In 1990, RUT initiated development of a mobile low-temperature treatment system for a range of soil contaminants
(such as volatile  hydrocarbons, chlorinated solvents, and mercury) with the following aims: to reduce overall
treatment costs compared  with high temperature treatment; to  reduce plant transport costs in order to increase
applicability of on-site treatment to smaller remediation projects; to improve performance and cost effectiveness of
thermal treatments for fine-grained and especially clay-rich soils; and to reduce project initiation times by producing
a process that did not require regulatory permitting for any off-gas or effluent emissions.

2. PROCESS DESCRIPTION

2.1 Pilot Plant

The pilot plant has a  capacity of 150  kg/hr and  is housed in  two containers.  The process,  which is  shown
schematically in Figure 1, involves heating soil (<15 mm grain size) in a gas-tight evaporator at temperatures in the
range of 200-330°C (lower temperatures can be employed for BTEX compounds—see below) under continuous agita-
tion for about 45  minutes. The soil is indirectly heated via  thermal oil, which acts as the  heat transfer medium.
During the first five minutes, the  soil is dried and then heated to the required temperature. Contaminant removal is
aided by steam injection.

The exhaust gases from the evaporator contain a large proportion of steam. Volatilized contaminants are collected
from the off-gases by a multi-step condensation system. Contaminants such as chlorinated solvents and mercury are
separated from the condensed steam, which is cleaned and recycled. Further contaminant separation may be carried
out to produce low boiling point and high boiling point products to increase their recycling value. The cooled off-gas
(at about 5°C) is further cleaned using an activated carbon filter for organic  contaminants (this is regenerated using
steam) and an unspecified chemical adsorption unit for  inorganics  (such as mercury) before discharging to the
atmosphere. The treated soil is cooled to about 60°C and water added to raise soil moisture content to 7-10%.

2.2 Concept for a Commercial Scale Plant

By the end of 1995, the basic engineering for  a commercial plant for treating mercury-contaminated soil was
completed. The plant has a capacity of 6-7 tonnes/hr. Its relatively large size  results from the need for safety devices
to comply with environmental protection regulations and worker safety protection rules. Plants for treating oil-
contaminated soil, however, can be smaller. One major factor is that  they do not require the safety devices needed
for mercury. All in all, the plant  requirements may be reduced to  a level of approximately  65%, with consequent
reduction in operating costs.

                                                   61

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
Figure 1: RUT TERRA »POR™ mobile low temperature soil cleaning system
        contaminated soil
            6-8t/h
                                                                   activated carbon filter
        residual air
                                                                 soil cooler
The concept of the commercial plant to treat mercury-contaminated soil is illustrated in Figure 2:

Figure 2: Flow sheet for full scale plant
  Utilities:
  -  Water
  -  Oil
  -  Fuel
  -  Electrical power [
  -  Pressurized air
  -  Nitrogen
  -  Chemicals
    Contaminated
    soil
    ?           fr
    35.700 t/a
                            .(12)
                                                                      Condensation /
                                                                        separation
  •  Prior to continuous feeding to the thermal treatment unit, contaminated soil or rubble is freed from any nibs or
    metallic fractions.

  •  In the  evaporator,  the contaminants are evaporated by indirect heating  and water vapor stripping. Then, the
    treated soil is cooled and remoistened with purified process water. The contaminated vapors are de-dusted with
    a cloth filter.
                                                    62

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
  •  The heat gradient required for the thermal treatment of the contaminated material is created in a thermo-oil
    heater. This unit houses the fuel oil tank and the facilities for the intermediate storage, feeding, and emptying
    of the heat transfer medium.

  •  The process steam plant is where the stripping steam is raised and superheated. The unit also contains facilities
    for the softening, degassing, and preheating of the feed water.

  •  Condensation involves a multi-step cooling of the hot and de-dusted vapors from the evaporator. The residual
    gas is finally purified with the help of activated carbon and then blown off. The contaminated residues, a mixture
    of water and mercury, are separated by gravimetric settling. The mercury is salvaged, and the water is cooled
    and used in the condensation process  as quench water.  Any excess water is fed to the process water treatment
    unit.

  •  The resulting process water contains mercury—both in  dispersed and ionic forms. Process water treatment
    ensures physical and chemical purification of the water to meet the required effluent standard(s). The purified
    water is used for moistening the cleaned soil. Excess water is discharged.

3. RESULTS  OBTAINED ON PILOT-PLANT

Pilot-scale trials (Table 1) were conducted during 1994 and  1995 to evaluate process performance and to aid design
of a commercial-scale plant. The effect of steam injection was to enhance volatilization of certain contaminants at
a particular temperature. Results showed  that at  270°C and low steam injection rates, a treated soil sample still
contained up to 35 mg/kg of poly cyclic  aromatic hydrocarbons (PAHs) with 5 or 6 aromatic rings. At higher injection
rates,  but  at the same temperature,  this residual concentration dropped to about 10  mg/kg.  In many cases,  the
observed concentration of  contaminants in the process water was low or below detection limits.

                              Table 1:  Summary Results of Pilot-Scale  Trials.
Contaminant
Mineral oil in clay soil
Heavy mineral oil
Light tar oil (PAH C2 - C4)
Tar oil (PAH C2 - C6)
Mercury in sandy soil
in clay soil
in clay soil
TNT
Temperature
(°C)
240
270
270
240
280
330
320
210
Stripping
steam
(kg/hr)
6
5
6
6
4
6
6
8
Input
concentration
(mg/kg)
38,000
22,000
4,500
5,000
300
900
11,000
180
Output
concentration
(mg/kg)
60
1,000
2
20
5
25
85
4.5
An additional set of experiments was set up to study to what extent the contaminant residues in treated soil are
available to the biosphere. For this purpose, eluate analyses were performed. The experiments with mercury showed
that although there was significant variation in treatment results (5-85 mg/kg), eluate concentrations never exceeded
the detection limit.

Figure 3 illustrates the effect of temperature. In soil contaminated with PAHs (2,000 mg/kg), contamination was
reduced to 70 mg/kg at 210°C, whereas it was reduced to 30 mg/kg at 270°C. For 2- and 3-ring aromatics, treatment
has almost reached its maximum efficiency at 210°C, whereas for higher-boiling aromatics, significant improvement
was observed at temperatures up to 270°C.
                                                   63

-------
NATO/CCMS Pilot Study, Phase II

Figure 3: Results in relation to temperature
                Final Report (Appendix IV)
           100   --
    CD
            50
  Input concentration: 2.000 mg/kg PAH

  Soil throughput: 135 kg/h

  Stripping steam: 2 kg/h

  Residence time: 50 min
                                                                                5 - 6 ring PAH
                                                                                   4 ring PAH

                                                                                2 - 3 ring PAH
                                   H-	1	h
                                   210                 240                270
                                       Soil temperature [°C]
Figure 4 shows how increasing the amount of steam at a fixed temperature (270°C) increased the effectiveness of
the process. Thus, the same effectiveness can be achieved at a lower temperature.

Figure 4: Results in relation to amount of steam
          100   --
   CD
           50   --
Input concentration: 2,000 mg/kg PAH

Soil throughput: 135 kg/h

Soil temperature: 270 °C

Residence time: 50 min
                                                                                5 - 6 ring PAH
                                                                                   4 ring PAH
                                                                                2 - 3 ring PAH
                                             468
                                      Mass  of stripping steam  [kg/h]
Figure 5 illustrates the impact of steam on PAH removal  at various temperatures. The effect of steam was
insignificant between 240 and 270°C, whereas a clear effect occurred at 210°C. Application of 8 kg/hr steam at this
temperature resulted in a similar degree of PAH removal as at 270°C without steam.

In 1996, additional experiments were carried out on a variety of PAH-contaminated soils from a number of coking
plant and gas works sites in the Ruhr area. The soils were treated under otherwise identical test conditions (residence
                                              64

-------
NATO/CCMS Pilot Study, Phase II
                                                                 Final Report (Appendix IV)
Figure 5: Results in relation to temperature and amount of steam
                                                             CD

                                                             CD
                                                             CD
                                                          Soil temperature: 270 "C
                                                          Soil throughput 115 kg/h

                                                          Soil temperature 240 °C
                                                          Soit throughput 1OO kg/h

                                                          Soil temperature 21O "C
                                                          Soil throughput: 67 kg/h
                                   246
                                      Mass  of stripping steam  [kg/h]
time and quantity of steam) at temperatures between 240°C and 270°C. Analytical results of the treated soils were
very similar (see Figure 6), thus confirming the wide applicability of this process. In addition, soils contaminated
with lignite tar oil could be treated successfully at 270°C in the pilot plant. The results are presented in Table 2.

Figure 6: Results for different PAH-contaminated soils
  CD
  .*:
  73)
  E
         4O- -
30- -
         20- -
         1O- -
                                                Coking plant 1 (PAH = 1.000 mg/kg)
                                              Coal mining area (PAH = 2 000 rng/kg)
               Coking plant 2 (PAH = 700 mg/kg)
                            Gasworks (PAH = 150 mg/kg)
                           24O
                                                                         27O
                                       Soil temperature [°C]
Treatment was extended to include bulk material and activated carbon loaded with BTEX compounds (benzene,
toluene, ethyl benzene, and xylenes). Treatment of activated carbon was investigated to determine if the process
might be a cost-effective alternative for the well-established regenerative process. The presence of BTEX compounds
and activated carbon required special explosion protection measures. The experiments were performed with high
steam loadings (8-12 kg/hr), and the dust filter was de-dusted with the help of nitrogen, instead of compressed air.
                                                65

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
                                      Table 2: Test results for 1996
Material treated
Soil with lignite tar oil
Bulk material with
BTEX
Activated carbon with
BTEX from air
treatment
Activated carbon with
BTEX from water
treatment
Conditions
(residence time: 40-60 mins.)
Temperature
(°C)
270
90
270
270
Stripping steam
(kg/hr)
8
8
12
12
Input
(mg/kg)
MKW 1,500
KW 17,000
4,500
7,600
45,200
Output
(mg/kg)
MKW 80
KW750
40
45
850
The low-boiling BTEX compounds were successfully removed from the bulk material when steam stripped, even
at treatment temperatures of 90°C (Table 2). Such lower treatment temperatures save energy and also preserve the
physical structure of heat-sensitive bulk material. Quantitative separation of BTEX compounds from activated carbon
required a temperature of 270°C (Table 2). The considerably higher temperature, compared to the bulk material,
results from the higher adsorption power of activated carbons.

4. COSTS

According to RUT, the costs of cleaning soil using a full-scale plant based on this process would be about 150-250
Deutsche Marks (U.S.$90-160) per metric ton, depending on the  contaminant(s). The cost depends on the plant
equipment required and waste disposal costs (which vary according to the contents of the harmful substances and
the composition of the soil).

An important cost-related factor is the quantity of soil to be treated at a particular site. A mobile plant cannot really
compete with stationary plants when only a small amount of soil is to be treated. However, for larger quantities of
soil, treatment of the above contaminants in the mobile plant is less expensive, because  on-site cleaning does not
involve any transport costs for the soil and does not require any expenditure for obtaining permits. Another advantage
of the mobile plant is that it may be installed rapidly and operated for up to 12 months under German law  without
the necessity for time-consuming applications for approvals to set  up and operate.

5. REFERENCE

    Fortmann, Jtirgen, Peter Jahns, and Klaus  Kossmann. Mobile low temperature thermal soil treatment process.
    Paper presented at NATO/CCMS meeting, Golden, Colorado, 1997.
                                                  66

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
   Project No. 22
                  Environmental Evaluations of Former Soviet Military Bases in Hungary
   Technical Contact:
   Pal Varga
   National Authority for the Environment
   Fo u.44
   H-1011 Budapest

   tel: +36/1-457-3530
   fax: +36/1-201-4282
   E-mail: vargap@kik.ktm.hu
Country:
Hungary
Project Status:
Accepted by study
Interim/final report
1993
    19
    94
1. INTRODUCTION

After the withdrawal of the former Soviet army from Hungary in 1990, the Hungarian Ministry for Environment
conducted an environmental assessment and damage survey. Hydrocarbon soil contamination was found to have
affected 2.7-3.0 million m3 of soil and between 1.0-1.2 billion liters of groundwater.

Tokol airbase was one of the  sites identified as needing prompt remediation since it is located just 600 m from the
Danube River, and on top of the aquifer that supplies Halasztelek with about 5% of its municipal drinking water.
Site investigations determined that the groundwater was contaminated by free-phase and dissolved hydrocarbons from
a jet fuel storage area. Contaminated groundwater from the airfield was migrating slowly towards the abstraction
water boreholes for Halasztelek.

Remediation at Tokol was accomplished in phases between August 1991 and June  1993 in cooperation with the
Danish Environmental Protection Agency. Free product was separated from contaminated groundwater by depressing
the water table to speed flow of groundwater to extraction wells, pumping, and on-site water/oil separation.

2. BACKGROUND

After the withdrawal of the former Soviet army from Hungary in 1990, the Hungarian Ministry for Environment
conducted an environmental assessment and damage  survey following a method acceptable to both the Hungarian
and Soviet governments. One hundred seventy-one garrisons, 340 settlements, 6,000 major buildings, and 46,000 ha
of land were included in the survey. Hydrocarbon soil contamination was found to have affected 2.7-3.0 million m3
of soil and between 1.0-1.2 billion liters of groundwater. During the assessment, between 5.5 and 6.5 million liters
of hydrocarbon product was found as non-aqueous phase liquids (NAPLs) floating on the surface of groundwater.

Due  to time and funding constraints, work was begun in 1991 at the 20 most-polluted bases, with an objective to
contain the pollution. By the end of 1993, eight sites were fully remediated, while the remaining 12 major sites were
completed in 1994. The goals of the  1993-94 assessment were  to identify those remaining sites that required
remediation and those sites where only monitoring is necessary.

Site investigations at Tokol airbase determined that the groundwater was contaminated by free-phase and dissolved
hydrocarbons from a jet fuel storage area. Contaminated groundwater from the airfield was migrating slowly towards
the abstraction water boreholes for Halasztelek.

3. TOKOL AIRBASE

3.1 The site

The Tokol airbase is situated in the northwestern part of Csepel Island, about 1 km from Halasztelek and about 1.2
km from Szigethalom and Tokol (Figure 1). The airbase is about 2.5  km long and contains  a 60-m wide concrete
runway.
                                                  67

-------
 NATO/CCMS Pilot Study, Phase II

 Figure 1: Location of Tokol Airbase (not to scale shown)
Final Report (Appendix IV)
                           •  Observation wells (waterworks)

                           O  Water works supply wells

                           A  Other abstraction wells

                           E3 Housing
                               1  Location of Tokol Airbase, Halasziekk and Tokoi Village (1:40.0001
Forty-two hangars are arranged in three blocks. Between the hangars are shelters and ammunition stores. A fuel
storage and filling station (capacity approximately 6,000 m3) is located in the southwestern part of the airbase.

The airbase overlies  a major aquifer (capacity about 600,000 nrVday) that supplies  drinking water to the local
population. The hydrogeological structure of the aquifer consists of a clay layer about 15-20 m below the surface
overlain by a  10-15 m sand and gravel layer. The water abstraction wells of the Halasztelek well field are drilled
and filtered in this layer. Fine sand and sandy silt, between 2-5 m in thickness overlie the sand and gravel layer. The
groundwater level is about 4-5 m below ground level. Water levels are influenced by the level in the Danube River,
by precipitation rates  and by abstraction from the aquifer. Groundwater flows in a west and northwest direction at
a rate of about 40-50 m/year.

3.2 Investigation, Assessment,  and Remediation

The Hungarian Institute for Environmental Management began investigating the airbase in August 1991. On the basis
of the investigations,  12 recovery wells and 46 infiltration wells were drilled in the  second half of 1992. Recovery
operations were continued in four locations until June 30, 1993.

Recovery of free product and groundwater containing dissolved hydrocarbons was carried out by:

  •  separating water  and free product  using 150 mm small-diameter  Filter Scavenger pumps in large-diameter
    recovery wells.

  •  depressing the groundwater table in the vicinity of recovery wells to  speed up the flow of free product.
                                                    68

-------
NATO/CCMS Pilot Study,  Phase II
                                                                     Final Report (Appendix IV)
From November 1990 to June  1993, approximately 280,000 m3 of contaminated groundwater were pumped out.
Approximately 224,000 liters of free product were recovered. By April 1994, the volume of recovered jet fuel had
reached about 700,000 liters.

3.3 International Collaboration

A group of Danish firms were commissioned by the Danish National Environmental Protection Agency to assist the
Hungarian authorities in the investigation and remediation of Tokol airbase. The Danish input was  made in close
cooperation with the Hungarian authorities and using Hungarian sub-contractors, such as laboratories and drilling
companies. Danish input was provided in two phases:

Phase 1:      investigations and preventative water abstraction; pumping tests; procurement, delivery and installation
              of pumps; sampling  and analysis of water samples.

Phase 2:      support through five workgroups: the Risk Group, the Recovery Group, the Monitoring Group,  the
              Model Group and  the Bio Group. The work of the five workgroups is summarized below:
 Workgroup
                  Area of Work
 Risk Group
 Recovery Group
 Monitoring Group
 Modeling Group
 Bio Group
Assessment of potential human health and environmental impact of contamination and the need for and extent of any
remedial action. Work has included collection and assessment of data, monitoring, hazard assessment, and review of
international and Hungarian  standards for contaminated soil and groundwater. Assessment indicated that the main
potential impact of the airbase was on the Halasztelek aquifer; ecological effects were judged to be insignificant.
The main purpose of this group was to minimize further dispersion of contamination and maximize free-phase recovery
of jet fuel. Work has  involved data collection, delivery and installation of equipment, training, evaluation of recovery
strategy and associated data, and maintenance of equipment.
The purpose of this group was to monitor the status  of the contamination present at the airbase,  together with  any
trends over time. Data obtained on groundwater levels,  thickness of free-phase product, and contaminant concentrations
in soil and  groundwater samples  are fed into a database from which graphic representations  of the  subsurface
conditions can be made.
Two separate  models have been  used: the SHE model for groundwater flow and contaminant transport; and the
ARMOS model for migration of free-phase product. The SHE model was used primarily to predict the dispersion of
contaminated  groundwater to local receptors (e.g., Danube and abstraction wells at Halasztelek). The ARMOS model
was used primarily to  optimize the free product recovery operation.

Model predictions (conservative) suggested that the dissolved hydrocarbon contamination (from Zone B of the airbase)
would reach the water catchment area with 6-8 years and from Zones C and D within 12  - 18 years; and  that
contamination from Zone  A will reach the Danube in 12 - 18 years.

The potential  for biodegradation of dissolved hydrocarbons in groundwater has also been investigated using the SHE
model.
The purpose of this group was to explore the potential for biodegradation  and to create a basis for implementation of in
situ biological  remediation. Work activities have included the establishment, operation and monitoring of a pilot scale
in situ biological treatment  plant; laboratory  tests and field measurements to  identify an appropriate test field;
monitoring of pilot scale operations and recommendations on the development of a full-scale biological treatment plant.
3.4 Costs

The total cost for remediating the Tokol site has been estimated to be U.S.$600M (1994).

3.5 Conclusions

  1.  Large quantities of free-phase jet fuel have been recovered from the Tokol airbase and immediate risks to  a
     major potable water supply source at Halasztelek have been reduced.

  2.  However, free-phase jet fuel cannot be recovered by pumping operations alone, and contamination adsorbed onto
     soil in the unsaturated zone is also present at the airbase, posing a long-term source of water contamination.
                                                         69

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

 3. Pilot scale in situ biological treatment was successful and was recommended as an effective method for securing
    further reduction of the risks to groundwater quality posed by the airbase.

4. REFERENCES

1.   Reiniger R. and Z. Horvath. Environmental problems  at former Soviet military bases in Hungary
    Paper  to NATO/CCMS meeting 1993.

2.   Anon. Danish assistance in the remediation ofTokol airbase near Budapest, Hungary, Phase 2
    (Final report: summary, prepared for Danish Environmental Protection Agency, June 1994).
                                                 70

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 23

                             Modeling and Optimization of In situ Remediation
    Technical Contact:
    Hennie  van Vree
    Tauw Milieu bv
    P.O. Box 133
    7400 AC Deventer
    tel: +31/55-549-3922
    fax: +31/55-549-3410
    E-mail: hw@Tauw.nl
Country:
The
Netherlands
Project Status:
Accepted by study  1992
Interim report       1993
Final report         1996
1. INTRODUCTION

In situ remediation of contaminated soil and water is widely perceived to offer the greatest potential for enhancing
performance and reducing the cost of treatment. However, its commercial implementation has been limited by the
perception that current methods are unreliable and their treatment duration unpredictable. The unpredictability of full-
scale in situ treatment is not solely due to the complex heterogeneity of field sites, but also due to economic and time
constraints on the preliminary collection of field data. Treatment design and predicted performance are often based
on bench-scale studies used in combination with models incorporating subjective default data values. The results are
often  misleading and overly optimistic.  This Pilot Study project reports on several aspects of an investigation into
the modeling of in situ treatment with an overall goal of enhancing and optimizing treatment performance.

2. THE MODELS STUDIED

2.1 Soil Vapor Extraction

A predictive  model for soil vapor extraction (SVE) was the first model discussed (Budapest meeting, 1992).  A
literature  survey  revealed  that although several  groundwater models  existed, there had  been  relatively   little
development  of soil air models for predicting the performance of  SVE. The presentation focused on a spreadsheet
model which had been developed by Tauw Milieu to predict SVE performance for sites contaminated with up to three
volatile organic compounds (VOCs). The  model consisted of differential equations linked to a mass balance based
spreadsheet. Model outputs included  the cumulative  amount of contaminants extracted and soil contaminant
concentration (both as a function of time). The model output was  used to determine expected treatment durations.
Parameters and expressions within the model accounted for equilibrium sorption to organic matter, transfer from
liquid to vapor phase and contaminant interactions, diffusion, biodegradation, and time dependence. Four case studies
were evaluated using the predictive model: one laboratory-based study, two on-site studies, and another conducted
in situ.

2.2. In situ Soil Flushing

A model developed to predict the treatment duration of in situ soil flushing has also been discussed (Quebec City
meeting,  1993). This model examined change in contaminant concentrations in extracted water as a function of time
which is an important parameter in determining treatment cost and equipment requirements for above-ground water
treatment. The model methodology reportedly improved the handling of uncertainty within the site and contaminant
invariability in data  across a site. Predictive  results for a remediation scenario where monochlorobenzene had
contaminated soil and groundwater at a model site were presented. A simple predictive model (a first-order approach)
combining transport factors with an expression for the linear adsorption of organic compounds by soil organic matter
was linked with a statistical simulation to model heterogenous site distributions of organic matter and adsorption
coefficients. Assuming a homogenous distribution of site parameters, soil flushing duration was predicted to achieve
the remedial target for groundwater contaminant concentration in  three years. However, modeling site parameters
more realistically  using an heterogenous distribution increased treatment times to the order of decades.  The authors
stressed that  a  higher quality  of soil investigation was  necessary in order to  generate a realistic prediction  of
remediation duration and cost.

                                                   71

-------
NATO/CCMS Pilot Study,  Phase II                                           Final Report (Appendix IV)

2.3 In situ Bioventing

A model developed to investigate the effectiveness of in situ bioventing was presented at the NATO/CCMS Oxford
meeting in 1994 and the Adelaide meeting in 1996. A mechanistic model, dubbed ECOSAT, was developed and
calibrated with data from laboratory batch and column bioventing tests. The advantage of mechanistic models over
simplified empirical models is that they allow for scenario calculations. Scenario calculations, in turn, may help to
develop site specific soil investigation strategies, to evaluate different combinations of remediation techniques, and
to design and optimize the process  control of the selected techniques.

The calibrated model was used to predict bioventing performance and the predictions were  compared to ten full-scale
soil venting projects. Ideal  assumptions were  made: (1) the aquifers were homogeneous and sandy; (2)  the
contaminants mobile and easily biodegraded;  (3) non-aqueous phase liquids were absent;  (4) that average measured
values represent degree of contamination and permeability; (5) retardation factor and Kd  are constant; and (6) zero
or first  order decay adequately describes  the kinetics of the process. Comparison with field data  revealed that
predictions based on averaged properties are too optimistic—the duration of the remediation is longer than predicted.
In full-scale remediations, the initial gas phase concentration is somewhat higher than estimated, and after a first
rapid  concentration decrease, the rate of the venting process changes considerably.  As  a result, the remediation
process  stagnates. The deviation between the  estimated and observed duration of remediation differs by at most one
order  of magnitude. The stagnation of biotreatment suggests that a significant fraction of contaminants is unavailable
to the treatment process.

Following up on the results presented at the  Oxford meeting, the authors used the ECOSAT model to more fully
investigate the stagnation phenomenon. The model was used to investigate whether availability of the  substrate was
reduced by non-linear contaminant sorption, heterogeneity or spatial variability in site parameters, or non-equilibrium
phenomena. The study concluded that non-equilibrium phenomena, particularly diffusion-controlled  mass transfer
from  areas of the soil where convective flow is absent, were the dominant factors in treatment stagnation. This
finding was supported by the anecdotal field observations that vapor-phase concentrations of contaminants often
increase when vapor extraction is interrupted.  The model was also used to compare continuous vapor extraction with
intermittent extraction  (i.e.,  for one day in  ten). Results showed that for either method, treatment times were
approximately the same since they  were dependent upon the  slowly diffusing contaminant fraction. However, the
intermittent extraction technique provided oxygen to the bioavailable water-soluble fraction of contamination much
more  efficiently, promoting biodegradation as the dominant contaminant removal process. In contrast, the continuous
extraction method rapidly reduced the concentration of available contaminant through physical volatilization and not
through biodegradation.

3. CONCLUSIONS

The use of average values as input variables  for remediation  models has been shown quite often to underestimate
the duration and cost of the  remediation program. By properly accounting for variabilities in site properties, such
as permeability, organic content, contaminant distribution, etc., more realistic  predictions of cost and duration can
be achieved. A high quality of soil investigation is necessary in order to generate  realistic predictions.
                                                   72

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)
    Project No. 24

           Combined  Remediation Technique for Soil Containing Organic Contaminants: Fortec®
   Technical Contact:
   Jan Bovendeur
   Heidemij Realisatie
   P.O. Box 5068
   5140 AR Waalwijk
   tel: +31/4160-44044
   fax: +31/4160-44080
Country:
The
Netherlands
Project Status:
Accepted by study  1993
Interim report           1994
Final report             1996
1. INTRODUCTION

In some cases, biological remediation of soils contaminated with petroleum hydrocarbons and polycyclic aromatic
hydrocarbons (PAHs) is only partially successful.  In order to  improve  biotreatment effectiveness for these
contaminants, Heidemij  Realisatie  is developing an innovative treatment system called Fast Organic Removal
Technology (Fortec*). The Fortec* process is a combination of several reportedly proven techniques for soil and
sediment treatment. This Pilot Study project described the development of the Fortec* process from bench- to full-
scale.

2. THE PROCESS

The objective of the Fortec* process is to improve treatment effectiveness through increased biodegradation rates;
decreased residual contaminant concentrations; and increased applicability to recalcitrant pollutants. Fortec* consists
of three unit operations  combined in series:  (1) multi-step hydrocyclone  separations; (2) UV/hydrogen peroxide
photochemical pretreatment;  and (3) slurry-phase bioremediation.

In many cases, organic contamination is found to be concentrated in the finer soil and sediment fractions, such as
in silt and clay. Hydrocyclone separation is used to separate the coarser sands and gravels, which are treated as
"clean," from the finer silts  and clays to  form a concentrate for further treatment. Heidemij claim efficiency of
separation is improved by using a series of hydrocyclones in a multi-stage configuration. Photochemical oxidation
of the fines concentrate is an optional process. The combination of ultraviolet (UV) light and hydrogen peroxide is
intended to partially oxidize organic contaminants in the soil slurry to enhance subsequent bioremediation, especially
of the more recalcitrant pollutants  such as higher molecular weight PAHs. The slurry is treated batch-wise within
a UV/hydrogen peroxide reactor  module.  Organic contaminants  (and their  breakdown intermediaries from
photochemical oxidation) are degraded  in a slurry-phase bioreactor operating at between 10-25°C with a slurry
containing from 3-15% solids. The bioreactor is aerated using a compressor, which injects air into the reactor through
a perforated plate at the bottom of the treatment chamber. Nutrients are mixed into the slurry prior to entering the
bioreactor at a  level appropriate to the concentration of contaminants present. Mixing within the bioreactor is
achieved by pumping the slurry from the bottom to the top. The bioreactor is operated on a batch-wise basis, with
typical residence times in the order of 3-20  days. After treatment, the slurry is transported to a settling tank and
dewatered by a belt filter press. The coarse and fine treatment products may be recombined, if appropriate.

3. RESULTS

Development of the Fortec* process has resulted from bench-scale, pilot-scale (25 and 50 m3 volume bioreactors),
and demonstration (300 m3 volume reactor) tests. All stages of scale-up have been supported by further laboratory
studies to measure biological activity during testing. Bench- and pilot-scale tests have focused on biodegradation with
and without photochemical pretreatment for soils contaminated with mineral oils, PAHs,  chlorophenol and  the
pesticide lindane ((3-HCH). Mineral oil contaminants were strongly degraded by the bioreactor (reductions from 5,000
mg/kg to <100 mg/kg in 3-8 days)  with  no  significant photochemical oxidation observed  (presumably the high
bioreactor degradation rates were an important factor). However, photochemical oxidation was found to be highly
effective at enhancing subsequent PAH biodegradation. PAH concentrations were reduced from 30 mg/kg to 5 mg/kg
in 15 days. Biodegradation without chemical pretreatment showed no PAH degradation over the corresponding

                                                  73

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

period. It was reported that enhanced degradation was due to the breakdown of the soil organic matter on which the
PAHs  were sorbed (thereby  increasing availability) rather than oxidation of the PAH molecules themselves.
Biodegradation of lindane was ineffective with or without photochemical pretreatment. It was suggested that lindane
was too recalcitrant to either chemical or biochemical reactions.

