COMMITTEE ON
THE CHALLENGES OF
MODERN SOCIETY
EPA 542-R-98-003
May 1998
www.epa.gov
www.clu-in.com
www.echs.ida.org
NATO/CCMS Pilot Study
Evaluation of Demonstrated and
Emerging Technologies for the
Treatment of Contaminated Land
and Groundwater (Phase III)
1998
SPECIAL SESSION
Treatment Walls and
Permeable Reactive Barriers
Number 229
NORTH ATLANTIC TREATY ORGANIZATION
-------
-------
NATO/CCMS Pilot Study
Evaluation of Demonstrated and
Emerging Technologies for
the Treatment of Contaminated
Land and Groundwater — Phase III
SPECIAL SESSION ON
Treatment Walls and
Rermeable Reactive Barriers
Harald Burmeier, Chairman
University of Vienna
Vienna, Austria
FEBRUARY 22 - 28,1998
May 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 (U.S. EPA). The document was funded by U.S. EPA's Technology
Innovation Office under the direction of Dr. Michael Kosakowski. The Annual Report was edited and
produced by Environmental Management Support, Inc., of Silver Spring, Maryland, under U.S. EPA
contract 68-W6-0014. Mention of trade names or specific applications does not imply endorsement or
acceptance by U.S. EPA.
-------
Treatment Walls
NATO/CGMS Pilot Study Phase lt(
CONTENTS
Introduction iii
General Overview 1
Harald Burmeier
Permeable Reactive Barrier Research at the National Risk Management Research Laboratory,
U.S. Environmental Protection Agency 3
Robert W. Puls
Technical Construction of Treatment Watts
Permeable Treatment Walls—Design, Construction, and Cost 6
Eberhard Beitinger
Development of Iron-Based Reactive Barrier Technologies for Remediation of Chlorinated Organic
Contaminants in Groundwater 17
Robert W. Gillham
Practical Solutions for the Treatment of Polluted Groundwater 22
Gerard Evers
Reactive Materials
Degradation of TCE at Zero-Valent Iron: Chemical Processes Effecting the Design and Performance
of Permeable, Reactive Fe(0) Walls .30
Wolfgang Wiist, O. Schlicker, and A. Dahmke
The Treatment of Groundwater with Mixed-Wastes: Reductive Dechlorination of
TCE and Reductive Precipitation of Uranium 36
Liyuan Liang and B. Gu
Bioprocesses in Treatment Walls: Bioscreens 44
Huub H. M. Rijnaarts
Novel Catalyses for Reactive Barriers 48
Timothy M. Vogel
Full-Scale Projects
Funnel-and-Gate Systems for In Situ Treatment of Contaminated Groundwater at
Former Manufactured Gas Plant Sites 56
Hermann Schad and Peter Grathwohl
i.
Reactive Treatment Zones: Concepts and a Case History 66
: ,,, Stephan A. Jefferis and Graham H. Norris
-------
Treatment Walls
NATO/CCMS Pilot Study Phase HI
Horizontal Treatment Barriers of Fracture-Emplaced Iron and Permanganate Particles 77
Robert L. Siegrist, Kathryn S. Lowe, Lawrence W. Murdoch,
Traci L. Case, Douglas A. Pickering, and Thomas C. Houk
In Situ Remediation Research in a Complexly Contaminated Aquifer: The SAFIRA Test Site
at Bitterfeld, Germany .. 84
H. Weiss, F.-D. Kopinke, P. Popp, and L. Wiinsche
Summary and Conclusions 92
Harald Burmeier
ABOUT THE AUTHORS 95
NATIONAL CONTACTS 98
PARTICIPANTS 102
11
-------
Treatment Y*aMs
NATO/CCMS Pilot Study Phase 111
Introduction
The Council of the North Atlantic Treaty Organization (NATO) established the Committee on the
Challenges of Modern Society (CCMS) in 1969. CCMS was charged with developing meaningful programs
to share information among countries on environmental and societal issues that complement other
international endeavors and to provide leadership in solving specific problems of the human environment.
A fundamental precept of CCMS involves the transfer of technological and scientific solutions among
nations with similar environmental challenges.
The management of contaminated land and groundwater is a universal problem among industrialized
countries, requiring the use of existing, emerging, innovative, and cost-effective technologies. This
document provides a report from the first meeting of the Phase m Pilot Study and is designed to share
information among countries on innovative treatment technologies. The United States is the lead country
for the Pilot Study, and Germany and The Netherlands are the Co-Pilot countries. The first phase
successfully concluded in 1991, and the results were published in three volumes. The second phase, which
expanded to include newly emerging technologies, concluded in 1997; final reports documenting 52
completed projects and the participation of 14 countries will be published in 1998. Through these pilot
studies, critical technical information has been made available to participating countries and the world
community.
Phase HI focuses on the technical approaches for addressing the treatment of contaminated land and
groundwater. This includes issues of sustainability, environmental merit, and cost-effectiveness in addition
to continuing to draw on the merits of emerging remediation technologies. The objectives of the study are
to critically evaluate technologies, promote the appropriate use of technologies, use information technology
systems to disseminate the products, and to foster innovative thinking in the area of contaminated land. The
first meeting of the Phase ffl Pilot Study on the Evaluation of Demonstrated and Emerging Technologies
for the Treatment and Clean Up of Contaminated Land and Groundwater convened in Vienna, Austria, on
February 22-27,1998, with representatives of 20 countries attending. Each participating country presents
case studies or projects to the Pilot Study for discussion.
Also at this first Phase m meeting, the first special technical session was convened on treatment walls and
related permeable reactive barrier technologies, under the chairmanship of Prof. Dipl.-Ing. Harald Burmeier
of the University of Northeast Lower Saxony. This report contains the proceedings of that special session.
A companion publication, the first Annual Report of the Phase m studies, contains abstracts of the first 15
demonstration projects selected, reports on the legislative, regulatory, programmatic, and research issues
related to contaminated land in each participating country, and the statement of purpose for the Phase m
Pilot Study. General information on the NATO/CCMS Pilot Study may be obtained from the Country
Representatives listed at the end of this report, or—for each paper—from the authors themselves.
Stephen C. James
Walter W. Kovalick, Jr., Ph.D.
Co-Directors
in
-------
-------
Treatment "Watts
NATO/CCMS Pilot Study Phase lit
General Overview
Harald Burmeier1
Remediation of groundwater is most commonly performed by "pump-and-treat" methods: the contaminated
groundwater is extracted from the aquifer and subjected to above-ground-treatment before being reinjected
or discharged. Multi-component systems are designed according to the particular contamination by taking
physical, chemical, and biological treatment processes into consideration.
Pump-and-treat methods require continuous energy for pumping water from the extraction wells and
operating the water treatment systems. Besides that, periodic maintenance and monitoring has to be
performed. Generally, these systems have to be operated for a long time. As a consequence, remediation
costs are substantial. Despite the wide spectrum of technologies available for above-ground treatment,
residual contaminants frequently remain at undesirably high levels in the subsurface.
Alternatives to these methods are passive working systems that depend on the usage of the groundwater's
natural hydraulics. There is a wide spectrum of possible solutions ranging from intrinsic remediation to
permeable reactive or adsorptive barriers.
As early as 1982, the idea to treat contaminated groundwater by installing a permeable reactive barrier was
shown in a figure in a remediation handbook issued by the U.S. EPA. But the 1980s were not the time for
further development of such innovative in-situ clean-up methods.
In 1989, the potential of Permeable Reactive Barriers (PRB) was recognized more clearly and was developed
further by the University of Waterloo, Canada. Since that time a lot of laboratory- and bench-scale investi-
gations led to the first full-scale in situ demonstration at Borden, Ontario, Canada.
Now, in 1998, over 500 project studies world-wide reflect the interest in this technology. The small number
of so far only 20 commercial applications shows that it is a long way from the pilot study to the field
application.
Today, a lot of experts in North America and Europe are working on several research issues concerning the
PRB technology, as there are:
• processes in the media and aquifers (hydraulics, geochemistry, etc.),
• suitable materials for reactive walls,
• design and construction,
• long-term effectiveness,
• treatment of contaminant-mixtures, and
• standards for construction, monitoring, etc.
Although many questions remain to be answered, a significant cost reduction potential of more than 30
percent is expected when PRBs are used instead of pump-and-treat methods.
In this Special Session of the Phase III Pilot Study on the Evaluation of Demonstrated and Emerging Tech-
nologies for the Treatment and Clean Up of Contaminated Land and Groundwater (Phase III), a dozen
1 University of Applied Studies and Research, University of Northeast Lower Saxony, Herbert-Meyerstrasse 7,29556
Suderburg, Germany, tel: 49/5103-2000, fax: 49/5103-7863, e-mail: h.burmeier@t-online.de
1
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
papers are presented on the international state-of-the-art concerning technical construction issues, reactor
materials, and full-scale projects. This report summarizes those proceedings and the discussions following
them, research and development needs, and recommendations for further action items.
I would like to extend my special thanks to all experts for preparing reports and giving presentations during
this Pilot Study Meeting in Vienna. Last-but-not-least, I would like to gratefully acknowledge the input and
substantial support provided by Steve James, John Moerlins, and Volker Franzius during the preparation
and coordination of this session.
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
Permeable Reactive Barrier Research at the
National Risk Management Research Laboratory,
U.S. Environmental Protection Agency
Robert W. Puls1
Much of the current research on ground-water remediation has focused on the removal of contaminated
water from the subsurface and treating it at the surface. While the removal of the contaminants is desirable,
the costs are often prohibitive and rarely are contaminant concentrations lowered to the required regulatory
levels. This has been particularly evident for standard "pump-and-treat" approaches. In situ chemically
reactive permeable walls are being considered as a low-cost and effective alternative for the treatment of
contaminated waste sites. The chemical form of the contaminant in question is transformed via oxidation,
reduction or precipitation reactions to an immobilized or non-toxic form. The application of in situ
approaches to subsurface remediation will increase the emphasis on adequate site characterization and
thorough understanding of the subsurface system targeted for remediation as well as the geochemical
mechanisms controlling contaminant transformations.
Background
Research into the use of zero-valent metals to remediate ground-water contaminated with mixed wastes
(inorganic and organic) has been ongoing at the U.S. Environmental Protection Agency's National Risk
Management Research Laboratory (NRMRL) in Ada, Oklahoma since 1991. The primary emphasis has been
on inorganic constituents such as chromate, arsenic, nitrate and sulfate and chlorinated organic compounds
(e.g., trichloroethylene, cis-dichloroethylene, and vinyl chloride). Laboratory research conducted by
scientists at the NRMRL conclusively demonstrated the effectiveness of using zero-valent metals to
remediate chromate and chlorinated organic compounds in groundwater and this research was scaled up to
a pilot-scale demonstration in September 1994, near an old chrome plating facility on the U.S. Coast Guard
(USCG) Support Center near Elizabeth City, North Carolina. Results indicated that complete treatment of
chromium and the chlorinated organic compounds in the groundwater was possible using this technology.
Chromium concentrations were reduced to less than 0.01 mg/1, much less than drinking water limits. The
chromate was reduced via corrosion of the elemental iron to the nontoxic and insoluble chromic ion (Cr3"1"),
which forms an insoluble mixed chromium-iron hydroxide phase. The chlorinated organic compounds were
similarly degraded to below maximum contaminant levels set for groundwater by the U.S. EPA.
Full-Scale Systems
A full-scale demonstration of a permeable reactive barrier to remediate groundwater contaminated with
chromate and chlorinated organic compounds was initiated at the USCG site by researchers from NRMRL
and the University of Waterloo in 1995. A continuous wall composed of 100% zero-valent iron was installed
in June, 1996, using a trencher that was capable of installing the granular iron to a depth of 8 m. The
trenched wall was approximately 0.6 m thick and about 60 m long. The wall begins about 3 feet below the
ground surface and consists of about 450 tons of granular iron. The installation of the wall was completed
in less than a day. The total installation cost was $500,000, with the cost of iron representing approximately
35% of the total. Total installation, site investigation, design feasibility tests, general contracting, and post-
installation monitoring costs have been less than $1 million. Savings compared to traditional pump-and-treat
1 U.S. Environmental Protection Agency, R.S. Kerr Environmental Center, P.O. Box 1198, Ada, OK 74821-1198, USA,
e-mail: puls.robert@epamail.epa.gov
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
exceed $5 million over a 10-year period primarily due to reduced operation and maintenance costs due to
the passive nature of this technology.
The continuous trenching equipment used for the installation was similar to a large "Ditch Witch." It uses
a large cutting chain excavator system combined with a trench box and loading hopper. As the trenching
machine moved along, the cutting chain excavation removed the native soil, while a front end loader filled
the hopper of the trench box. The granular iron flowed through the hopper and out the back of the trench
box into the just-excavated trench. The walls of the trench box extended to the width and depth of the trench
maintaining the opening until the iron was emplaced.