Based  on the initial studies, a full-scale Fortec* Bioreactor was designed and built. This was integrated with the
Heidemij fixed full-scale soil washing plant at Moerdijk (capacity 30-50 tonnes per hour) for demonstration tests.
A quantity (unspecified) of sandy soil contaminated with heavy crude oil (up to 14,500 mg/kg) was tested using the
soil washing plant and bioreactor. Soil washing separated two "clean" fractions: a coarse fraction (10.5% by soil
weight) and a sand fraction (87.5% by soil weight with an average contaminant concentration of 130 mg/kg). The
fines fraction (12.5% by soil weight) had an initial contaminant concentration of 110,000 mg/kg, which was reduced
to 22,000 mg/kg after treatment in the bioreactor (residence time of 40 days). After treatment, the soil fractions were
recombined. In a second demonstration, sediments contaminated with up to 2,000 mg/kg of PAHs were treated with
a final concentration of <300 mg/kg. No cost  data was included for either demonstration.
                                                   74

-------
NATO/CCMS Pilot Study, Phase II
                           Final Report (Appendix IV)
    Project No. 25
                                     Slurry Reactor for Soil Treatment
    Technical Contact:
    Rene H. Kleijntjens
    Biotechnology Integrated Research &
    Development (BIRD) Engineering
    De Brauwweg 13
    3125 AE Schiedam

    tel: +31/10-415-7822
Country:
The
Netherlands
Project Status:
Accepted by study 1993
Interim report       1994
1. INTRODUCTION

This project focused on the treatment of excavated clay soils and sediments contaminated with organic compounds,
such as mineral oil and polycyclic aromatic hydrocarbons (PAHs). The treatment technology described is the Slurry
Decontamination Process (SDP), which combines separation processes with a microbiological slurry reactor. The
objectives of this project were to:

  •  design and construct a pilot plant;

  •  establish all controlling process parameters;

  •  determine a cost per tonne of clay; and

  •  design a full-scale plant and secure funding to build it.

2. THE PROCESS

The flowsheet for the SDP system begins with soil pretreatment.  Vibrating grizzly bars are used to remove debris
(particle sizes >60 mm). The  undersize particles from the bar screen are fed to a micro-pulverizer. After grinding,
the soil is transported to a sieve where the coarse gravel fraction (>4 mm) is separated. In this wet screening section,
both recirculation water and the nutrient addition are used as washing water.

The first bioreactor is a DITS (Dual-Injected-Turbulent-Separation) reactor. In addition to biological degradation,
the slurried feed soil (particle sizes between 1 and 4,000  um) is separated into coarse and fine fractions.  Slurry is
injected at the bottom of the reactor creating a controlled turbulent mixing within the suspended slurry.  The coarse
fraction settles at the bottom of the reactor while the fine  fraction is kept in suspension. Slurry from the top of the
reactor is pumped to a solid-liquid separator where particles larger than 100 um are removed, and the treated slurry
is re-injected at the reactor bottom.

From the base of the first reactor, the coarse fraction from the DITS reactor is fed directly to a dewatering plant
where it is used to aid subsequent filtration of the treated fines. Fine particles from the top of the reactor are further
treated in a cascade of suspension bioreactors. Approximating plug flow, a cascade of three equally-mixed tanks are
optimized for first-order kinetics of contaminant degradation.  Due  to the removal  of the coarse fraction at the
beginning of the process, the  designers do not expect any settling problems within the cascade reactors, which are
agitated by injecting air through the base.

After being treated in the cascade, the slurry is dewatered by  means of filtration. As a filter  aid, the coarse fraction
from the DITS-reactor is used. From the dewatering section,  percolated process water is recycled to the beginning
of the process. Approximately 75% of the process water is recycled. Additional water input  comes from the supply
of nutrients to the system.  Process water temperature is maintained at 30°C by transfer of excess heat from the air
compressor. Therefore, biodegradation within the bioreactors is optimized by careful control of temperature, air and
nutrient supply, and slurry pH. To prevent escape of fugitive  air emissions from the treatment plant, air outlets are
directed to a biofilter.
                                                    75

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

3. RESULTS

The pilot-scale plant has demonstrated a reduction of 3,000 mg/kg oil to 200 mg/kg oil in 60 days; the remediation
proceeded very rapidly with concentrations reduced to 500 mg/kg after 25 days.

4. CONCLUSIONS

The designers have concluded that this technology has the potential to treat polluted wet clay and peaty soils from
either a mobile or fixed installation. However, the technology is restricted to the treatment of organic contaminants
and will cost more than other biological technologies such as landfarming.

[Note: No final report was provided on this project].
                                                  76

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 26
                                Treatment of Creosote-Contaminated Soil
                               (Soil Washing  and Slurry Phase Bioreactor)
   Technical Contact:
   James Berg
   Colifast Systems ASA
   P.O. Box 31
   1324 Lysaker

   tel: +47/67-59-9600
   fax: +47/67-58-1658
   E-mail: james.berg@colifast.no
Country:
Norway
Project Status:
Accepted by study
Interim report
1992
    1993
1. INTRODUCTION

This Pilot Study project reports on an investigation into the remediation of soil contaminated with three- and four-
ring poly cyclic aromatic hydrocarbons (PAHs) using a process that combines soil washing (especially froth flotation)
with slurry-phase bioremediation. The project consisted of two phases of work: bench scale treatability studies and
pilot scale remediation trials.

The Norwegian State Railways cooperated with the Norwegian Pollution Control Authority to support the technology
evaluation and develop the project. The Railways owns several former wood preserving facilities and wished to
evaluate options for treating possible  creosote contamination where total polycyclic aromatic hydrocarbon (PAH)
concentrations exist at levels between 270 and 51,000  mg/kg.

2. BENCH-SCALE STUDIES

Bench-scale biological treatability studies consisted of two steps: (1) screening of isolated "native" soil organisms
to determine their tolerance and ability to degrade PAHs; and (2) experiments to determine optimum conditions for
biodegradation of selected isolates in aerated slurry reactors  (1 dm3 in  size). SBP Technologies, Inc and the U.S.
Environmental Protection Agency (USEPA) in Gulf Breeze, Florida characterized the microbial populations. Five
different test soils contained 270 to 51,000 mg/kg  (dry weight) PAHs. These soils were investigated at 20°C under
pH-controlled conditions. PAHs were extracted from soils by four  serial extractions using boiling hexane.  The
investigators noted that the PAH data reported here may  be 2-5 times greater than those derived from the 16-
component standard used by the USEPA SW-846  Method 8100.

Studies to isolate enhanced PAH degraders showed that the  five soils contained degraders for phenanthrene and
fluoranthene, and these were closely related to organisms isolated earlier in the U.S. Tests to evaluate the effect of
added nitrogen and phosphorus revealed that at optimal concentrations  of these nutrients, acclimated  microbial
cultures degraded 94% of the PAHs within 7 days. In contrast, the control achieved only 20% degradation in the
same period.

The bench-scale  soil washing investigations  showed that the most effective  combination of one of the cationic
collectors and a foamer removed 90-95% of PAHs for sandy  soils. Increased water temperature or pH provided no
significant benefits and were incompatible with downstream biological  treatment.

3. PILOT-SCALE STUDIES

The pilot-scale plant set up at the Norwegian State Railways site in Lillestrom, Norway combined froth flotation as
a pretreatment for soil washing, a 1 tonne/hour soil washing  plant and  a 454-liter bioslurry reactor. This phase of
work examined four excavated soils made up of sand, silt, clay, and sawdust/sand from the railways site, as well as
two soils from another location. Initially, the soil was screened at 2 mm to remove potentially "clean" coarse material
and as a pretreatment for the  froth flotation process.  Evaluated variables for froth flotation included single and
                                                  77

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

blended commercial anionic and cationic surfactants used with a foamer, while soil washing variables consisted of
temperatures between 10-50°C and pHs between 7 and 11.

At the pilot scale, soil washing at a loading rate of 550-859 kg/hour removed 20-90% of PAHs from two clay soils.
The cleaned fraction contained 15-1500 mg/kg PAH. In terms of overall performance, clayey soils impaired the
efficiency of soil washing and resulted in less than optimum performance.

The pilot-scale biotreatment phase of the study tested indigenous microorganisms as well as patented degraders.
Amendments for bio stimulation included nitrogen and phosphorous, pH, aeration, surfactants, and temperature. Tests
were conducted with five 60-liter batches of sludge resulting from soil washing two clayey soils. During treatment
of these sludges containing 14-20% solids, PAH concentrations fell by up to 97% after 6  days.  Oxygen uptake
decreased during this time and corresponded to a lowering of bioavailable PAHs in the bioreactor. An interim report
noted that the native populations of PAH degraders "appeared to be sufficient to achieve residual PAH concentrations
ranging from 55-200 mg/kg after a 6-day retention time." The investigators attributed pilot-scale results—better than
those  achieved at the bench scale—to better mixing and aeration and a higher operating temperature (28-32°C).

A Microtox 15-minute bioassay revealed that the combined soil washing and biological treatment reduced toxicity
by a factor of  10. Soil washing itself reduced the relative toxicity only slightly. The bioreactor treatment accounted
for most of the reduced toxicity.

4. COSTS

Costs (in U.S.$ per cubic meter) for various aspects of the remediation were estimated at $160 for excavation, sorting
and backfilling; $300  for washing; and $530 for biological treatment.
                                                   78

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 27
                  Soil Washing and Chemical Dehalogenation of PCB-Contaminated Soil
   Technical Contact:
   James Berg
   Colifast Systems ASA
   P.O. Box 31
   1324 Lysaker

   tel: +47/67-59-9600
   fax: +47/67-58-1658
   E-mail: james.berg@colifast.no
Country:
Norway
Project Status:
Accepted by study  1994
Interim report       1996
1. INTRODUCTION

Leakage from PCB-containing insulator oils from stored transformer equipment contaminated soils with up to 2,000
mg/kg of PCBs  at a site in Oslo, Norway. The shallow distribution of contamination meant that groundwater
pollution had not occurred  and that the  most heavily polluted soils  (>10 mg/kg PCBs) could be excavated for
treatment and disposal. In Norway, the cost of incinerating PCB-contaminated materials ranges from U.S.$5,000-
$7,500/tonne; thus, soil pretreatment techniques  such as soil washing were viable economic options. This report
summarizes the Pilot Study project on the development of an on-site treatment for PCB-contaminated soils that
combines soil washing (as a volume reduction step) with chemical dehalogenation.

The three-year project included bench-scale studies on soil washing (completed in 1993), followed by pilot-scale and
full-scale washing in 1994 and 1995, respectively. Dehalogenation investigations began at the bench-scale in 1995
and ended at the full-scale in 1996.

The Pilot Study focused on a combined system of water-based soil washing and dehalogenation because of its unique
application to Norway. Both processes were new  in that  country,  and reductive dehalogenation was considered
innovative.

2. TECHNICAL CONCEPT

2.1 Soil Washing

The principal objective of this project was to treat the 1,288 tonnes of excavated PCB-contaminated material. Volume
reduction was achieved by separating debris (>0.05 m) from soil and by treating the soil in a washing plant. A total
of 11.9 tonnes of scrap iron, plastic, and wood were separated from the excavated material and hand sorted into
porous and non-porous components. Non-porous debris was surface washed to remove contamination (<5 ug PCB
/100 cm2) and disposed in an on-site landfill. Porous material was sent for incineration. Soil was treated in a washing
plant whose design was based on bench- and  pilot-scale testing. The aim of soil washing was  to achieve a
contaminated volume reduction of 70%. This objective was based on size characterization  of the soil,  which
suggested that the heavily contaminated silt and clay fraction (<0.1 mm size) represented 30% by volume of the total
soil.

The washing plant had an operating capacity of 1.5 tonnes per hour and comprised three basic processes for particle
size classification and surface cleaning. Contaminated soil was first slurried with water and a cationic surfactant in
an attrition mill (operating at a pH of 10-11 and a temperature of 35-50°C). Size classification was carried out using
a spiral separator, which produced a clean coarse fraction (>0.1 mm) and a contaminated fines concentrate (<0.1 mm)
for further treatment. The fines concentrate was treated in a double air flotation cell using a proprietary collector and
methylisobutylcarbanol as flotation reagents with the floated concentrate collected, coagulated, thickened, and filtered
for further treatment.  All coarse "clean" fractions were sampled and re-treated if contamination exceeded 10 mg/kg
PCB. After treatment, the coarse fraction was returned to the original excavation area which was subsequently
capped. Process water from the effluent plant was treated,  as necessary, before discharged to the sewage system.
                                                   79

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

The full-scale washing was completed in 1995 with the concentrated fine contaminated sludge from the flotation
process being stored until the chemical dehalogenation plant was ready. The biggest problem with full-scale washing
at the site resulted from the occurrence of higher than expected levels of fines (<0.1 mm) in the excavated soil as
a result of initially unrepresentative site characterization. The presence of up to 63% fines in the input stream caused
significant material handling problems for the washing plant and made the 70% volume reduction target impossible
to achieve. The final volume reduction was claimed to be 60% with 780 tonnes of clean soil and 400 tonnes of PCB-
enriched sludges being produced. Treatment lasted approximately 6 months.

2.2 Chemical Dehalogenation

Three chemical dehalogenation systems were evaluated at bench scale by technology vendors, and one was selected
for pilot-scale operation at the site. Details of the treatability study and the  exact nature of the treatment technologies
were not presented. However, all three bench-scale systems were based on the use of a reductive reagent that acted
on the contaminated sludge (<10% moisture content) in a closed chemical reactor. Results for soils containing around
250 mg/kg (dry weight) PCBs showed that for two  of the three processes tested at the bench scale, residual soil
levels of PCBs were less than 1 mg/kg. The unsuccessful third vendor attributed the failure to interferences by the
soil washing additives.

Treated soil  contained residual PCBs, byproducts formed during treatment, and  chemicals used to promote
dehalogenation of PCBs. All of these compounds posed possible environmental hazards. To evaluate these hazards,
a suite of ecotoxicological tests was conducted on leachate generated from samples treated according to the U.S.
Environmental Protection Agency's Toxicity Characteristic Leaching Procedure (TCLP). The interim report did not
contain results of these tests to evaluate the risk associated with clean and treated soil to be redeposited at the site.

The dehalogenation technology tested at the pilot scale involved first drying soil to 10% moisture, then adding a
nucleophilic agent and exposing the mixture to a reducing environment. The final report documenting results was
not available.

3. COSTS

The capital costs of the soil washing plant were reported to be about U.S.$750,000 with operating costs in the order
of U.S.$380 per tonne, which included all material  handling and chemical analysis. Operations include handling
materials and analyses (approximately 20% of operational costs).  Costs for dehalogenation were not available.
                                                   80

-------
NATO/CCMS Pilot Study, Phase II
                           Final Report (Appendix IV)
    Project No. 28
                Use of White-Rot Fungi for Bioremediation of Creosote-Contaminated Soil
   Technical Contact:
   James Berg
   Colifast Systems ASA
   P.O. Box 31
   1324 Lysaker

   tel: +47/67-59-9600
   fax: +47/67-58-1658
   E-mail: james.berg@colifast.no
Country:
Norway
Project Status:
Accepted by study  1994
Interim report       1996
Interim report       1997
to be continued in Phase III
1. INTRODUCTION

Lignin is resistant to degradation by many types of natural microorganisms. An exception to this are white rot fungi,
which are able to degrade lignin and other natural recalcitrant compounds through production of powerful free radical
extra-celluar enzymes. The enzymes produced by white-rot fungi are non-specific in their degradation mechanisms,
and their potential to degrade difficult to degrade contaminants such as polycyclic aromatic hydrocarbons (PAHs),
pentachlorophenol, and poly chlorinated biphenyls (PCBs) has been the increasing focus of research. This Pilot Study
project reports on the evaluation of two species of white-rot and one species of brown-rot fungus for their potential
to degrade high molecular weight PAHs in soil.

2. BACKGROUND

From 1994-95, samples for initial and bench scale testing were collected from an abandoned wood preservation site
in Southern Norway, which is contaminated with creosote. The initial total PAH concentration was  reported to be
2,400 mg/kg (based  on 40 PAH compounds and heterocyclic compounds). The proportion of PAHs that were 2- or
3-ring compounds was less than 16%, reflecting the age of the contamination. The remainder was accounted for by
4-, 5-, and 6-ring compounds.

Pilot-scale testing was scheduled to take place between 1995-96 using soils from an abandoned mine in Norway, but
has been delayed.

3. TECHNICAL CONCEPT

The contribution that fungi appear to make to bioremediation is in the degradation of poorly-soluble high molecular
weight compounds by extra-cellular lignolytic systems. The fungi of interest are primarily basidiomycetes, and most
studies have been carried out using the white rot fungi. The ability of these fungi to degrade these compounds (which
were reported to include DDT, TNT, PCBs and PAHs) derives from the non-specific nature of their attack on wood.

Wood is comprised largely of lignocellulose in cell walls around xylem tissue,which is used for water transport and
structural support by plants. For white rot fungi, hyphae are able to grow in the xylem lumen, but are unable to grow
into the cell walls. The cell walls are also resistant to penetration by large molecules such as enzymes. The main
energy source for these fungi is cellulose in the cell walls. Consequently, the degradation of wood is initiated by low
molecular weight compounds such as hydrogen peroxide, metal ions, and veratryl alcohol. However, the cellulose
in the cell walls is associated with lignin, which is a polymer that does not have a regular structure. Hence, no
specific  enzymes have evolved to degrade it. Rather, degradation of lignin takes place via a nonspecific array of
enzymes and free radicals. As the fungal hyphae cannot (at least initially) penetrate the cell wall, and because of the
highly damaging and non-specific nature of the degradation system,  the enzymes and other compounds are excreted
outside the fungal hyphae to do their work. Thus, there are advantages of fungal lignolytic degradation for treating
toxic compounds because:

    the degradation system is not specific so many compounds are susceptible to attack; and
                                                  81

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

    the degradation takes place outside the cell so less soluble and larger compounds can be degraded—compared
    with bacterial degradation, which is generally an intracellular process.

Two outstanding fundamental questions for fungal degradation are:

    Do fungal types other than white rot fungi have similar effectiveness (hence the test of a brown rot fungus)?;
    and

    What are the products of the degradation process?

The technical approaches being considered by the testing encompassed:

    fungal degradation activity;

    the impact of compost—for example, the possibility of immobilizing PAH in the  humic fraction; and

    pre-treatment of samples to  allow degradation of readily-degradable materials by  indigenous types prior to
    inoculation with fungi.

4. INITIAL FINDINGS

Initial  studies  were designed to investigate the manipulation  of parameters  such as  substrate  type, humidity,
temperature, and pH in order to  optimize fungal biodegradation of contaminants. Lignin-degrading enzymes are
generated by the fungus only when other nutrient sources are limiting and when lignin is also present as a substrate.
Nitrogen availability is a key  determining factor.

Screening tests were conducted using the white-rot fungi Pleurotus ostreatus and Trametes versicolor and the brown-
rot fungus Lentinus leptinus. The tests were used to  evaluate different lignin substrates, such as wheat straw, wood
chips (birch or pine), and newspaper, and to determine the influence of pH, aeration, and adding compost. Tests were
carried out either on  petri dishes or in 1-L Erlenmeyer flasks. Following these tests, Pleurotus ostreatus was
investigated further using wheat straw as the lignin  substrate.

Contaminated soils were incubated with straw and fungus for 8 weeks at 20°C. A range of supplements were added
(peat, compost, and potato pulp). Degradation performance was dependent upon the number of aromatic rings in each
PAH compound, ranging from up to 70% degradation for 3-ring PAHs to less than 35% for 5-ringed compounds
(such as benzo(a)pyrene).

Further bench studies  (using 60-liter batches) investigated the impact by degradation of indigenous microbial
populations as a pre-treatment, followed by fungal inoculation. The pre-inoculation incubation period was two
months. After inoculation samples were incubated for a further three months. Samples  incubated with bark prior to
fungal  inoculation exhibited the greatest PAH degradation, and this benefit was reduced if fertilizer was also added
during  the pre-inoculation incubation period. Degradation of  PAHs was greater at  20°C than 8°C. However, the
benefit of preliminary incubation with bark was  greater for the lower  temperature. Compared with the controls,
degradation of PAHs was enhanced by the fungal inoculation,  but degradation of heterocyclic compounds was not.

Radiolabeling  of the PAH  compounds enabled tracing  of treatment metabolites. The results produced  the
contradictory findings that indicated no accumulation of intermediate metabolites, yet  low rates mineralization.
                                                   82

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 29

     Sorption/Solidification of Selected Heavy Metals and Radionuclides onto Unconventional Solvents
    Technical Contact:
    Prof. Resat Apak
    Istanbul University
    Faculty of Engineering
    Avcilar Campus
    3840 Istanbul

    tel: +90/212-591-1998
    fax: +90/212-591-1997
Country:
Turkey
Project Status:
Accepted by study
Interim report
Final report
1992
    1994
    1997
1. INTRODUCTION

The project involved a laboratory-scale investigation of the use of solid wastes, such as red muds and coal fly ashes,
to absorb toxic  heavy metals and radionuclides from water (a fixation or  stabilization process) followed by
solidification of the metal-loaded solid wastes in a cement-based system.

2. BACKGROUND

Cadmium (II) (Cd2+), lead (II) (Pb2+) and copper (II) (Cu2+) are toxic heavy metals that pose a serious threat to the
fauna and flora  of receiving water bodies when discharged in industrial wastewater.  Cesium-137  (137Cs) and
strontium-90 (90Sr), with half lives of 30 years and 28 years, respectively, pose significant threats to the environment
as a result of fallout from nuclear bomb tests and reactor accidents. In recent years, land burial of radioactive wastes
has become a common practice, also posing a radioactive contamination risk to groundwater. In Turkey, 137Cs became
a matter of public concern after the Chernobyl accident, especially due to contamination of tea-growing areas on the
Black Sea coast.  Milk products and other biological materials containing Cs were also extensively investigated for
possible 90Sr contamination.

Various treatment technologies have been developed for the removal of these metals from water. Processes available
include ion exchange, electrodialysis, reverse osmosis, membrane filtration, sludge leaching, electrowinning,  solvent
stripping, precipitation, and common adsorption. The cost of adsorptive metals removal processes is relatively high
when pure sorbents, such as activated carbon or hydrated oxides, are used. Consequently, there is an increasing trend
toward substitution of pure adsorbents with natural byproducts or stabilized solid waste materials. Such materials may
also be useful for constructing "natural  barriers" around radioactive waste disposal  sites, especially for facilities
involving shallow-land burial of low-level wastes , in order to prevent the leakage of radionuclides from the  facility
to the environment.

The aim  of this Pilot Study project was to develop cost-effective unconventional sorbents, preferably metallurgical
waste solids, for heavy metal and radionuclide removal from contaminated water. The removal capacities of heavy
metals (Cd, Pb, and Cu) and radionuclides (137Cs and90Sr) and the sorption modeling of red muds and fly ashes were
studied in this regard.  The irreversible nature of sorption needs to be shown so as to  guarantee non-leachability of
metals from the metal-loaded sorbents.

3. SCOPE OF THE STUDY

Metal uptake (sorption) and release (desorption) were investigated by thermostatic batch experiments using bauxite
red muds and coal fly ashes. The distribution ratios of metals between these solid sorbents and the aqueous solution
were determined as a function of sorbent type, equilibrium aqueous concentration of  metals, and temperature. The
breakthrough volumes  of the heavy metal solutions were  measured by  dynamic column experiments so as to
determine the saturation capacities of the sorbents . The sorption data were analyzed and fitted to linear adsorption
isotherms.
                                                   83

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

The materials used were red muds (alkaline leaching wastes of bauxite from the Bayer process for the manufacture
of alumina) and fly ash from a coal-fired power station. The red muds comprise a mixture of minerals, including
sodium aluminosilicates, kaolinite, chamosite, iron oxides  (hematite), and iron hydroxides. 94% of the  mud had a
grain size less than 10  um. About 99%of the fly ash passed a 200-mesh sieve (74 um grain size).

Both materials were strongly alkaline  when leached with water. The red muds were washed with water to a neutral
pH, and then dried and sieved before use. Portions  of the  red mud were subjected to acid treatment. A portion of
this was, in turn, subjected to heat treatment at 600°C for 4 hours. The raw fly ash was washed with distilled water
5-6 times (washing removed about 40% of the initial fly ash by weight), filtered, and dried. A portion of the water-
washed ash was acid-treated, washed  with water, and then dried at 100°C.

The heavy metal solutions contained up to 10,000 mg/L of Cd, Pb, or Cu made from the corresponding nitrate salts.
No further pH adjustment was made. Solutions containing 137Cs and  90Sr were made up by  dilution of flacons of
standard  solutions supplied by Amersham International Ltd.

Desorption studies  were carried out using distilled  water,  saturated aqueous  carbonic  acid  (pH=4.75),  and
H2CO3/NaHCO3 buffer solutions  (pH=7.0) so as to simulate  carbonated groundwater conditions for the  purpose of
risk analysis around waste disposal sites.

4. SOLIDIFICATION OF THE HEAVY METAL-LOADED SORBENTS

The sorbents may also serve as effective fixation agents for removing heavy metals from water prior to solidification
for disposal. Although stabilization tests were conducted on the fly  ash and red muds, as well as on the  adsorbents
after they were loaded  with contaminating metals, the detailed results are not described  in the final project report.
Solidification of the red muds and fly  ash was accomplished by addition to a mixture  of cement, standard sand, and
carefully measured water. When metal-loaded solid waste was added (up to 20% by mass) to Portland cement-based
formulations, the fixed metals did not leach out from the solidified  concrete blocks over extended periods with the
exception of Cu2+ which reached  a concentration of 0.4  mg/kg after 8 months in water of pH  8-9.  In concentrations
below 20%, the compressive strengths and shear strengths of the doped concrete did not significantly differ from the
control concrete. However, there is a critical weight percentage of  10-20% additives  above  which the strength
declines dramatically.

5. CONCLUSIONS

The bauxite wastes of alumina manufacture (i.e., red muds) are hydrous oxide-like adsorbents capable of removing
the radionuclides 137Cs and90Sr as well as other heavy metals from water. Radiocesium and strontium adsorption on
red mud may be represented by B.E.T.-Langmuir isotherms.  Acid- and heat-treated red muds are  more effective in
Cs removal than water-washed red muds, but heat-treatment is detrimental to the surface -SOH sites important for
ion-exchange sorption of 90Sr.  Cesium uptake is predominantly irreversible and exothermic and increases with the
specific surface area of the sorbent, though some retained Cs+ may  be released with H+ and Na+ ions from the
surface. A rise in pH favors the exchange sorption of Sr, while the specific adsorption of Cs  is negatively affected.

Coal fly  ashes  of thermal  power  plants  are composite adsorbents  capable of removing  radiocesium  and
radiostrontium, as well  as other heavy metals from water. The adsorption of these radionuclides may be represented
by Langmuir isotherms as an approximation of multilayered adsorption expressed by the  B.E.T. equation. Although
acid treatment of fly ash has shown no  improvement in adsorption capacity, it is recommended to prevent trace
pollutant leach-out from the adsorbent into water.

In investigating  the possibility of using metallurgical solid wastes as sorbents in heavy metal removal  from
contaminated water, red muds and especially fly ashes have been shown to exhibit a high capacity for heavy metals
with the  sorption sequence Cu > Pb >  Cd in accord  with the order of insolubility of the corresponding metal
hydroxides. An empirical Langmuir approach could approximate isotherm modeling of metal sorption. Essentially,
the metals were held irreversibly  and would not leach out into carbonic acid or bicarbonate-buffered solutions.

The metal-loaded solid wastes could be solidified to an environmentally safe form thereby serving the two-fold aim
of water treatment and solid waste disposal.  A two-fold aim of heavy metal fixation and metallurgical solid waste


                                                   84

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

disposal would then be achieved with the constraint that fly ashes better serve the purpose of heavy metal fixation
than red muds.

Red muds and  fly ashes,  along with  other metallurgical solid wastes and clay minerals, may be utilized for
constructing "natural barriers" around shallow-land burial sites of low-level radioactive wastes and heavy metal-
containing products.

6. REFERENCE:

Apak, R., G. Atun, K. Guclu, E.  Tutem, M. Hugtil, J. Hizal and S. Ayhan. Sorption/solidification of selected heavy
metals and radionuclides from water, Final Report to NATO/CCMS meeting Golden, March 1997.
                                                  85

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 30

            Using Separation Processes from the Mineral Processing Industry for Soil Treatment
   Technical Contact:
   Michael  Pearl
   Planning Directorate
   UKAEA
   Harwell
   Did cot
   Oxfordshire
   0X11 ORA

   tel: +44/1235-435377
   fax: +44/1235-436930
   E-mail:,  michael.pearl@ukaea.org.uk
Country:
United Kingdom
Project Status:
Accepted by study      1992
Interim report           1993
Final report             1994
1. INTRODUCTION

This report summarizes a project carried out to investigate the general application of particle separation techniques
from the mineral processing industry to the treatment of contaminated soils from the U.K.

Physical separation techniques can be used to divide contaminated soil into fractions that can be relatively easily
treated.  Physical treatment can also act as an early process in a treatment train involving subsequent chemical,
biological, or thermal processes to reduce the volume of material requiring subsequent remediation and to convert
the material into a more suitable form (such as a slurry).