The system was designed to meet lower concentrations of contaminants below maximum concentration
limits (MCLs) for groundwater set by U.S. EPA of 0.05 mg/1 Cr(VI), 5 ug/1 TCE, 70 ug/1 DCE and 2 ug/1
Results
To-date, there has been two years of post-installation performance monitoring performed by these same
researchers. For all but one quarterly sampling event, 15 multilevel samplers (7 to 11 sample ports per
sampler) and 9 to 10 compliance (5-cm PVC) wells have been sampled. In addition to on-site sampling of
the full suite of geochemical indicator parameters listed in the site work plan, samples have been collected
for laboratory analysis of the following constituents: TCE, cis-DCE, vinyl chloride, ethane, ethene, methane,
major anions, and metals. In addition, numerous vertical and angle cores have been collected to examine
changes to the iron surface over time and to evaluate the formation of secondary precipitates which may
affect wall performance over time. Coring was done vertically (perpendicular to ground surface) and on an
angle (30°). The former method provided continuous vertical iron cores, while the latter provided a
transverse core through the wall with the aquifer-iron interfaces intact (front and back of the wall). These
cores continue to be under study. Inorganic carbon contents increase dramatically at the up-gradient aquifer
sediment-iron interface and decrease within the wall moving down gradient, reaching background levels
within 10 cm down gradient from the down gradient iron-aquifer sediment interface. Total inorganic carbon
content has increased over time within the wall.
Results of geochemical sampling on site indicate that iron corrosion is proceeding within the wall. There
are significant reductions in Eh (to >-400 mv), increase in pH (to >10), absence of DO, and decrease in
alkalinity. Down gradient of the wall (1.5 m), pH returns to near neutral and Eh is quite variable with depth.
Over time there have been indications that a redox front is slowly migrating down gradient within the first
few meters of the wall. Water levels indicate little difference (<0.1 m) between wells completed and
screened at similar depths up gradient and down gradient of the wall, indicating that the wall continues to
effectively function as a "permeable" reactive barrier.
Sampling results for chromate indicate that all chromate was removed from the groundwater within the first
15 cm of the wall as expected. No chromate is detected down gradient of the wall either in the multilayer
samplers or in the 5-cm compliance wells located immediately behind the wall.
The vast majority of the multi-layer sampling ports show reduction of the chlorinated compound
concentrations to less than regulatory target levels. Only one port (ML25-1) continues to show levels above
target concentrations. This is the deepest port in the middle of the wall where the organic solvent compound
concentrations are highest.
-------
Treatment Wa»s
NATO/CCMS Pilot Study Phase III
Other Inorganic Contaminants
Other inorganic contaminants currently under study include nitrate and arsenic. These inorganic
contaminants are transformed via oxidation-reduction reactions to less mobile and/or less toxic forms and
immobilized as insoluble precipitates or through essentially irreversible adsorption reactions. Zero-valent
iron can effectively transform nitrate to reduced forms as confirmed with mass balance determinations. The
use of different organic materials also further reduces the nitrate and experiments are in progress to account
for all nitrogen forms produced. Experiments have also explored the use of iron-organic material mixtures.
Experiments with arsenic are currently underway and we are exploring conversion to reduced and oxidized
forms for immobilization in a reactive barrier.
Disclaimer
Although the research described in this article has been funded wholly or in part by the United States
Environmental Protection Agency, it has not been subjected to the Agency's peer and administrative review
and therefore may not necessarily reflect the views of the Agency and no official endorsement may be
inferred.
Discussion
Zero-valent iron walls can affect the pH of chlorinated solvent plumes. The pH of a plume can increase to
10 to 11 after passing through a 100% iron wall. This change is always observed due to the nature of the
corrosion reaction that takes place and occurs regardless of whether the plume contains chlorinated solvents
or chromium. However, the iron can be mixed with pyrite or native aquifer sediments (e.g., 50% iron and
50% sediment) to buffer the pH in the walls to less than 7.5. The buffered walls exhibit an environment
favorable for microbial activity, so there is a lot of sulfide is produced. Additionally, hydrogen is produced
during the corrosion reaction at the iron surface. The production of hydrogen is significant—up to 1,000 nM
of dissolved hydrogen has been measured at the Elizabeth City site.
Puls's data documented a buildup of chromium (2% by weight after one year) in the PRB installed at
Elizabeth City, and Puls expects it to increase in the future. This expectation is based on two-year column
tests in which over 3,000 pore volumes were pumped through under accelerated conditions. Although a
buildup occurred that reduced the permeability of the column, it did not impact the reduction of chromate
in the groundwater. The pressure buildup caused by the precipitate occurred after pumping approximately
500 pore volumes. It is difficult to extrapolate these laboratory results to field-scale tests, and Puls said that
they will continue to monitor for long-term reactivity in the wall.
The residence time for pore water in the fastest part of the barrier is two days. The hydraulic conductivity
of the aquifer varies over its 24-foot thickness, and the highest hydraulic conductivity occurs within a 2-
meter interval. This interval contains the highest concentrations of Cr6*, but most of the TCE is found in
deeper zones. The reduction of chromium causes no inhibitory or competitive effect on the degradation of
TCE, because the chromium and TCE are present in different intervals. While the hydraulic conductivity
of the aquifer ranges from 2-5 inches/day, the hydraulic conductivity of the wall is 10-12 inches/day.
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Permeable Treatment Walls—Design, Construction, and Cost
Eberhard Bellinger1
Abstract
The Design of Permeable Reactive Barriers (PRB) is discussed with reference to preventing effects such as
inflow of fine soil particles, precipitation of carbonates, iron and manganese, and loss of effectiveness by
uncontrolled growth of biomass.
Several construction methods are presented and evaluated in terms of design objectives, installation
methods, implementation, effectiveness and cost.
1. In Situ Groundwater Remediation with Permeable Reactive Barriers
According to EPA, "...a permeable reactive barrier (PRB) is a passive in situ treatment zone of reactive
material that degrades or immobilizes contaminants as groundwater flows through it. Natural gradients
transport contaminants through strategically placed treatment media. The media degrade, sorb, precipitate,
or remove dissolved organics, metals, radionuclides, and other pollutants. These barriers may contain
reactants for degrading volatile organics, chelators for immobilizing metals, nutrients and oxygen for
microorganisms to enhance bioremediation, or other agents." [1]
The following report is focused on construction methods that prevent negative effects that may reduce the
effectiveness of the treatment media, negatively change the hydraulic conditions, or reduce the long-term
PRB performance. This includes the application of knowledge and experience from groundwater well design
and the evaluation of innovative technologies to remove and/or reactivate the treatment media in situ.
Permeable Reactive Barriers have commonly been, or are designed to be, installed through excavation and
replacement. This method is limited to approximately 8 m in depth, and the costs becomes prohibitive at
greater depths. Alternative installation methods for greater depths include slurry wall construction, high
pressure jetting, deep soil mixing, and hydro fracturing in the United States [2]. In Germany large diameter
vertical borings, slurry methods and a modified deep wall construction are under evaluation or investigation,
but have not yet been tested in the field.
2. Design Objectives for Permeable Treatment Walls
The hydraulic behaviors of the two major permeable treatment wall design types—funnel and gate systems
or continuous reactive walls—are based on the hydraulic permeability of the whole construction system,
including permeability of filter layers, screens and the treatment media itself. The system permeability of
the wall construction should be at least twice the permeability of the aquifer. It might be better to choose
a factor of 10 times higher permeability for the wall system because of to all the limiting parameters that will
reduce permeability with time. The major limiting factors are expected to be as follows:
• inflow and settling of fine-grained soil particles, which will block the pore spaces and reduce the
permeability;
1WCI Umwelttechnik GmbH, Sophie Charlotten - Str.33, 14059 Berlin, Germany, tel: 49/30-3260-9481, fax- 49/30-
321-9472
-------
Treatment Walls
N ATQ/CCMS Pilot Study Phase 1»
• precipitation of carbonates such as calcium or magnesium carbonates, iron oxides/hydroxides, and
ferrous carbonate or other metal precipitates in the filter layer or treatment media;
• uncontrolled growth of microorganism, such as bio-clogging; and
• other, mainly long-term and unknown effects, which may reduce the permeability.
The following design objectives are focused mainly on the use of activated carbon as the treatment media.
But most of these design objectives may also be useful if zero-valent iron (ZVT) or other media are chosen
as reactive filling materials.
Major design objectives are to:
• Prevent the inflow of fine-grained soil particles by installing a filter layer between the adjacent soil and
the treatment media according to the well-known filter criteria;
• Prevent changes to the physical and chemical properties of the groundwater such as temperature,
pressure, pH, oxygen content (redox potential), nutrients, or be aware that these changes will result in
precipitation or microorganism activity that might be of advantage in specific cases;
• Design a wall-system that allows the removal and replacement of the treatment media after a period of
several years or decades;
• Design a pipe-system in the wall to allow the injection of water or air for flushing to eliminate
precipitates or sludges or to remix the media by turbulence; and
• Design openings for inspection, removal or sampling of the treatment media.
The above list of design objectives may not be applicable in every case. Based on the specific site
conditions, the nature of contamination and the remediation goals, additional design objectives have to be
considered.
3. Design Criteria
The design of Permeable Reactive Barriers needs to carefully address the following design criteria. In
general, investigations to obtain design data should be planned with reference to the specific design
parameters for a long-term performance of the underground treatment wall.
Design criteria are:
• Geology
• Hydrology
• Depth to groundwater
• Thickness of the aquifer
• Groundwater flow direction and gradient
• Permeability
• Grain-size distribution
• Pore volume
• Groundwater chemistry, such as pH-value, redox-potential, electrical conductivity, oxygen content,
temperature, hardness, iron and manganese concentration, sulfates, and nutrients
• Biological activity (microorganisms) '
• Pollutants (concentration, spread, transfer conditions)
• Site conditions, topography, and access
• Sensitivity of aquifer regarding potential users
• Damage to existing buildings regarding settling problems
• Available time for remediation
• Effectiveness of treatment
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
• Costs
• Monitoring.
4. Construction Methods
During the last five years, a variety of construction methods have been developed in Canada, the United
States and in Europe. Some of these construction methods have been implemented or tested in field scale.
Others are proposed for implementation but have not yet been adapted.
4.1 Peat-Filled Trench
The University of South Carolina [1] obtained a patent October 15,1991, for a method for in situ removal
of hydrocarbon contaminants from groundwater using a permeable barrier comprising a peat material or,
an immobilized nutrient layer and peat material layer in series.
As shown in Figure 1, it is assumed that an open trench would be excavated and filled with the reactive
media. The construction method is simple and inexpensive.
The specific costs for open trenches are mainly
the costs for excavation and disposal, including
costs for lowering groundwater and, if neces-
sary, stabilizing the open ditch.
The actual costs, without stabilization, may
range between US$10-100/m2. Costs for stabili-
zation will vary within US$50-100/m2.
4.2 Solid-Free Trench
GROUND SURFACE
LEAKING
UNDERGROI
STORAGE
TANK
PEAT-FILLED
TRENCH
CONTAMINANT-
PLUME
Figure 1. Peat-filled trench [1]
emission control
In February 1992, a German patent [2] was granted to
Dr. Haldenwang and Dr. Eichhorn for a solid-free
trench construction in which the walls of an open
trench are permeable and stabilized and the
groundwater flows through it (Figure 2). Additionally,
piping for aeration or circulation of water within the
open trench was proposed including measures for
capping and emission control.
For field testing, a 6 m deep, 1.4 m wide and approxi-
mately 10m long trench was established on a refinery
site and tested for several years.
No solid media is used as reactive media but
according to its function as an underground permeable
barrier this technology is mentioned as an alternative
within the group of permeable reactive barriers. To
stabilise the wall, permeable sheet piles have been Figure 2. Solid free trench [2]
proposed. The specific costs are estimated to range
between US$200-400/m2 not including any preparation or other aeration/de-aeration installations.
aquitard
Figure 2: Solid-free trench [2]
-------
Treatment Walls
NATO/CCMS Pilot Study Phase HI
Phaaal: Excavation
PhaM 2: Computed
aqutaid
4.3 Open Trench Construction
In several cases, open trenches filled with reactive
media have been planned or executed. Vertical sheet
piling was used to stabilise the open excavation
ditch and pulled after filling of reactive media was
completed. The maximum width of the reactive wall
is dependent on the distance that the sheet pile walls
are driven into the ground (Figure 3).
Figure 3. Open trench construction of Permeable
Reactive Barriers
One of the major disadvantages is that there is no
filtration layer between the adjacent soils and the
reactive media to help avoid inflow of fine-grained soil particles. Also recovering the filling material
(reactive media) is not possible without destroying the structure itself. The specific construction costs will
vary (without reactive media) in the range of US$100-200/m2.