The project was included in the NATO Pilot Study because it aimed to design and operate a soil washing plant to
treat contaminated soil which had attributes unique to the U.K. The  Warren Spring Laboratory3 carried out the work
on behalf of the U.K. Department of the Environment and the U.S. Environmental Protection Agency.

The study comprised two phases:  laboratory characterization tests,  and  pilot-scale tests. These are described in
Sections 2 and  3, respectively.

2. LABORATORY  CHARACTERIZATION TESTS

Laboratory characterization tests were developed to ascertain treatability potential by physical segregation alone or
in combination  with other treatment technologies. Soil samples (5-20 kg) were collected from various UK locations
for characterization.  Sites included former  gasworks, coke works, canal dredgings,  chemical  works,  metal
reprocessing and pickling plants with principal contaminants including heavy metals, inorganic cyanides, polycyclic
aromatic hydrocarbons (PAHs), and phenols.

Bench-scale laboratory characterization tests were carried out to determine the preferential distribution of the majority
of the contamination to specific particle fractions of the soil,  leaving the bulk of the soil depleted in contaminants.
These tests were based on differences in grain-size distribution; settling velocity (hydrocyclone separation); abrasion
by using attrition scrubbing to abrade contaminant coatings, followed by hydrocycloning; specific gravity distribution
using sink-float tests with heavy liquids; surface chemical properties, specifically hydrophobicity using froth flotation;
and magnetic properties  assessed using the Franz Isodynamic Separator.

Results from the laboratory tests showed that most of the examined U.K. soils displayed some degree of preferential
distribution of contaminant to particular soil fractions but:
        In 1994 Warren Spring Laboratory was merged with the Environmental Division of AEA Technology to form the National
        Environmental Technology Centre. AEA Technology has more recently been privatized.
                                                   86

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

        the level of contaminants in the depleted fractions was not always sufficient for the fractions with the lowest
        concentration of contaminants to be considered suitable for reuse as inert fill; or

        the proportion of soil in the contaminant depleted fraction was not always large enough for commercial
        exploitation.

The results from two of the soils from the laboratory treatability studies were used to design treatment circuits, which
were tested at pilot scale.

3. PILOT-SCALE TESTS

Pilot-scale tests were conducted with the Warren Spring Laboratory's National Environmental soil washing plant.
Tested contaminated soils (up  to fifty tonnes) came from the  sites of a former  metal reprocessing works and
gasworks. Material from both sites consisted of building rubble, wastes,  and soil. A relatively high clay content of
around 30% made handling this material difficult.

The throughput of the pilot plant was 0.5 tonnes of soil per hour with the selected remedial scheme based on the
exploitable separation features ascertained in the  characterization tests. This plant  compared froth flotation and a
Multi-Gravity Separator for separating fine particles. Altogether, treatment processes consisted of particle release and
coarse sizing, further sizing, attrition scrubbing and classification, and contaminant concentration. The plant featured
process water containment and recirculation, and carbon filters to extract volatile contaminants from emissions.
Wastes and sludges from the plant received further treatment at downstream processes or were shipped to a licensed
landfill.

The level of contaminant reduction, when compared to the original soil, measured the effectiveness of the process.
In the pilot-scale tests using froth flotation, residues with low levels of contaminants (/'. e., treated material) made up
around 48% by weight of the original feed. For tests with the Multi-Gravity Separator  instead of froth flotation,
residues with low levels of contaminants totaled around 50% of the original feed. Compared to their respective feeds,
both residues contained  markedly lower levels of PAHs, petroleum hydrocarbons, cyanide,  lead,  and arsenic.
Comparing the residues themselves, those derived from froth flotation contained higher levels of PAHs and petroleum
hydrocarbons but lower levels of lead and cyanide. Arsenic levels did not differ much in the two.

Samples from the former metal reprocessing works produced a small portion of dense material (specific gravity >2.8)
with marked concentrations of contaminants. Despite this separation into the dense fraction, the remaining bulk of
treated material contained unacceptable levels of contaminants. According to a report, rather than consider absolute
concentrations only, researchers evaluated Toxicity Characteristic Leaching Procedure (TCLP) data to assess this soil
in terms of mobile and non-mobile contaminants. The researchers reconsidered the role of physical treatment to
reduce hazards presented by this soil. An option would be to combine physical separation with a mild  acid leach to
reduce the  mobile contaminants in the  soil to below  an indicative leachability  criteria  (in  this case, TCLP).
Concentrates from the physical separation and precipitates from downstream processing of metal-rich solutions could
then be potentially stabilized or treated to recycle the metals.

With a soil from a former  gasworks, the  maximum volume reduction would be  expected  to be 60% with an
optimized treatment circuit. The authors suggested that the reason for this was that the  clay and fine silt fraction
(O.01 mm) of the soil represented approximately 30% of the total weight of material. Treatment of the clay and silt
fraction using a slurry phase bioreactor could  potentially  reduce the organic contaminants in the soil. Physical
separation within this treatment train would have acted as  a pretreatment  stage not only to reduce the volume of
material  requiring downstream treatment, but also to present the material to the downstream process  in a more
appropriate form, that is for example, as a slurry not as a mixture of contaminated rubble combined with clay.
                                                    87

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 31
                               Decontamination of Metalliferous Mine Spoil
   Technical Contact:
   Ivor Richards
   Richard, Moorehead and  Laing Ltd.
   55 Well Street
   Ruthin, Clwyd
   LL15  1AF

   tel:  +44/1824-704366
   fax: +44/1824-705450
Country:
United Kingdom
Project Status:
Accepted by study  1992
Final report         1993
1. INTRODUCTION

Metals in discarded spoil can be  solubilized by rain or shallow groundwater, or can be dispersed into sensitive
environments when dry spoil is blown. Traditional reclamation systems focus on containment, not decontamination.
These systems rely on impermeable strata below waste, and/or a cap consisting of low-permeability materials like
clay.

The objective of this Pilot Study was to determine if current metal-processing techniques for extracting metal-rich
ores from less valuable waste material could be used to reprocess old metalliferous spoil materials to recover metals
of some value and to reduce the environmental impact of the spoil. For a period of three years starting in 1990, work
was conducted to evaluate methods for decontaminating metalliferous spoil from abandoned mines in central and
north Wales. The investigation focused on lead-zinc mining in Wales—these metals are of particular significance
in current and former mining operations across Europe.

The Water and Environment Division of the Welsh Office commissioned the study. Richards, Moorehead and Laing
Ltd. undertook the work in association with the School of Engineering (Materials and Minerals) within the University
of Wales College, Cardiff.

2. THE STUDY

Laboratory-scale studies were conducted to evaluate the mineral processing techniques on lead-zinc mine spoil
materials from five unreclaimed former metal mines. The evaluation  of techniques represented the first phase of a
two-part study to identify promising treatment options to be examined  in detail during the second phase of the study.
During the study, researchers first characterized spoil material and  then examined techniques involving gravity
separation, froth flotation, chemical leaching, and biological extraction.

The mine spoil was characterized to determine the  mineralogy of contaminated particles and the distribution of
contamination according to particle size. Concentrations of lead and zinc in the whole spoil samples were up to 20%
and 15% by mass, respectively, but copper and cadmium concentrations were in the range of parts per million. Lead
was observed to concentrate in the finer particle sizes of the spoil.

The nature of the spoil influenced minerals  produced by weathering. A fine-grained, compact, clay-like spoil
saturated with water created anaerobic conditions that inhibited oxidation of the minerals. Under these conditions,
the minerals retained their sulfide forms. In contrast, a sandy spoil allowed extensive oxidation to occur and convert
lead and zinc minerals to their more easily leached sulfate forms.

Mineral processing techniques exploit physical and chemical differences between contaminated and uncontaminated
particles. In this study, differences in particle density and surface chemistry for the separation of spoil particles were
evaluated using dense media (so called "sink and float") and froth flotation tests.

A multi-gravity separator (MGS) was used to further evaluate density separation at the pilot scale. Laboratory-based
density separation resulted in consistently reduced metal concentrations (<2% by weight) in the lighter spoil fractions

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

representing over 90% of the total sample weight. The heavier concentrate contained up to 32% lead and 5% zinc
by weight. Although significantly cleaner than original spoil, these levels still greatly exceeded U.K. soil guidelines.
The pilot-scale MGS treatment produced similar results on the less than 0.5 mm fraction of spoil. Results showed
that the MGS can concentrate almost 70% of the metals into a mass of less than 10% of the original spoil, leaving
a combined metal assay of 2.5% in the residue.

For most of the tested spoils, froth flotation performed poorly but achieved reasonable performance where spoils
remained unoxidized. Additives showed little promise in terms  of producing marketable concentrates or tailings with
significantly lower metals contents. The  effectiveness of this technique depended critically upon the mineralogy of
the spoil.  Unweathered material  that was  rich in sulfides showed better segregation of contaminants into the
concentrate.

Leaching of metal  contaminants  from the  spoil by  a variety of  chemical agents  such  as  sulfuric  acid, sodium
hydroxide,  and diethylenetriamine, and by using ferric bacteria inoculum was evaluated on unprocessed spoil and
treated fractions from the density  and froth flotation tests. Effective chemical leaching depended on the degree of
spoil weathering. Generally, sulfuric acid and diethylenetriamine, leached 2-33% of the lead, and 12-64% of the zinc.
Sodium hydroxide leaching of weathered spoil mobilized 25-92% of the lead and 3-23% of the zinc in unprocessed
spoil.

To evaluate the extent to which metals remaining  in reprocessed (by gravity separation or froth flotation) residues
would be leached and, hence, be  of environmental concern, the residues were subject to chemical and biological
leaching. This leaching removed only 2-5% of the remaining contaminants, indicating a resistance to  leaching in
residual material. Bacterial leaching of lead proved to be ineffective, but significant movement of zinc from the spoil
was recorded.

It was recommended that future work focus  on the applicability of alkaline leaching  and its integration with gravity
separation methods  for the treatment of this  type of metalliferous waste. Further work on the pilot-scale assessment
and optimization of the MGS was also suggested.
                                                    89

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 32
                                   CACITOX™ Soil Treatment Process
   Technical Contact:
   Mr. Ken W. Brierely
   BNFL Engineering Ltd.
   The Victoria
   Harbour City
   Salford  Quays
   Manchester M5 2SP

   tel. +44/161-952-6354
   fax. +44/161-952-6091
   E-mail:  ken.brierely@beltd.com
Country:
United Kingdom
Project Status:
Accepted by study  1992
Interim report       1993
Final report         1994
1. INTRODUCTION

CACITOX"" is a multi-component leaching process using a proprietary combination of highly selective and environ-
mentally tolerant reagents. It is used in conjunction with a soil washing pretreatment and effluent recovery system
called the EXCEL*CR™ Soil Treatment Plant. The CACITOX™ reagent comprises low concentrations of carbonate,
oxidants, and complexing agents. This mixture, used at near neutral pH and at ambient temperature, reacts with
contaminated materials and  converts insoluble or absorbed contaminants  such as  heavy metals and transuranic
elements into soluble complexes. The low reagent concentrations and their high selectivity are designed to minimize
secondary waste.

The leaching process is considered an emerging technology because it represents an innovative way to chemically
treat contaminated soils. One of the  objectives of the process is to treat soils with significant clay and silt content
for  which  conventional  soil washing techniques are not cost-effective.  The  technology can potentially treat
contaminated soils and sediments, without depositing unacceptable by-products into the soil matrix or destroying the
matrix itself.

2. THE PROCESS

The project consisted of three phases: (1) evaluation of the process with heavy metals; (2) evaluation with radionuc-
lides; and (3) engineering design for a mobile ex situ  soil treatment plant. Early testing using spiked synthetic soils
demonstrated the effectiveness of the  CACITOX"" process in reducing contaminant concentrations to regulatory levels,
even in the high-clay soils. There are four stages to the process: (1) size classification,  where oversize materials are
sorted out and contaminated soil volume is reduced; (2) soil leaching using the CACITOX"" reagent, with laboratory-
optimized parameters for the particular soil and contaminant conditions of the site; (3) soil/leachate separation using
a bank of hydrocyclones, followed by dewatering with filters; and (4) leachate treatment using precipitation and ion
exchange processes. Precipitated contaminants are ultimately treated by containerization,  encapsulation, or dewatering.
The pilot-scale transportable plant has a capacity  of 10 kg/hour, with easy  scale-up to 100 kg/hour.

3. PERFORMANCE

Performance evaluation was accomplished at bench scale using synthetic test soils spiked with aqueous solutions of
the nitrate form of commonly encountered metals.  Seven heavy metals and six radionuclides were evaluated at "low"
and "high" concentrations. Since clay soils proved the most intractable, most testing was conducted with fine-particle
soils. In general, all heavy metals partitioned in a similar manner in the various  soil size fractions.  While the  clay
and silt fractions contained the bulk of the spiked contamination, 26% of the contaminants were contained within
the coarse sand/gravel fractions. For the "high" spiked samples, the coarse fraction exceeded the Dutch "B" Values8
or Canadian residential levels for all metals except cadmium and arsenic.  Thus, conventional soil washing alone,
        These values have now been superseded.
                                                   90

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

which relies on particle size classification, would be ineffective in treating these soils. Soil pH did not appear to
significantly affect contaminant removal. While the bench-scale tests failed to meet regulatory levels for soils spiked
with 1,000 mg/kg cadmium or arsenic, it was reported that further optimization of the reagent formulation would
enable target values to be achieved. Similarly, the addition of organic contaminants had little effect on initial leaching
efficiencies, and  98%  of the  organic contaminants  were removed from  the  soil.  Removal efficiencies were
comparable to leaching with mineral acids, but CACITOX"" dissolved less than 10% of the soil matrix (compared
to 40% for acid washing).

During the leaching trial for removing radionuclides from test soils, a single contact with the CACITOX"" reagent
mixture removed 52-76% of the plutonium, americium, neptunium, strontium and radium, but only 2% of the cesium.
Limited data for removal of radionuclides from a contaminated site showed how repeated contacts could improve
removal.

4. CONCLUSIONS

In summary, the CACITOX"" technology can process materials containing high levels of fine-grained particles like
silt  and clay and can selectively dissolve actinides and heavy metals from other wastes. Because the CACITOX""
process  treats inorganic contaminants, it requires other technologies for treating mixed wastes. The final report
mentioned that using low concentrations of mild chemicals resulted in low costs, but did not provide supporting data.
                                                   91

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 33
                       In-Pulp Decontamination of Soils, Sludges, and Sediments
   Technical Contact:
   Giles Gillett
   Kvaerner Water
   P.O. Box 37
   Ashmore House
   Stockton-on-Tees
   TS18 3RE

   tel:  +44/1642-341000
   fax: +44/1642-341001
Country:
United Kingdom
Project Status:
Accepted by study  1992
Interim report       1993
Final report         1994
1. INTRODUCTION

This project reports on a collaborative program of bench-scale and demonstration plant-scale development of a
treatment technology by Davy and Kommunekemi of Denmark. The project receives funding from the U.K. and
Danish governments under EUREKA EUROENVTRON, and has been  accepted into  the  U.S. Environmental
Protection Agency's Emerging Technology Program. The NATO/CCMS accepted this project  in 1992 for its novel
application to "in-pulp" adsorption of heavy metal contaminants.

The process combines conventional ex situ soil washing techniques with an innovative chemical treatment stage. A
major advantage expected for this approach is the ability to remove contaminants from fine-sized soil fractions such
as silt and clay. Two techniques are used to extract contaminants:

    (1)  leaching using acidic or alkaline reagents followed by adsorption to activated carbon or ion exchange resin;
        and

    (2)  adsorption by activated carbon or cation exchange resins in direct contact with a soil slurry.

Contaminants may  then be  desorbed from the recovered  activated  carbon or ion exchange resin, which is
subsequently recycled.

2. TEST PROGRAM

Test materials consisted of contaminated soils and sediments. The soils came from:

  •  a reclaimed site contaminated with arsenic;

  •  a gasworks facility  containing zinc and lead;

  •  a wood preserving site contaminated with copper, chromium and arsenic; and

  •  chlor-alkali and gas metering sites with mercury contamination.

In addition, a sediment sample containing zinc, lead, and iron was collected from Hamilton Harbour in Ontario. Zinc,
lead, iron, and manganese levels in the  sediment exceeded Ontario's  "severe effect" guidelines, while copper,
chromium, and nickel levels ranged between the "severe effect" and "limited effect" levels.

This summary highlights work done on samples from the wood preserving site, the chlor-alkali site, and the harbor.
Project reports did not contain information on the other samples.
                                                  92

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

Initial tests focused on the physical and chemical characterization of the contaminated materials. This characterization
included determining the contaminant distribution according to soil particle size. Since contamination existed in all
fractions of soil collected from the wood preserving site, leach tests were carried out on samples of the whole soil.

The leaching work was based on small pilot-scale experiments. These examined various inorganic and organic acids,
alkalis, and chelating reagents under various conditions. A 300-g sample of soil was stirred constantly with 1  liter
of leaching fluid for a period of 24 hours with the concentration of contaminants in the leachate and soils periodically
monitored. It was concluded that mild leaching was inadequate for obtaining target contaminant concentrations in
the solid residue.  Sulfuric acid was the most effective leaching agent, removing 90-97% of the contaminants from
the soil. Copper, zinc, and chromium were removed to below target levels, but arsenic concentrations exceeded the
target of 30  mg/kg.

Adsorption tests on leachate from soils from the wood preserving site were carried out with various ion exchange
resins, activated carbon, and magnetite to treat the leachates produced by the leaching procedure. Compared with
the other adsorbents, the ion exchange resins were shown to remove the most copper, chromium, and zinc from the
leachate. Acid resins adsorbed from 75-100% of the copper, chromium, and zinc from the leachate but only 20-60%
of the arsenic. It was suggested that  arsenic was difficult to  treat using resins since at the  low pH values where
arsenic acid  is most soluble, it does not normally exist as ionic species. At high pHs, the ion exists in solution but
adsorbs  onto iron  precipitates found under these conditions. A further complication was the competitive adsorption
of bulk metal ions, such as calcium and iron, which necessitated a change in flow sheet design to include a chemical
precipitation step  to remove these ions prior to adsorption.

A 10-kg sample of the  soil from the wood preserving  site was treated at bench scale using  the unit operations of
physical separation and chemical leaching that  had been identified from the experimental studies. The metal content
in the treated soil was lower than in the feed material: copper (22 vs. 360 mg/kg), chromium (74 vs. 621 mg/kg),
zinc (68 vs.  414 mg/kg) and arsenic (102 vs. 1,204 mg/kg). Only arsenic exceeded its preliminary treatment target
of 30 mg/kg. Further work on arsenic removal has been conducted on a different soil sample. Focusing on protocols
that take account  of arsenic speciation, chelating agents were shown to leach a maximum of 52% of arsenic from
the soil. A combination of flotation, screening, and hydrocycloning achieved 60% removal in 80% of the soil, but
this still exceeded the target level. Multiple acid leaching was able to reduce arsenic from 650 to 22 mg/kg, but at
a relatively high cost. The degree to which using such strong acids destroyed the soil matrix was not mentioned but
is likely to be a significant effect. An alternative suggestion involved raising the leachate pH to co-precipitate arsenic
acid with  iron, but no test results for this technique were reported.

Trials with the mercury-contaminated soil used oxidative and complexing conditions such as nitric acid, hydrochloric
acid, and sodium hypochlorite. To improve extraction of mercury, the approach included size separation to  remove
fines and was operated at higher temperatures. Ion exchange resins developed for mercury adsorbed the metal slowly.
As a result of difficulties with this approach, a thermal option was investigated. Preliminary tests  revealed that by
heating  contaminated materials to around 800°C, treated soils achieve regulatory targets for  mercury.

Work with Hamilton Harbour sediment aimed to demonstrate the technology, obtain necessary data for pilot plant
design, and show that the Ontario guidelines for sediment could be met. Leaching tests were conducted with mineral
and organic acids. During leach screening tests, strong mineral acids dissolved contaminants, as well as calcium and
iron, in 90 minutes.  These two metals and organic contaminants would compete with the metal contaminants during
adsorption. A two-stage leaching approach was tested as a way to eliminate this competition. This approach involved
a mild acid  to dissolve iron first, followed by a strong acid to attack the contaminants. The presence of organic
contaminants had  little impact on the removal of metals.

Tests  with several adsorbents pointed to chelating resins as the likely candidate. These resins proved to be more
selective than activated carbon or magnetite for the contaminant ions over the other metal ions. The  high level of
iron in solution inhibited the adsorption of contaminants, and thus required pretreatment such as magnetic separation
to remove iron from the sediment before leaching. Precipitation to remove metals from the leachate was somewhat
ineffective because some  metal remained in solution and required further processing.
                                                   93

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

3. RESIDUALS AND EMISSIONS

The process generates a variety on residuals and emissions, some of which require further treatment or disposal.
Decontaminated solids may require final treatment, such as pH adjustment and dewatering, before being disposed.
For most metal extractions, acid is used to remove metals from the resin,  resulting in a concentrated solution of
metals. These  solutions may be treated by precipitation, reduction, or electrowinning to recover metals, or may be
disposed at  a  secure disposal site or encapsulated. In the case of sediments,  the leach stage may produce gas
emissions if anaerobic activity produces sulfides.  These emissions, noted during leach tests with sediments from
Hamilton Harbour, may require gas scrubbing in a commercial plant. Decanted water from excavated sediment may
also have to be treated. Finally, cleaned soils and sediments may introduce anions to their disposal site.

4. COSTS

Little information is available on cost. If the process demands aggressive leaching with oxidative and complexing
agents as well as elevated temperatures for enhanced leaching, the treatment plant will have to be built of unique
materials. Under these conditions, treatment costs are expected to be high (no data or expected costs were provided).

"In-pulp" processing, used alone or combined with other physical processing techniques, offers an option to treat
soils chemically to achieve low absolute values of metals. If a teachability standard is required, a chemical treatment
may be attractive provided the leaching agent is more  severe than the standard requires. Thus, applicability of the
technology would depend on its ability to achieve regulatory requirements.

5. CONCLUSIONS

Although applicable to metal contamination, the technology may be adapted to handle organic contaminants by using
carbon as an additional adsorbent. Plans to develop the technology further include investigating surfactants, solvents
and other agents to extract organic contaminants and to combine this with adsorption by activated carbon.
                                                   94

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 34
                     Chemical Fixation of Soils Contaminated with Organic Chemicals
   Technical Contact:
   Neil  McLeod
   Envirotreat Limited
   Suites 2/3, LCP House
   The  Pesnett Estate
   Kingswinford
   West Midlands, DY6 7NA

   tel: +44/01384-288876
   fax: +44/01384-288876
Country:
United Kingdom
Project Status:
Accepted by study  1993
Interim report       1996
Final report         1997
1. INTRODUCTION

Conventional stabilization/solidification methods, such as those based on cementitious and pozzolanic materials, have
been applied with limited success to treating soils with organic contaminants. Hydrocarbon chemicals can have
detrimental effects on cement hydration and structure formation. This project involved the development of a range
of modified organophilic clays with the aim of overcoming these limitations in an advanced in situ stabilization/
solidification application based on cementitious  binders. This technology has been tested at field scale on a site
contaminated with a variety of inorganic and organic substances.

2. PROCESS DESCRIPTION

The Envirotreat process is an in situ technology developed with the financial support from the U.K. Department of
the Environment's Environmental Technology Innovation Scheme (ETIS). The research and development has been
carried out in collaboration with the former Warren  Spring Laboratory,  the University of Birmingham, and May
Gurney Ltd., a U.K.-based civil  engineering contractor.

The stabilization/solidification process uses specialized modified smectite clays that contain reactive species.  These
clays  are claimed to have diverse potential environmental applications  and to be compatible with conventional
stabilization/solidification process materials. The clay technology is primarily targeted at organic contaminants, but
also has the capability to treat both cationic and anionic heavy metal species by a combination of  ion-exchange
processes and chemical interaction with pillaring agents and other intercalatants.

The technology uses modified continuous-flight auger drilling and injection techniques. The treatment materials are
injected in a slurry form into the contaminated ground and mixed with the soil in situ. The primary challenge is to
ensure that the contaminated soil and treatment media are mixed homogeneously. Envirotreat is now evaluating the
in situ process as an "active containment system" and is considering a number of options, including the application
to permeable barriers/reactive treatment walls.

3. BACKGROUND

Commercially available  organophilic clays are typically made by substitution of quaternary ammonium salts into
phyllosilicate clay matrices (e.g., montmorillonite). The clays are commonly employed in the paint and oil industries
as sorbents  for organic chemicals.  The primary objective of  this  project was  to  modify these clays through
intercalation (i.e., substitution of the cations between clay layers) to improve stabilization of organic  contaminants
by:

    (1)  increasing the size of the interlamellar spacing within the clays to  accommodate larger-sized molecules such
        as polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs), the effective surface
        area and the potential reactivity by the introduction of selected pillaring agents;
                                                   95

-------
NATO/CCMS Pilot Study, Phase II
                                                                              Final Report (Appendix IV)
    (2) optimizing the polarity of the interlamellar environment in order to increase absorption and adsorption of
        organic contaminants;

    (3) providing a  reactive environment  on  interlamellar clay  surfaces to  chemically  bind contaminants
        permanently; and

    (4) providing an active medium for effective treatment of heavy metals (both cationic and anionic) and other
        inorganics by cation exchange processes and/or interaction with the intercalated species (which can then be
        immobilized within the cementitious matrix following alkaline precipitation).

Certain intercalating agents, such as transition metal complexes and  organometallic cations, are  known to have a
pillaring effect on clays, acting as molecular props or "pillars" between clay lamellae to create a two-dimensional
porous interlayer structure. This improves clay stability to excessive dehydration and widens interlamellar spacing
to allow larger  organic molecules such as PAHs  to enter. By using iron- and aluminum-based pillaring agents, a
variety of chemically reactive sites can be created inside the  interlayer spacing.  For example, the presence of
aluminum ions  (A13+)  adjacent to the interlamellar spacing provides an effective site for Lewis acid/base reactions
with chlorinated and  heterocyclic hydrocarbons.  In addition, numerous other forms of chemical bonds may be
possible between the  intercalated organophilic clay and the  contaminant, including  van der Waals,  dipole, and
hydrogen bonds. Diagrammatic representations of the generic modified clays are shown in Figures 1 and 2.
              R


            M3+
                     QAS
                                          QUATERNARY  AMMONIUM
                                          N+   (R)4
                                                                    - ALIPHATIC
                                                                      AROMATIC
                        3*      2+
                     AI
                        3+
                             (M2+)  -     TRANSITION  METAL

                                      "    HYDROXY  ALUMINIUM  PILLAR
Figure 1: Schematic of modified clay
An advantage of using tailored organoclays is  that a clay can be  manufactured to  treat a specific group of
contaminants by manipulating the type and amount of intercalating agents used. By selective use of available smectite
clays and intercalation of appropriate quaternary ammonium ions in conjunction with other relatively low-cost agents
(principally Fe and Al compounds), it is possible to produce highly-efficient and multifunctional treatment media.
Envirotreat's aim was to produce a range of clays with varying hydrophobicity and chemical composition to deal
with five groups of organic compounds:
                                                   96

-------
NATO/CCMS Pilot Study, Phase II
                      Final Report (Appendix IV)
       A1
           7+
A1
    7+
                                                    A1
                                                       7+
                                                                 HYDROXY ALUMINIUM PILLAR
                                                                 AI3+ - AI203
                                                                 Al7* - AI1304 (OH)24 (H20)127+
                         LOCALISED CHARGE ZONES
    ADDING EXCESS AI3+/AI7+ PRODUCES POSITIVELY CHARGED CLAY
    CATION EXCHANGE WITH Al3* NOT ASSOCIATED WITH PILLAR
    ORGANICS ATTRACTED TO N*R4 ANIONS ATTRACTED TO PILLAR / CLAY (+ve) I
Figure 2: schematic of modified clay (positive)
  •  PCBs, dioxins, chlorinated hydrocarbons;

  •  PAHs;

  •  benzene, aryl compounds, benzene derivatives;

  •  hetrocyclics, organophosphorus/sulfurous compounds; and

  •  polar compounds, phenolics.

It was acknowledged that this approach would not cover all eventualities. However, it was felt that most organic
contaminants likely to be encountered at sites would react to some extent with one of the modified clays. Where
there is more than one type of organic contaminant at a site, the intention is to use more than one clay formulation,
as deemed appropriate.

4. INITIAL LABORATORY ADSORPTION TESTS

Adsorption tests were carried out on three representative organic compounds: benzene, orthodichlorobenzene, and
pyridine. Benzene  and orthodichlorobenzene have low solubilities in water and were well adsorbed by all clay
formulations. In contrast, the results for pyridine indicated that the  more hydrophillic (inorganic)  clay formulations
are generally more effective in removing organic compounds of high solubility. The results confirmed the need for
a range of modified clays.

Further sorption tests  were carried out at the University of Birmingham. This research focused on the  ability of
pillared clays to sorb humic acids, which are common organic constituents of groundwater. When groundwater is
chlorinated, humic acids can react to formation  mutagenic or carcinogenic substances. It was  concluded that
aluminum and aluminum+phosphoric acid pillared clays were most effective in sorbing fulvic acid (a light humic
acid) from water. Iron pillared clay was generally less effective, and a combined iron/aluminum pillared clay was
the least effective (probably due to its low surface area).
                                                 97

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

Conventional smectite clays have cation exchange properties that can be used in the treatment of cationic species
(e.g., toxic heavy metals) under controlled conditions. They are not, however suitable for the treatment of problematic
anionic and complexed species, in particular chromates and dichromates, arsenates, and ferriferrocyanides, etc.
However, modified pillared clays were shown to be effective for the treatment of hexavalent chromium solutions.
Untreated bentonite was ineffective in absorbing CrO42". The modified clay formulation provided a reactive medium
for the treatment of both cationic Cr3+ and anionic CrO42VCr2O72".

5. AUGER DEVELOPMENT

The four main challenges of auger development were:

    the piling rig should remain largely unaltered;

    the standard injection system for conventional flight augers need to be modified to inject smaller volumes at
    higher pressures;

    the auger system needs extensive modification to inject smaller volumes and needs to be adapted to ensure that
    the in situ mixing process produces homogeneously treated soil without significant displacement of the soil to
    the surface; and

    the auger system and supporting ancillary equipment have to be "purpose designed" to accommodate the injected
    cementitious material containing the modified clays.

Several prototypes of the auger system were developed and tested for mixing efficiency prior to the field trial of the
stabilization/solidification system.

6. FIELD TRIAL

5.1 The Site and the Program

The technology was tested at field scale on a contaminated U.K. Ministry of Defense site located in West Drayton,
Middlesex. A variety of chemical plants has existed on the site since the turn of the century, including a chemical
waste "quarantine store," chemistry laboratory, flammables  store, battery bank, engine testing areas, underground
storage tanks and nucleonic laboratory. Likely contaminants include, flammable materials, solvents, concentrated
sulfuric acid, oils, petrol, and radioactive species. Following investigation, the principal contaminants of concern were
judged to be PAHs (7-12 mg/kg) and, in particular, benzo(a)pyrene.

Prior to field operations, a laboratory program was carried out to determine the type of intercalated clay to be used,
the cement/fly ash/clay ratio for the grout, and the optimum soil/grout ratio.

The grout was injected using a prototype auger fitted to a conventional piling rig that produces 900-mm diameter
columns either individually or overlapping. For this  demonstration, columns with 50% overlap  were chosen and
several different optimized  mixes were evaluated in adjacent columns.

Cored material aged 50 days (taken at 44 days), 70 days (taken at 57 days) and 1 year (presumed to have been taken
at 57  days) from each column were tested in the laboratory  against the treatment criteria outlined below.

6.2 Site Characterization

Previous chemical analysis records provided some data on the concentration  of a number of heavy metals and
referred to general organic  contamination with a specific reference to acid tars  in the  vicinity of the trial pit.

The ground consisted of 1.7 m of fill material (sand and sandy-clay containing fragments of bricks, timber, metal,
concrete, and glass), overlying 0.3 m gravelly-sand underlain by sandy gravel to the base of the trial excavation. Soil
samples were collected from depths of 0.5, 1.5 and 2.3 m from the trial pit. The analytical results are summarized
in Table 1.


                                                    98

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

              Table 1: Concentrations of PAHs and heavy metals in soil samples from field site
Depth of soil sample
pH
Cadmium
Copper
Chromium
Lead
Nickel
Zinc
Total of 6 PAHs (a)
m

mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
0.5
7.5
3.4
460
ND®
1,560
84
564
7.4
1.5
7.1
4.4
310
45
2,300
37
600
10.2
2.3
7.3
4.2
635
62
3,330
53
785
12.0
Accuracy

+/- 0.2
+/-40
+1-2
+/- 160
+/-4
+/-40
+/- 0.02
Note: (a) fluoranthene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene,
benzo(g,h,i)perylene, and indeno(l,2,3-cd)pyrene
(b) not determined
The  results  indicated there were potentially  17 PAHs present in the soil.  Those investigated in detail were
fluoranthene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, benzo(g,h,i)perylene and indeno( 1,2,3-
cd)pyrene, which are specified by the World Health Organization (WHO) and the U.S. Environmental Protection
Agency (USEPA) as proven or suspect human carcinogens. The total concentration of the six WHO PAHs ranged
from 7.4-12 mg/kg of dry soil with the most abundant PAH being fluoranthene with a concentration of up to 4.3
mg/kg.

6.3 Performance Criteria

The  effectiveness of the various treatments in the laboratory and field was assessed against:

  •  unconfined compressive strength (UCS) by an ASTM method (greater than 350 kPa after 28 days curing);

  •  teachability, based on the USEPA Toxicity Characteristic Leaching Procedure (TCLP), which uses an aggressive
    acid leaching medium, with respect to six WHO PAHs (total PAHs less than 10 ug/L and benzo(a)pyrene less
    than 0.5 ug/L) and chromium, copper, lead, and zinc with target levels of 50 times U.K. drinking water limits;
    and

  •  a final leachate pH in the TCLP test (between 8-10 following 28 days curing) to ensure low metal solubility (this
    was later amended to a pH of 7-11).

Durability (freeze/thaw and wet/dry)  and permeability tests (less than lxlO"9m/s) were used as secondary evaluation
criteria. ASTM test procedures were used for the initial freeze/thaw tests on laboratory prepared materials, but the
test method was subsequently modified as it was judged to harsh in relation to typical U.K. weather conditions.

6.4 Treatment Optimization in the Laboratory

6.4.1 Preparation of modified clays

Seven  modified clays were prepared from a conventional Ca2+ bentonite clay (premixed with 5% w/w sodium
carbonate):

  •  A13+pillared;

  •  Fe3+ exchanged and A13+ pillared;

  •  Fe3+ exchanged, A13+ pillared and then treated with benzyl quaternary ammonium  salt (QAS);
                                                   99

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

  •  Fe3+ exchanged, A13+ pillared and then treated with 2HT (aliphatic) QAS;

  •  A13+ pillared and Fe3+ exchanged (spiked);

  •  A13+ pillared and Fe3+ exchanged (spiked) and then treated with benzyl QAS; and

  •  A13+ pillared and Fe3+ exchanged (spiked) and then treated with 2HT QAS.

X-ray diffraction analysis of the clays showed that all had been effectively pillared/intercalated. The modified clay
was produced as a suspension water (0.5 g clay/10 g of liquid).

6.4.2 Tests on synthetic soils

The treatment/optimization trials were conducted in two stages: first on synthetic soils intended to model site soils
and then on actual site soils. The model system comprised a mixture of sand and clay with pyrene added at a
concentration of 20 mg/kg, this being the average PAH content of the site soils. The pyrene was added to the pre-
mixed clay and sand as a powder because of its low water solubility. The non-carcinogen pyrene, which was present
on the site, was judged a satisfactory substitute for the  six WHO PAHs.

The following treatment materials were used:

  •  ordinary Portland cement (OPC);

  •  power station fly ash (known in the U.K. as pulverized fuel ash);

  •  modified clays; and

  •  quick lime (as an alternative to cement for raising the pH).

The treatment materials were combined in a grout with the  clay suspension as the source of water. The grout mixes
were added to the prepared soil and thoroughly  mixed to obtain a pumpable consistency (by inspection).

Initial development work focused on OPC/fly ash mixes. Parameters that affect the properties of the soil-grout mix,
such as  cement/fly ash ratio, soil/grout ratio, water/cementitious materials (OPC+fly ash) ratio and the amount of
modified clay added were  considered.  Uncontaminated soil was used  as a control  to determine if contaminants
affected performance/properties.

6.4.3 Tests on site  soils

Following the tests  on the model soils,  a limited number of soil-grout mixes were prepared using site soils. These
latter tests indicated that the real soil/grout mixes produced treated soils with a UCS of about 50% of the model soils.
Physical examination showed a similar degree of homogeneous mixing for all three different soil-grout ratio mixes
tested. Consequently, a decision was made to proceed with further tests using a soil/grout ratio of 5:1, thereby using
the minimum amount of grout to achieve the overall objective with regard to UCS.

Two options were considered for raising the pH: (1) increasing the amount of OPC in the grout; and (2) adding quick
lime to the grout. Following initial trial and error experiments six mixes were selected for detailed evaluation. The
addition of quick lime retarded strength development and significantly reduced strength at 28 days. Examination of
the pH data led to the target range be amended from 8-10 to 7-11. Leachate results for pyrene were  all satisfactory.

On the basis of these results, four mixes were selected  for further testing with a view to use in the field trial.
                                                   100

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

6.5 On-site Trial

6.5.1 Installation

An area of 6 m2 to a depth of 2.5 m was treated in the trial. As indicated above, four mixes were selected following
extensive laboratory development work, and two further mixes were added to extend the range of material combina-
tions (see  Table 2).  One was replicated.

The clay formulation used was A13+ pillared and Fe3+  exchanged (spiked) and then treated with benzyl QAS. The
modified clay was produced as a suspension at Birmingham University and transported to site  in steel drums. The
clay concentration was 20 g/L (2.5 times more dilute than used in laboratory treatment trials).

A grid treatment plan was prepared as shown in Figure 3. The approach adopted with the six soil-grout mixes was
to designate specific columns for each mix, with a minimal overlap of nominally 50% to determine the rate of spoil
generation in the worst-case scenario. The site trial took place over two days. The solid line columns in Figure 3
were installed on day one, and the dotted line overlapping columns were  installed on day two.
Figure 3: Plan of the soil-grout overlapping columns showing the mix used in each column

The columns were formed by advancing the auger into the soil to the required depth of the column, mixing the soil
in place during the auger descent and then injecting the grout slurry into the soil with simultaneous mixing on the
auger withdrawal. The overall volume increase was measured at the end of the second day to be about 1 m3, which
is an increase of about 6.5%.

6.5.2 Sample collection

The properties of the stabilization/solidification soil-grout material were investigated for single as well as overlapping
column areas to ensure that homogeneous mixing had taken place and that no weak areas or areas of preferential
flow had been created during column formation.

Cores were collected at 44 and 57 days (logistical problems prevented sampling at 28 days as originally planned).
UCS, pH, teachability, durability and permeability testing were conducted on the cored material.  The total curing
times, including storage at Birmingham University before testing, were 50 and 70 days, respectively.
                                                   101

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
                            Table 2: Composition of Mixes Used in Field Trial
Mix
Fly ash kg
Cement kg
Lime kg
Bentonite kg
Water
clean L
clay L
suspens L
ion
Total
Soil:grout (solids+water) w:w
Grout (solids only): soil %
Modified clay kg
Total solids kg
Modified clay: solids g/kg
A
80
30
-
-

26
20
46


5
14
0.4
660
0.61
B
80
30
1
-

26
20
46


5
14
0.4
661
0.60
C
80
25
4
-

26
20
46


5
13.5
0.4
654
0.61
D
80
25
4
-

30
16
46


5
20
0.32
654
0.49
E
80
20
-
-

23
17
40


5
14
0.34
600
0.57
F
-
80
-
8

110
30
140


5
8
0.6
528
1.14
Mix
Soil in column kg
Dry grout in column kg
Fly ash in grout column kg
Cement in grout column kg
Lime in grout column kg
Bentonite in grout column kg
Water in grout column
clean water L
clay suspension L
Total L
modified clay /column kg
modified clay /kg soil g
A
1,200
240
125
46
-
-

40
30
70
0.6
0.5
B
1,200
240
120
46
1.5
-

40
30
70
0.6
0.5
C
1,200
240
125
40
6
-

40
30
70
0.6
0.5
D
1,200
340
175
55
10
-

70
30
100
0.6
0.5
E
1,200
240
140
35
-
-

40
30
70
0.6
0.5
F
1,200
240
-
85
-
85

120
30
150
0.6
0.5
D*
1,200
340
175
55
10
-

40
30
70
0.6
0.5
Notes: 1. The density of the soil taken to be 1,700 kg/m3.
2. The column is 0.6 m in diameter and 2.5 m high.
3. The concentration of the modified clay suspension is 1 kg clay/50 liters.
4. Replication of mix D is denoted as D*'
6.5.3 Results

The results of testing are summarized in Table 3. All UCS values exceeded the target value of 350 kPa and were
greater than the equivalent (28-day) laboratory results. This was particularly true for the mixes containing quick lime.

The leaching results were all satisfactory. All mixes satisfied the pH range of 7-11 and tended to be at the alkaline
end of the range. The total concentrations of the six WHO PAHs in the leachates were all below 2 ug/L and often
below the 0.02 ug/L detection limit.  The individual leachate values for benzo(a)pyrene were also within the target
                                                   102

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
value of less than 0.5 ug/L. All results for chromium, copper, lead, and zinc were satisfactory. Subsequently, more
detailed analyses also showed satisfactory performance with respect to a wider range of metals.

The samples survived a modified freeze-thaw test—the ASTM method being considered too aggressive following
the laboratory trials for U.K. conditions. The wet-dry test results were also satisfactory.

Only  one sample satisfied the permeability criterion of less than 
-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
Samples from all mixes showed an increase in UCS cured for 12 months, compared to those cured for 70 days.
Permeability values decreased as expected due to continuing cement hydration processes. Wet-dry durability testing
gave similar results after 12 months, and freeze-thaw testing was largely inconclusive (all samples failed at very low
temperatures as previously experienced). The results for UCS and permeability are given in Table 4. In general, the
results indicated that physical properties improve with aging.
                              Table 4: Results for 2 months and 14 months(a)
Mix
A
B
C
D
E
F
D*
Compressive strength
(kPa)
2 months
990
1,332
1,231
1,274
1,480
1,335
1,365
14 months
2,325
2,812
2,180
2,570
1,712
ND®
1,875
Permeability
(xlO'9 m/s)
2 months
2.64
0.69
2.56
2.21
1.99
0.64
0.70
14 months
0.39
0.18
0.10
0.52
1.30
0.19
0.12
Notes: (a) cores taken at 57 days and stored in the laboratory under humid conditions for 12 months
(b) not determined
7. REFERENCE

N. A. McLeod. Chemical fixation of soils contaminated with organic chemicals (final report for NATO/CCMS Pilot
Study), Envirotreat Limited, Kingswinford, U.K., 1997.
                                                  104

-------
NATO/CCMS Pilot Study, Phase II
                       Final Report (Appendix IV)
   Project No. 35
                          In Situ Soil Vapor Extraction within Containment Cells
                   Combined with Ex Situ Bioremediation and  Groundwater Treatment
   Technical Contacts:
   Paul Theile
   MEL Limited
   20 Bond Street
   Wakefield
   West Yorkshire WF1 2PQ

   tel: +44/1924-361816
   fax: +44/1924-363144

   Harry Shipley
   Department of Planning
   Gateshead Metropolitan Borough Council
   Civic Centre
   Gateshead NE8 1HH

   tel: +44/191-477-1011
   fax: +44/191-478-3491
Country:
United Kingdom
Project Status:
Accepted by study     1994
Final report           1996
1. INTRODUCTION

The former Derwenthaugh Cokeworks site is located adjacent to the River Demerit, near Gateshead, Tyne, and Wear.
Investigation of the site identified a 7.9-ha area that was significantly contaminated with coal carbonization wastes
from the original plant including BTEX (benzene, toluene, ethylbenzene, and xylenes), polycyclic aromatic hycrocar-
bons (PAHs), phenols, heavy metals, and cyanides. The site was underlain by fill, sand, and gravel to a depth of 2-4
m. Groundwater contamination was severe and included the presence of free floating product shown to be seeping
into the adjacent river. This Pilot Study project reported on the remediation scheme adopted at the site; the project
was funded by a Land Reclamation Grant from the U.K. government.

In summary, the remedial  scheme involved four stages of work: (1) installation of a cut-off wall to protect the
adjacent river from further pollution and to allow safe excavation (to a depth of 5 m) of the contaminated ground;
(2) installation of wells for dual-phase vapor extraction (DVE) to remove volatile organic compounds (VOCs) and
free-phase product; (3) use of wells to extract contaminated groundwater for surface treatment; and  (4) excavation
and treatment of contaminated soil.

2. BACKGROUND

Built in 1928-29, the  former Derwenthaugh Cokeworks site is situated about 2 km southwest of the confluence of
the Rivers Tyne and Derwent in Northeast England, and lies between the River Derwent and the A694 Rowlands
Gill to Blaydon Road. The site lies within the valley of the River Derwent and follows the course of the river for
approximately 2.5 km, covering a total area of about 55-hectares. The cokeworks carbonized coal with by-products
including tar, ammonium sulfate,  benzole, and town gas. Prior to the closure of the plant, 475,000  tonnes of coal
per year were being used to produce foundry coke.

Site investigation works indicated the area to be underlain by variable fill material overlying fine-to-medium sands
and a  persistent gravel layer, usually  at about 2-4 km below  ground level.  Groundwater was encountered in
monitoring wells installed in the gravel during the initial investigation.  The contours suggested a general hydraulic
gradient running from south to north with a fall in level of about 3 m across the site. The original site investigation
indicated that a high level of contamination was present in both soil and groundwater. In particular, the groundwater
was found to be contaminated with phenol, oil, and PAHs  across the entire site. A floating layer of free oil was
                                                  105

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

evident at certain locations, and oil was observed to be seeping into the River Derwent at several places along the
river bank. Benzene and other VOCs were also detected in significant concentrations.

3. TECHNICAL CONCEPT

3.1 Overall Scheme

A deep cut-off slurry wall was installed, enclosing the area to be treated in order to allow safe excavation to up to
5 m depth (with minimal dewatering) and to prevent outflows of contaminants directly into the river. The cut-off
wall was taken down to a sufficient depth to form an aquiclude to minimize: (1) river water entering the site; and
(2) the flow of pollutants from the site to the river.

A well point system was installed. Dual vacuum extraction was conducted, to dewater the  site and remove vapor
contamination (including benzene and VOCs) and free-phase product. Extracted groundwater was treated prior to
discharge. The treated liquid had to conform to strict default criteria (devised by the National Rivers Authority (now
the Environment Agency) before discharge into a soakaway adjacent to the river was permitted.

Following dewatering, the solid material was excavated. Solids unsuitable for treatment were placed directly into an
on-site encapsulation facility. Other material was screened and either air  dried and reused or treated by ex situ
bioremediation to remove oil and PAH contamination. Suitably treated materials were used as backfill to reinstate
the excavated contaminated area.

3.2 Slurry Wall

The first stage of treatment was to install a cutoff wall using a vibratory emplacement technique along the boundary
of the remedial works.  Laboratory investigations optimized the slurry  used  (about 34% solids), which consisted of
bentonite, ordinary Portland cement, ground granulated blast furnace slag (a cementitious material), and water.  The
average thickness of the emplaced wall was 0.132 m, and it had a permeability of at least IxlO"8 m/s and strength
of 200 kN/m2.

3.3 Dual Phase Vapor Extraction  (DVE)

Following testing, DVE was implemented by dividing the area into a series of treatment cells. Initially 30 0.1-m inner
diameter wells were installed to a depth of 4.5-5 m below ground level, but this was increased to 37 m following
discovery of complex strata differences between the whole site and the original field test area. DVE uses an applied
subsurface vacuum to draw contaminant vapors and free product to the surface for separation and treatment.  At the
surface, free product was separated from the groundwater in a separator  tank and volatile organic  vapors were
adsorbed onto activated carbon filters. In order to facilitate free product removal, dewatering trenches were  dug in
some  sectors of the site (see below).

3.4 Groundwater Treatment System

Extracted groundwater  was generated at a rate of 75 m3 per day. Groundwater treatment consisted of several phases
designed to sequentially  treat contaminant  components using a succession of holding lagoons and  reactors: (1)
reception/buffer  lagoons  (floating  product  removed by skimming);  (2)  heavy metal removal by hydroxide
precipitation; (3) cyanide oxidation;  (4) sulfide oxidation; (5)  pH correction; (6) sand filtration; (7) activated carbon
filtration; and (8) biological oxidation of ammoniacal nitrogen. Treated groundwater was discharged to a soakaway
along the Derwent.

3.5 Treatment of Excavated Materials

The excavated solid material varied in composition. Excavated materials were classified on the basis of past work
on the site based on physical appearance, location, and depth of the excavation, and previous analytical data for that
site sector. Material judged unsalvageable was sent to an on-site encapsulation area. The remainder of the material
was screened at 100 mm and left in piles to air dry. Oversized material was crushed and reused on site, if analysis
indicated this would be  acceptable. Undersized material was sent for biological treatment using "biopiles." The  design


                                                   106

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

of the biopiles was based on preceding laboratory and field scale tests (see below). The biopiles were constructed
over a high density polyethylene (HDPE) membrane and clay liner. Runoff was collected via an HDPE-lined ditch
and was reapplied to the biopiles. Aeration was provided via a network of air injection pipes. The biopiles were
clearly identified and delineated to allow determination of process effectiveness by material batch. Forced aeration
was provided on a continuous basis. Effectively treated material was reused onsite. A total of 28,000 m3 of material
was treated in four batches of 7,000 m3. Treatment duration was 2-3 months, depending on the time of year.

4. FINDINGS

4.1 DVE Process

The DVE process was operated for 46 weeks, during which time the total mass of VOCs recovered was in excess
of 150,000 kg.  Severe weather conditions often caused problems during DVE treatment because of freezing of
supporting  pipework and  equipment.  The  objective of the treatment was to remove  free  product  and reduce
concentrations of VOCs—particularly  benzene—to safe levels for excavation rather than as a complete remedial
treatment in itself. This objective was achieved; however, the contractor regarded the following points as noteworthy:

  •  Pipework requires protection from severe weather conditions.

  •  The variable ground conditions meant that assumptions made based on a field trial in one sector were not valid
    for the whole site, and further wells had to be retro-fitted.

  •  The volume of liquid requiring extraction exceeded the capacity of the DVE  system, and slit trenches with
    surface pumps were required to remove groundwater and free product. The pumps were more effective for this,
    and the vacuum effort was then focused on the removal of VOCs.

  •  Over the project, there emerged greater-than-anticipated requirements for process monitoring,  in part because
    of the complexity of the site conditions. Much of this was carried out by hand. It would  have been far more
    efficient to have installed capability for centralized or remote process monitoring when works were begun.

4.2 Groundwater Treatment

Consent limits for water treatment discharge was one of the major regulatory controls on this project. A particular
difficulty was coping with seasonal variation in water volumes, ranging from conditions with low treatment volumes
containing high concentrations of pollutants, to high volumes with low concentrations. The variability of groundwater
quality required a higher degree of monitoring than originally was envisioned. Furthermore, at times the volume of
groundwater requiring treatment exceeded the design capacity of the system, necessitating its storage in "buffer"
lagoons.

The treatment process generated a number of waste streams, some of which were disposed of offsite,  others of which
were encapsulated onsite.

The biological treatment plant, which  consisted of a nitrification lagoon, was a retro-fit forced by changes in the
composition of the ammoniacal nitrogen in the groundwater as work on site progressed. Initially,  this nitrogen was
found as NH3. However, as time  went on, free ammonia levels fell and NH4+ concentrations increased. Inoculum
from a local sewage works was found to perform far better than commercially available  inocula.

A reed bed is planned to treat residual contamination following site reinstatement. It is envisioned that the reed bed
will need to accommodate a groundwater flow of 30-50 m3 per day.

4.3 Excavated Materials

Over 130,000 m3 of solid materials were excavated during the remediation. Of this total,  51,000 m3  were able to be
reused as fill following screening and drying. 55,000 m3 were targeted for bioremediation, of which 28,000 m3 were
reportedly treated. Of this 28,000 m3,  4,000 m3 failed to reach the required standard and were encapsulated (the
required treatment targets were not reported). The balance of the  materials (20,000 m3) was also encapsulated. A


                                                  107

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

significant proportion of the material sent to biological treatment was demolition rubble. 30% of the treatment volume
was removed by screening at 100 mm prior to placement in the biopiles.

4.4 The Biopile Trials

Preliminary investigations based on plate counts implied that an adequate range of PAH-degrading bacteria was
already present  in  the site  soil.  Aeration and addition of nitrogen and phosphorous were  found to enhance
biodegradation rates in solid phase laboratory-scale tests. Throughout the tests, levels of naphthalene did not greatly
decline, whereas levels of higher molecular weight PAHs  (e.g., acenaphthene and fluorene)  did decline. It was
suggested that naphthalene was being produced as an intermediate breakdown product from the degradation of higher
PAHs.

Field-scale tests  were carried out using small-scale biopiles with forced aeration to test the effect of covering them
with insulating materials  and the effect of water and supplement addition. The results of these trials indicated no
significant benefit from either nutrient addition or insulation. Degradation rate  did appear  to decline with lower
ambient temperature.

Following the main works, the following points were noted:

  •  the main factors affecting the degradation process were aeration and ambient temperature. Attempts to replicate
    laboratory conditions that were indicated as beneficial showed no cost effective benefits at full scale;

  •  significant reductions in the contaminant levels were achieved through redistribution and volatilization;

  •  often these processes alone were enough to achieve the specified criteria, with no further action required;  and

  •  a key contribution to success is careful management of material batches  and screening.

5. MATERIALS TREATED

The materials treated have been described above and include:

  •  rubble that was separated by screening and dried;

  •  undersized material,  which was treated biologically  or sent  for on-site encapsulation;

  •  extracted groundwater, treated by a series of physical, chemical,  and biological processes; and

  •  extracted soil air treated by activated carbon filtration.

6. PROCESS RESIDUALS

6.1 Range of Residuals

Residuals included the materials encapsulated onsite and material deemed to  have failed to meet treatment targets
following its biopile treatment, which was also encapsulated on site.  Treatment process residuals from the air and
water treatments were:

  •  oil from the reception lagoon,  which was disposed offsite;

  •  sludge from the settlement lagoon  (following hydroxide dosing), which was encapsulated onsite;

  •  oil from the DVE separator, which was disposed offsite;

  •  sludge from the DVE separator, which was encapsulated onsite;
                                                   108

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

  •  sludge following pH correction, which was encapsulated onsite; and

  •  spent carbon, which was encapsulated onsite.

The sludges  typically contained 3% solids.

6.2 Water

The  final report presented  a typical  groundwater composition  prior to treatment,  discharge consents,  and
concentrations following treatment as documented number of groundwater monitoring reports (Table 1).
                   Table 1: Composition of Groundwater Prior to and Following Treatment

Analyte

Cyanide
Chloride
Sulfates
Phenol
Sulfide
PAH
Lead
Cadmium
Zinc
Arsenic
Iron
NH4
Suspended solids
COD
pH
total hydrocarbons
Concentration (ppm)
Typical Pretreatment
Values
3.8
373
1,520
21.8
71.7
-
0.02
0.01
0.9
0.01
-
260
-
-
6.1-9.3
6,220
NRA Effluent
Discharge Criteria
0.1
-
-
1
5
1
0.02
0.002
0.09
0.01
5
5
60
500
5-9
~
Range in Reports

O.I (free)
-
-
0.05
O.02 (free)

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 36
                       Enhancement Techniques for Ex Situ Separation Processes
                                 Particularly with Regard to Fine Particles
   Technical Contact:
   Mike Pearl
   Planning Directorate
   UKAEA
   Harwell
   Did cot
   Oxfordshire OX11  ORA

   tel: +44-(0)1235-435377
   fax: +44-(0) 1235-436930
   E-mail: michael.pearl@ukaea.org.uk
Country:
United Kingdom
Project Status:
Accepted by study  1994
Final report         1996
1. INTRODUCTION

Ex situ soil separation processes (often referred to as soil washing) are widely used in Northern Europe and North
America for the treatment of contaminated soil. They mostly use separation processes from the mineral processing
industry to remove contaminated concentrates from soils and produce relatively uncontaminated soil fractions that
may be suitable for reuse as relatively "inert" fill. At present, commercial soil separation is mostly restricted to the
removal of  fine  fractions  (for example,  <0.063 mm), which normally contain the highest concentration of
contaminants, leaving the coarse fractions (>0.063 mm) relatively uncontaminated. In this approach, the separation
of the contaminated fines often results in a net cost saving to the overall treatment process since, for example, it can
significantly reduce the volume  of contaminated material requiring landfill disposal.  However, in cases where
contaminated soils have a high silt and clay content (>30-40%), the financial benefits of volume reduction are
diminished to the point where conventional soil washing is considered uneconomical. For such soils, enhancement
of soil separation processes and the development of downstream treatment of fine fractions  are felt necessary to
increase the range of soil types treatable by soil washing and to reduce the volume of secondary wastes that require
disposal.

The study was carried out under a contract  to the  U.K. Department of the Environment by AEA Technology's
National Environment Technology  Centre.

2. THE STUDY

This Pilot Study project investigated the feasibility of extending ex situ soil washing processes to soils that contain
a high proportion of fine particles through  the use of physical separation techniques and equipment normally used
for the treatment of fine mineral ores and industrial minerals.

A laboratory study and pilot-scale study were conducted on the treatment of two fines-rich soils  containing organic
contaminants. One soil from a former industrial site  contained 62% soil particles less than 0.063 mm and was
contaminated with diesel fuel. The other soil, from a former gasworks facility, had up to 43% soil particles less than
0.063  mm and was contaminated with polycyclic aromatic hydrocarbons (PAHs), total petroleum hydrocarbons
(TPH), and complexed cyanides. The project examined key processes such as disaggregation of agglomerated fines,
particle sizing and classification, attrition scrubbing to remove fine particles attached to surfaces of coarser particles;
density concentration, dewatering,  and  integration of the aforementioned processes  with slurry-phase biological
treatment.

Results of the investigation showed that both soils could be disaggregated using a tumbling mill partially filled with
steel balls  or pebbles. Subsequent attrition scrubbing of silt- and sand-sized particles from both soils had little effect
on the contaminant content of these particles, even after extended periods of time and after the addition of various
dispersant reagents (to prevent fine  particles re-coating larger particles). The surfaces of these particles in both soils
were therefore thought to be relatively free of persistent fine coating contaminants.
                                                   110

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

During screening and classification, the contaminants in the diesel-contaminated soil were markedly concentrated
in the fraction less than 0.002 mm. Removal of entrained or misplaced ultrafines (<0.002 mm) from the coarse soil
fractions was found to be essential to achieve low levels of contamination in the gravel, sand, and silt-sized fractions.
Froth flotation of the fraction less than 0.01 mm did not significantly reduce contamination in the non-floating
product.