4.4 Treatment Wall Construction by WCI-
Umwelttechnik GmbH
WCI Umwelttechnik GmbH has developed a special
method of wall construction that allows the
adsorbing or other reactive media to be recovered
without the need to demolish and rebuild the wall
structure [3]. The patented system includes filtration
layers to prevent inflow of fine soil particles and
measures to avoid precipitation by oxidation of iron
and manganese. As shown in Figure 4, the main
design components are:
• a filter layer, consisting of gravel or sand pack,
located between the trenched aquifer and the
interior wall elements;
• interior wall elements, made of specially
designed brick elements or pre-cast concrete
shells to define an interior filling space for the
treatment media, including openings for the
groundwater through-flow and horizontal
elements to stabilise the earth pressure;
• a clay seal to prevent inflow of stormwater and
contact with atmospheric oxygen;
• a cover plate as an opening device for the
recovery and replacement of the treatment media
—additionally these covers can be made
waterproof and airtight.
Some typical dimensions are shown for the cross
section of the wall construction in Figure 5.
Figure 4. Permeable Treatment Wall Construction
aquifer
sheet pile wall
dimensions wall construction
(cross section) detail
' scale 1:10
will m
Figure 5. Cross-Section—Treatment Wall
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Depending on the usually low groundwater flow velocity, the retention time of groundwater in a 0,30 m
thick treatment media layer is between several hours and one or two days.
In order to dimension the retention time and media thickness, new approaches must be developed based on
test results in the field. In particular for the use of activated carbon, the retention time, surface capacity and
the height or width (thickness) of media layers in the permeable treatment walls will differ widely from
those dimensioning criteria used in conventional aboveground activated carbon filters (Table 1).
Table 1: Dimensioning Criteria
Dimensioning Criteria
Retention time
Surface load capacity
Height/width filter layer
PRB-adsorbing
5 - 30 hrs
0.02 - 0.06 m/h
0.3 - 0.5 m
Conventional
Activated Carbon Filter
l-2hrs
10 -15 m/h
2m
Installation methods
The installation methods are described in detail in Figures 6 to 11.
In phase 1 (Figure 6) sheet pile walls will be driven into the earth. The distance between the piles will be
1 m or more, according to the chosen system width. The sheet piles will be installed 2 m into the bottom
layer of the aquifer.
In phase 2 (Figure 7) excavation will commence within the two sheet pile walls. An open ditch will be used
to lower the groundwater table below the excavation level.
I I
>hM< pa* driving
ph«M2: neovitioa and dMMIMlng of th*tlw«twftll construction
Figure 6. Construction Phase 1
Figure 7. Construction Phase 2
After completion of the excavation (Figure 8), a concrete base layer will be cast in situ and dewatering will
be continued. The horizontal earth pressure will be supported with beams.
Now the installation of the forms and placing the filter gravel can begin on the concrete layer (Figure 9).
As the construction grows, groundwater level can be allowed to rise again.
10
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
phflsa 3: complete excavation of tb» trench, stiffening and construction
oi a concrete bottom layer, dtwattrlng
Figure 8. Construction Phase 3
forms.
phasaS: phHiwniMofbulkmtlwMlnfbnnwoffctlwmnt*,
compUH ramanl ol >bMt plto mH
topvtow:
phma4: pUetmtraoffennworit e)em«nt«, ptawmtnt of gravel pack,
parti*] removal of shoot pB» waU. recovery of groundwatcr tewl
Figure 9. Construction Phase 4
As the wall construction continues (Fig. 10), sheet
piles can be raised accordingly and the treatment
media will be filled into the interior space of the
Finally, the wall cover, the clay seal and some other
measures can be installed to complete the
construction (Figure 11). In addition, bank
protection (if located near a shore as in the given
example), monitoring wells and a roadway for the
tank trucks (to remove and replace the treatment
media) may be installed.
ptuuxS: niific.coin|il«Ion of tKMtnMnt wall with etay M«I
•nd •mb*nkm«nt, coniMctkm with capping of landfill,
InxUMttton ol nwnHoring wrtl
Figure 10. Construction Phase 6
Figure 11. Construction Phase 6
11
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Figure 12 shows an approximately 30-m long constructed longitudinal section. All the necessary installation
phases from excavation to covering the completed wall are illustrated. For longer walls, several construction
sections may be useful.
The construction costs for permeable treatment walls with a depth of 8 m are estimated to range from DM
850 to 1400 per square meter (US$500-900/m2) without the treatment media (an additional DM 300/m2
[US$ 180/m2] for an activated carbon layer of 0.3 m thickness).
placement of kxmwork elements
driving of sheet pile watts
placement of gravel pack
dewatering
pUcwnml
Mngnwlei
material
Figure 12. Construction Progress in Longitudinal Direction
4.5 Construction Patented by Dr. Steffen Ingenieurgesellschaft
Another construction method based on a slurry
wall technology was patented (June 27,1992)
by a German consultant, Dr.-Ing. Steffen
Ingenieurgesellschaft mbH, Essen [4]. The
patented idea is to press a sheet pile or a
double-T-formed steel beam down into the
ground and to inject activated carbon by
pulling the beam out again (Figure 13). The
overlapping process of pressing and injection
will form a continuous wall of reactive media
with a width of approximately 0.10-0.15 m.
Depending on geology, depths to 20 m or more
can be reached. When the adsorbent capacity
has been exhausted, a second wall is planned
upstream.
Figure 13. Dr.-Ing. Steffen—Barrier Construction Phases
There is no filter layer and the recovery of the media is not anticipated. Such a wall construction may be
blocked easily by inflow of fine soil particles or by precipitation effects. The cost range is estimated, to be
US$200/mz without the media itself.
12
-------
Treatment Walls
N ATO/CCMS Pilot Study Phase (((
4.6 Large Diameter Borehole Construction Method
This construction method is used in the U.S. as well as in Germany. Mull und Partner, Hannover, Germany,
has proposed a continuous overlapping iron-filled wall in Nordrhein-Westfalen to prevent downstream
contamination of chlorinated hydrocarbons emitted by a laundry.
Usually, borehole diameters from 0.8-1.6 m are
chosen. Individually located bore holes, established
in several lines are known as an "iron-fence." The
method is applicable for continuous walls, if the
bore holes overlap, or for gate constructions. Depths
to 20-25 m may be reached with no problems. By
overlapping the bore holes, a loss of approximately
15-20 % of the filling material has to be allowed.
Depending on the geology, in most cases steel pipes
have to be used for stability of the borehole during
excavation. This construction method is shown in
Figure 14.
The costs for this method are US$300-500/m2,
excluding filling materials.
4.7 Method developed by Sax+Klee, Germany
The German construction company, Sax+Klee
GmbH, Mannheim, has developed a modified large
diameter borehole construction method for a funnel-
and-gate PRB system. The innovative approach
includes using bore holes filled with impermeable
clay to connect the funnel sheet pile walls with the
gate structure. Also new is the proposed installation
of an interior well screen with four wings to create
a filler zone between the surrounding soils and the
reactive media (Figure 15).
The costs "for a 40 m wide and 19 m deep gate
construction are estimated to be approximately
US$l,400/m2.
A filling material of petroleum activated carbon is
proposed for a site with groundwater contaminated
with petroleum hydrocarbons and aromatics
(BTEX). The site is a former U.S. Army tank farm.
4.8 "Funnel and Gate" Patented by University of
Waterloo
The funnel and gate method (Figure 16) has been
patented worldwide since April 23, 1992, for the
University of Waterloo, Canada [5].
Figure 14. Large Diameter Borehole
Construction Method
Layout
bcrahoto
•dbmeUr jr 1200 mm
Figure 15. Sax+Klee GmbH, Germany
Figure 16. Funnel and Gate System,
University of Waterloo
13
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
"Contaminated groundwater is treated in situ, by funnelling the water through a gate or gates in a watertight
in-ground wall. Treatment material in the gate breaks down the contaminant, or otherwise removes the
contaminants from the flowing water. A removable caisson is first driven into the ground, excavated, and
then a receptacle, for the treatment material, is lowered into the hollow interior."
The patent itself shows a wide variety of construction elements, such as caissons made of steel (receptacles),
slurry and sheet pile waterproof walls, removable baskets in the receptacle, removable caissons, etc.
Costs for the construction methods proposed by the University of Waterloo may vary in the range of
US$200-l,000/m2.
4.9 Bio-Polymer Slurry Trenching Technique
Geo-Con, a subsidiary of Woodward-Clyde, reported (December 23,1997) the performance of a funnel-and-
gate project for the U.S. Department of Energy, Oak Ridge, as follows:
Geo-Con recently completed what we believe to be the first permeable reactive wall installed using
the bio-polymer slurry trenching technique. The reactive gate consists of a 30-foot (900 cm) iron
filing "gate," with two 100-foot (30 m) wings composed of pea gravel. The system is 2 feet (60 cm)
wide and extends down to bedrock, which is approximately 26 feet (780 cm) below ground surface.
The purpose of the permeable reactive wall is to collect a plume of groundwater contaminated with volatile
organic compounds (VOCs). The trench intercepts the groundwater, funneling it through the pea gravel to
the iron filing gate. As the groundwater passes through the filing gate, the VOCs are removed.
This project was originally contracted to another company, to be built by excavating dry and backfilling with
the filings and pea gravel. Upon nearing completion of the first cut-approximately 20 feet (600 cm) deep,
the trench collapsed. Geo-Con proposed to complete the project using a bio-polymer slurry to maintain
trench stability during gravel and iron filling installation. Geo-Con was able to complete the project in three
days, expediting the project schedule and providing a considerable cost saving over other installation
methods. [6]
The author estimates costs for this technique to be in the range of US$300-500/m2, not including the filling
materials.
4.10 Other Construction Methods
The following listing of other construction methods under discussion or developed for the installation of
Permeable Reactive Barriers is not complete; it reflects the knowledge of the author about actual
information.
• Enviro Wall is a trademark of Barrier Member Containment Co. Ltd. and Argonne National Laboratory.
By help of a guide box installation, an HOPE geomembrane will be installed, and a groundwater
collection and redistribution system will guide groundwater to a pass through module. The open trench
will be refilled with site soil and capped with bentonite after installation of membranes and horizontal
collection systems.
• The PRB Action Team reports studies using high pressure jet grouting systems to create either
permeable grout diaphragms or columns [7]. Overlapping these structures creates longer, continuous
barriers. In a field study, two grout formulas have been tried: a resin/granular cast iron slurry and a
14
-------
Treatment Walls
NATO/CCM5 Pilot Study Phase ffl
kaolinite clay/iron slurry. Also, deep soil mixing and hydrofracturing (horizontal and vertical) are under
discussion in the U.S. as alternatiye installation methods.
• Envirotreat Ltd., UK, is currently working on the development of active containment systems using
special clay materials. Addition of reactive materials to clay barriers is also under investigation.
Overlapping large diameter bore holes are one of the construction methods is considered to be
applicable.
• Several German engineering companies have proposed the installation of gates in the form of concrete
structures for horizontal or vertical through-flow (Peschla und Rochmes, Edenkoben, and WCI
Umwelttechnik, Essen). These are conventional underground concrete structures, caissons, or chambers
that contain removable reactive media. These buildings are expensive and limited to use as single gate
structures.
• WCI-Umwelttechnik GmbH is working on a deep wall construction method using modified slurry wall
techniques with additional caissons as containment for the reactive media.
Finally, the injection of reactive material into the unsaturated or saturated zone, not forming a defined
structure, is a potential method, but not discussed in this report.
5. Costs
Table 2 presents a rough overview to costs for the several construction techniques mentioned. They do not
include the filling material. They also depend on national markets and on actual/future experience and
developments. Some of the actually methods may disappear in the near future, and others may show
economical advantages. Generally, we need more field testing and full-scale experience than we now have.
Table 2: Cost Overview
Construction method
4.1 Peat-filled trench
4.2 Solid-free trench
4.3 Open trench construction
4.4 WCI-treatment wall
4.5 Dr. Steffen's patent
4.6 Large diameter borehole
4.7 Sax+Klee system
4.8 Funnel and Gate
4.9 Bio-polymer slurry
4.10 Other construction methods
Max
Depth [m]
5
10
10
12
20
20-25
20
20
30
10
20
30
Cost Range
[US$/m2]
10-100
200 - 400
100 - 200
500 - 900
200
200 - 500
1000 - 1500
200 - 1000
300 - 500
100 - 300
200 - 1000
500 - 1500
6. Conclusions
Li our estimation, it is too early to evaluate construction methods in general. They strongly depend on the
required depth, hydrogeology, contaminants, remediation targets, etc. Most of the construction methods
discussed are unproven in relation to PRBs, or are under testing.
15
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
A wide variety of construction elements are available. The future will show which systems are most
applicable to shallow and deep aquifers and that are economically sound.
Full-scale applications and field tests will be helpful in establishing the innovative PRB technology as a
proven and available remediation alternative. Besides the research and development activities regarding
effective reactive media, we have to deliver inexpensive, durable construction techniques to install
(recoverable) reactive media in the contaminated groundwater plume.