The contaminants in the gasworks soil were more evenly distributed in the various size  fractions, and  removal of
entrained soil particles made little difference in the contaminant content of the fractions. With this soil, other physical
separation processes could substantially reduce PAH and TPH contamination, but not complex cyanide contamination.
For the fine particles, a combination of classification at 0.01 mm with froth flotation or specific gravity concentration
was reported to substantially reduce the contaminant content of the treated product. Froth flotation proved best on
the 0.01-0.63 mm fraction, producing a non-floating fraction weighing 69% of the untreated material. Three stages
of froth flotation reduced  PAH contamination by 61%, TPH by 54%, and cyanide by 39%. By comparison,  specific
gravity separation on the 0.01-0.63 mm fraction reduced PAHs, TPH and cyanide contamination by 68%, 76%, and
14%,  respectively.  Despite  substantial  reductions, however,  the  products  still contained  significant  residual
contamination, which made them unlikely to be suitable for re-use.

In-slurry biological treatment of the contaminated fine fractions from the diesel-contaminated soil removed over 80%
of the TPH contamination present. Despite this reduction during the 28-day treatment, treated fines retained relatively
high levels (2,300 mg/kg) of TPH. It was suggested that further work  was  required to reduce the time period for
treatment and the level of residual contamination in the treated product so that the process could be more effectively
integrated with physical pretreatment. With the fine fraction from the gasworks soil, in-slurry biological treatment
for 28 days achieved only  a 40-50% reduction in PAHs, 25% reduction of TPH, and virtually no cyanide degradation.
The operation of the bioreactor, especially system pH, was found to be difficult to control and further work was
suggested to improve treatment performance.

3. RESIDUALS AND  EMISSIONS

Residuals from the  combined process include treated soil materials and  others that  did not achieve remediation
guidelines. Clean material could be reused onsite,  while partially clean materials would have to be dewatered and
disposed. Water  from soil washing and bioslurry treatment would also  have to be treated before disposal.

4. COSTS

The report detailed fixed costs as well as costs for maintenance, administration, etc., but provided an operational cost
of £UK 29-43/tonne for the soil separation process. In practice, the overall operational costs would also  include the
cost of disposal or the further treatment of the contaminated concentration. Thus, if disposal of the concentrate costs
£100/tonne and the concentrate amounts to 35% of the original mass, the overall treatment costs would be  £74-887
tonne. In a U.K.  context, considering typical disposal cost of 50/tonne, the overall treatment cost would be  £47-61.
Thus, the process would be only marginally cost effective, unless a smaller amount of contaminated concentrate were
produced, or a cheaper means of treating the concentrate were technically feasible.

5. CONCLUSIONS

The main conclusions of the study were that physical separation processes may be appropriate for soils containing
a high fines content provided that a substantial fraction of the soil of grain size greater than 0.002 mm is contaminant
free, and that material less than 0.002 mm comprises less than 30-35% by weight of the  soil. It was recommended
that the separated fraction less than 0.002 mm should also be dewatered to minimize disposal costs.
                                                   Ill

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 37
               Bioventing of Hydrocarbon-Contaminated Soil in the Subarctic Environment
   Technical Contact:
   Alison Thomas
   U.S. Air Force AL/EQW
   139 Barnes Drive, Suite 2
   Tyndall AFB,  FL  32403-5323

   tel: +1/904-283-6303
   fax: +1/904-283-6064
   E-mail:alison_thomas@
       ccmail.aleq.tyndall.af.mil
Country:
United States of
America
Project Status:
Accepted by study  1992
Interim report       1993
Final report         1996
1. INTRODUCTION

Petroleum hydrocarbons such as JP-4 are generally considered biodegradable, if indigenous microorganisms receive
an optimal supply of oxygen and nutrients. Bioventing is a technique that aerates subsurface soils using a series of
injection and extraction wells in order to stimulate in situ biodegradation on sites where oxygen is limiting the
degradation rate. Bioventing in this instance is outwardly similar to soil vapor extraction (SVE) except that the air
flow is controlled to reduce the physical vaporization of contaminants. Its objective is to avoid the costs of off-gas
treatment, which may occur with SVE (estimated as U.S.$33-$39/m3 of contaminated soil). So far as stimulating in
situ biological activity is concerned, bioventing has two advantages over aqueous oxygen delivery systems (/'. e., pump
and treat). First, on a mass basis, less air is required than water to deliver an adequate amount of oxygen. Second,
the permeability of soil to air is much larger than the permeability of soil to water. This  Pilot Study project reports
on the application of bioventing in cold climates where the basic approach is combined with soil warming.

2. EXPERIMENTAL DESIGN

Eilson Air Force Base (AFB) is an operational base located in the  Alaskan interior. The climate is subarctic, with
an average annual temperature near 0°C (ambient temperatures range from -30°C to +30°C). The technology demons-
tration was conducted on a 0.4-ha area, designated  Site 20, which had been contaminated to a depth of nearly 3 m
with JP-4 from suspected leaking pressurized fuel lines. The site soil consisted of a mixture of sand and gravel, with
increasing silt content to 3 m, which was the  depth of the  water table. TPH levels ranged from  100-3,000 mg/kg.
A field evaluation of bioventing was undertaken at Site 20 to determine whether and to what degree soil warming
can enhance the effectiveness of bioventing of JP-4 and to  determine whether soil warming promoted a higher rate
of biodegradation all year round.

In the summer of 1991, a bioventing system was installed at Site 20 consisting of an air blower plumbed into air
injection/extraction wells spaced at 10-m intervals. Air movement in bioventing can be achieved by blowing air into
the ground (injection bioventing) or by creating a vacuum to draw air out of the soil (extraction bioventing) with the
latter approach more commonly adopted. The majority of this study used injection bioventing with a blower forcing
air into the ground at a  rate of 0.015 m3 per minute. The demonstration area was divided into four test plots where
the bioventing system was combined with different soil warming techniques:

    (1)  Warm Water System  test plot:  Groundwater collected by an extraction well was pumped through an
        electrical heater and warmed to around 35°C before reinfiltration. Insulation was placed over the ground
        surface to retain heat;

    (2)  Heat Tape System test plot: Strips of tape heater were buried to a depth of 1 m in the test plot and warmed
        at a rate of 16  W/m2. Insulation was  placed over the plot to retain heat;

    (3)  Solar test plot:  Insulation placed over the ground during the winter months was replaced with plastic mulch
        sheeting during the spring and summer to  capture solar heat and passively warm the soil; and
                                                  112

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

    (4)  Control test plot: No soil warming. Both the warm water and heat tape systems were operated for two years
        from summer to  summer over the demonstration period (1991-94). The solar and control plots were
        monitored for 3 years.

The monitoring program consisted of three major elements: (1) an evaluation of system performance, which included
measurement of soil  gas  composition,  soil temperature, microbial  respiration and  contaminant levels;  (2)  an
evaluation of environmental impact of measuring of surface air emissions; and (3) a cost evaluation.

3. RESULTS

Bioventing significantly increased oxygen levels in the soil gas with a corresponding decrease in carbon dioxide and
hydrocarbon concentrations. Soil temperature in the warm water and heat tape test plots were consistently 10°C
higher than in the solar heated and control plots. The average biodegradation rates in the warm water and heat tape
treated plots were between 3 and 4 times higher than those in the solar and control test plots. It was interesting to
note that after switching off the heating systems, biodegradation rates in these plots decreased below that of the
control suggesting that the microorganisms may have adapted to the higher temperatures. TPH removal in the
warmed plots was an order of magnitude higher than in the solar-heated and control plots.

4. EMISSIONS

In general, air emissions of benzene in the control plot were higher when the bioventing system was on than when
it was off. Emission rates were 0.00038 and 0.0001 kg/day with and without air injection. Although bioventing
increased emissions, they were well below regulatory limits. Using the data provided by the demonstration, costs
were evaluated for remediation of a 5,000 m3 site with an average TPH level of 4,000 mg/kg using bioventing with
or without a soil heating system.

5. COSTS

The evaluation showed a trade-off between cost and treatment time with a basic bioventing system costing U.S.$25.57
m3 over 9.4 years operation, compared with a cost of U.S.$24-26/m3  for a bioventing system using either a warm
water or heat tape system operating for up to 3.4 years.
                                                  113

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 38
                              Demonstration of Peroxidation Systems, Inc.
                             Perox-Pure™ Advanced Oxidation Technology
   Technical Contact:
   Norma Lewis
   U.S. Environmental Protection Agency
   NERL-MCEARD
   26 M.L. King Dr.
   Cincinnati, OH 45268

   tel:  +1/513-569-7665
   fax: +1/513-569-7620
   E-mail:  lewis.norma@epa.gov
Country:
United States of
America
Project Status:
Accepted by study  1992
Final report         1992
1. INTRODUCTION

The Peroxidation Systems, Inc., perox-pure™ is a chemical oxidation technology that destroys dissolved organic
contaminants in water using an oxidation process based on ultraviolet (UV) radiation and hydrogen peroxide. The
basic concept of the treatment system is that a variety of organic chemicals can be  effectively oxidized by the
combined use of two oxidants such as UV radiation and hydrogen peroxide, UV radiation and ozone, or ozone and
hydrogen peroxide. The most  direct method for the generation of hydroxyl radicals is through the cleavage of
hydrogen peroxide. Photolysis of hydrogen peroxide produces two  hydroxyl radicals, which react rapidly after
formation with organic contaminants present.

This summary highlights  the technology demonstration at the Lawrence Livermore National Laboratory site in
California,  which was  completed under the  Superfund  Innovative Technology  Evaluation  (SITE)  program.
Participants in this SITE demonstration included the U.S. Environmental Protection Agency (USEPA) Risk Reduction
Engineering Laboratory, USEPA Region 9, Lawrence Livermore National Laboratory, Peroxidation Systems, Inc.,
the German Federal Ministry of research and technology, and the PRC Environmental Management, Inc. SITE team.
Peroxidation Systems Inc. was responsible for delivering, setting up, operating, and demobilizing the  treatment
system. The PRC SITE team and others collected and analyzed samples. Under a U.S.-Germany bilateral program,
the German Federal Ministry of Research and Technology reviewed the demonstration plan and asked for additional
analyzes.

2. THE PROCESS SYSTEM

The technology consists of four modular units that are skid-mounted and portable: the UV/oxidation unit, hydrogen
peroxide feed module, acid feed tank, and the control panel unit.  Two cartridge filters are placed upstream of the
system to remove suspended solids present in the groundwater.

Several operating parameters influence the performance of the system including hydrogen peroxide dose, influent
pH, tube cleaner operation, and influent flow rate (hydraulic retention time). During the demonstration, each of these
operating parameters was varied, and its impact on system performance observed. The authors claimed that a unique
feature of the perox-pure™ system is the circular wipers attached to the quartz tubes housing the UV lamps (5 kW
mercury vapor lamps). The wipers are designed to  improve system  performance by  minimizing the loss of UV
radiation caused by the formation of a chemical coating on glass surfaces from contaminant oxidation.

3. THE DEMONSTRATION PROJECT

3.1 Introduction

During the demonstration at the Lawrence  Livermore  National Laboratory, testing was carried out on shallow
groundwater contaminated with volatile organic compounds (VOCs). The  32-ha site was part of the laboratory's
General Services Area. Trichloroethene  (TCE) and tetrachloroethene  (PCE)  were  the principal groundwater
                                                 114

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

contaminants  with  concentrations  of about 1,000 and 100  ug/L,  respectively. Information from  a remedial
investigation and feasibility study and data from treatability studies were used to identify three primary objectives
for the demonstration: (1) determine the VOC removal efficiencies of the perox-pure™ system; (2) determine whether
treated groundwater meets applicable discharge limits at the 95% confidence level; and (3) estimate treatment costs,
including process chemical dosages and utility requirements. A secondary  objective was to obtain information on
the type of by-products formed during treatment.

3.2 Results of the Demonstration  Project

The perox-pure™ system generally produced an effluent with TCE and PCE  levels below detection limits, and 1,1,1-
trichloroethane (TCA) and chloroform levels slightly above detection limits. The system achieved maximum removal
efficiencies of greater than 99.9, 98.7, and 95.8 % for TCE, PCE, and 1,1-dichloroethane (DCA), respectively. The
system also achieved removal efficiencies of up to 92.9 and 93.6 % for TCA and chloroform, respectively. The
treated effluent met California drinking water action levels and Federal drinking water maximum contaminant levels
for all VOCs at the  95% confidence level.

3.3 Residuals

Gas chromatograph and mass spectrometer analysis of influent and effluent samples for VOCs indicated that no toxic
intermediate organic compounds were created during treatment. Although the toxicity of the effluent to water flea
and fathead minnow was found to  be greater than for the influent, this was assumed by the authors to be due to
residual hydrogen peroxide being present.

4. OTHER PROJECTS

Information on how  the technology performed at three other facilities provide additional performance data to support
findings of the SITE project. For these studies, the technology was used to treat wastewater containing acetone and
isopropyl alcohol (IPA), groundwater contaminated with TCE, and groundwater contaminated with pentachlorophenol
(PCP).

The first case  study at the Kennedy Space  Center in Florida tested the oxidation system as a replacement for the
existing carbon adsorption treatment  for wastewater containing acetone  and IPA. The replaced system could not
achieve the required discharge level of 0.5 mg/L. In contrast, effluent from the new oxidation system met all of the
discharge criteria, including the  demineralization discharge standards  in less than the specified 24-hour maximum
treatment time. The  system was efficient enough to allow treatment of the wastewater in a flow-through rather than
a batch mode at a flow rate of 5  U.S. gallons (18.9 L) per minute (gpm), a hydrogen peroxide dosage of 100 mg/L,
and 10 kW of power for a period of 20 hr/day.

For the second case study, the perox-pure™ technology was used to treat well water containing 50-400 mg/L of TCE.
Because the well was located in the middle of a large residential area in Arizona, the treatment was chosen because
of its  low-visibility and quiet operation. When treatment was conducted at a flow rate of 135 gpm (510 L/min) and
15 kW of power, TCE was consistently treated to a level below the analytical detection limit of 0.5 ug/L.

A full-scale perox-pure™ system treated groundwater contaminated with PCP at levels up to 15 mg/L for the third
case study  conducted on the property of a chemical manufacturing company  in Washington in  1988. Continued
operation confirmed that the perox-Pure™ system could destroy the PCP to below the target level of 0.1 mg/L. To
obtain this  result, the system operated at a flow  rate of approximately 70 gpm (265 L/min), a hydrogen peroxide
concentration of 150 mg/L, and a power requirement of 180 kW. High levels of iron required a pretreatment system
to oxidize and remove this metal. To  reduce the  tendency for scaling, acid  was added to lower the pH of water to
around 5, and the oxidation system featured automatic devices to keep the  lamps clean.

5. COSTS

Cost analysis of the demonstration project indicated that the groundwater remediation cost for a 190 L/min perox-
pure™ system would range from U.S.$2-3/m3  (1992), depending on contaminated groundwater characteristics. Of
this total cost, the perox-pure™  system direct treatment cost would range from U.S.$0.80-1.30/m3 (1992).


                                                  115

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

Treatment costs for three of the other four projects ranged between U.S.$3.60 and U.S.$5.00 per 1,000 U.S. gallons
(U.S.$0.95-1.3/m3). Treating well water for the  second case  study cost only  U.S.$0.28 per 1,000 U.S. gallons
(U.S.$0.07/m3). These costs exclude those for capital.
                                                  116

-------
NATO/CCMS Pilot Study, Phase II
                                            Final Report (Appendix IV)
   Project No. 39
                          Management of Soil Vapors at the Basket Creek Site
   Technical Contact:
   R. Donald Rigger
   U.S. Environmental Protection
   Agency, Region 4
   345 Courtland street, NE
   Atlanta, GA 30365

   tel: +1/404-347-3931
            Country:
            United States of
            America
Project Status:
Accepted by study  1992
Final report         1997
1. BACKGROUND

The Basket Creek Surface Impoundment site (Basket Creek) is located in Douglasville, Georgia, USA. The site was
contaminated during the 1960's when it was used for the illegal disposal of hazardous wastes.  At that time, an
intermittent stream bed was dammed with soil to form a surface impoundment. The impoundment measured 35 feet
(10.7 m) by 50 feet (15.2 m). It ranged in depth from 6-12 feet (1.8 to 3.7 m).

Liquid refinery and other hazardous wastes were reportedly disposed in the impoundment over a number of years;
however, detailed information on the quantity and specific composition of the wastes was not available. The wastes
were accidentally ignited in July  1970 and burned for  several days.  Subsequently, local officials required the
landowner to cease disposing waste and cover the impoundment with soil.

In 1989, the EPA conducted a Hazard Ranking System evaluation for Basket Creek, but the site did not qualify for
the National Priorities List. Emergency action, however, was initiated in 1991. The site was found to contain high
soil concentrations of lead, mercury, and organic compounds (trichloroethene, toluene, methyl isobutyl ketone and
methyl ethyl ketone).

Several different remedial strategies were evaluated. Off-site remediation was  rejected on the grounds of cost.
Treatability studies were carried out for two on-site treatment technologies: in situ soil vapor extraction (SVE) and
low temperature  thermal desorption. In situ SVE was  ruled  out on the basis of the soil's low permeability. Low
temperature thermal desorption was rejected after the  soil sample ignited during trials.

The selected technology was ex situ SVE, which involved excavation, processing, and stockpiling of contaminated
soil on the site. The stockpiled soil was then treated using  SVE. Soil excavation and treatment was carried out under
cover of a large ventilated enclosure; exhaust gases were passed through a thermal oxidizer (fume incinerator) before
discharge to atmosphere.

2. TECHNICAL CONCEPT
The treatment technology comprised four main components:
 (1)  Metal  enclosure,  60  ft
      (18.3 m) x 120 ft (36.6 m)

 (2)  Soil handling plant
Designed to totally enclose the  impoundment and  stockpiled  soil and to
contain processing operations.

Track-mounted excavation equipment and power screen. Soil was excavated
and placed directly onto the power screen, which  shredded  soil clumps.
Screened soil was transported to the stockpile using a  covered  stacking
conveyor.
                                                  117

-------
NATO/CCMS Pilot Study, Phase II
                                                     Final Report (Appendix IV)
 (3)  SVE system
 (4)  Air   emission
      equipment
control
Comprised seventeen,  4-inch (100-mm) diameter slotted well screen strings
positioned horizontally in the soil stockpile. Chemical resistant screens were
placed over the  well  screen to prevent soil from clogging the  slots. The
vacuum for each well screen row was supplied by a vacuum pump fitted with
a filter canister and silencer.
Vapors  drawn out of  the SVE system were routed via PVC piping  to the
thermal oxidizer. This was a three-chamber, propane-fired unit designed to
treat  10,000  ft/min3  (283 nrVmin)  of vapors with a greater  than 99%
destruction and  removal efficiency. In  addition, a  moveable fume  hood
collected vapors at the excavation point and at the power screen. The vacuum
was provided by a draft blower located outside the  enclosure. A baghouse
filter plant was used to remove particulates (down to 0.5 microns) from the
air stream. Air exiting the baghouse was routed to the thermal oxidizer.
3. MATERIALS PROCESSED
Samples of material from the impoundment were collected in March 1990, May  1991, and  January  1992  and
analyzed for organic compounds and metals. Samples collected in May 1991 were subjected to the USEPA Toxicity
Characteristic Leaching Procedure (TCLP).  The total concentrations and TCLP concentrations of the primary
contaminants are listed in Table 1.

                            Table 1: Concentrations of Primary Contaminants
Contaminant
Trichloroethene
Toluene
Methyl ethyl ketone
Methyl isobutyl ketone
Lead
Mercury
Concentration
Total (mg/kg)
BDL - 8,600
9,300 220,000
BDL - 23,000
1,400 - 66,000
668 - 4,400
38.2 3,554
TCLP (mg/L)
11.0
BDL
280
BDL
32.6
0.1
BDL = below detection limit
Key characteristics of the host matrix were:

        Particle size distribution of 16.4% clay, 34.4% silt, 40.8% sand, and 8.4% gravel

  •      Moisture  content of 16.9%;

  •      Air permeability of 1.5 x 10"7 cm/s;

        Porosity of 0.316;

        Bulk density of 112.5 lbs/ft3 (1,800 kg/m3); and

        pH of 5.46.

4. SYSTEM OPERATION

A trial burn of the thermal oxidizer took place in October 1992. Soil in the impoundment was exposed and stirred
to liberate volatile organic compounds (VOCs). The vapor was routed to the thermal oxidizer which was operated
at approximately 1,600°F (870°C). Mass emission rates were calculated for VOCs, semivolatile compounds, dioxins
and furans, and were reported to the Agency for Toxic Substances and Disease Registry (ATSDR). The ATSDR
determined that predicted emissions from the oxidizer would not pose a threat to public health.
                                                 118

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

Full-scale operations began in November 1992. The contaminated area was divided into 48, 10 ft x 10 ft (3 m x 3
m) grids. Excavation was carried out in individual blocks to minimize the surface area of exposed soil. Rocks and
debris larger than 2 inches (50 mm) were rejected by the power screen and placed in roll-off boxes. Excavation (up
to 14 feet (4.3 m)) was halted when solid homogeneous rock was encountered.

The  SVE system was  operated continuously, and that associated with excavation and screening was  operated an
average of 25-30 hours/week. In February 1993, excavation and backfilling of the 48 grid sections was complete.
The  SVE system was  operated for an additional three weeks after excavation to complete treatment of the soil.
Approximately 100 U.S.  short tons (90 tonnes) of rocks and debris from power screening,  18 yd3 (13.8 m3) of
excavated metal and crushed drums and approximately 4,250 gallons (16.1 m3) of decontamination water from health
and safety activities were disposed at appropriate off-site facilities.

Site restoration involved dismantling and removing the enclosure and process equipment. The site was then regraded
using site soils and seeded with  grass. Site restoration was completed in August 1993.

5. RESULTS

The  action memorandum covering remediation activities at the  site defined treatment targets for treated soils as
shown  below. In addition to these targets, the thermal oxidizer was required to be at least 95% efficient for VOC
destruction.
                      Parameter                           Soil Treatment Target
                      Trichloroethene (TCLP)                       0.5 mg/L
                      Tetrachloroethene (TCLP)                     0.7 mg/L
                      Benzene (TCLP)                             0.5 mg/L
                      Methyl ethyl ketone (TCLP)                   200 mg/L
                      Lead (TCLP)                                5.0 mg/L
                      Mercury (TCLP)                             0.2 mg/L
                      Total halogenated organic compounds         1,000 mg/kg
Analytical data for the treated soil showed that soil treatment targets were met for all contaminants after the six
months treatment period. Total VOCs in treated soil ranged from 0.142-1,571 mg/kg, and approximately 72,000 Ibs
(32,688 kg) of total VOCs were recovered from the soil. Toluene accounted for the largest proportion of recovered
VOCs (approximately 80%) with methyl isobutyl ketone accounting for the next largest (11%). The destruction
efficiency of the thermal oxidizer was at least 95%, and reached 98% for three months of system operation.

Airborne concentrations of VOCs inside the enclosure occasionally approached 500 ppm (the stop work condition)
and regularly ranged between 200-400 ppm. However, no VOCs were detected during hourly air monitoring surveys
outside the enclosure or in off-site high volume air samplers.

6. COSTS

Total project costs  were approximately U.S.$2.2 million  broken down (according to  the  inter-agency  Work
Breakdown Structure, WBS) are set out in Table 2. Costs directly attributable to treatment activities (U.S.$660,000)
correspond with a treatment cost of U.S.$413/yd3 ($540/m3) of soil treated (1,600 yds3 or 1,223 m3 of soil in surface
impoundment) or U.S.$275/U.S. ton ($27I/tonne) of soil treated and U.S.$9.20/lb ($20.28/kg)  of VOC removed.

7. HEALTH AND SAFETY

One of the main health and safety concerns of the project was the potential for the build-up of VOCs inside the
enclosure from soil excavation and screening activities, and the possibility of a fire or explosion. Various protective
measures were taken including:


                                                  119

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

•       A site safety officer was present inside the enclosure at all times during excavation.

•       Level B  personal protective equipment  (PPE) for workers  (supplied breathing apparatus, fire  resistant
        overalls,  disposable outer suits and boots, and hard hats).

•       Reusable PPE was decontaminated onsite.

                                        Table 2: Costs Summary
         Cost Element
                                                        Estimated Costs (U.S.S)
         (i)
Before Treatment Costs
                        Monitoring, sampling, testing, and analysis
                        Site work (excavation and soil preparation

                        Air  Pollution/gas   collection   and  control
                        (enclosure,   air   handling   system,   part   of
                        incinerator)
                                                               $260,000
                                                               $390,000

                                                               $650,000
                                                         Subtotal
                                                              $1,300,000
         (ii)
Treatment Costs
                        Operation (short term, up to 3 years) (operating
                        costs and personnel)

                        Cost  of ownership  (SVE   system,  part  of
                        incinerator)
                                                               $130,000


                                                               $530,000
                                                         Subtotal
         (iii)
After Treatment Costs
                        Disposal at a commercial facility

                        Site restoration

                        Demobilization
$660,000


$130,000



 $22,000

 $68,000
                                                         Subtotal
                                                               $220,000
The atmosphere inside the enclosure was monitored using a combustible gas indicator. A limit of 10% of the lower
explosive limit (LEL) was set as an automatic cease-work condition, and workers were required to evacuate the
enclosure. A photo-ionization Detector (PID) was also used to monitor airborne contaminants as total hydrocarbons
in air. A limit of 500 ppm total hydrocarbons was  also set as a cease-work condition.

Potentially explosive atmospheres were also monitored in the air handling system (inside and outside the enclosure)
including vapor extraction piping, flexible duct work and steel duct work.  Detectors signalled an audible alarm at
15% of LEL  (excavation stopped);  a  reading of 20% of LEL caused automatic  interlocks to activate, which shut
down the blower. The 15% LEL level in duct work was exceeded several times per day during  excavation of the
                                                   120

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

highly contaminated areas. Several times during the project, the 20% LEL level was exceeded. Air monitoring was
also conducted at the perimeter of the enclosure and using off-site high volume samplers.

8. CONCLUSIONS

        Data for samples collected from the soil stockpile showed that soil treatment targets were met in all 14
        sampling grids after six months of treatment although there were variations. For example, total VOC data
        showed a range over four orders of magnitude (0.142-1,571 mg/kg).

        A total  of 72,084 Ibs (32,730 kg) of total VOCs were recovered of which 75-80% were from the soil
        stockpile and 20-25% from excavation and screening.

        VOC recovery varied during the course of treatment, peaking during weeks 7, 8 and 9 at 11,000 Ibs/week
        (4,994 kg/week). Variations are thought to reflect those in soil concentrations within the impoundment.

        Recoveries varied for different VOCs with toluene and methyl isobutyl ketone (measured at the highest
        concentrations prior to remediation) accounting for approximately 80% and  11% of the total  recovered
        VOCs, respectively.

        The destruction efficiency of the thermal oxidizer was at least 95% over 12 weeks of operation, reaching
        98% during 3 months when monthly average inlet concentrations were greater than  100,000 ppmv.

        Although VOC concentrations inside the enclosure were elevated, they  never exceeded the stop-work
        thresholds of 10% of the LEL or 500 ppm total hydrocarbons. No VOCs were detected outside the enclosure
        or at off-site sampling locations.

        Although no pre-treatment data for stockpiled soils were available, TCLP concentrations for trichloroethene
        and methyl ethyl ketone in treated soil were lower than those in pre-excavation samples. Concentrations of
        lead and mercury in soils were also lower following treatment, although it is believed that pre-excavation
        sample results for these metals may not be representative  of the contaminated area as a whole.

        The choice of ex situ soil vapor extraction resulted in lower costs than would have been expected for off-site
        incineration of soil. The  cost is estimated between U.S.$6-6.7 million.

9. LESSONS LEARNED

        Excavation and power screening greatly increased soil permeability.

        There were limitations associated with the construction materials used for the soil stockpile. Spacing of
        horizontal wells was difficult because they were made of PVC and would bend under the weight of the soil.
        Carbon  steel pipes would have been more rigid, but more expensive.

        It is important to oversize the air handling system for  venting the enclosure. This helps  prevent the
        development of explosive conditions in the enclosure and ductwork.

        Excavation within the enclosure took much longer than it would have done outside due to space constraints.
        Excavation within the enclosure at Basket Creek took 3 months compared to an estimated 4 days had it
        taken place outside.
                                                  121

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 40

       An Evaluation of the Feasibility of Photocatalytic Oxidation and Phase Transfer Catalysis for
           Destruction of Contaminants from Water (In Situ Treatment of Chlorinated Solvents)
   Technical Contact:
   Mark Smith
   U.S. Air Force
   EOARD
   223/231 Old Marylebone Road
   London NW1  5TH

   tel:  +44/171-514-4285
   fax: +44/171-514-4960
   E-mail:  msmith@eoard.af.mil
Country:
United States of
America
Project Status:
Accepted by study      1992
Interim report           1993
Final report             1994
1. INTRODUCTION

Photocatalysis is an emerging treatment technology that uses an advanced oxidation process based on generating
hydroxyl radicals using UV light in the presence of a semi-conductor catalyst, such as platinum-coated titanium
oxide.  The Hand D process  uses  a fixed-bed catalyst with the  semi-conductor fixed to a silica gel support.
Pretreatment of groundwater by removing suspended matter and inorganic ions is conducted by a series of filters and
ion-exchange columns. Dissolved oxygen levels are increased, as necessary, to supply oxidants for the destructive
oxidation process. The UV light is provided by natural sunlight.

The project researched development of highly photoactive catalysts and a fixed-bed photocatalytic process to destroy
toxics  in air and water.  It also researched  development of a treatment process  using adsorption to remove
contaminants and advanced oxidation processes (AOPs) to regenerate spent adsorbents. The developed technologies
were field tested at Tyndall Air Force Base (AFB), K.I. Sawyer AFB, and the  Wausau Water Treatment Plant.

2. LABORATORY STUDIES

Some of the photocatalysts used in the laboratory studies were modified on their surfaces with noble metals, or
changed by doping with transition metals to extend the photocatalyst's response to visible light. Artificial light and
UV sources and solar radiation were used to evaluate the photoactivities of these catalysts for destroying model
compounds. Results showed that platinum-coated Aldrich titanium oxide (Pt-Aldrich-TiO2)—a surface-modified
catalyst—performed best for destroying  hydrophobic  compounds. The  laboratory-developed, platinum-coated
Michigan Technology University TiO2  catalyst performed best for hydrophilic compounds.

The supports tested for fixed-bed photocatalysts consisted of random packing and structured materials. These supports
were chosen for their adsorption capacity, UV transmission, and mass transfer properties. Silica-based materials were
included to test how they performed at  destroying organic compounds, and some of them were surface-modified to
increase their adsorption capacity. Investigations excluded electron-rich materials that could scavenge reactive radicals
and diminish the efficiency of photocatalysis. Researchers developed a unique procedure involving heating and
annealing to prepare supported catalysts for fixed-bed reactors. During solar experiments, the fixed-bed processes
were optimized with respect to the type of catalyst and dosage, support type and size, and preparation methods.
Destruction of a model compound was tested under various UV irradiance, influent concentration, pH, and hydraulic
loading. Tanning lamps tested on the same fixed-bed process tested the destruction of several other compounds in
air and water. Two reactor designs consisted of a catalyst added as a slurry and passed through a lighted reactor, and
a catalyst attached to a support in a lighted fixed-bed reactor.

Tests with the fixed-bed photocatalysts revealed that Pt-Aldrich-TiO2 supported on silica gel completely mineralized
TCE in water (8  mg/L) in one contact time of 1.3  minutes. This destruction rate was  16 times better than observed
for an optimized slurry of a commercially-available photocatalyst (Degussa P25).
                                                  122

-------
NATO/CCMS Pilot Study, Phase II                                            Final Report (Appendix IV)

Increasing the adsorption capacity of a silica-based support improved the overall destruction kinetics. Silica gel
modified to increase surface hydrophobicity  and the  adsorption capacity  for non-polar organic water pollutants
increased the adsorption capacity for trichloroethene (TCE) in water by more than a factor of five. When used as
a support for  Pt-TiO2, the  modified silica gel showed faster overall  degradation kinetics for TCE than did the
unmodified silica gel.

For experiments with contaminants in the gas-phase, results showed faster destruction kinetics than in the aqueous
phase. This suggests the option to strip volatile compounds followed by gas-phase destruction by photocatalysis. For
trichloroethane (TCA), a relative humidity of 25% yielded the best destruction rate. On the other hand, toluene
destruction increased with increasing vapor content. The fixed-bed approach provided high  light efficiency. The
reported ratio  of organic  molecules destroyed to  UV photons required was 40%—much higher than the 5% value
commonly reported. Phosgene and carbon monoxide, two major toxic by-products in air-phase photocatalysis, were
not found above regulated levels.

Developing the combined processes of adsorption followed by regeneration consisted of using a fixed-bed system
adsorber to remove and accumulate organic compounds, and regenerate the spent adsorbent using homogeneous AOP,
photocatalysis, or a combination of steam and photocatalysis. For homogeneous AOP, hydrogen peroxide/ozone and
UV light/hydrogen peroxide were used for destructive adsorbent regeneration. Both of these options consumed three
to six times more oxidants in the regeneration process than would be needed to destroy the contaminants in water
directly by conventional AOP. Regeneration appeared to be limited by adsorbate desorption from the interior to the
exterior adsorbent surface. Neither option is feasible because the desorption rate is too slow, and therefore too much
oxidant is required to regenerate the adsorbents.

In testing photocatalysis for destruction of adsorbed contaminants and regeneration of adsorbents simultaneously, the
catalysts were  impregnated onto the adsorbent before being used to adsorb organics. The first test consisted of using
UV illumination to test photocatalysis alone for regenerating the spent adsorbents and destroying the contaminants.
Results showed that desorption of adsorbates from the interior to the exterior of an adsorbent limited the regeneration
process. Thus, temperature played a leading role in photocatalytic regeneration, and increasing temperature enhanced
regeneration rate much more effectively than increasing light intensity. Using heat to increase the desorption rate and
match the photocatalytic  oxidation rate is one way to maximize the photolytic regeneration efficiency.

Saturated steam was used to overcome this problem of slow desorption rates. Heating promoted the kinetics of
photocatalysis and AOP processes by desorbing organic contaminants on the interior of the adsorbents and moving
them to the exterior for oxidation. Thus, steam  followed by photocatalysis was found to be an effective way to
regenerate spent adsorbents and to clean up the regeneration fluid (off-steam or steam condensate).

3. FIELD TRIALS

At Tyndall AFB, a solar photocatalytic process was used to remediate fuel-contaminated groundwater containing
BTEX compounds (benzene, toluene, ethylbenzene, and xylenes) at greater than 2 mg/L. Two options  were tested:
(1) a solar photocatalytic fixed-bed process using Pt-TiO2 supported on silica; and (2) fixed-bed process with Pt-TiO2
impregnated adsorbents alone or combined with Pt-TiO2 supported on silica gel. During option 1, ionic species fouled
the catalysts and inhibited destruction, so the water was pretreated to remove suspended particulates and ionic species
and to increase dissolved oxygen. Following pretreatment, catalyst photo-activity continued undiminished  after 25
days of operation. The BTEX compounds were destroyed with 6.5 minutes of empty bed contact time on rainy days.
Test results with various flow  rates,  reactor diameters, influent concentrations, solar irradiances and  weather
conditions confirmed the  potential application of the process. Treatment cost was estimated at  U.S.$5.52/1,000 U.S.
gallons (U.S.$1.46/m3).

During option 2,  two different reactor design configurations were tested. The first design was a continuous flow
configuration with three fixed-bed reactors in  series. The first and third reactors were packed  with platinum-coated
TiO2 supported on silica gel supports. The second reactor was packed with a photocatalyst-impregnated adsorbent.
In daylight, the incoming organic compounds were destroyed in the first reactor. At night, the organics were adsorbed
onto photocatalyst-impregnated adsorbent in the  second reactor; the adsorbed organics were  either mineralized or
partially mineralized in the second reactor during daylight. The remaining organics or destruction by-products were
destroyed in the third reactor. Because organic compounds can be adsorbed during periods of insufficient sunlight,


                                                   123

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

the process can treat water continuously. Like option 1, option 2 was also affected by catalyst fouling. After four
days of operation, the process was apparently unable to destroy any more BTEX compounds. An ion exchange unit
was added to the system and satisfactory BTEX destruction was resumed; however, the third reactor still did not
perform well, presumably due to fouling species desorbed from the second reactor.

In the second design, a reactor packed with photocatalyst-impregnated adsorbents was used to remove the organics
during darkness. During  daylight, the adsorbent was taken off-line and regenerated by passing heated water through
the reactor. The hot water was then passed through a fixed-bed reactor packed with platinum-coated TiO2 supported
on silica gel supports to destroy any residual desorbed organics or by-products. The strategy behind this design was
to have an 18-hour adsorption period followed by a 6-hour regeneration period during which the solar irradiance is
strong enough to destroy most contaminants.

The spent adsorbents were regenerated in the presence of sunlight while passing hot water (90°C) counter-current
to the  flow  direction during the adsorption process. The  desorbed organics from the regeneration process were
destroyed in the subsequent fixed-bed photoreactor. The process was examined for 10 adsorption and regeneration
cycles. The efficiency (ratio of organics removed and destroyed to organics adsorbed) of the process to regenerate
the adsorbents was examined for each cycle. Overall, 21% of the total influent BTEX was destroyed,  and 98% of
the BTEX was removed from the waste stream. The detention times required for 99.9% destruction of TCE (5 mg/L),
TCA (5 mg/L),  and toluene (115 mg/L) were 2.15 seconds, 11.2 seconds, and 40 seconds, respectively. While the
process appears to be effective for 10 cycles, more cycles are required to determine whether steady-state is achieved.

At K.I. Sawyer AFB, the unit used a fixed-bed photocatalyst and a solar panel made up of 80 tubular reactors to treat
chlorinated compounds in groundwater. The design included a water pretreatment unit, which included a turbidity
filter, a bubble-less oxygen contactor, and ion-exchange columns. The reactor was a modified solar thermal reactor
panel with 80 plastic tubes mounted in parallel.  The capacity for each reactor panel was  designed as 0.25 U.S.
gallons per minute. On a sunny afternoon, the panel destroyed 95% TCE, which was present in groundwater at 100
ug/L, within a two-minute contact time.

At the Wausau Water Treatment Plant, fixed-bed adsorption removed and accumulated organic compounds. The spent
adsorbents were regenerated off-line with steam followed by photocatalysis. The groundwater contained chlorinated
compounds and BTEX. Direct photocatalytic oxidation was not effective for regenerating spent adsorbent loaded with
contaminated groundwater because nuisance substances in the water fouled the catalyst. Tests combining steam
regeneration followed by photocatalysis of steam condensate showed that carbon adsorbent was not effectively
regenerated and lost a significant amount of capacity.  Background organic matter in the water may  have caused this
loss of efficiency. Based on the  results of chloride yield with all the tested adsorbents, steam regeneration did not
appear to destroy significant amounts of the sorbed chlorinated compounds.
                                                   124

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 41
                                         In Situ Microbial Filters
   Technical Contact:
   Richard  Knapp
   Lawrence Livermore National
   Laboratory
   L206
   P.O. Box 808, LLNL
   Livermore, CA 94550

   tel: +1/510-423-3328
   fax: +1/510-422-3118
Country:
United States of
America
Project Status:
Accepted by study  1993
Final report         1994
1. INTRODUCTION

This technology exploits the capability of resting-state methanotrophic bacteria to  degrade trichloroethene  (TCE)
oxidatively. Bacteria are produced in bioreactors and injected through a borehole into the subsurface where they
attach to the solid rock matrix creating an inoculated subsurface zone. Contaminants dissolved in groundwater are
biodegraded by bacteria as  the groundwater  passes through the "biofilter." The Pilot Study followed a field
demonstration at the Kennedy Space Center, Florida where groundwater is contaminated with TCE. In the field
demonstration, the biofilter was established by injecting bacteria into the contaminated aquifer through a borehole,
and the filter function was tested by withdrawing groundwater through the same borehole.

2. TECHNICAL CONCEPT

Three major engineering issues were considered for the performance of the in situ biofilter:

    (1) Groundwater residence times in the microbially enhanced zone must be sufficient to allow degradation to
        take place;

    (2) Nutrients are not supplied to the  bacteria, and the  oxidative transformation of contaminants  consumes
        cellular resources, which are not replenished; therefore, a fixed mass of contaminants can be biodegraded
        by each resting-state bacterial population, which must be periodically replaced;

    (3) The  efficiency of  microbial attachment to subsurface soil and  rock matrix  depends on site-specific
        parameters.

Laboratory study of these considerations led to development of empirical mathematical models for predicting site-
specific in situ conditions. Saturation, attachment rates, entrainment rates, and enzymatic longevity functions were
included, providing quantitative data for qualitative limits on engineering parameters.

3. RESULTS

This technology has been tested at field scale at the Kennedy Space Center on groundwater contaminated with TCE.
Indigenous TCE-degrading bacteria were collected from the site groundwater and grown in culture to  increase their
numbers. Bench-scale testing revealed that the degradation capacity of bacteria at the site was about 0.25 g of TCE
per gram of bacteria.  When sufficient bacteria were produced, the organisms were mixed with site groundwater and
reinjected into the subsurface around the borehole. An estimated 10% of injected bacteria attached to subsurface
rocks around the borehole. Contaminated groundwater was repeatedly flushed through the biofilter zone by extraction
and  reinjection through the same  borehole to greatly increase contact  residence times between  bacteria  and
groundwater. Nutrients were not supplied so that the contaminants were the only potential carbon source.

The thickness of the established biofilter was found to vary according to the rate of groundwater flow and the TCE
concentration, with values ranging from under 10 mm to over 500 mm at 20°C  and 108 cells/gram of sand.  It was
                                                  125

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

concluded that a 100-mm thick biofilter established at the site provided complete breakdown of TCE for a period
of 8 weeks, and had a reduced degradation capacity for an additional 8 weeks.

The demonstration was reported as being "successful" but regulatory guidelines for TCE were not met in the treated
water.  The  authors  concluded  that this was  due  to  a  high concentration of co-contaminants,  including
chlorofluorocarbon and methane,  which were preferentially degraded to TCE.

4. COSTS

Remedial costs were claimed to be lower than pump and treat (up to 50%), with the most economical problems being
associated with large fast-flowing plumes with low contaminant concentrations (<10 mg/L).

5. CONCLUSIONS

The authors suggested limitations  to the potential use of this technology: a minimum subsurface pore size of greater
than 10 um to permit microbial transport; groundwater pH near neutral; concentrations of TCE in groundwater should
not exceed 25 mg/L, since higher levels are toxic to the microbial degraders; and dissolved oxygen levels must be
in the range 0.37 to 1.28 mg/L to support the microbial community, depending on the specific contaminant and its
concentration.

6. REFERENCE

R. T. Taylor, et al. In Situ Bioremediation ofTrichloroethylen-Contaminated Water by a Resting-Cell Methanotrophic
Microbial Filter; Hydrology Science Journal, 38(4), pp. 323-342, 1993.
                                                  126

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 42
                       Integrated Pneumatic Fracturing and In Situ Bioremediation
   Technical Contact:
   Stephen James
   U.S. Environmental Protection Agency
   National Risk Management Research
    Laboratory
   Martin Luther King Dr.
   Cincinnati, OH  45268

   tel:  +1/513-569-7877
   fax: +1/513-569-7680
   E-mail: james.steve@epa.gov
Country:
United States of
America
Project Status:
Accepted by study  1994
EPA Reports       1993,1995
1. INTRODUCTION

This in situ technology incorporates extending and creating fractures to increase permeability in tight soils to enhance
in situ treatments such has bioremediation. The U.S. Environmental Protection Agency (USEPA) conducted pilot-
scale demonstrations under the USEPA Superfund Innovative Technology Evaluation (SITE) program(l). USEPA
reviewed the research and development of hydraulic and pneumatic fracturing technologies to remove contaminants
from soil and groundwater(2).

2. BACKGROUND

A field-scale pilot study conducted under the SITE Emerging Technology Program investigated using pneumatic
fracturing to enhance microbial processes. Fracturing and nutrient injections took place over a one-year period. This
study  was completed in March 1995.

3. TECHNICAL CONCEPT

Tight  soils hamper in situ technologies such as soil vapor extraction, bioventing, air sparging, and other air and
nutrient injection techniques. Hydraulic and pneumatic fracturing are enhancement technologies to increase treatment
efficiency of in situ techniques. Hydraulic fracturing utilizes pressurized water, while pneumatic fracturing utilizes
pressurized air to create cracks in low-permeability and over-consolidated sediments. By increasing the permeability
of the soil to liquids and vapors,  removal of contaminants  by  soil vapor extraction and biodegradation can be
accelerated. Aerobic processes dominate at the fracture interfaces and, to a limited distance, into the soil away from
the fracture.

Initial site characterization indicated low-permeability soils and benzene, toluene, and xylene (BTX) contamination.
Over a one-year period, the site was pneumatically fractured, and periodic injections of nitrate and ammonium salt
to enhance aerobic and anaerobic biodegradation were performed.  Offgases from monitoring wells were  analyzed
for BTX  compounds along with carbon dioxide, methane, and oxygen to evaluate treatment efficiency. Additional
soil cores were analyzed to measure change in the extent of site contamination during remediation. Carbon mass
balances were performed to evaluate treatment efficiency.

4. RESULTS AND CONCLUSIONS

Fracturing increased subsurface permeability by up to 40 times within an effective radius of approximately 20 feet
(about 6 m). Results from soil sampling at the end of the demonstration showed a 79% reduction in soil-phase BTX
concentrations. Cores from three  distinct depths taken before and after remediation showed the total  mass of BTX
removed  to be 22 kg.  Based on periodic soil gas sampling, the mass of BTX removed during soil vapor extraction
was computed to be 3.1 kg. Other BTX losses accounted for 0.8 kg of BTX. The total BTX mass removed due to
biodegradation during the 12-month period was calculated to be over 82%.
                                                 127

-------
NATO/CCMS Pilot Study, Phase II                                         Final Report (Appendix IV)

References

1.   U.S. Environmental Protection Agency. Hydraulic Fracturing Technology Application Analysis Report
    National Risk Management Research Laboratory, Cincinnati, OH, September 1993, EPA/540/R-93/505.

2.   U.S. Environmental Protection Agency. In Situ Remediation Technology Status Report: Hydraulic and Pneumatic
    Fracturing, Office of Solid Wasted and Emergency Response, Technology Innovation Office, Washington, DC,
    April 1995, EPA542-K-94-005.
                                                 128

-------
NATO/CCMS Pilot Study, Phase II
                              Final Report (Appendix IV)
   Project No. 43
                     Multi-Vendor Bioremediation Technology Demonstration Project
   Technical Contact:
   James Harrington
   New York State Department of
       Environmental Conservation
   50 Wolf Road, Rm 260A
   Albany, NY 12233-7010

   tel:  +1/518-457-0337
   fax: +1/518-457-9639
   E-mail: jharrin@gw.dec.state.ny.us
Country:
United States of
America
Project Status:
Accepted by study  1994
Interim report       1996
Final report         1997
1. INTRODUCTION

The "Multi-Vendor Technology Demonstration" project took place at the "Sweden 3 Chapman" site, Sweden, New
York, which was an abandoned hazardous waste disposal area containing very high concentrations of chlorinated and
non-chlorinated solvents such as trichloroethene (TCE), tetrachloroethene (PCE), 2-butanone (MEK)  and toluene.
Specific technical objectives of this study were: to determine whether the use of naturally-occurring microorganisms
can effectively remediate volatile organic compounds (VOCs) present in unsaturated soils at the site; generate field
data for simultaneous evaluation of different biological processes; permit an evaluation of in situ and ex situ
approaches; and evaluate performance of each demonstrated technology in meeting site specific remedial objectives.
A protocol for conducting pilot-scale demonstrations of bioremediation technologies at the Sweden site was prepared
at the University of Buffalo, and this was used as the core of the invitation given to technology vendors. Of thirteen
technical proposals submitted, three different technologies were selected for demonstration:

    (1)  Biovault (ENSR Consulting and Larsen Engineers): an ex situ biopile-based treatment;

    (2)  In  Situ Field Bioremediation Treatment System (R.E. Wright Environmental): an in situ bioventing and co-
        metabolic bioventing technology.

    (3)  In  Situ UVB (Vacuum Vaporized Well) Process (SBP Technologies/Environmental Laboratories [ELI]): a
        system-based treatment within and around wells.

2. SITE BACKGROUND

The  Sweden 3 Chapman site is in a rural residential area, where homes are serviced by private wells.  The site
operated from 1970-78 and received industrial, municipal and construction wastes. In 1991 a State Superfund Interim
Remedial Measure removed over 2,500  drums of waste and 2,270 tonnes (2,500  U.S. tonsa) of contaminated soil.
The treated area was graded and capped. However, unacceptable amounts of VOC contamination  remains on site.
An estimated 9,800 m3 of contaminated soil remained in the demonstration area.

The site is planned to be remediated by ex situ treatment using low-temperature thermal desorption, along with some
"limited" groundwater treatment. Demonstrated technologies were to be considered as alternatives, if their trials were
successful.

The site consists of a clay capping layer over a mixed fill layer, comprised of the disposed materials. Beneath the
fill material is a thin silty-clay soil layer (0.4 to 2.1 m) with numerous fine silty sand lenses. Beneath this layer lies
sandy-silt glacial till. Work during the trial revealed that a plastic liner had been placed between the fill material and
the natural  soil, creating a perched water table.
        The original US text refers to "tons, " which the editors have interpreted as short tons (2,000 pounds).
                                                  129

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

3.0 TRIAL CONCEPT

The project was a partnership of State and Federal agencies, universities and technology vendors. The technical
objectives of the project were to:

  •  demonstrate that the use of naturally-occurring microorganisms could effectively remediate VOCs onsite;

  •  generate field data to allow the simultaneous evaluation of different biological processes;

  •  permit an evaluation of ex situ and in situ approaches; and

  •  evaluate the performance of each technology in meeting the established clean-up goals.

Clean-up goals were set by the N.Y.  State Department of Environmental Conservation for six VOCs: acetone, 200
ug/kg; MEK, 600 ug/kg; methyl isobutyl ketone (MIBK), 2,000 ug/kg; PCE, 2,500 ug/kg; TCE, 1,500 ug/kg and
dichloroethene (DCE) 600 ug/kg soil. 90% of tested samples were to meet these targets to achieve  compliance
(success). Toluene concentrations were also monitored.

Prior to selection of demonstration technologies, a detailed test protocol was drawn up by a panel of experts and used
as part of the technology tendering. Sampling and analysis was carried out by a university and an independent
consultant was retained by the U.S. Environmental Protection Agency (USEPA). Three sampling events were carried
out. The  demonstrations started in August 1994 and were completed for two of three projects by December 1994.
The UVB technology was  demonstrated over a longer time  period (approximately 14 months).

Each vendor was allocated a trial area to carry out a demonstration-scale test. However, large differences between
contaminant concentrations in the material treated complicated comparisons between technologies. Interpretation of
results was further complicated by higher than expected detection limits for target analytes because of interference
in the measurements from  other VOCs. In some cases, detection limits exceeded treatment targets.

4. TREATMENT TECHNOLOGY DESCRIPTIONS

4.1 RE.  Wright In Situ Treatment

This technology was essentially a co-metabolic bioventing approach consisting of extraction wells linked to a central
blower (Figure 1). The output of the blower was connected to the injection wells via ports to allow inputs of: ambient
air, methane and anhydrous ammonia. The methane  was  intended to boost  co-metabolic processes degrading
chlorinated VOCs. The ammonia was intended to boost available nitrogen. The blower operated for approximately
2 hr/day for about five  months. The system was designed to treat an area of 535 ft2  (50 m2) to a depth of 6 ft (1.8
m). The radius of influence of each well was estimated to be 10 ft  (3 m). The treated material turned out to be less
contaminated, with many  samples meeting treatment targets in  advance of treatment. The installation of the
technology was complicated by the presence of the plastic liner and perched water, which were not anticipated based
on available site investigation information.

4.2 ENSR/Larsen Biovault

Biopiles were constructed on twin layers of high density polyethylene (HOPE) supported by earth berms (Figure 2).
The HDPE was covered by a layer of sand for protection and drainage. Approximately 100 yd3 (76 m3) of graded
material was placed on a geofelt liner over the sand. The area of the pile was around 40 by 40 ft (12 x 12 m), and
its height was about 3 ft (0.9 m). The surface of the pile was covered by a similar felt, sand, and HDPE construct.
Aeration was via extraction pipes in the bottom sand layer connected to a fan with passive air injection pipes in the
top sand layer. Liquids could be percolated into the biopile via a  sprinkler system.  Air and liquid movement was
intended to optimize conditions for biological activity within the piles.

Two piles were  set up, one with continuous aeration and one with discontinuous aeration intended to allow the
development of alternating periods of aerobic and anaerobic activity within the piles. The materials treated within
the piles were the  most grossly contaminated site materials. The treatment was made  more difficult by the fine-


                                                   130

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
Air Phase
Carbon Units
< r^ r^
< ^j LJ
r~L
-OOJ
i / — ,
/
V
Water Phase i —
Carbon Units /

Atmospheric —
Intake




/
Extraction
Manifold






/°\ D
09°
(5?\ .
l^y o

09°
(S?)o


09°
Amendment
^~ Injection port Extraction/Injection
^ — Unit
^ — -~ ~~


/ Air Flow Adjust Valve
-i ^^ Sample Port
^Q^Q^Q^Q^Q^Q
(5?)°
^QiQ^Q^piQiQ^

^^Q^Q^Q^Q^Q^Q
(5?)o
^Q^Q^Q^QiQipi
(5J)o
""iQApApip, ±Q±Q
Injection ^ — Extraction
Well Well
Figure 5.  Schematic of R.E. Wright In Situ Treatment


grained texture of the soil being treated which impeded aeration and water movement. The vendor also claimed that
the intermittent aeration pile did not achieve conditions suitable for anaerobic activity.

4.3 UVB Process

A schematic of the process is given in Figure 3. The basic unit was a 16-inch (400-mm) diameter steel water well
with upper and lower screens inserted to a depth of 26 feet (8 m) below the ground surface. A submersible pump
in the bottom zone of the well pumped ground water through an in line bioreactor packed with granular activated
carbon. The treated water was discharged back into the soil from the top zone of the well. The bioreactor unit was
integrated with an air stripping function, both to  remove  VOC, and to oxygenate  the groundwater prior to its
discharge. The discharge (coupled with the pumping of water from the lower zone) is intended to circulate water
around the well and so create a zone of enhanced stripping and biodegradation activity around the well. The assembly
included monitoring wells and nutrient addition.

One well assembly was tested during the trial. The unit was designed to treat approximately 1,000 cubic yards  (760
m3), covering an area of about 280 yd3 (230 m2). The radius of action of the well was estimated to be 35 ft (11 m).
Installation of the well was complicated by the plastic liner and an unexpectedly shallow water table, due to  high
rainfall.

4.4 General

All three treatments included the capacity for treating offgases  and  liquid effluents generated.
                                                  131

-------
NATO/CCMS Pilot Study, Phase II
                                                            Final Report (Appendix IV)
                                          Top View
                                                                      WATOT PIPING (TOP)

                                                                                   TOE WALL AROUND PAD
                                                                   AM PIPING (BOTTOM AND TOP OF BWVAULT)

                                                                    SUMP TO LAIS
                                            40'-Q"
                                        Cross Section
              HOPE COVER -\
                4" SAND
                                      SITS. son.
         J
                   SITE SOL
BERM
•\_T
                   4*SAND
                   4'SAM)
                                                                  • SUMPTOLMS

                                                                   AIR PIPING TO
                                                                   AMBIENT AIR
                                                                  , WATER PIPING
                                                                   FROMLMS
                                                                                  -*• AIR PIPING
                                                                                     TO CMS
                                           DERM
X—
LEAK
DETECTION
                                      1-2% SLOPED
                        40 MB. HOPE
                           LINER
                                                        -— LEAK DETECTION
                                                          PROBE
     Figure 2: ENSR/Larsen biovault system: plan view and cross-section
5. RESULTS

VOC removal results for each technology are  summarized in Table  1. Note that the results are still subject to
verification by the USEPA.

5.1 R E Wright In situ System

The system achieved the compliance target. However, contaminant levels—particularly for MIBK and PCE—were
already low in the test area, and data for acetone and MEK could not be interpreted because of difficulties with their
detection limits. However, residual concentrations of toluene remained high. The estimated starting mass of toluene
in the test volume was estimated to be 14 kg, and the finish mass to be 7 kg. MEK and MIBK data were not used
in compliance assessment.

5.2 ENSR/Larsen System

Neither the continuously or intermittently aerated biopiles achieved compliance (79%  of samples met targets),
although both achieved substantial degradation  of VOCs. As well as the difficulties with soil texture and aeration
regime  already mentioned, low ambient temperatures during the latter part of the treatment period may also have
                                                    132

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
                                  Withdrawal
                                                             Withdrawal
     Figure 3: Schematic of SBP/ELI in situ UVB process
reduced biopile effectiveness. Interpretation of acetone and MEK data was complicated by difficulties with detection
limits. Compliance assessment for other contaminants was based on 6-9 usable data points.

5.3 UVB

After 5 months,  when the other trials were completed,  samples from the UVB treatment area did not indicate
significant contaminant reduction. Compliance was achieved for PCE only. A longer treatment period was considered
appropriate for this technology. After 14 months compliance was also achieved for MIBK and TCE, (based on 45
usable data points).

                                 Table 1. Overview of %VOC Removals
Treatment
R. E. Wright: bioventing1
ENSR Larsen: intermittent
aeration biopile
ENSR Larsen: continuous
aeration biopile
ELI/SBP: UVB
(after 14 months of
operation)1
Acetone
-
72
83

MEK
-
59
71

MIBK
-
60
79

DCE
-
45
57

TCE
-
88
99

PCE
-
22
45

Toluene
-
43
51

Note: 'Data not provided
5.4 General

Insufficient evidence was collected from any of the treatments to conclusively determine the relative contributions
of biodegradation and volatilization to contaminant removal. However, monitoring of indicators of biological activity
implied that degradation was taking place. Significant amounts of DCE, TCE, and toluene were collected from the
offgases from the biopile and bioventing systems.  Mass balance determinations indicated generation of DCE,
presumably through biotransformation of TCE during both of these treatments. Smaller amounts of TCE and DCE
were collected for the UVB treatment, and DCE accumulation was not detected.  Evidence of in situ biodegradation
                                                  133

-------
NATO/CCMS Pilot Study, Phase II
Final Report (Appendix IV)
was also more scant for the UVB treatment. The system design does not permit easy on line monitoring of O2 use
and CO2 generation.