7. References
[1] United States Patent Number 5,057,0227, Cohen, University of South Carolina, Columbia, S.C.,
October 15,1991.
[2] PatentschriftDE41 05 987 C2, Haldenwang, Eichhorn, February 26,1991 "Vorrichtung und Verfahren
zur in ,s//H-Behandlung verunreinigter Grundwasser".
[3] Patentschrift DE 44 25 061 Cl, WCI-Umwelttechnik GmbH, July 15, 1994 "Permeables
Behandlungsbett zur Reinigung kontaminierter Grundwasserstrome in situ".
[4] Patentschrift DE 4221198 C2, Dr.-Lig. Steffen Ingenieurgesellschaft mbH, June 27,1992 "Verfahren
zum Entfernen von wasserloslichen sorbierbaren Schadstoffen aus einem abstromenden Grundwasser
in der Umgebung eines Kontaminationsherdes".
[5] International Patent W093/22241, Cherry, Vales, Gillham, University of Waterloo, Canada, April 23,
1992 "System for Treating Polluted Groundwater".
[6] Geo-Con, internal company information, not published, December 1997.
[7] Permeable Reactive Barrier Action Team, Summary of the Remediation Technologies Development
Forum, Virginia Beach, Virginia September 18-19, 1997.
Discussion
It was suggested that treatment walls may not need to intercept all of the contaminated groundwater, due
to the expense of constructing the walls, and could be designed to treat (for example) only 90% of the
groundwater flow. This would make the technology cheaper and more appealing. Beitinger agreed that such
walls can be constructed (e.g., an iron fence or non-overlapping borehole reactors), but the decision is
usually based on whether the regulating authority would permit continued flow of untreated water.
16
-------
Treatment Walls.
NATQ/CCMS Pilot Study Phase Ml
Development of Iron-Based Reactive Barrier
Technologies for Remediation of Chlorinated Organic
Contaminants in Groundwater1
Robert W. Gillham2
Abstract
Granular iron was not recognized as an effective reductant for promoting the dechlorination of halocarbons
in aqueous solution until the late 1980s. Furthermore, the suggestion that granular iron be used, for in situ
remediation of groundwater containing halocarbons was initially met with a high degree of skepticism. In
the intervening years, the use of granular iron for groundwater remediation has emerged as a new and
significant environmental technology. This paper outlines some of the major contributions that have led to
the growing scientific and market acceptance of the technology.
Introduction
Recognizing the limitations of pump-and-treat methods for groundwater remediation (NAS, 1994),
considerable attention has now turned to the use of in situ permeable reaction barriers (PRBs). An early
review of this topic (prepared in 1992) is given in Gillham and Burris (1997). In this concept, a permeable
"wall" containing the appropriate reactants is constructed across the path of a contaminant plume. As the
contaminants pass through the reactive material under passive groundwater flow, they are removed by
chemical and/or physical processes. Proponents of PRBs cite reduced capital cost, low operating and
maintenance costs, and conservation of water and energy among the potential advantages over pump-and-
treat systems.
The use of granular iron for removal of chlorinated organic contaminants has led the interest in PRBs. With
initial recognition in 1989, this concept is now emerging as a significant technology for groundwater
remediation. This paper summarizes the major steps in both the scientific understanding and acceptance,
and in the commercial development of the technology.
Scientific Development
In studies of the effect of sampling-well materials on sample integrity, (Reynolds et al., 1990) it was
observed that the concentration in aqueous solution of several chlorinated organic contaminants declined
when in contact with certain metals. Though this observation was made in 1984, it was not until 1989 that
the potential significance with respect to groundwater remediation was recognized. Further testing
confirmed the earlier results and revealed various aspects of the process. In particular, based on our early
understanding of the reactions, the passive in situ use of granular iron was proposed as an alternative for
remediation of groundwater containing halocarbons. Response to publication of the early experimental
results and concepts (Gillham and O'Hannesin, 1992) generally varied from skepticism to disbelief.
A degree of credibility and the attention of the scientific community was gained through two papers
published in 1994. Gillham and O'Hannesin (1994) showed a wide range of contaminants to degrade, in
the presence of granular iron, at rates that are several orders of magnitude greater than natural abiotic
degradation rates reported in the literature. Evidence was presented to indicate that the reaction was abiotic
1 Submitted for publication in: Proceedings, American Society of Civil Engineers, 1998.
2 University of Waterloo, Department of Earth Sciences, 200 University Avenue W., Waterloo, ON N2L 3G1, Canada,
tel: +519-888-4658, fax: +519-746-7484, e-mail: rwgillha@sciborg.uwaterloo.ca
' /•
17
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
reductive dechlorination and followed pseudo-first order kinetics. Matheson and Tratnyek (1994), from
detailed studies of the degradation of carbon tetrachloride, showed results consistent with those of Gillham
and O'Hannesin (1994), and further proposed three possible reaction mechanisms. All involved sequential
dechlorination through one-electron transfer. Both Matheson and Tratnyek (1994) and Gillham and
O'Hannesin (1994) proposed two simultaneous redox reactions 1) oxidation of the iron coupled with
reduction of water and 2) oxidation of iron coupled with reduction of the chlorinated organic compound.
Reaction (1) results in the production of H2 and Fe2+ and an increase in pH as a consequence of the
production of OH'. Reaction (2) causes a further release of Fe2+, the release of Cl" and production of a less
chlorinated organic compound. Based on generally low concentrations of dissolved iron observed in batch
and column tests, it was proposed that iron precipitates as Fe(OH)2, limiting the pH increase to values in the
range of about 9 to 10.
The scientific aspects of the technology "came of age" through an American Chemical Society symposium
"Contaminant Remediation with Zero-Valent Metals" organized by Drs. Paul Tratnyek and Marline
Reinhard (209th ACS National Meeting, Anaheim, California, April 2-7, 1995). As an indication of the
increasing interest, over 40 abstracts were submitted for presentation. Information was presented indicating
that the reactions occur on the solid surfaces rather than in free solution (Weber, 1995 and Totten and
Roberts, 1995); based on relatively low production of intermediate degradation products, the one-electron
sequential degradation pathway was brought into question (Sivavec and Homey, 1995 and Orth and
Gillham, 1995). Totten and Roberts (1995) showed that two-electron transfers can occur, and through this
and subsequent work by Roberts et al. (1996) and others, it is now widely accepted that at least for the
chlorinated ethenes, trichloroethene and tetrachlorothene, there are two competing pathways, single- electron
transfer with DCE isomers and vinyl chloride as intermediate products and two-electron transfer where
chloroacetylene is the intermediate. Though both pathways are followed simultaneously, in most situations
it appears that the latter is dominant. This is favorable of course since chloroacetylene is highly unstable,
while the DCE isomers and vinyl chloride generally degrade more slowly than their parent compounds (TCE
and PCE).
Other important contributions concern the affects of inorganic constituents (other than Fe°). Sivavec and
Horney (1995) noted that the iron surfaces are generally covered with an iron oxyhydroxide coating,
suggesting a two-step degradation process, sorption onto the surface followed by electron transfer. Noting
that electron transfer must occur through the surface coatings, the nature of the coatings and their ability to
conduct electrons becomes an important issue. Precipitation of Fe(OH)2, a consequence of both reactions
(1) and (2) as presented previously, could form a protective coating thus resulting in a loss of activity over
time. In subsequent studies, however, (Odziemkowski et al., 1998) it has been shown that Fe(OH)2 is
converted to magnetite, which is conducting to electrons, thus maintaining the activity of the iron surfaces.
It is also recognized that the increase in pH caused by the reduction of water can have a profound influence
on the inorganic chemistry of natural waters. In particular, in response to increasing pH, bicarbonate in
solution is converted to carbonate, which can result in the precipitation of carbonate minerals such as siderite
and calcium carbonate. These have the potential to form surface coatings that could reduce activity and
ultimately clog pores, resulting in loss of permeability. In pilot-scale laboratory tests reported by Mackenzie
et al. (1995), there was no apparent loss of reactivity of the iron, though pore blockage was identified as an
issue requiring further investigation. A further contribution of the 1995 ACS symposium was the
identification of materials with much higher reaction rates than iron. For example, Korte et al. (1995), and
later Liang et al. (1997), showed that a bimetallic consisting of palladium plated onto the iron surface could
increase reaction rates, by up to two orders of magnitude.
Current areas of active investigation include fundamental studies of reaction mechanisms and pathways, and
degradation rates and pathways for a wider range of organic contaminants. Work continues to identity
effective bimetal couples and concerns remain regarding the long-term integrity of the catalytic
18
-------
Treatment Walls
NATO/CCMS Pilot Study Phase \\\
enhancement. Certainly the most pressing practical issue is the long-term performance of zero-valent iron
in natural subsurface environments. Various research groups are continuing to investigate the changes over
time in the surface characteristics of iron in contact with various types of groundwater, and the nature and
consequences of precipitates that may form.
Commercial Development
As discussed in NAS (1997) and summarized in Macdonald (1997), many companies developed for the
purpose of marketing new environmental technologies have had a remarkably poor record of success. In
particular, the requirement for regulatory acceptance of new technologies and the fact that the incentives
for site owners generally encourage the delay of remedial action are significant contributors to a sluggish
environmental market. The in situ iron technology is marketed by EnviroMetal Technologies Inc. (ETI)
under a licensing agreement with the University of Waterloo. ETI was incorporated in 1992, at a time when
there was a high degree of skepticism. In addition, because capital costs of installing an in situ iron treatment
system are not particularly advantageous, the main financial incentive is in greatly reduced long-term
operating and maintenance costs. Clearly, in 1992, it was not possible to demonstrate these savings.
The first important contribution to market access was the development, in 1992, of an in situ demonstration
at Canadian Forces Base Borden. The early results of that test showed that the dechlorination reaction would
indeed proceed in situ, and each passing year of consistent performance, raised the level of confidence in
the efficacy of the technology. As reported in O'Hannesin and Gillham (1998) this facility performed
consistently and effectively over the five- year duration of the test. The second major contribution to market
development was the first in situ treatment system at a commercial site Yamane et al., 1995). This occurred
in late 1994 and early 1995, at a time when general knowledge of the technology was at a low level and
long-term in situ performance remained as a significant unresolved question. The interest of the client in
new technologies, and willingness to support the testing required to develop the design was critical to this
initial application, and a significant contribution to market development.
Another important development was the identification of large quantities of granular iron available at
reasonable cost. Currently there are three major suppliers, all of whom recover cuttings from metal
machining and fabrication operations, treat and grade the materials and resell the product for a variety of
commercial uses. These materials have proven to be highly effective as the reductant in the dechlorination
process.
While regulatory concerns can slow technology implementation, regulators, particularly the EPA, have
contributed in a major way to the growing acceptance of the iron technology. In addition to favorable results
reported by EPA researchers, an above-ground demonstration installed in New Jersey in 1995 was
monitored under the EPA SITE program, resulting in a favorable report (US EPA, 1997). Similarly, an in
situ demonstration in New York State promises to result in a favorable report under the SITE program
(anticipated in 1998). Inclusion of PRBs, as a Working Group under the Remediation Technology
Development Forum has also played an important role in disseminating information concerning PRBs
generally and the use of granular iron in particular. ,
ETI currently reports the installation often demonstration facilities and twelve full-scale treatment facilities.
While the numbers remain small, the trends suggest a growing awareness and acceptance of the technology.
In particular, all recent installations have been full-scale, the time between initial contact and
implementation is decreasing and the number of inquiries continues to increase.
19
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
References
Gillham, R.W. and O'Hannesin, S.F., 1992. Metal-catalyzed abiotic degradation of halogenated organic
compounds. IAH Conference: In Modern trends in hydrogeology, Hamilton, Ontario, May 10-13 pp
94-103.
Gillham, R.W. and O'Hannesin, S.F., 1994. Enhanced degradation of halogenated aliphatics by zero-valent
iron. Ground Water, Vol.32, No.6, pp. 958-967.
Gillham, R.W. and Bums, D.R., 1997. Recent developments in permeable in situ treatment walls for
remediation of contaminated groundwater. Chapter 21: Subsurface Restoration, Eds. Ward, C.H.,
Cherry, J.A. and Scalf, M.R., Ann Arbor Press, Inc., Chelsea, Michigan, pp 343-356.
Liang, L., Korte, N., Goodlaxson, J.D., Clausen, J., Fernando, Q. and Muftikian, R., 1997. Byproduct
formation during the reduction of TCE by zero-valent iron and palladized iron. Ground Water
Monitoring and Remed., Winter, pp.122-127.
Macdonald, J.A., 1997. Hard times for innovative cleanup technology. Environmental Science and
Technology, Vol.31, No. 12. pp.560-563.
Mackenzie, P.D., Baghel, S.S., Eyholt, G.R., Horney, D.P., Salvo, J.J. and Sivavec, T.M., 1995. Pilot-scale
demonstration of reductive dechlorination of chlorinated ethenes by iron metal. In proceedings of 209th
American Chemical Society National Meeting, Anaheim, CA, Vol.35, No.l, pp.796-799.