The heterogeneous nature of the  site posed particular problems in both monitoring and performance assessment.
General site information, such as direction of groundwater flow, did not always reliably predict the local conditions
encountered the test plots.

Over the course of the demonstration, elevated levels of vinyl chloride (a degradation product of TCE) were noted
in the groundwater downgradient of the test sites. It was  not clear whether or not this was associated with the in situ
treatment demonstrations taking place.

6. COSTS

The cost and  duration of treating the estimated 9,800 m3 of contaminated material at the Sweden 3  Chapman site
were estimated by the vendors as follows, based on the demonstration tests (Table 2):

                               Table 2. Demonstration Test Treatment Costs
Technology
R.E. Wright
ENSR Larsen
ELI/SBP UVB
Total Cost
(U.S.S)
For 8,100 m3,
$381,000
For 9,800 m3,
$695,000
-
Cost per m3 (U.S.S)
$52 ($39.50/yd3)
$71 ($54.40/yd3)
$240 ($185/yd3)
Duration
12 months
40 weeks
14 months
The cost basis was not the same for each estimate. For example, the R.E. Wright estimate included costs associated
with permitting and the UVB estimate did not.
                                                  134

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 44
                 Enhanced In Situ Removal of Coal Tar: Brodhead Creek Superfund Site
   Technical Contact:
   John Banks
   U.S. Environmental Protection Agency
   Region 3 (MS 3HS-22)
   841 Chestnut Street
   Philadelphia, PA  19107

   tel: +1/215-566-3214
   fax: +1/215-566-3001
   E-mail: banks.john-d@epa.gov
Country:
United States of
America
Project Status:
Accepted by study  1993
Interim report       1994
Final report         1996
1. INTRODUCTION

Past waste management practices at a former gasworks site at Brodhead Creek resulted in the contamination of soil
and groundwater with polycyclic  aromatic  hydrocarbons  (PAHs) and  BTEX  compounds (benzene,  toluene,
ethylbenzene, and xylenes). During plant operation, waste coal tars were disposed onsite in an open pit. Over time,
the  coal tars migrated into the subsurface where they collected in a natural depression formed at the geological
boundary between coarse gravels and silty sands. Site investigation identified a layer of free product coal tar covering
an area of 0.12 ha which was shown to be polluting an underlying aquifer and a nearby river. A larger 3-acre (1.22-
ha)  area of residual contamination contains dissolved contaminants.

Remediation efforts have focused on the free product because it  was believed to represent the major source of
groundwater pollutants. After construction of a slurry cutoff wall to protect the river from further pollution, it was
decided to remediate the free product area. Due to unspecified problems, excavation of the contaminated ground was
not  considered practical, and an innovative in situ treatment was selected. This Pilot Study project reported on the
remedial operations that were initiated in July 1995 with an expected completion date of March 1996. The site was
remediated  under the U.S. Environmental Protection Agency (USEPA) Superfund Program.

2. PROCESS CONCEPT

The selected technology known as  Contained Removal of Oily Wastes (CROW) is a thermally-enhanced in situ
recovery process. It is a technique developed from the recovery of petroleum from oil  shale deposits. It uses hot
water injected into the contaminated region to reduce the density and viscosity of tar deposits to a level where they
can be pumped to the surface for further treatment and disposal. The rate at which the heated water is injected into
the ground is used to control the displacement and temperature of the tar material so that it is forced towards a series
of extraction wells. Lateral containment is achieved by carefully controlling injection and extraction rates to isolate
the affected area hydraulically.  A layer of cooler water above the area of active recovery prevents mobilized fluids
from migrating vertically. This cooler water also absorbs volatile contaminants if they are released.

3. TREATABILITY STUDIES

Preliminary  treatability  studies on  coal tar samples were used  to  determine  the optimum  temperature  and
injection/extraction rates required for full-scale operation. These  studies indicated that coal tar  in soils could be
reduced to a residual saturation of 60-70%.

4. FULL-SCALE APPLICATION

Full-scale operation was based  on a pattern of six injection wells surrounding two extraction wells in the center of
the  contaminated zone. Screens were used in each well  to focus injection and  extraction flows at the depth of
contamination. Plans in the original design included injection of heated water (93°C) at a rate of 378 L/min for
recovery of water/coal tar mixtures at a rate of 435  L/min. However, shortly after full-scale operation began, much
                                                  135

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

lower injection rates were observed than had been anticipated. Iron precipitate that formed around the injection well
screens  significantly reduced  operational performance.  Although this problem was partially solved by installing
agitators in the injection wells, flow rates remained low (114-132 L/min). The lower flow delivered less energy than
planned to the aquifer, so the system operated at 71°C instead of the expected 93 °C. A water heater designed for
high flows did not heat the water to the required temperatures at the lower flows. Replacing the heater solved this
problem. Also, the reduced flow rate and lower operating temperature extended the treatment times from four to eight
months. A higher injection temperature of 96°C was achieved in January 1996 (presumably due to installation of the
new heater).

Specifications for performance required treatment to continue until results  showed that there was less than 0.5%
additional recovery of coal tar per pore volume of water flushed through the contaminated zone. During December
1995, a  cumulative flow of 3.1 times pore volume of water resulted in recovery of 602 L of tar. This was based on
previous operational experiences at a similar site where achieving this specification meant that 98.5% recovery of
recoverable tar had been attained.  Planned groundwater sampling will assess the  effectiveness of the treatment.

The report did not mention how extracted liquids were treated or disposed. However, the proposed treatment involved
pumping recovered waste product to storage tanks for further treatment such  as flotation and settling to recover light
non-aqueous phase liquids (LNAPLs) and dense  non-aqueous phase liquids  (DNAPLs). Most of the clarified water
was to be reheated and reinjected, but the rest was to be treated biologically, then passed through a granular activated
carbon filter before being discharged to the surface.

In summary,  hot  water was injected into the subsurface over 7 months to  remove free coal tar.  The project was
considered  successful  although it did not achieve  expected operating conditions  and  recovery  of contaminants.
Sampling groundwater in the future will assess the  overall effectiveness of  the treatment.
                                                   136

-------
NATO/CCMS Pilot Study, Phase II
      Final Report (Appendix IV)
   Project No. 45
                           This was an accidental replication of Project 15
   Technical Contact:
                                               Country:
Project Status:
                                               137

-------
NATO/CCMS Pilot Study, Phase II
      Final Report (Appendix IV)
   Project No. 46
                                         Project withdrawn
   Technical Contact:
                                               Country:
Project Status:
                                              138

-------
NATO/CCMS Pilot Study, Phase II
                           Final Report (Appendix IV)
    Project No. 47
                                In Situ Electro-osmosis (Lasagna   Project)
   Technical Contact:
   Dr. B. Mason Hughes
   Monsanto Company
   800 N. Lindberg Blvd (02A)
   St. Louis, MO  63167
   Washington, DC

   tel: +1/314-694-1466
   fax: +1/314-694-6507
   E-mail: b.mason.hughes@monsanto.com

   Dr. Michael Roulier
   U.S. Environmental Protection Agency
   National Risk Management Research Laboratory
   26 W. Martin Luther King Drive
   Cincinnati, OH 45268

   tel:  +1/513-569-7796
   fax:  +1/513-569-7879
   E-mail: roulier.michael@epa.gov
Country:
United States of
America
Project Status:
Accepted by study 1995
Interim             1997
1. INTRODUCTION

The Lasagna™ process is intended to overcome the significant technical problem of delivering treatment agents to
contaminants in low-permeability soils, and render in situ treatments more feasible in such soils. The technology
provides an integrated in situ remedial process by combining electroosmosis to move dissolved contaminants, and
treatment zones to degrade or adsorb the contaminants.

In 1994, a consortium consisting of Monsanto, E.I. Dupont de Nemours (Dupont) and General Electric signed an
agreement with the USEPA to jointly develop an integrated, in situ remedial technology referred to as the Lasagna™
process. In 1995,  funded in part by the U.S. Department of Energy, the vertical configuration of the process was
tested at the Paducah Gaseous Diffusion Plant in Paducah, Kentucky.

The project was accepted into the NATO Pilot  Study to recognize the project's partnership of government and
industry participants and to show the potential of a relatively low-cost technology to treat water-soluble organic
compounds, inorganics,  and mixed wastes occurring in low-permeability soils.

2. TECHNICAL CONCEPT

This in situ process can be operated in horizontal or vertical configuration but does not require injection or extraction
wells. The vertical configuration consists of outer layers that act as either positively- or negatively-charged electrodes.
Vertical treatment zones lie between the outer electrodes. Sheet piling,  trenching and slurry walls can be used to
create vertical treatment zones. In the horizontal configuration, treatment layers installed by hydraulic fracturing or
related methods occur between the upper and lower layers that make up the electrodes. These electrodes may contain
graphite or other granular electrically-conductive  materials.

The Lasagna™process depends on electroosmosis, treatment zones, and switching electrical polarity. Electroosmosis
relies  on a  low voltage electrical current applied  to the electrodes to create an electric field to flush contaminants
from low-permeability soil into or through the treatment zones. These zones, installed in close proximity through the
contaminated material, contain appropriate materials (e.g., sorbents, catalytic agents, microbes, oxidants, buffers, etc.,)
to sorb or degrade contaminants.  Placing the  treatment  layers close to each  other minimizes  the  time for
electroosmosis  to move liquids from one  layer to the next. Laboratory tests show that  water movement by
                                                   139

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

electroosmosis is very slow, typically about 25 mm/day. Switching electrical polarity reverses liquid flow and appears
to increase the efficiency of contaminant removal as well as allowing complete sorption and degradation by passing
contaminants through the treatment zones several times. Reversed flow minimizes the high pHs and low pHs that
occur at the cathode and anode respectively when the system is operated in one direction.

Characteristics of the site and contaminants determine the configuration. In general, the vertical configuration is more
applicable to shallow contamination within around 50 ft (15 m) of the surface. The horizontal configuration works
best for deeper contamination.

3. CCMS CASE STUDY

3.1  Overall Program

The NATO/CCMS case study was limited  to the field demonstration of electroosmosis,  using  the vertical
configuration, to transport and adsorb TCE at the Paducah Gaseous Diffusion Plant. The case study was carried out
in two phases:

  •  Phase  I,  conducted  between January and May  1995, evaluated the overall effectiveness of  coupling
    electrokinetics and carbon adsorption treatment zones.

  •  Phase Ha, commercial-scale demonstration conducted in 1996  examined iron filings in the treatment zone to
    dehalogenate TCE. This work was conducted to depths of up to 15m and with a wider spacing between the
    treatment zones.

The full-scale Phase II cleanup  (not included in the CCMS study)  will then take  place if the Phase Ha tests
successfully reduce TCE levels in soil to  5.6 mg/kg.

3.2  Study Site

The site covers an  estimated area  of 6,000 ft2 (557 m2)and extends to a depth of 45 ft. TCE concentrations are as
high as  1,523  mg/kg, but  average around  84 mg/kg.  Concentrations increase with depth, and the highest
concentrations are believed to occur at depths between 20-30 ft (6.1-9.1 m). The contamination resulted from cylinder
drop tests conducted at the site in the  1960s  and 1970s  to test the structural integrity of steel cylinders used to
transport and store uranium hexafluoride.  TCE was used  as a refrigerant during the tests contaminated the soil.

3.3  Phase I Study

The field study for remediation of chlorinated solvents, conducted at the Paducah Gaseous Diffusion Plant, lasted
for  120 days and ended in May 1995. A key objective was to successfully demonstrate coupling electroosmosis to
flush TCE from the clayey soil (hydraulic conductivity <10"7 cm/sec) and adsorption to remove the contaminant from
the  pore water. The test site was 15 ft x 10 ft (about 4.6 x 3 m) and 15 feet deep  (4.6 m). A control area was built
next to the test area and isolated from it hydraulically. The vertical configuration tested at the site consisted of steel
panel electrodes and treatment zones made of wick drains containing granular activated carbon.

A direct current of around 0.4 volt/cm applied to the electrodes caused groundwater to flow from the anode to the
cathode at a rate of about 13 mm/day. The induced pH gradient caused problems  such as soil drying and cracking,
and metal and mineral deposits at the cathode. Pumping water from the cathode to the anode reduced these problems.

During the 120-day demonstration, the area was covered  with plastic. A vent fan directed soil offgas to an in-line
filter for TCE analysis. Operating parameters during this  period were:

  •  Power after one month:           105 volts, 40 amperes

  •  Electroosmotic flow rate:         4-5 L/hour

  •  Average soil temperature:         25-30°C


                                                  140

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

Soil samples collected throughout the demonstration site before and after the test showed that the process removed
98%-99% of the TCE from the tight clay. TCE levels in the soil were reduced from 100-500 mg/kg to an average
of 1 mg/kg. Sampling and analyzing the carbon revealed how much dissolved TCE was adsorbed in the treatment
zones and provided mass balance data. These carbon samples accounted for around 50% of the original TCE. The
unaccounted TCE may be attributed to passive diffusion (5%), evaporation (5%), in situ degradation of TCE, non-
uniform distribution of the contaminant in the soil, or incomplete extraction of the compound from the activated
carbon before analysis.

Results suggest the process is effective for removing residual dense non-aqueous phase liquid (DNAPL) as well. At
most soil sampling locations with TCE concentrations of greater than 225 mg/kg indicative of residual DNAPL in
soil pores the process  reduced these levels to less than 1  mg/kg.

3.4 Phase Ha Study

The success of the Phase I study led a more extensive field investigation incorporating reactive treatment zones such
as iron filings to  destroy TCE in situ. The  first part (Phase Ha)  of this  two-stage  (Phase Ha and  Phase II)
investigation conducted  on 20  times  more soil than was treated in the preliminary field investigation will try to
resolve scale-up questions, verify cost estimates  for treatment, and evaluate how the zero-valent iron  performs.
Preliminary results of the Phase Ha demonstration at the site in Kentucky show that treatment zones with iron filings
can dechlorinate TCE, producing relatively innocuous end-products such as chloride ion, ethane and ethene. Other
potential intermediate products like dichloroethene and vinyl chloride are associated with the surface of the filings.

4. POSSIBLE TECHNICAL LIMITATIONS

Extended operation of electroosmosis could dry out soil, change pH in the soil, and precipitate metals and minerals
near the negatively-charged electrode eventually stopping the process. Reversing the polarity of the electrodes causes
the water to flow in the opposite direction and may resolve these problems. Pumping the  high pH water to the
surface and recycling it to the anode may provide a convenient way to maintain neutral conditions at the electrodes.

5. RESIDUALS AND EMISSIONS

Residuals and emissions include offgases resulting from evaporation, and the treatment layers. At the Kentucky test
site, TCE losses  by evaporation accounted for 5% of the mass balance. These off-gases did not require treatment.
If treatment layers are determined to be a hazardous waste, removing and disposing of them may be an  issue.

6. COSTS

An engineering evaluation and cost analysis for  the vertically configured process estimates a treatment cost of
U.S.$40-90/yd (U.S.$52-118/m3) of clay soil containing TCE at a depth of 40-50 feet and over an area of 1-2 acres
(0.4-0.8 ha). With optimized electrode spacing, improved ability to install treatment zones and electrodes at closer
spacing, and mass-produced prefabricated materials resulting from wider use  of the technology, costs are expected
to fall to U.S.$20-40 per cubic yard U.S.$26-52/m3). These costs exclude those for analysis, waste disposal, etc.

1. FUTURE DEVELOPMENTS

Work on the horizontal configuration has focused on developing long-lasting electrical and fluid connections between
the electrodes and the treatment zones and preventing gases from building up in the electrodes. The University of
Cincinnati is conducting laboratory and field research on hydrofracturing and biodegradation to develop the horizontal
configuration. Laboratory investigations are focusing on optional treatment processes for other types of contaminants,
such as DNAPL, heavy  metals, and mixed wastes.

In a horizontal unit in the field, the treatment zone consists of granular activated carbon seeded with a microorganism
and nutrients. This microorganism, isolated from a group of organisms, was chosen for its ability to degrade TCE.
This test, conducted in the spring and summer of 1996, examined how well methanotrophic microorganisms survived
in a treatment zone under the conditions required for electroosmosis to occur.
                                                   141

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

8. CONCLUSIONS

The Lasagna™ process offers promise for treating water-soluble organic and inorganic contaminant, and mixed wastes
in low-permeability soils as well as in groundwater. For highly non-polar contaminants, surfactants introduced into
groundwater or incorporated into treatment zones will solubilize the organic compounds. The process has been shown
to be effective in treating residual DNAPL as well. Larger scale demonstrations need to be conducted to confirm the
effectiveness of using zero-valent iron to  degrade  contaminants. The horizontal configuration using  biological
treatment zones needs to be evaluated as well.
                                                  142

-------
NATO/CCMS Pilot Study, Phase II
      Final Report (Appendix IV)
   Project No. 48
                                         Project withdrawn
   Technical Contact:
                                               Country:
Project Status:
                                              143

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 49
                  Characterization of Residual Contaminants in Bioremediated Soil and
                                       Reuse of Bioremediated Soil
    Technical Contact:
    Bernhard Hammer
    Federal Office of the Environment, Forests
    and Landscape (BUWAL)
    Federal Department of the Interior
    Laupenstrasse 20
    3003 Bern

    tel: 41/31-322-9307
    fax: 41/31-382-1546
Country:
Switzerland
Project Status:
Accepted by study  1995
Final report         1997
1. INTRODUCTION

Certain residual pollutants remain after soils contaminated with oil products have been bioremediated. The emission
levels to be expected during the reuse of remediated soil were estimated by means of laboratory and field tests. These
were so low that the risk of harm to protected materials may be considered extremely low. Nevertheless, the validity
of this claim would need to be checked in each individual case.

2. BACKGROUND AND PROJECT OBJECTIVE

Mineral oil products are major pollutants found at contaminated sites in Switzerland. The revised Umweltschutzgesetz
(USGrev, the Environmental Protection Act) and the relevant ordinances (e.g., Technische Verordnung uberAbfalle,
the Technical Waste Ordinance) require excavated material contaminated with pollutants to be treated, wherever
technically feasible, ecologically beneficial and economically viable. Thus, whether the soil has to be treated or
dumped depends largely on the quality of the products obtained from the treatment process and the options for reuse.
Biological treatments leave residual contaminants in the soil whose potential risk to the environment is largely
unknown.

Since the unsaturated zone of subsoil has not been defined as a protected material in Swiss legislation, no tolerance
values or limits are provided in the corresponding ordinances for pollutants in the unsaturated zone.  As a result,
Swiss environmental legislation can only  offer limited guidance in  the evaluation of soil contaminated with oil
products.  The  limit  for  hydrocarbons in  drinking  water  is specified  as  0.02  mg/L  in  the  Fremd-und
Inhaltsstofjverordnung  (FIV, the Ordinance  on Foreign Substances and Constituents) and the Schweizerisches
Lebensmittelbuch (SLB, the Swiss Food Book). As regards the dumping of waste, the Technical Waste Ordinance
does not specify any limits for hydrocarbon levels in dumped material, but does set limits for the leachate of inert
substances (0.5 mg HC/L) and residual substances (5.0 mg HC/L). The quality target for running waters is 0.05 mg
HC/L. Wastewater with an HC content of up to 10 mg/L may be introduced into surface waters, while those with
an HC content up to 20 mg/L have  to pass through a wastewater treatment plant.

The lack of a legal basis for  evaluation  has prompted a number of cantons to  define their own tolerance limits for
the assessment of excavated material. Thus, for example, a tolerance limit of 50  mg HC/kg applies to uncontaminated
excavated material in the  canton of Zurich. Material with HC levels  exceeding 500 mg/kg may not be dumped in
inert material landfills. However, the HC residual content of bioremediated material can range from 100-5000 mg/kg,
depending on the oil product involved, the initial content, the degree of weathering,  and the soil type. This means
that the soil cannot usually be freely reused after bioremediation and  must therefore be sent to a landfill. But since
bioremediation with subsequent dumping of the treated material is neither ecologically beneficial nor economically
attractive, biological remediation technologies are only used in  rare cases. The soils contaminated with mineral oil
products are much more likely to be treated by a soil washing process, used as an aggregate in cement works, or
depending on the canton involved, sent  directly to a landfill.
                                                  144

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

The purpose of this project was to investigate the environmental acceptability of bioremediated soil by means of a
risk analysis. The project was subdivided into the following operational stages:

  •  physical and chemical characterization of the residual contaminants;

  •  environmental behavior of the residual contaminants;

  •  effects of the residual contaminants on the environment; and

  •  evaluation of the environmental acceptability by means of a risk assessment.

3. CHARACTERIZATION OF RESIDUAL CONTAMINANTS

First of all, various extraction methods were evaluated with the aid of reference substances. Soxhlet extraction with
tetrachloromethane proved to be the most effective method and was therefore used for all subsequent investigations.
Infrared spectroscopy (according to DIN 38'409-H18) and gas chromatography were used to quantify the HC content.

        Infrared spectroscopy (according to DIN 38'409-H18) (1);

        Gas chromatography with a hydrocarbon mixture as the calibration solution; and

        Gravimetric assay (according to DIN 38'409-H17)  (2).

Similar total hydrocarbon (THC) contents in bioremediated soil material were obtained for all three methods. Infrared
spectroscopy was subsequently used routinely, while the other quantifying methods were employed primarily for
comparison purposes.  With each measurement a differentiation was made between the total solvent extractable
material (TSEM) and the total petroleum hydrocarbon (TPH) content. The TPH was obtained by using (alumina) to
separate all the compounds with polar functional groups (e.g., acid-alcohol groups) contained in the TSEM.

The remediated soil material used for all tests originated from various minor  oil spills (primarily involving EL
heating oil) that was combined before being sent to a bioremediation plant. The  contaminant content in the fine
material (<2 mm) was 780 mg/kg of TSEM and 430 mg/kg of TPH, in  each case  related to the dry content of the
sample. The content of polycyclic aromatic hydrocarbons (PAHs) was below 2.8 mg/kg and was not subsequently
measured.

The residual contaminants in the bioremediated soil were chemically analyzed and characterized. These residual
contaminants were then separated into individual compound classes on the basis of their polarity using two different
methods: column chromatography on  silica gel and a two-dimensional  separation by means of high performance
liquid chromatography coupled with gas chromatography. Most of the residual contaminants consisted of apolar, low-
volatility compounds (boiling points >280°C). Small quantities of polar compounds (e.g., fatty acids and long-chain
alcohols) were also detected in individual fractions. On the whole, however, the compounds were very hydrophobic,
with a very high octanol/water partition coefficient (Kow) of over 106.

4. ENVIRONMENTAL BEHAVIOR OF THE RESIDUAL CONTAMINANTS

The second phase of the project involved the investigation of the environmental behavior of the residual contaminants
in the bioremediated soil after being applied as topsoil. The processes investigated were biodegradation, teachability
due to precipitation, and volatilization.

4.1 Leachability of the Residual Contaminants

The teachability of the residual contaminants through percolating water was investigated in laboratory tests by means
of shake and column leachate tests and by means of a lysimeter test outdoors. The initial THC content was 0.09 mg
THC/L, as measured by both the column and lysimeter tests, and rapidly declined thereafter. After approximately
60% (column tests) and 95% (lysimeter test) of a theoretical annual precipitation, the THC content in the percolating
water was already below the drinking water limit of 0.02 mg/L. The leached quantity of pollutants corresponded to


                                                  145

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

0.1% of the THC content in the soil material. Extrapolation of these measurements indicated that only approximately
1% of the residual contaminants would leach out even after one hundred years' precipitation.

Similar percolating water contents were  determined for a given water/solids ratio in both test designs. With the
German DIN-S4 leachate test, aTPH content of 0.1 mg/L was obtained with a water/solids ratio of 10:1. This figure
was above the percolating water contents of approximately 0.09 mg TPH/L measured at the start of the column tests.
It was also  approximately 10 times  higher than the average THC content in percolating water with the same
water/solids ratio. This difference is probably due largely to the break-up of lumps of soil caused by shaking in the
DIN-S4 leachate test and the resulting greater leachability of the residual contaminants. The DIN-S4 leachate test
can therefore be considered as a highly  conservative test that greatly overestimates the actual percolating water
contents in the field.

Another important parameter is the dissolved organic carbon content (DOC). This includes all the polar and apolar
organic compounds dissolved in the water. Whereas the DOC of 4.3-15 mg/L in the bioremediated soil was similar
to that found in gravel (6 mg/L), the DOC  of plant earth was much higher at a level of 107 mg/L. These results
indicate that a large part of the soluble compounds in all the investigated materials consisted of polar compounds
(lipids, humic substances, etc.). Experience has shown that the DOC is  an indicator of the degree  of microbial
degradation of a material, as demonstrated by the high DOC levels in compost, sewage  sludge, or spinach. A low
DOC, on the other hand, means that low oxygen depletion can be expected  in groundwater.

4.2  Air-borne Emissions

Air-borne emissions were investigated by gas chromatography  (atmospheric measurements). Since the residual
contaminants exhibited a very low vapor pressure (<10"6) any air-borne emissions can be considered to be negligible.

4.3  Degradability of the Residual Contaminants

The degradability of the residual contaminants was investigated in laboratory degradation tests. Under optimized
conditions, the remediated soil material was incubated, and both the TPH content and CO2 formation were monitored.
During the test period of 3 months, the residual contaminant content decreased by approximately 20%; approximately
85% of this decrease in TPH content was converted into CO2.

A CEC degradation test was  conducted  in  order to investigate possible  further biodegradation of these residual
pollutants in aqueous solution. After 21 days, the  quantity of residual contaminants present at the start of the test had
declined by about 20%. This indicates that the bioavailability was either greatly reduced as a result of the low water
solubility or else that these residual contaminants were poorly degradable.

The long-term biodegradation was  investigated in a 28-month field trial under near-natural conditions. Accordingly,
approximately 2,000 m3 of remediated soil was homogenized and applied as a 1-meter-thick layer. Three repetitions
of four types of cultivation were investigated: (1) planting with red clover; (2) planting with rye-grass; (3) fallow;
and (4) soil tillage with regular plowing. Twenty samples from 0-25 cm and 25-50 cm were taken from each plot
of land and composited. The average reduction in concentration of the residual contaminants over these 28 months
was approximately 13%, and no significant differences were observed between the individual cultivation types. Since
only small quantities of residual contaminants leached out or were emitted into the air over the entire trial period,
the  reduction in the TSEM and THC  contents can be attributed to biodegradation.

5. EFFECTS OF THE RESIDUAL CONTAMINANTS ON THE  ENVIRONMENT

5.1  General

Relatively little  is known about the mammal toxicity of individual hydrocarbon compounds. Given the metabolism
of hydrocarbons in mammals, n-hexane can be considered as a relevant substance in lexicological terms; the toxicity
of other alkanes (e.g., n-heptane, i-octane, etc.) is much less relevant. Since low molecular weight hydrocarbons (e.g.,
n-hexane) are no  longer present in bioremediated  soil, the residual hydrocarbons are probably  not very relevant
lexicologically. Higher molecular weight hydrocarbons and other compounds observed in remediated soil (fatty acids
                                                   146

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

and alcohols) are also naturally occurring compounds (e.g., in plant material) and are therefore not considered to be
lexicologically relevant.

5.2 Effect on Plants and Soil Organisms

The germination and growth of 36 plant species on both remediated soil and plant earth (control) were investigated
in greenhouse trials. The plants developed normally on both soil materials and showed no phenotypical changes of
any kind.

In a second series of tests the effect of the  residual contaminants on the plant yield of eight crops and eight wild
herbs grown with plant earth that had been  deliberately contaminated with residual pollutants (TSEM dissolved in
1,1,2-trichloro-trifluoroethane and sprayed on the soil material). Uncontaminated material and material treated with
solvent were employed for reference purposes. After 30 test days the plants were harvested and the yield (fresh
weight) determined gravimetrically. On average,  a 25% increase in plant yield was observed in the contaminated
plant earth, compared to the control plants. Although the effect was not statistically significant given the considerable
variation, these results show that the contaminants added to the  soil did not inhibit growth.

Since a wide variety of plant species was used, adverse effects on other plant species are unlikely. This was further
confirmed by the diverse plant communities that grew on the fallow fields. Normal growth was also  observed in the
plants employed in the field trials, and the typical root nodules of the red clover were well developed. Uptake of the
residual contaminants by the plants was not  investigated since similar compounds also occur naturally in plants (see
above) and higher molecular weight hydrocarbons are absorbed  by the plants only to a very limited extent.

While the effect on soil organisms was  not specifically investigated, a high worm density was observed in the field
trial, particularly on the cultivated plots.

5. 3 Effect of soil leachates on water-fleas and bacteria

Dilution series of DIN-S4 leachates and percolating water samples from the above-mentioned lysimeter tests were
used to investigate the effect of aqueous soil leachates on the water-flea (Daphnia magna) and on the bioluminescent
bacteria Vibrio fischeri.  However, no significant acute toxic effects on these aquatic organisms were  observed.

6. RISK ASSESSMENT FOR REMEDIATED  SOIL

A risk assessment based on the  data from the tests described above was performed. A mass flow  analysis was
conducted in order to determine the emissions expected during the first year after the application of bioremediated
soil material as topsoil.  Leaching into  groundwater, volatilization, uptake by cultivated plants, and losses due to
biodegradation were all taken into account.  Over 99% of the measured reduction in THC content from 430 mg/kg
to 407 mg/kg was attributable to biodegradation. Leaching produced the highest emission of approximately 0.4%.
However, since the leaching rate showed a clear decline in the second year after application of the  remediated soil
material, almost no emissions would likely be detected after the first year. Those residual contaminants that are
soluble in percolating water should be more  readily  degradable than  the  insoluble residual contaminants, thus
resulting in reduced groundwater contamination.