Matheson, LJ. and P.O. Tratnyek, 1994. Reductive dehalogenation of chlorinated methanes by iron metal.
Environ. Sci. Technol., Vol.28, No. 12, pp. 2045-2053.
National Academy of Sciences, 1994. Alternatives for ground water cleanup. Report of the National
Academy of Science Committee on Ground Water Cleanup Alternative. Washington, D.C., National
Academy Press.
National Academy of Sciences, 1997. Innovations in Groundwater and soil cleanup: From concept to
commercialization. National Research Council,
Committee on Innovative Remediation Technology, National Academy Press, Washington, D.C.
Odziemikowski, M.S., T.T. Schuhmacher, R.W. Gillham, and E.J. Reardon, 1998. Oxide Film Formation
on Iron in Simulating Groundwater Solutions: Raman Spectral and Electrochemical Studies. Corrosion
Studies (in press).
O'Hannesin, S.F. and Gillham, R.W., 1998. Long-term performance on an m situ "iron wall" for
remediation of VOCs. Ground Water, Vol.36, No.l, pp. 164-170.
Orth, S. and Gillham, R.W., 1996. Dechlorination of trichloroethene in aqueous solution using Fe°.
Environmental Sci. and Technol., Vol.30, No.l, pp.66-71.
Reynolds, G.W., Hoff, J.T. and Gillham, R.W., 1990. Sampling bias caused by materials used to monitor
halocarbons in groundwater. Environmental Science and Technology, Vol.24, No.l, pp.135-142.
20
-------
Treatment Walls
NATO/CCMS Pilot Study Phase HI
Roberts, A.L, Totten, L.A., Arnold, W.A., Bums, D.R., and Campbell, T.i, 1996. Reductive elimination of
chlorinated ethylenes by zero-valent metals. Environmental Science and Technology, Vol. 30, No.8,
pp.2654-2659.
Sivavec, T.M. and Horney, D.P., 1995. Reductive dechlorination of chlorinated ethenes by iron. In
proceedings of 209th American Chemical Society National Meeting, Anaheim, CA, Vol.35, No.l,
pp.695-698.
Totten, L.A. and Roberts, A. L., 1995. Investigating electron transfer pathways during reductive
dehalogenation reactions promoted by zero-valent metals. In proceedings of 209th American Chemical
Society National Meeting, Anaheim, CA, Vol.35, No.l, pp.706-708.
Weber, E.J., 1995. Iron-media reductive transformations: Investigation of reaction mechanism. In
proceedings of 209th American Chemical Society National Meeting, Anaheim, CA, Vol. 35, No.l,
pp.702-705.
Yamane, C.L, Warner, S.D., Gallinatti, J.D., Szerdy, F.S., Delfmo, T.A., HanMns, D.A., Vogan, J.L., 1995.
Installation of a subsurface groundwater treatment wall composed of granular zero-valent iron. In
proceedings of 209th American Chemical Society National Meeting, Anaheim, CA, Vol.35, No.l,
pp.792-795.
United States Environmental Protection Agency, 1997. Metal-enhanced dechlorination of volatile organic
compounds using an aboveground reactor: Innovative technology evaluation report, Office of Research
and Development, Washington, D.C., EPA/540/R-96/503.
Discussion
Cores collected from two commercial sites showed little change in bacterial activity over the upgradient
interface; however, no information was collected for the downgradient side of the treatment wall.
Although there were numerous monitoring wells situated throughout the Borden site, there was no
noticeable affect on the hydrologic flow regime. Most are small-diameter shallow wells that were installed
for another purpose prior to constructing the wall.
Treatment walls can be used in low-permeability or fractured-bedrock aquifers if there is a measurable
gradient. Fractured bedrock is more difficult, although Gillham mentioned a treatment wall currently being
installed at a site with fractured bedrock. The fractures are being sealed by jet grouting, and reactive
treatment materials are being jetted into the fractures. Hydrofracturing can be used to widen fractures before
injecting materials.
There is a sharp pH gradient within the iron wall. The pH-neutral groundwater typically increased to pH 9-
10 within the first 10-15 cm of the upgradient side of the wall.
21
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Practical Solutions for the Treatment of Polluted Groundwater
Gerard Evers1
The remediation of polluted sites and its treatment constitute a complex problem, and every case is special.
Many solutions are proposed to clean or to isolate pollutants, such as biological treatment, soil washing or
flushing, electrochemical treatment, vacuum extraction, stabilization, or containment.
Matching a site and a remediation technique depends on many parameters. Design studies for the
improvement of these sites take into account not only the chemical context, but also the geology and
hydrogeology as well. In many cases, it is necessary to preserve the groundwater flow.
Soletanche Bachy developed several innovative solutions in this field. Drainage trenches, used either alone
or in combination with cut-off walls, allow controlling and extracting pollutants while the local groundwater
regime is maintained.
Extensive research over the last ten years about the behavior of dissolved pollutants, organic and mineral,
ended up with a variety of materials able to trap pollutants in drainage conditions (ECOSOL and IRIS
materials), and in a process of underground water seepage control and treatment called the "drain panel
process."
Trapping mechanisms
Several physical and chemical mechanisms are involved to trap pollutants in groundwater. Precipitation,
adsorption, and ionic exchange are the most important.
Precipitation
Chemical precipitation is mainly used to eliminate dissolved heavy metals, such as iron, nickel, copper, lead,
trivalent chromium, and hexavalent chromium.
These cations are often precipitated in the form of metallic hydroxides in chemical reactions controlled by
the pH. In cement-based materials, lime solubility dictates a pH of 12-12.5. However, this reserve of alkali
in cements, although large, is not infinite, and alkaline buffers may be incorporated in trapping materials
to extend their precipitation ability with time. In some cases, specific reagents are employed to react with
pollutants and to precipitate into a insoluble mineral forms.
Adsorption
Adsorption is a physical mechanism based on the properties of some porous materials to fix molecules on
their surface. Specific surface governs this mechanism. Attractive forces have different origins:
• physical bonds by pores of similar sizes to the target molecules;
• Van der Waals forces (electrical trapping) is applicable to polarized molecules;
• surface affinity—hydrophobic organic molecules have affinities for sites on activated carbon.
1 6 Rue De Wattford, F92000 Nanterre, France, tel: 33/14-7764-262, fax: 33/14-9069-734, e-mail: gerard.evers®
soletanche-bachy.com
22
-------
Treatment Waffs
NATO/CCMS Pilot Study Phase »)
Adsorption on activated carbon is a widely used method for removing organic pollutants dissolved in waste
water. Adsorption capacity values vary depending on the compounds to be removed. For instance,
adsorption capacity of chlorophenol is six times higher than that of butylacetate in same conditions. Some
clays are able to adsorb cations such as cadmium, strontium, mercury, nickel, and zinc.
Ion exchange
This process is employed for removing dissolved anions in polluted groundwater. Removal method can be
percolation through specific resins or clays. This mechanism is well adapted to eliminating cyanides. Two
processes are described here to implement these pollutant trapping drainage materials: the drain panel and
the drainage trench.
The Drain Panel Process2 ,
In order to avoid spreading an industrial or accidental pollutant, watertight confinement barriers are often
installed together with a pumping system, to keep the water table within the confined area below the outside
water table. A water treatment installation will clean the pumped water before returning it to the ambient
groundwater. But this classical concept has several inconveniences, which can be solved by the drain panel
process.
The drain panel process consists of drainage panels separated by watertight cement slurry panels. A conduit,
at the bottom of the watertight panel, establishes the hydraulic continuity between the two drainage panels
on either side. A valve, operated from the surface, can be installed to regulate this hydraulic continuity;
The drain panel can be built in combination with the classical confinement barrier, or in combination with
a partial barrier, which will then form the so-called "funnel-and-gate" solution. The drain panel system can
be used to considerable depths (30 meters or more).
Construction Method
The watertight cement slurry panel is excavated first. When the excavation is completed, a horizontal
conduit (200 to 400 mm diameter), linked to a vertical standpipe, is installed at the base of the panel. This
conduit has destructible parts closed off with caps at both ends. A temporary inflatable plug avoids any flow
through the conduit. __„_„„__„__„ „
When the slurry has set, the adjacent drainage panels can
be excavated using a biodegradable drilling mud. The
sealing slurry, vertically covering the destructible parts
located at the ends of the conduit, is excavated, and the
destructible parts are cut, either by the excavation tool or
by another suitable method.
?2| When the excavation of the drainage panel is completed,
it is filled with appropriate drainage material, and
conductivity between drainage panels is established by
withdrawal of thd temporary inflatable plug (next three
figures).
1ECOSOL and DRAIN PANEL are patented.
23
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
| Drainage Material
Drainage panels can be tilled with trapping materials
adapted to the nature and content of organic or mineral
pollutants (figure below).
If required, successive drainage panels with different
retention filters can be installed to clean the ground-
water. In this case, appropriate positioning of the
conduits in the slurry panel may regulate the flow and
create an efficient volume for the active barrier.
Filter Cartridges
In this case, the whole treatment system, with conduits,
filter retention tanks and monitoring equipment, is
installed in the cement slurry (see figure at right).
Upstream, intermediate, and downstream piezometers
allow permanent monitoring of the pollutant concen-
tration at each point of the treatment system. The filter
retention tanks are designed for an easy exchange of
the filter cartridges when they are saturated.
24
-------
Treatment Walls
NATO/CCMS Pilot Study Phase \\\
Other Applications for Drain Panels
Underground construction often requires temporary, watertight, cut-off walls around the excavation pit. But
without precautions, such cut-off walls will create a permanent perturbation in the groundwater flow, and
consequently will have a lasting influence on the hydraulic environment.
The drain panel system will probably reestablish the
hydraulic continuity between the upstream and down-
stream aquifers.
For a temporary cut-off wall, drain panels can be
installed at different points along the perimeter. Each
drain panel is composed of a central watertight cement "—*••»*«»«-**•"«*•
slurry panel and two drainage panels (one at each side of
the cut-off wall). While lowering the water in the
excavation pit, hydraulic connectivity through the drain
panel is avoided by temporarily plugging the conduit.
When construction is completed, the hydraulic
connectivity between upstream and downstream can be Walorlablare<^b)lslMKluponcomp,ellonofworits
reinstated by withdrawing the plug (figures at right).'
A possible solution for permanent works is shown on the
figure below. Drain panels are installed at each part of
the work, and the hydraulic continuity is reinstated, by
installing a drainage pipe at slab level and connection
with the drainage panels. - . . . •-
Horizontal
pipe
Slurry wall
I
- Closed during work phase i
- Open to ro-ortabllth watar I
tablo circulation I
• Ctoaod for pollution alert .
PhnWtw
Ptanvfew
Treatment of Chrome Pollution by Pollutant-Trapping
Drainage Trench
The purpose of the pollutant-trapping drainage trench is to
. | intercept polluted groundwater and treat it in situ. The
trench does this without maintenance, staff, or energy
input. The example given here after was applied to solve a
chromate-polluted groundwater problem.
Pollution Source and Type
The pollution was caused by materials containing hexa-
valent chrome being used in the construction of the A22
motorway in northern France, near Lille.
The toxicity of the material, which was known at the time
of construction, prompted engineered safeguards in the fill.
Unfortunately, the design of the central platform did not
completely control seepage. Soon after the road was
25
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
opened, there were indications of chrome (VI) in the seepage water at some spots.
In 1987, a company located below the motorway (downgradient of groundwater flow) reported chrome (VT)
in its drainage water, and this pollution was accompanied by bare patches in the vegetation of the motorway
bank.
Site Characteristics
The ground under the road embankment consists of 4-5 meters of sandy silt overlying Flanders clay. The
water in this silt flows gently towards a banked slope some 10 m downgradient of the fill, directly
overlooking the company car park.
Over time, rain infiltrating into the fill leached into the soil. This water, contaminated with chrome (VI),
percolated into the aquifer despite the layer of rolled shale at the base of the fill. Water in the toe ditch below
the bank had the characteristic lemon yellow coloring of chromate pollution. In dry weather, patches of
yellowish efflorescence appeared on the slope. Seeps of polluted water that killed off the vegetation could
be seen on the slope above the company car park.
The site was surveyed to explore the extent of the pollution. Auger holes were sunk for soil samples.
Chemical analysis revealed serious soil pollution, with chrome (VI) present in concentrations of up to
1,000 mg/kg, but disappearing beyond a depth of 5-6m
because of the impervious Flanders clay bed. *~
Groundwater samples taken from piezometers yielded
very high chrome (VI) contents, practically up to
2,000 mg/1. Chrome (VI) is soluble and highly toxic,
and the maximum permitted level in drinking water is ^
0.05 mg/1. oof
•*?
Engineering Design
The proposed solution consists of intercepting the
polluted groundwater with an Ecosol drainage trench
designed to trap chrome (VI). More precisely, a trench
is dug and backfilled with a pervious, porous,
pollutant-trapping material (figure at right).
The required material is obtained by mixing sorted
gravel with a special slurry from the Ecosol range.
These slurries are special formulations developed and
*t*
Sfijjs *,
26
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
patented by Soletanche Bachy to suit the exact nature of the pollutants in question: radioactive elements,
heavy metals, or organic compounds.
The gravel grains coated with pollutant-trapping slurry bond together so when the mix hardens, they form
a rigid, porous structure through which water readily circulates. As the water seeps through, the chemical
pollutants are trapped on the film of hardened Ecosol formulation adhering to the grains of gravel.
Preparation of Ecosol Material
Several crucial parameters must be controlled when
preparing the Ecosol material (see figure at right):
• Slurry/grovel weight ratio. This ratio must not be
too high, or the structure will not be sufficiently
porous. Nor must it be too low, or the gravel grains
will not be completely coated.
• Gravel size. For any given ratio between a unit
volume of slurry and the weight of gravel, a finer
gravel size will increase the active surface area of
the structure and reduce the thickness of the hardened slurry coating the gravel grains. This speeds up
the trapping reactions. On the other hand, it reduces the permeability of the structure, with the attendant
risk of the pores becoming clogged.
• Selection of gravel and slurry. The bond between the film of hardened slurry and the gravel skeleton
must be sound and durable. The use of cement slurry and compatible,gravels ensures durability,
provided the same precautions are used as when designing standard concrete mixes.
• Rheological properties of slurry. The rheology of the freshly-mixed slurry paste must be designed to
form a stable film coating the gravel. Therefore, apart from the specific pollutant-trapping substances,
the mix must be designed so that the initial stiffness of the paste when the gravel is incorporated is not
too great, nor too low.
• Slurry formulation. All types of Ecosol slurry can be used, provided their rheological properties are
suitable. The exact formulation of the slurry depends on pollutant types and concentrations and required
trapping capacity. The mix design is done in the laboratory. In this instance, the cumulative quantity of
chrome (VI) trapped per cubic meter of material could be determined. The Ecosol material is prepared
by adding the various reagents in the composition at the mixer. This slurry is then used for coating the
gravel. This is done in a truck mixer, so that the sorted gravel is delivered, the slurry added, the mix
blended and the material delivered to the point of use all in one vehicle.
Trenching
On site, the trench is dug by a machine capable of
digging to a depth of six meters. The mobile frame
carries a chain and square-sided chute to guide the
backfill material from the top hopper into the trench.
The Ecosol-coated gravel is brought in by the truck
mixer to the trench hopper, and the blended material
flows like concrete. The result is a trench backfilled with
Ecosol-coated gravel.
27
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
There were two zones: (A), where groundwater is
58 intercepted by two trenches, because this area has the
highest chrome (VT) concentrations; and (B), where the
flow is intercepted by a single trench.
The 0.25-m wide drainage trenches were widened at
the top. This extra thickness of trap potential deals with
surface water contaminated with chrome (VI) liable to
seep into the aquifer.
Trench Performance Monitoring
Three piezometer profiles perpendicular to the trench monitor the trend in chrome (VI) concentration up-
and downgradient of the of the trench. In total, 11 piezometers were installed, and each month, their water
levels are recorded and water samples taken for laboratory analyses.
Analysis of these data confirm the existence of an hydraulic gradient, perpendicular to the trenches, from
the embankment downwards to the car park location.
The following figure shows the chrome concentration upstream and downstream of the barrier, during the
18-month observation period following the completion of the construction. The figure shows the good
performance of the curtain: the chrome concentration remains at a high—but variable—level upstream of
the curtain, while negligible concentrations are recorded downstream.
The measurements are still going on, and they confirm these findings. The Ecosol material confers a high
trap potential, around 15 kg of chrome (VI) per linear meter of trench.
Evolution of the croma Cr6+ concentration
upstream (PZ15/16) . between (PZ17) and downstream (PZ18)
10
28
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
Discussion
In response to a question, Evers explained that the treatment wall was excavated using a trenching machine
capable of excavating only to a depth of 10 m. The overall problem with the permeable treatment wall is that
the reactive material is expensive (US$800/m3), but not all of it is receiving contaminated groundwater.
Because the soil is not uniform, there are preferential flow pathways, and approximately 10% of the wall
handles 90% of the contaminated groundwater. The advantage of the funnel-and-gate system is having
successive reactive materials that receive contaminated groundwater flow.
There are at least three options for constructing a treatment wall to optimize cost-efficiency: (1) construct
one continuous trench and fill it with a mixture of iron and sand; (2) use pea gravel screens upgradient to
average the flow rate through the wall; or (3) construct a wall of varying thickness to account for variations
in groundwater flow. As long as substantial information on groundwater flow was available, plans can be
made for treatment walls having variable thicknesses. Generally, however, funnel-and-gate technology has
an advantage in evening out flow through the wall, but may magnify uncertainties.
There was some concern that if the hydrology of a site is not understood, use of a funnel-and-gate system
can retard normal groundwater flow, causing an increase in upgradient thickness of the water. The
hydrology of a site must be modeled carefully prior to constructing a treatment wall. Modeling and the
design of a funnel-and-gate system can be very expensive, perhaps accounting for up to 20% of the overall
construction costs (compared to only 10% for slurry walls).
29
-------
Treatment Walts
NATO/CCMS Pilot Study Phase 111
Degradation of TCE at Zero-Valent Iron:
Chemical Processes Effecting the Design and Performance
of Permeable, Reactive Fe(0) Walls
Wolfgang Wiist, O. Schlicker, and A. Dahmke1
Permeable reactive walls present a cost-effective alternative to classical pump-and-treat technologies for the
removal of dissolved contaminants from groundwater plumes, provided that a fairly passive long-term
performance can be guaranteed. Many materials are under research for use in such reactors, but zero-valent
iron has been studied most intensely, since it is commercially available at low costs, and a variety of
contaminants can be immobilized by reduction reactions (Dahmke et al, 1997). Based mainly on the results
of our own studies, we describe the degradation of chlorinated aliphatic hydrocarbons, the most important
application of Fe(0) reactors, and in this context, we discuss the chemical processes effecting the design and
performance of permeable reactive Fe(0) walls.
Degradation of chlorinated hydrocarbons with zero-valent iron is known to be a heterogeneous surface
reaction, and thus dechlorination can only take place after the chlorinated compound has been associated
at the iron surface (Matheson and Tratnyek, 1994). Kinetics is described as pseudo-first-order (k^ [1/h])
with respect to dissolved concentrations, and k^ is reported to be proportional to the iron surface (BET)
concentration (Gillham and OUannesin, 1994; Sivavec and Horney, 1995). Different commercially
available irons show a broad variety in costs as well as physical properties (Figure 1).
Figure 1. Comparison of different irons
Iron
source
A
B
C
D
Spec, surface area
[m2/g]
0.12
0.5
0.032
0.63
Porosity
[%J
37
58
39
0.74
Costs
[DM/t]
800
450
500
300
Costs
[DM/m3]
3,400
1,400
2,300
600
Testing iron from different producers, we found
considerable differences in reactivity with respect
to TCE degradation. Differences in iron surface
area concentrations could not sufficiently account
for it, since we normalized k,,,,, to this parameter
(Figure 2). The normalized first-order rate
constants were in the same range as reported for
different iron sources in North America (Johnson
et al., 1996).There a various other parameters like
foreign elements and grain geometry that may
affect reactivity. Even iron from the same source
showed differences in reactivity.
Degradation of TCE at different com mercial irons
AF. [mVg|
Figure 2.
In low mineralized waters competitive, cathodic
reactions may occur that can affect the performance of permeable, reactive Fe(0) walls. Oxygen
1 Institut fiir Wasserbau, Lehrstuhl fur Hydraulik und Grundwasser, University of Stuttgart, Pfaffenwaldring 61,70550
Stuttgart, Germany, tel: 49/711-685-4714, fax: 49/711-685-7020, e-mail: ww@iws.uni-stuttgart.de
30
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
Gas development at different commercial irons (a, b, d)
in column experiments (O2-saturation, low min. water)
Figure 3.
Composition of gases from iron source b
(initial pH 7, low m in. water)
Sum HC CH4
C2H4 03H8 C3H6 C4H10 HC>C4
Figure 4.
Composition of gases from iron source d
(initial pH 7, low min. water)
d (no TOE) Sd (TOE in: 0.1 mMol)
I II II
Sum HC CH4
C2H4 C3H8 C3H6 C4H10 HC>C4
Figure 5.
consumption was found to be fast in our
column experiments, leading to a zone of
Fe(in) precipitates at the column inlet,
where hydraulic conductivity was reduced
considerably (Dahmke et al. 1997). Effects
on the degradation kinetics in Fe(0) reactors
were not detected, since only a few
centimeters are affected by aerobic
corrosion. Formation of hydrogen by
anaerobic corrosion is thermodynamically
favorable, but it depends strongly on the
voltage differential. For different irons, we
found dependence of reactivity on the iron
source, and pH at the inflow ranging from
no anaerobic corrosion to a strong gas
production (Figure 3), which affects
hydraulic conductivity and discharge of
gaseous TCE from the column. Hydrogen is
the dominating compound in the gas phase,
and low-weight aliphatic hydrocarbons (total
HC) consisting exclusively of saturated HC
account for less than 1 % mol gas, fairly
independent of the iron source (Figure 4,
Figure 5). TCE, at a concentration of 1
mMol in the column feed, increases the
molar fraction of the total HC to 1-3 %, and
short-chain unsaturated compounds (ethene,
propene) appear in the HC pattern,
accounting for TCE degradation. Further
column experiments discussed below were
performed with iron from a source that was
found to have the highest voltage differential
with respect to anaerobic corrosion, and a
fairly high normalized
Degradation kinetics of TCE for different inflowing concentrations
1500
• 1000
Looking at the dechlorination kinetics in more detail,
we found that the degradation rate was levelling off
at a higher concentration due to saturation of reactive
sites at the iron surface. An enhanced model
accounting for first-order sorption, desorption, and
reaction of sorbed TCE (Figure 6, Figure 7) was
fitted to the data with good success by the
AQUASM model. Parameters for zeroth- and first-
order kinetics with respect to dissolved TCE-
concentration are estimated. Since the dominant
chlorinated intermediate, cis-DCE, did not exceed
1% of the initial TCE concentration, first-order
Figure 6. build-up and degradation was found to be sufficient
to fit the measured data. Neither the intermediate cis-DCE nor the kinetics of parent TCE were significantly
affected by a 5-fold increase in flow velocity (Figure 8). Cis-DCE as parent compound was degraded faster
than TCE both in the zeroth-order and first-order region (Figure 9) in the column experiments. This is the
20
time[hl
40
31
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Kinetic model for the parent compound (TCE)
(1st-ordersorption, desorption and reaction)
Zerolh-order region:
[TCE,]»Ki/2^>kQ,
First-order region:
[TCEi]«Ki,2 *> ku
[ftMol / Lh]
[1/h]
ikot, k*t rate constants for first-order sorption, desorption and
reaction of the sorted compound [1/jiMol h], [1/h], [1/hJ
t concentration of reactive sites at FeQiMol/kg]
kon mass concentration [kg/L]
, kn zeroth-order and first-order degradation rate constants with
respect to UquW TCE [nMol/h], [1/h]
half max, conversion rate parameter [nMoi]
Kinetic model for the chlorinated intermediate
face fraction of sorbing TCE contributing to build-up of cis-DCE
t first-onter rate constant with respect to dissolved cis-DCE [1/h]
Figure 7.
Concentration of parant TCE and intermediate cis-DCE at
commercial iron (a) at different flow velocities
time [h]
Figure 8.
First-order degradation of TCE and cis-DCE
in single and binary component experiments
0 10 20 30 40 [h]
-1 •
1 -2-
1-3.
-4-
-5-
: ^5^^^_
^^^•••Cr " ^T^8
• TCE '•-,_
• dsJTCE ''i. tj^oCE
»TCE:DCE=1:1
• DCE:TCE=1:1 >'•.. .
• u
"1 3
2 S
-3 I
s.
=
•5
Figure 9.
reverse of the kinetics reported for these compounds
(Johnson et al., 1996). In the binary component
experiment, degradation rates of both compounds were
reduced and competition between TCE and cis-DCE
will be discussed in more detail (Wust et al, in
preparation). Since groundwater plumes contaminated
jfAwj./ui.uvAv/ii/. v^juiww gii/uiiuwait/i piuiu&a L/uiiiaiiiiijialC'U
by chlorinated hydrocarbons usually contain a variety of compounds, competition between these single
compounds has to be considered.
We were also testing the effect of different groundwater constituents (chloride, phosphate, sulfate, chromate,
nitrate) on the degradation of TCE in column experiments, where the dissolved concentration of the
respective groundwater constituent is gradually increased.