To conclude, these results can be illustrated quantitatively taking the example of n-hexane  as a "worst case" scenario.
Assuming that the  measured emissions  in percolating water relate to n-hexane, Table 1  shows that leachate levels
of 0.11-0.15 mg/L would be expected. This would exceed the SLB drinking water threshold of 0.02 mg/L. However,
if the maximum permitted level of  n-hexane is consumed by an adult in drinking water (70 kg bodyweight; 2 L
drinking water/day intake) with a reference dose (RfD) level of 0.06 mg/kg/day) this gives a permitted concentration
of 2.1 mg/L. This corresponds to one hundred times the SLB threshold and more than ten times the measured initial
values in percolating water. These calculations demonstrate that this evaluation of potential risks based on leachate
tests and compared with drinking water values is very conservative, particularly since the  actual concentrations
decline rapidly (Figure 1) and this percolating water undergoes further dilution before being used as drinking water.

The exposure to humans and the environment from emissions is so low that, considering the toxicity of these residual
contaminants,  no adverse effects can be expected  for humans or the environment.


                                                   147

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

Estimates of potential emissions for various applications (backfilling, filling of trenches, etc.) and groundwater
conditions indicate that protected goods would not be impaired, even if reused in sensitive locations (e.g., protected
watersheds).

7. CONCLUSIONS

The residual contaminants in successfully bioremediated soil are hardly soluble, difficult to volatilize, and undergo
further biodegradation only over prolonged periods. Since the anticipated emissions are very low, no adverse effects
to humans or the environment are expected. Nevertheless, bioremediated soil should not be evaluated on the basis
of the residual content of THCs. Rather, maximum emissions should be used as a basis for evaluation, and these can
be determined according to the planned reuse of the soil material. Leachate tests and gas volume measurements are
suitable for determining  such emissions  and should include,  in addition  to the aliphatic hydrocarbons,  other
contaminants of possibly  greater relevance  (e.g., PAHs).

After undergoing  quality  control measures bioremediated soil might  be  used for a variety of purposes, although
sensitive applications (e.g., use in agriculture or as fill in groundwater areas) would require prior thorough testing
for all relevant contaminants.

8. ACKNOWLEDGEMENTS

This project was supported by the Swiss National Science Foundation (NF Project No. 5001-35299) as part of the
Priority Programme Environment. The company NUVAG AG made available the bioremediated soil material and
the land for the field trials. The  greenhouse trials were made possible thanks to the support of Sandoz  Agro AG.

9. REFERENCE

Daniel Angehrn*, Rene Galli*, Mathias Schluep*,  Andreas  Haner* and Josef Zeyer**
Bioremediated Soils from Oil Spills: Waste or Product?

  *     BMG Engineering AG, Ifangstrasse  11, CH-8952 Schlieren,  Switzerland
 **     Soil Biology, Institute of Terrestrial Ecology, Swiss Federal Institute of Technology (ETH), Grabenstrasse
        3, CH-8952  Schlieren, Switzerland
                                                  148

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 50

     Integrated Rotary Steam Stripping and Enhanced Bioremediation for In Situ Treatment of VOC-
    Contaminated Soil (Cooperative approach to application of advanced environmental technologies)
   Technical Contact:
   Paul M. Beam
   U.S. Department of Energy
   EM 451 - Cloverleaf Center
   19901  Germantown  Road
   Germantown, MD 20874-1290

   tel:  +1/301-903-8133
   fax: +1/301-903-3877
   E-mail: paul.beam@em.doe.gov
Country:
United States of
America
Project Status:
Accepted by study  1996
Interim/final report   1997
1. INTRODUCTION

This project involved technology demonstrations at the Department of Energy's Pinellas Northeast Site located in
Largo, Florida, USA. Concentrations of chlorinated VOCs in the sandy surficial aquifer at the site range from 10-
1,000 mg/L. In January 1995, the following three technologies were selected for pilot tests at the Pinellas site:

    (1)  groundwater pump and treat with a pervaporation system to remove VOCs from the pumped groundwater,
        which eliminated air emissions and the need for costly groundwater pretreatment;

    (2)  in situ rotary steam and air stripping to treat the highest concentrations of VOCs in soil and reduce them
        to a level of 100 mg/kg; and

    (3)  nutrient injection to enhance in situ anaerobic bioremediation of soil with VOC concentrations of 100 mg/kg
        or less.

A potential cost savings of U.S.$5-10 million was anticipated over the proposed baseline remedial design, which was
a standard 30-year pump-and-treat system using  groundwater recovery wells and an air stripper. Furthermore, the
proposed baseline remedial design was not expected  to reduce the most concentrated areas of VOCs to below
drinking water standards.

2. PERVAPORATION TECHNOLOGY EVALUATION

Evaluation of the pervaporation pilot test was conducted from 1995-1996. Two recovery wells pumped groundwater
to the system, which used membranes for preferential permeation of VOCs.  Transport of VOC vapors through the
membrane was induced by maintaining a lower vapor pressure on the permeate side of the membrane than on the
side of the influent groundwater. This pressure difference was achieved by cooling the permeate vapor to make it
condense.

The pilot system was capable of treating 1-2 U.S. gallons (3.8-7.6 liters) of groundwater per minute. Approximately
6,250 U.S. gallons (23.6 m3) of groundwater with VOC concentrations ranging from 500-1,000 mg/L were treated
during the pilot test. Effluent contaminant concentrations  were reduced  to 1-4 mg/L under optimum operating
conditions; however, the efficiency of the system was lowered due to membrane fouling caused by the precipitation
of iron gel. Moderate success was achieved by modifying the system and adding chemicals  to reduce membrane
fouling.  The only wastes produced during the pilot test were the permeate and the spent filters used to inhibit
membrane fouling. VOCs were successfully concentrated in the permeate, eliminating air emissions. Full-scale system
capital costs for the pervaporation system to treat 20 gallons per minute are expected to range from U.S.$200,000-
$275,000; operating costs are expected to range from U.S.$10-$20 per 1,000 U.S. gallons ($2.6-$5.2/m3) of treated
water.
                                                 149

-------
NATO/CCMS Pilot Study, Phase II                                          Final Report (Appendix IV)

3. ROTARY STEAM AND AIR STRIPPING TECHNOLOGY EVALUATION

Evaluation of the rotary steam and air stripping pilot test is currently being conducted.  The stripping system injects
steam and hot air through a rotating auger to volatilize VOCs sorbed onto soil particles. The off-gases are transported
by the injected steam and air to a metal hood at the ground surface where they are treated using a catalytic oxidation
system for the destruction of contaminants and an acid-gas scrubber to eliminate air emissions.

Thus far, the system has effectively reduced the concentrations of chlorinated VOCs in soil from 1,000-6,000 mg/kg,
to 100-300 mg/kg, which was the goal for this site. However, several operational problems were experienced with
the total system. In particular, the  catalytic oxidation  system  was unable  to handle the quantity of the vapors
generated by the stripping system. As a result, some vaporized  VOCs initially escaped from the metal hood. This
problem was subsequently controlled by reducing the injection pressures. The only wastes generated by the system
were from the scrubber. The initial results of the pilot test indicate that the operational costs will range from U.S.$70-
200/yd3 ($90-260/m3).

4. NUTRIENT INJECTION TO ENHANCE 77V SITU ANAEROBIC BIOREMEDIATION

The in situ anaerobic bioremediation system recently began operation and has not yet been evaluated. The system
involves a series of horizontally installed wells and infiltration galleries to control the hydraulic gradient and supply
nutrients to the contaminated zone.
                                                  150

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 51
                             Sobeslav, South Bohemia Wood Treatment Plant
   Technical Contact:
   Marek Stanzel
   KAP s.r.o
   Skokanska 80
   169 00 Prague 6
   Czech Republic

   tel: +420/2-2431-3630
   fax: +420/2-5721-1255
   E-mail: kappraha@login.cz
Country:
Czech Republic
Project Status:
Accepted by study  1996
Interim report       1997
to be continued in Phase III
Since the 1870s, a wood processing plant has been operating on the outskirts of the town of Sobeslav, located 120
km south of Prague, Czech Republic. Wood products such as railway sleepers and telegraph poles were treated at
the Sobeslav wood processing plant either by immersion in boiling tar ("black" impregnation) or by coating with a
mixture of heavy metal salts ("white" impregnation).

In 1990, following a decision to install and upgrade existing process equipment, a site investigation was performed
to determine the extent of soil and groundwater contamination resulting from industrial operations at the plant. This
investigation was completed  in December  1996.  Soil and groundwater contaminated with polycyclic  aromatic
hydrocarbons (PAHs)  and metals were identified in an area of approximately 1-2 km2. Contamination at the site
resulted from on-site disposal of tar sludges  in poorly lined pools, discharge of untreated effluent into the Luznice
River, and discharges  of chemicals from treated products staged at the site  prior to transportation.

Initial concerns about site pollution centered on the foul odors emanating from local drinking water wells. The odors
prompted an investigation by the local environmental health authority, which examined the results of biochemical
oxygen demand (BOD), chemical oxygen demand (COD), and non-polar extractable organic analyses of surface water
and groundwater samples. The more comprehensive investigation initiated in 1990 included a detailed site walkover
survey,  aerial imaging, inspection of plant life, installation and logging of monitoring wells, chemical sampling of
soil and groundwater,  and pump testing to determine subsurface  permeability. Although originally  intended to be
completed in 14 months, the investigation was extended to 35 months because of financial, analytical, and regulatory
difficulties.

During the walkover survey, the principal sources of contamination were identified as a leaking creosote oil storage
tank and seepage from unlined disposal lagoons in the southeast part of the site. Nearly 100 boreholes were drilled,
and  sandy to sandy clay soils overlying sedimentary clays were found to underlie the  site.  Monitoring of  the
groundwater indicated a light non-aqueous phase liquid (LNAPL) layer up to 0.7 m thick at  the site. Extensive
analysis of soil and groundwater samples confirmed elevated concentrations (unspecified) of PAHs, heavy metals,
and phenols.

An initial investigation of remedial options was carried out by a German consultant, which studied the effectiveness
of biotreating the contaminated soils. Results showed that an initial decrease in PAH concentrations was  followed
by a greater increase  to  concentrations destroying the degrading organisms.  The study concluded that microbial
activity had caused rapid desorption of contaminants and increased bio-available concentrations to intolerably high
levels. Although more success was achieved with bench-scale testing using in situ groundwater biotreatment,  the
technique  was  unsuccessful during  field-scale tests.  As a  result, it  was concluded that soil encapsulation in
combination with pumping and treating groundwater using gravel filters was the most cost-effective remediation
solution.
                                                  151

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 52
                                        Permeable Treatment Beds
   Technical Contact:
   Prof. Dipl.-lng.  Harald Burmeier
   University of applied  Studies and Research
   Herbert-Meyerstrasse 7
   29556 Suderburg
   Germany

   tel: 49/5103-2000
   fax: 49/5103-7863
   E-mail: h.burmeier@t-online.de
Country:
Germany
Project Status:
Accepted by study  1996
Interim             1997
to be continued in Phase III
Conventional methods for groundwater treatment have often been based on pump and treat systems. Pump-and-treat
involves  groundwater extraction,  above-ground treatment (using established effluent treatment  processes), and
subsequent discharge of the treated water to the subsurface or to a local sewage drain. There is a growing consensus,
however, that pump-and-treat systems often incur excessive treatment costs because of their potential for continuous
operation over long periods of time. This Pilot Study project reports on the development of an alternative in situ
treatment approach, which may provide lower operating costs and enhanced treatment performance  over pump-and-
treat.

WCI Umwelttechnik GmbH are developing an in situ permeable treatment bed that can be used on a site to prevent
off-site migration of contaminated groundwater or to provide long-term source treatment of the contaminated soil
or non-aqueous phase liquid (NAPL) in groundwater. This type of treatment system has also been investigated by
the U.S. Environmental Protection Agency, the University of Waterloo (Canada),  and  the U.K. Department of the
Environment, and has also been called "active containment."

In the WCI  Umwelttechnik GmbH concept,  a permeable treatment bed is installed within  an  aquifer as  either  a
permanent, temporary, or exchangeable structure perpendicular to the migrating contaminated groundwater. As the
contaminated groundwater passes through the treatment bed, contaminants are transformed, destroyed, or immobilized
within the bed matrix. Beds based on transformation or destruction of contaminants, such as the dehalogenation of
organic solvents using zero-valent iron, are said to be reactive.  Beds that immobilize contaminants are called sorption
beds. The material from which the permeable bed is constructed is termed the filling. The physical and chemical
nature of this material is highly dependent upon the  type of contaminant and hydrogeological environment to be
treated.

The WCI Umwelttechnik GmbH system consists of an excavated ditch to which the appropriate filling is added. An
open ditch may be suitable for treatment  depths up to  12  m, while a diaphragm wall with steel caissons may be
required for  depths up to 50 m.  The  filler material can be exchanged (for  example when the sorbent is exhausted)
without the ditch collapsing by the construction of a double-wall within the ditch into which the filler material and
monitoring equipment can be placed.

A proposed system has been suggested by  WCI Umwelttechnik GmbH, although no pilot- or field-scale studies have
been carried out or proposed. Excavation of  the ditch to the appropriate depth  (usually keyed into an aquiclude) is
carried out conventionally with drawdown of the local water table as necessary. A concrete base to the  ditch is
installed  to  act as a clean and strong platform for  bed wall installation. The bed  itself is constructed from  a
prefabricated unit consisting of two vertical  slabs connected by  horizontal bars that are stacked until the height of
the ditch is reached. These prefabricated elements must have a hydraulic permeability  higher than  the surrounding
aquifer, a long-term stability to physical and chemical stresses, and low production costs. The gaps between the
installed  prefabricated units will provide sufficient overall permeability.

The dimensions of the proposed system are  governed by site-specific parameters, such as the rate of groundwater
flow and the contaminant concentration flux into the treatment  bed, as well as by intrinsic  parameters of the bed
itself, such as its sorption capacity and reaction rate. These characteristics are important to determine the residence
                                                  152

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

time required for complete  contaminant treatment. The monitoring requirements for an installed  system include
monitoring wells downstream of the treatment bed to determine performance effectiveness and periodic sampling
of filling to verify that physical and chemical properties (such as permeability and sorption capacity) still meet
operating guidelines.

Treatment cost for such a system has been suggested to lie in the range of 500-700 Deutsche marks per square meter
of bed over an operating lifetime of 50 years.
                                                  153

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 53
                        In Situ Bioremediation of Chloroethene-Contaminated Soil
   Technical Contact:
   Hennie van Vree
   Tauw Milieu bv
   Handelskade 11
   Postbus 133
   7400 AC Deventer
   The Netherlands

   tel: +31/570-699-561
   fax: +31/570-699-666
   E-mail: hw@Tauw.nl
Country:
The
Netherlands
Project Status:
Accepted by study  1996
Final report         1997
1. INTRODUCTION

In the Netherlands, between 20-30% of heavily contaminated sites have chlorinated hydrocarbons—in particular,
trichloroethene  (TCE) and tetrachloroethene (PCE)—as  a  principal  constituent. On some  sites, the depth to
contamination makes ex situ treatment technically difficult and uneconomical. Therefore, there is a need to develop
in situ remedial techniques for this type of contamination. Biotreatment is particularly  attractive since it offers the
potential for  remediation without  generating any secondary hazardous waste for further treatment or disposal.
However, biodegradation of chloroethenes is technically and microbiologically complex, and might prove difficult
to achieve under field conditions. This Pilot Study project reports on the development of an in situ biotreatment for
chloroethenes from bench-scale studies  to a pilot-scale field application.

2. TECHNICAL CONCEPT

The  Tauw Milieu approach to treating chloroethenes exploits  both aerobic and anaerobic microbial processes.
Although chloroethenes such as PCE are persistent under aerobic conditions, they can be dechlorinated sequentially
by anaerobic bacteria  to  TCE,  dichloroethene (DCE),  vinyl chloride, and ultimately  to  ethene.  Since the
dechlorination of DCE and vinyl  chloride  are the rate-limiting steps, these compounds are  accumulated under
anaerobic conditions as intermediate breakdown products of PCE. In  contrast to PCE, however, less-chlorinated
ethenes  can be co-metabolically mineralized by aerobic  bacteria. The Tauw Milieu process degrades PCE through
stimulation of bacteria in spatially separated aerobic and anaerobic zones. PCE is degraded anaerobically to TCE and
DCE, which are transported downstream to  an aerobic zone where they are mineralized to ethene.

3. LABORATORY STUDIES

By February 1996, the process had been demonstrated and verified for PCE at the bench scale. These studies showed
that under anaerobic conditions, using formate and methanol as carbon and electron-donor sources respectively, PCE
and TCE were readily dechlorinated to DCE and other intermediaries. In addition, using toluene or phenol as a co-
substrate, TCE, DCE, and vinyl chloride were readily degraded  aerobically. Flask studies with indigenous bacteria
from a proposed pilot-scale test site had similar results.

In order to evaluate degradation kinetics,  a  series of column experiments were performed using 0.7-L columns of
site soil through which 200 mL of site groundwater was recirculated. The soil column  was operated anaerobically,
spiked with PCE, and supplied with formate  or methanol  as a carbon source and electron donor.  PCE concentrations
initially  decreased  due to sorption onto column material, but dechlorinated breakdown products were noted
approximately three weeks after start-up. Although degradation  was observed using both formate  and methanol, it
was concluded that methanol would be used for the pilot-scale system because of its lower cost and more stable pH
profile.  The aerobic column was set up in a similar way to the anaerobic column. It was spiked with PCE and its
breakdown products (TCE, DCE, and vinyl chloride). The daughter  products were readily degraded with the addition
of phenol to the column, but PCE degradation was not observed.
                                                  154

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

3. PILOT-SCALE STUDY

The pilot-scale remediation study was  conducted from August 1995 to September 1996 at a former dry cleaning
facility near the town of Breda. Soil and groundwater was contaminated with PCE over an area of  1,800 m2 to a
depth of 10 m.  Concentrations of PCE in groundwater ranged up to 10,000  ug/L. The unsaturated zone was
remediated using soil vapor extraction, and the saturated zone and groundwater were remediated by a combination
of aerobic and anaerobic biodegradation.

The anaerobic biodegradation zone was created by the extraction and injection of groundwater (up to 250 mVday)
to which methanol and nutrients were added. A total of about 1,100  kg  of methanol was injected. By carefully
controlling the hydrogeological environment, anaerobic conditions were created within the recirculating "groundwater
loop." Downstream of this loop, aerobic biodegradation was supported by slow infiltration of phenol (acting as a co-
metabolite) through a series of wells. A well downstream of the anaerobic treatment zone was used to extract up to
150 m3 of groundwater per day, in order to prevent the contamination spreading and to provide water for the phenol
infiltration process. Extensive monitoring was conducted to determine the effectiveness and kinetics of the
degradation processes and to mitigate the migration of the contamination  plume. Methanol was detected in all
monitoring wells in the anaerobic zone, in concentrations ranging from  120 to 800 umol/L. As soon as the methanol
was detected, anaerobic degradation was observed. PCE was degraded rapidly to DCE; TCE,  vinyl chloride, and
ethene were present but did not accumulate significantly.

Under anaerobic conditions, PCE was dechlorinated at rates comparable to  those found in the laboratory, and
evidence of complete anaerobic degradation to ethene was found. The aerobic process also proved successful, with
complete degradation of phenol and DCE. The limiting factor associated with in situ biodegradation of chloroethenes
appears  to be the availability of suitable electron donors rather than the presence of microorganisms.

5. PROGNOSIS

The pilot study proved so successful that the site owner is continuing the remediation, which is expected to take two
to three  years.
                                                  155

-------
NATO/CCMS Pilot Study, Phase II
                          Final Report (Appendix IV)
    Project No. 54

               Treatment of PAH- and PCP-Contaminated Soil in Slurry Phase Bioreactors
   Technical Contact:
   Erik Backlund
   EkoTec AB
   Nasuddsvagen 10
   P.O.  Box 34
   932 21 Skelleftehamn
   Sweden

   tel: +46/910-333-66
   fax: +46/910-333-75
Country:
Sweden
Project Status:
Accepted by study  1996
Interim report       1997
to be continued in Phase III
1. INTRODUCTION

EkoTec, a Swedish engineering company, has carried out several full-scale remedial actions at contaminated sites
across Sweden. EkoTec currently is involved in two biotreatment projects that use slurry-phase bioreactors. This Pilot
Study project reports on the use of these reactors to treat 3,000 metric tons of creosote-contaminated soil containing
1,000 to 10,000 mg/kg polycyclic aromatic hydrocarbons (PAHs) and  100 metric tons soil contaminated with 500
to 1,000 mg/kg pentachlorophenol (PCP).

Small-scale bioreactors were used to treat approximately 1 m3 of each contaminated soil. The pilot-plant treatment
of PCP-contaminated soils lasted from October 1995 to February  1996. Treatment of creosote-contaminated soils
began in December 1995 and continued until March 1996.

The full-scale biological treatment of PCP- and PAH-contaminated soil  is scheduled for completion in 1996 and
1997, respectively. Full-scale treatment will be designed on the basis of the pilot-scale work and will use large
bioreactors with and operating capacity of 150 m3.

2. TECHNICAL CONCEPT

Biodegradation of PAHs is accomplished aerobically, while PCP  degradation uses  a combination of aerobic and
anaerobic  conditions. Reductive dechlorination of PCP is reported to occur under  anaerobic conditions, with an
aerobic phase required to mineralize the dehalogenated breakdown products. During each study, oxygen and nutrient
inflow to the reactor is carefully monitored  along with slurry pH and operating temperature. The concentration of
VOCs in the bioreactor headspace is also measured continuously to determine the extent of any gaseous emissions.
The process  water from the bioreactors is recirculated in order to maintain slurry moisture content with minimum
water amendments. The water is treated with an activated carbon filter, which will be incinerated after treatment is
completed. Where possible, operating parameters and contaminant concentrations are determined using field-portable
measuring equipment.

Full-scale  treatment will be  designed  on the basis of the pilot-scale work and will use large bioreactors with an
operating capacity of 150 m3. It is expected that soil pretreatment will be required and a vibratory screen will be used
to remove material coarser than 2-3 mm. Further screening and hydrocyclones will reduce material input for the
bioreactors to a uniform size distribution.

3. PILOT-SCALE STUDY

A 28-day respiration test followed by degradation tests at bench scale demonstrated the ability of slurry-phase reactor
to degrade of PAH compounds by more than  90%. Design and implementation of the 1 m3 pilot-scale bioreactor
concentrated on  evaluating different methods of  supplying  oxygen, stirring,  and dispersion.  Soil with PAH
concentrations of 859 mg/kg  were treated in the bioreactor,  using a commercial  mixture of nutrients and
microorganisms from the bench-scale tests. Sodium carbonate and potassium phosphate were added to control pH,
and water was added to maintain the moisture content. After 28 days, the PAH concentration decreased to 75 mg/kg.
                                                  156

-------
NATO/CCMS Pilot Study, Phase II                                           Final Report (Appendix IV)

A second phase test was run after the reactor was modified and equipped with a cooling system to maintain the
slurry at  an optimum temperature of 24-28°C. A degradation test was also carried out in a 450-L EIMCO Biolift
reactor, using the same soil and contaminants as in the EkoTec reactor. Although the contaminant concentration in
the EkoTec reactor was 313 mg/kg, compared to 164 mg/kg in the EIMCO reactor, residual PAH concentrations were
similar: 48 mg/kg for the EkoTec, and 32 mg/kg for the EIMCO. Residual concentrations declined to 22 mg/kg in
the EkoTec after  36  days and to  15 mg/kg in the EIMCO after 45 days.  Laboratory analyses documented an
anomalous increase in PAH concentrations during some phases of treatment.  This may have been  caused by
mechanical grinding that increased the surface contact area or by secondary formation of PAH compounds during
degradation.

A third pilot-scale test is underway using the EkoTec reactor and soil with an initial PAH concentration of 4,000
mg/kg. Treatment will be completed and evaluated during 1997. Results to-date have shown that the EkoTec reactor
is more reliable than the EIMCO reactor. Pyrene was more resistant to degradation than other PAHs, and a specific
growth substrate was added to the slurry to compensate. VOC emissions from the slurry were negligible. A residence
time of 28 days appears to be sufficient to achieve treatment of the PAH-contaminated soil.

Soil contaminated with 126 mg/kg of PCP was treated by composting in a static sheltered warehouse with controlled
pH. After four months, soil residual levels declined to 75 mg/kg. Soil with an initial PCP concentration of 630 mg/kg
was treated in a pilot-scale bioreactor, first anaerobically and then aerobically. After six weeks, the PCP concentration
dropped to 80 mg/kg. Results to-date demonstrate that PCP can be degraded either with composting or in a bioslurry.
Static  composting may  be accomplished  successfully	even in cold climates. Preheating process air and
encapsulating the treatment pile are important factors.
                                                  157

-------
NATO/CCMS Pilot Study, Phase II
                                 Final Report (Appendix IV)
   Project No. 55
                                     Czechowice Oil Refinery Project
   Technical Contact:
   Eric Lightner
   U.S. Department of Energy
   Cloverleaf Building (EM-53)
   19901 Germantown Road
   Germantown, MD 20874-1290

   tel: +1/310-903-7935
   fax: +1/310-903-7457
   E-mail: Eric.Lightner@em.doe.gov
Country:
United States of
America
Project Status:
Accepted by study  1996
to be continued in Phase III
This project reports on a working partnership between the U.S. Department of Energy (DOE) and the Institute for
Ecology of Industrial Areas (IETU), an independent organization under the Polish Ministry of Environmental Protec-
tion. The objective of this partnership is to foster cost-effective, innovative research and development activities to
address contamination problems in Poland, with particular emphasis on environmental characterization, risk assess-
ment, and remediation methodologies that may be beneficial to Eastern Europe. Each phase of the project aims to
demonstrate the technology and decision-making processes involved in site remediation.

The first project initiated by the partnership was the environmental characterization, risk assessment, and remediation
of the  Czechowice Oil Refinery,  which has operated for nearly 100 years near the city of Katowice in southern
Poland. The refinery uses a catalytic cracking process to refine crude oil. Wastes from the cracking process were
deposited in site lagoons, which are now filled with a thick viscous sludge. Leakage of the lagoons has contaminated
soil and groundwater with several organic compounds.

Environmental characterization of the refinery will consist of a two-phased expedited approach utilizing low-cost,
simple rapid-response technologies to obtain general site information, followed by quantitative sampling using direct
push technologies for data collection. Risk assessment will involve the development and refinement of potential
exposure scenarios. The scenarios will be combined with the expedited site characterization results to quantify
potential risks from the site to humans and the environment. The estimates of risk will be compared to appropriate
benchmark concentrations for the contaminants.

Based  on preliminary  site information, bioventing and biosparging have  been proposed to remediate the lagoons.
Bioventing involves the injection of oxygen and nutrients into the subsurface to aerobically stimulate the indigenous
microorganisms to degrade hydrocarbons to carbon dioxide and water. Vertical injection wells  will be installed
around the perimeter of the lagoons to aerate and remediate the contaminated  vadose  zone. Biosparging will be
conducted to treat the lagoon sludge. Biosparging is similar to bioventing except that the air and nutrients are injected
into a liquid, in this case a lined basin filled with process water. Sludge that has been pH-adjusted or mixed with
a surfactant will be added to the basin in batches. In the implementation of both remediation technologies, the level
of microbial activity, pH, contaminant concentrations, and rate of degradation will be carefully monitored.
                                                  158

-------
NATO/CCMS Pilot Study, Phase II
                                Final Report (Appendix IV)
   Project No. 56
                              Spolchemie a.s.—Mercury-Contaminated Site
   Technical Contact:
   Marek Stanzel
   KAP s.r.o
   Skokanska 80
   169 00 Prague 6
   Czech Republic

   tel: +420/2-2431-3630
   fax: +420/2-5721-1255
   E-mail: kappraha@login.cz
Country:
Czech Republic
Project Status:
Accepted by study  Sept. 1996
Interim study       1997
to be continued in Phase III
The town of Usti nad Labem lies in a region of North Bohemia known as the "black triangle," so named for the
severity of its environmental problems. The Spolek pro chemicke a hutni vyroba (Company for Chemical and
Industrial Production), also known as "Spolchemie," is one of the most contaminated sites in this region. Spolchemie
was founded  in 1856 for the production of chlorinated lime  and sodium bicarbonate. The company has since
produced pesticides, including DDT and "Agent Orange," and a variety of other chemicals. More than 30  separate
production facilities are known to have operated at the site, which occupies over about 500 hectares (1250 acres).

Mercury has been used in electrolysis operations at Spolchemie since 1890. Based on an environmental audit, it is
estimated that the total release of mercury at the  site has been greater than 500 tonnes. Liquid mercury has been
observed during soil excavations.

Site investigations  have been conducted near the electrolysis plant to assess mercury contamination in  soil and
groundwater. Mercury concentrations in nine groundwater samples collected in January 1996 ranged from <0.1 ug/L
to 154.1 ug/L, and concentrations in 39 soils samples ranged from <0.0001 mg/kg to 707 mg/kg. Concentrations in
all but one of the  groundwater samples exceed  0.1 ug/L, which is the Ministry of Environment's Category A
threshold concentration for mercury, and four samples exceed 5 ug/L, which is the Category C threshold.  Further-
more, nine of 39 soil samples exceed the Category A threshold concentration of 0.3 mg/kg for mercury  in soil; six
samples exceed the Category C threshold of 10 mg/kg.

Up to several  hundred micrograms per liter of chlorinated hydrocarbons and slightly elevated concentrations of zinc
and copper were also detected in samples at Spolchemie. Remedial options are being studied.
                                                 159

-------