TCE and nitrate reduction at commercial iron
0.8
— NO3 ---Mm -«-TCE -*-pH|
•10
Abiotic nitrate (Figure 10) and chromate (data not
shown) reduction at zero-valent kon were found to
be faster than the degradation of TCE. Nitrite as an
intermediate could be detected only in traces, and all
nitrate-N was reduced to the terminal N(-in).
Reduction of 50 mg/1 nitrate (German drinking water
standard) can result in 13.3 mg/1 ammonium entering
the aquifer. Since pH is driven up to values greater
than 10, ammonia is likely to form. High inflowing
nitrate concentration leads to inhibition of both the
reduction of nitrate and TCE by zero-valent iron.
Consequently, the TCE concentration front is Figure 10.
passing through the kon column ahead of the nitrate front (Figure 11). Passivation is due to the formation
of a considerable amount of kon precipitates (dissolved kon <0.01 mg/1). They are possibly more oxic due
to the higher redox potential both in solution and at the kon surface, and thus less conductive.
0.0
20
time [h]
40
32
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
Effect of nitrate reduction on the degradation of TCE
(PV: pore volumes permeated at 50 mg/L nitrate)
TCE(noNO3-) TCE(5PV)\TCE(10PV)
Effect of inflowing phosphate poncentration on the
degradation of TCE at Fe
100.00
1 10
PO» in [mg/L]
Figure 11.
Figure 12.
Phosphate can affect the performance of iron walls only at higher inflowing concentrations (>1 mg/1
phosphate). At 10 mg/1, we observed a decrease in both, zeroth- and first-order rate constant (Figure 12).
100-mg/l phosphate in the feed further reduced zeroth-order degradation but so far did not effect first-order
degradation in the back part of the column. Phosphate is sorbing, and forms vivianite, which was detected
by XRD in addition to magnetite. Formation of these precipitates could be predicted by the geochemical
equilibrium program PHREEQEC. In contrast to phosphate, loss of sulfate from solution could not be
observed. Considerable decrease in the degradation of TCE, however, was found at a concentration of 500
mg/1 sulfate (Figure 13). Higher sulfate concentration (1,000 mg/1) did not enhance the passivation of iron.
Effect of SO, at various Inflowing concentrations on the
degradation of TCE at Fe In column experiments
1000 mg/L SO4
100ms/LSO, soo mg/L SO.
>^^— ^. V
no SO, (po»l) , ^^^JSa^-.. ^--V>--.
no SO4 (anta)
time [h]
Effect of microbiological sulfate reduction
on the degradation of TCE
£.
.2
Pore v.olumaa slnca Infiltration of
1200 mg/L SO4a'+ 600 mg/L HCO»*+ 100 mg/L NH«Ci
Figure 13.
Figure 14.
Sorption of sulfate was found to be reversible, and the column fully recovered after its desorption from the
iron surface.
Abiotic sulfate reduction was not observed, but microbiological transformation to sulfide took place in one
of our columns to some extent. There we found degradation rates to increase (Figure 14), which we attribute
either to the enhancement of abiotic degradation because of sulfide formation (high electric conductivity)
or to co-metabolic TCE degradation. Enhancement of reactivity was also attained by palladizing zero-valent
iron. First-order kinetics was stable over a wide range of flow velocities and no maximum degradation rate
(zeroth-order kinetics) could be detected even at high inflowing TCE concentrations (Figure 15). Increasing
chloride concentrations confirmed the results known from aerobic anaerobic corrosion studies: the first-
order rate constant increased by 2 in the range of 0-100 mg/1 chloride (Figure 16).
In a low-mineralized grouhdwater environment, we expect long-term stability of dechlorination rates in
permeable, reactive Fe(0) walls and hydrocarbons as final products enter the aquifer. Description of the
degradation kinetics by a pseudo-first order model was not sufficient at higher TCE concentrations, where
zeroth-order kinetics was observed. We succeeded in fitting the data by a Monod-type model, Competition
33
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Dagridatlon of TCE at Palladizod Iron (0.01% Pd)
0 1 2 3 4 S • 7 S 0 10 11 IS 13 14 15
•19
*'
•to
.188
time [h]
Ellect of inflowing chloride concentration on the halt life
olTCE In Iron walla
£
£
O 11 32 41 44 50 61 09 69 74 77 70 SS
Figure 15.
Figure 16.
among different chlorinated compounds has to be expected as shown in the case of TCE and cis-DCE, and
has to be considered in the reactor design.
Redox-competitive inorganic groundwater constituents (oxygen, water, nitrate, chromate) can affect the
design and performance of reactive iron walls considerably. Products like hydrogen gas and ammonia can
be formed and enter the aquifer. Iron precipitates can reduce permeability and reactivity. Under certain
conditions, the concentration front of the chlorinated compounds is passing through the reactor due to
passivation of zero-valent iron. These reactions may limit the scope of permeable reactive Fe(0) walls.
Inhibition observed by phosphate and sulfate seems to be of minor importance. At ambient groundwater
concentrations, phosphate did not effect degradation kinetics of TCE, since it was precipitated at the front
of the column. No abiotic reduction was observed for sulfate (as for nitrate and chromate), but degradation
rates were enhanced by microbiological sulfate reduction.
We expect that microbiological reduction of nitrate and chromate in the groundwater plume can reduce the
impact of these anions on the performance of reactive iron walls. Passivated walls can possibly be
regenerated by chloride, which is known to break down passive films on zero-valent iron, and which we
found enhanced degradation rates in our experiments. We further assume that ascorbic acid will recover
degradation rates due to its reducing and complexing properties.
References
Dahmke, A., Schlicker, O. & Wiist, W., 1997. Literaturstudie ,,Reaktive Wande-pH-Redoxreaktive Wande"
Berichte LiU.
Matheson, L.J. & Tratnyek, P.O., 1994. Environ. Sci. Technol. 28, 2045-2053.
Gillham, R.W. & OUannesin, S.F., 1994. Ground Water, 32,958-967.
Sivavec, T.M. & Homey, D.P., 1995, Prepr. Ext. Abs. ACS Special Symposium, Anaheim, 695-698.
Johnson, T.L., Scherer, M.M. & Tratnyek, P.G., 1996. Environ. Sci. Technol. 30,2634-2640.
Wast W., Schlicker, O. & Dahmke A. in prep. Kinetics of the Stepwise Degradation of TCE and cis-DCE
at Commercial Iron in Batch and Column Experiments.
Dahmke, A., Schlicker, O. & Wiist, W., 1997. In: Grundwassersanierung 1997, IWS SR 28, 324-341.
Schlicker, O., Wiist, W. & Dahmke, A., 1998. TerraTech, 01,43-46.
34
-------
Treatment Waffs
NATO/CCMS Pilot Study Phase TO
Discussion
Paul Bardos asked for an explanation of Figures 3, 4, and 5, which show a system with just oxygenated
water and iron producing trace amounts of short-chained hydrocarbons. Wiist said that the iron contained
3 percent carbon, which may have been the source of the hydrocarbons. Alternatively, there may have been
organics in the gas phase that were reduced.
Hermann Schad also asked for an explanation of the apparent conversion of 100% of the nitrate to ammonia,
while column experiments using site groundwater showed that much less nitrate had been converted. Wiist
responded that the difference is probably due to microbiological reactions that reduce more of the nitrate.
35
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
The Treatment of Groundwater with Mixed-Wastes:
Reductive Dechlorination of TCE and Reductive Precipitation of Uranium
Liyuan Liang1 and B. Gu2
Abstract
The overall goal of this project was to perform a pilot-scale field evaluation of the reactive barrier
technology at the Bear Creek Valley, Tennessee. The groundwater plume at the site contains mixed wastes
of uranium (U), technetium, organic solvents and a high concentration of nitrates. This study was undertaken
to determine the effectiveness of zero-valent iron (Fe°) and several adsorbent materials for the removal of
radionuclides, nitrates and chlorinated volatile organic compounds from the contaminated groundwater.
Several types of reductive and adsorbent materials were evaluated, including Fe° filings, peat materials,
ferric oxides, and Cercona™ Bone-Char. Results indicate that Fe° is a promising barrier material to extract
U from contaminated groundwater and is more effective than the adsorbents, particularly at high U
concentration. A combination of Fe° and peat materials was effective in removing both nitrate and U from
the contaminated groundwater. Organic solvents, such as trichloroethylene (TCE), trichloroethane and
tetrachloroethylene, were all effectively degraded in the contaminated groundwater with Fe°
For uranium, almost 100% was removed through reactions with Fe° at an initial concentration of up to 84
mM (or 20,000 mg U/l). Results from the adsorption and desorption kinetic studies and spectroscopic
studies demonstrate that a reductive precipitation of U with Fe° is the major reaction mechanism. Only a
small percentage (<4%) of uranyl appeared to be adsorbed on the corrosion products of Fe°. Experiments
showed that the reduced U(IV) species on Fe° surfaces could be reoxidized and potentially remobilized if
re-exposed to atmospheric oxygen. However, the remobilization may not occur if the U is sequestered by
co-precipitation with ferric oxyhydroxides. Results of this study demonstrate that Fe° can be an effective
medium for U removal from mixed waste in passive groundwater treatment systems.
INTRODUCTION
In situ permeable reactive barrier technology (based primarily on zero-valence iron, Fe°) has been identified
as a potentially cost-effective, passive treatment technology for contaminated groundwater [1-2]. Interest
in the technology by both private industries and federal agencies has generated extensive research activities
in the last few years. Initial laboratory research has been carried out to determine the rates of the reactions
for both the destruction of chlorinated solvents, such as trichloroethylene (TCE) and the immobilization of
metals (e.g., Cr(VI), Tc(VH), and U(IV)) by Fe° [3-7]. The understanding of the kinetics of contaminant
removal is necessary for designing the residence time and sizing the reactive barrier.
Following the kinetic work, mechanistic studies have mushroomed in an attempt to understand the
pathways for dehalogenation and to identify daughter products [8-9]. Although detailed reaction pathways
and mechanisms have not been determined unequivocally, the reaction is believed to be a heterogeneous
surface reaction [7-10]. Reduction of halogenated compounds is known via hydrogenolysis, in which a
halogen ion is exchanged for a hydrogen ion, consuming 2 electrons [11]. From the known partially
dehalogenated daughter products (such as dichloroethylene (DCE) and vinyl chloride (VC) from TCE),
laboratory studies show that the stepwise dehalogenation via hydrogenolysis indeed takes place on the metal
surface [10, 12-13]. Further study has indicated that a reaction pathway via reductive beta-elimination
'Department of Earth Sciences, University of Wales, Cardiff, P.O. Box 914, Cardiff, CF1 3YE, U.K., e-mail:
liyuan@cardiff.ac.uk
2 Environmental Sciences Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge TN 37831-6036
36
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
occurs with Fe°, producing acetylene [14-15]. With a series of chlorinated ethene or methane, results show
that the less chlorinated compounds, such as DCE and VC, are more difficult to dehalogenate than are the
more highly chlorinated compounds [7]. Alternative bimetallic systems (e.g., plating of small amounts of
Ni, Cu, Zn, and Pd onto Fe°) have been found to accelerate the dehalogenation rates for various volatile
organic compounds. In some cases, the bimetallic systems dehalogenate the compounds that are ineffective
with Fe° alone [16-18].
While numerous studies have been conducted in the laboratory and field to determine the degradation
mechanisms and kinetics for organic contaminants by Fe°, relatively few studies have examined the
potential for using Fe° to remove radionuclides and heavy metals from contaminated groundwater. It is
believed that reductive precipitation is a major pathway for the removal of Cr(VI) and Tc(VII) [5,19-20].
However, for uranium (U) it is still unclear whether the removal of uranyl results from reductive
precipitation or from adsorption onto the corrosion products of Fe° [4,21-23]. Knowing which of these two
mechanisms dominates the reaction is necessary because they largely determine the mobility and fate of
these contaminants in a given geochemical environment. For example, if uranyl is largely adsorbed onto the
hydrous iron oxides or the corrosion products of Fe°, the adsorbed U would be readily desorbed by the
presence of competing ions and complexing agents such as CO32' and dissolved organic matter in
groundwater. Furthermore, a change in the groundwater pH would profoundly influence the adsorption and
desorption behavior of U on ferric oxide surfaces. Alternatively, if U is primarily removed by reductive
precipitation, the precipitated U would not be remobilized unless the groundwater redox conditions are
changed and the reduced U(IV) species are reoxidized.
At Bear Creek Valley, Tennessee, the groundwater is contaminated with uranium, technetium, organic
solvents (such as tetrachloroethylene, TCE, and trichloroehtane) and a high concentration of nitrates. The
overall goal of this project was to perform a pilot-scale field evaluation for treating the mixed contaminant
plume with the reactive barrier technology at the site.
The objectives of this study were to determine the effectiveness of zero-valent iron and several adsorbent
materials in removing radionuclides, nitrates and chlorinated volatile organic compounds from the
contaminated groundwater. Results for organic solvent removal have been discussed elsewhere [33], and
in this paper, we report on the reaction kinetics and mechanisms for the reduction or removal of uranyl [as
UO2(NO3)2] with various iron filings and adsorbent materials. The work was part of the remedial actions
using the permeable reactive barrier technology to intercept and treat U, nitrate, and other contaminants
migrating to the tributaries of Bear Creek at the U.S. Department of Energy's Y-12 Plant located in Oak
Ridge, Tennessee. In particular, there is a desire to remove U from the groundwater because it is a major
contributor to risk down-gradient of the Bear Creek.
EVALUATION OF REACTIVE MEDIA FOR U REMOVAL
Uranium Reaction Kinetics
The reaction rates for uranyl removal by Fe° filings were studied in batch experiments by mixing 2 g of iron
filings with 10-ml U solution as [UO2(NO3)2] at an initial concentration of 4.2 mM (1,000 mg U/l). Four
sources of iron filings were evaluated: (1) Master-Builder iron filings (0.4-1 mm) (Master-Builders, OH),
(2) Cercona cast iron foam (Cercona of America* OH); (3) Peerless iron filings (0.5-1 mm) (Peerless Metal
Powers and Abrasives, MI), and (4) coarse Peerless iron filings (3-12 mm). Aqueous samples were
withdrawn, filtered and U activity was then analyzed by a Liquid Scintillation Analyzer [24].
37
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
y
y
-3|g- Medium Peerlou Fe
me Peerleii Fe
--O-. Ceroona oait Fe
PalUdized Ceroona oaitF
0.0
10 20 30 40 50 60 500 10001500
All four types of Fe° filings studied were
effective in removing uranyl from the aqueous
solution (Figure 1). Over 97% U were removed
within 30 min by all types of Fe° filings and no
detectable amounts of U were left in the solution
phase after about 1 hour of reaction. However,
ferric oxide was not as effective as iron filings in
removing uranyl in solution; only about 15% of
the added uranyl was removed by the ferric
oxide. The reaction rate for uranyl removal by
ferric oxide appeared to be fast, reaching equilib-
rium in less than 1 min (only one data point
could be collected).
Time (min)
These observations indicate that U was removed t r-r :———: ——
K» £«•»•* nv;,u .««, i *i, u -f A Figure 1. Reaction kinetics between uranyl and zero-valent
by feme oxide largely through surface adsorp- iron filings or iron oxide powder |n yeous so|ution
tion, which often exhibits fast reaction kinetics.
In contrast, the reductive precipitation process usually takes longer because it involves the corrosion of Fe°
and an electron transfer process from Fe° to uranyl. Because the U solution was not deoxygenated before
the experiment, a time-lag would be required to establish a low redox potential. The partial U removal by
the ferric oxide could be attributed to the limited availability of surface sites; the adsorption process reached
its maximum when an excess amount of uranyl was added in the solution. Although the Fe° filings have a
relatively low specific surface area (about 1 m2/g) in comparison with ferric oxide powder (10.1 m2/g), the
filings removed almost 100% of uranyl in the aqueous solution after approximately 30 min. The relatively
Slow reaction rates and the relatively high U removal efficiency by Fe° compared to ferric oxide suggest that
reductive precipitation of uranyl is the dominant mechanism for U removal by Fe°.
U Removal Efficiency and Capacity
Because Fe° filings corrode in water, uranyl is expected to adsorb onto Fe°-corrosion products such as ferric
oxyhydroxides. To evaluate the partitioning of U in Fe° and its corrosion products, the following experi-
ments were performed. A 10-ml uranyl solution (42 mM or 10,000 mg/1) was equilibrated with 2 g of Fe°
for approximately 3 weeks on a shaker. The solid and solution mixture was then removed from the shaker,
and the supernatant was immediately decanted. The iron filings were thus separated from the corrosion
products (i.e., particulates) in the supernatant. The amount of U associated with the filterable particles was
then estimated by analyzing U concentrations
in solution before and after samples were
filtered through a 5 um filter.
The U removal efficiency and capacity through
the reductive precipitation and adsorption by
Fe° and several adsorbent materials (ferric
oxide, peat, and Cercona™ bone-char) were
further evaluated in batch studies. Figure 2
shows the typical adsorption isotherms of
uranyl on these adsorbent materials. The initial
slope and the plateau of the adsorption
isotherms define the adsorption affinity (or
efficiency) and capacity [25]. It is evident that
all these adsorbent materials were effective at
removing U at a relatively low solution
300 -
2000 4000 6000 8000
Equil. U Concentration (mg/L)
Figure 2. Uranyl adsorption by adsorbent materials
(Wards peat, ferric iron oxide, and Cercona1" bone Char)
38
-------
Treatment Walls
NATO7CCMS Pilot Study Phase 111
concentration (<50 mg U/l). The initial slopes (i.e., the initial partitioning coefficients) ranged from about
40 to > 10,000 ml/g, .with the Cercona™ Fe-bone-char being the most effective at a low initial concentration.
However, the Wards peat exhibited an adsorption capacity that was 2-3 times higher than that of Cercona™
Fe-bone-char and about 5-6 times higher than that of the ferric oxides on a weight basis. As the solution U
concentration increased, a relatively large percentage of U was left in the solution because of the limited
sorption capacity of the solid materials.
100
- 100
•
"
o
£
—D— MB Fe, With headspace
—A— MB Fe, No headspace
Peerless coarse Fe
n
3
o
<
n
Q.
20 -
5000 10000 15000
Initial U Concentration (mg/L)
0
20000
Figure 3. U removal by various zero-valent iron filings: Master-
Builder (MB) with headspace, MB without headspace,
and Peerless coarse iron filings
In contrast, reactions between uranyl
and Fe° resulted in almost 100%
removal of U in solution as shown in
Figure 3. Regardless of the initial U
concentration in solution (up to 84
mM or 20,000 mg/1), no detectable
amounts of U were found in solution
after reaction with Fe°. Note that
Figure 2 is plotted against the
equilibrium concentration of uranyl in
solution, whereas Figure 3 is plotted
against the initial U concentration
added because no detectable amounts
of U could be found in the equilibrium
solution. Analysis by the phospho-
rescence-lifetime instrument gave not
only a high sensitivity for detecting
uranyl (on the order of parts per
trillion) but also the valence states of
U because U(IV) does not phosphoresce [26]. In order to determine if the reduced U(IV) species were
present in the equilibrium solution phase, some selected aqueous samples were also exposed to the air and
treated with peroxide (H2O2) and HNO3 and re-analyzed by the phosphorescence-lifetime instrument. Again,
no detectable amount of U was found in the solution after the treatment. These results are therefore
indicative of a reductive precipitation of U by Fe° rather than an adsorption process because adsorption
would have resulted in a partitioning of U in the solution phase. No maximum loading capacity may be
defined for U removal as long as sufficient amounts of Fe° are present in the system to maintain a favorable
reducing environment. The adsorption process may dominate only when Fe° is consumed and the corrosion
products (Fe-oxyhydroxides) are formed in the system.
We further evaluated the percentage distribution of U in the Fe° and its corrosion products in suspension
because, if U was largely adsorbed on the corrosion products of Fe°, the adsorbed U may be desorbed and/or
co-transported with the suspended colloidal particles. Results (Table 1) indicated that only a small
percentage of U was associated with these suspended particles (or the corrosion products) after shaking for
~3 weeks in the batch experiments. The majority of added uranyl was precipitated on the Fe°. Additionally,
a desorption experiment with 0.1M NajCOj indicated that U associated with suspended particles (i.e., the
corrosion products) was readily desorbed (>64%). However, only small amounts of U (<0.21%) in Fe° could
be washed out with the carbonate solution (Table 1). It is known that CO32" complexes with uranyl to form
negatively charged U species such as UO2(CO3)22", which does not 'adsorb on the negatively charged ferric
oxyhydroxide surfaces at the given pH conditions (pH > 10) [27-28]. The zero point of charge of common
ferric oxyhydroxides is in the range of 6-8.5 [29]. Therefore, these results suggest that uranyl, which was
in the initial solution, was converted to less soluble U(IV) species by the Fe°. The large percentage of
extracted U from the suspended particulates implies that U was primarily adsorbed in its Vl-oxidation state
by the corrosion products, although these corrosion products only constituted a small percentage of the Fe°
mass. These observations are consistent with that of Grambow et al. [21], who found that a large percentage
39
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
of uranyl was adsorbed on the corrosion products of Fe° and that uranyl was only partially reduced to U(IV)
species. It is important to note, however, that we obtained these suspended particles of iron corrosion
products by vigorously shaking the Fe° solution and then immediately decanting the suspension. Under
static column flow-through conditions, no suspended particles could be observed by means of turbidity
measurements; this suggests that iron corrosion products may be cemented within the Fe° mass or
precipitated downgradient in the column. This cementation of iron corrosion products is desirable for
sequestering U or other contaminants in in situ groundwater treatment systems.
Table 1. Uranium partitioning in the Fe° and the suspended particles (i.e., the corrosion products
of Fe°) before and after washing with 0.1 M NaaCO9
UinFe0
U in suspended
particles
Peerless Iron Filings
Before wash
(mg)
96.09
3.91
After wash
(mg)
95.89
1.39
U desorbed
(%)
0.21
64.5
Master-Builder Iron Filings
Before wash
(mg)
97.23
2.77
After wash
(mg)
97.16
0.18
U desorbed
(%)
0.07
93.6
Identification of Reaction Products by Spectroscopic Studies
Reductive precipitation of uranyl to U(IV) by Fe° is thermodynamically favorable according to the
following reactions:
(1)
(2)
UO2+
2e' ~ Fe(0)
+ 4H+ + 2e- ~ U(IV)
2H2O
E"=- 0.440V
E = -0.07 V at pH 8.
480
520
560
600
640
The reduced U(IV) readily forms oxyhy-
droxide precipitates in solution [30-31] or
precipitates on Fe° surfaces. The fluorescence
Spectroscopic technique was employed in an
attempt to identify the valence state of U on
Fe° surfaces and in solution. The analytical
technique was also employed to evaluate the
possible re-oxidation processes of the reduced
U(IV) on Fe° surfaces. It is known that only
the oxidized U(VI) gives strong fluorescence
whereas the reduced U(IV) does not. In Figure
4, the fluorescence spectra are plotted for (1) a
uranyl aqueous solution, (2) a uranyl suspen-
sion containing ferric oxide powder (hematite),
(3) a uranyl solution in the presence of Fe°,
and (4) a background aqueous solution without
uranyl. The uranyl aqueous solution itself
showed strong intensity in fluorescence. In the
presence of hematite, the intensity was decreased, but no fluorescence spectra were observed after uranyl
reacted with Fe°. These results are consistent with the batch kinetic and equilibrium studies, which showed
that uranyl was reduced to U(IV) by Fe°. However, because uranyl is only adsorbed onto ferric oxide and
not reduced, a strong fluorescence spectra was observed in the ferric oxide systems as a result of the
adsorbed uranyl species and the uranyl remaining in solution.
Several lines of evidence presented above have demonstrated that, in the presence of Fe°, uranyl was
primarily reduced to U(IV), which was precipitated on the kon surface. This work has addressed how
Wavelength (nm)
Figure 4. Fluorescence spectra of uranyl before and after
reactions with ferric iron oxide (hematite) or zero-valent
iron filings in solution
40
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
480
520
560
600
640
Wavelength (nm)
Figure 5. Fluorescence spectra of U on zero-valent iron
surfaces, showing the adsorbed uranyl or the reoxidation
of precipitated U(IV) to U(VI) species
persistent these reduced U species are on the
iron surface, especially if they are exposed to
atmospheric oxygen. Figure 5 shows the
fluorescence spectra of a few solid Fe°
samples at different exposure to air after
reacting with uranyl solutions. A strong
fluorescence spectrum was observed after
uranyl was reacted with a rusted Fe° and then
dried in the air for 2 days. By comparison, a
much weaker fluorescence spectrum was
observed when uranyl reacted with Fe° and
was then dried in a Speed-Vac for 4 h. No
fluorescence signal could be identified for the
freshly prepared U- Fe° specimen. These
results demonstrate that the reduced U(IV)
species on Fe° surfaces could be re-oxidized.
Furthermore, the re-oxidation rate appeared to
be relatively slow (on the order of days) in comparison with its reduction process (on the order of minutes,
Figure 1).
Implications for groundwater remediation
Both batch experiments and spectroscopic studies showed that Fe° is effective in removing U from water
under reducing conditions. The major reaction pathway is via reduction of uranyl by Fe° to form insoluble
U(IV) surface species. Adsorption of uranyl by corrosion products accounts for a small percentage of total
uranyl removal. The overall removal rates are fairly fast, and the half-life of the reaction is on the order of
a few minutes, assuming a pseudo-first order kinetic reaction. Experiments also show that re-oxidation can
occur when reduced U(IV) is exposed to atmospheric oxygen.
These results imply that using iron in a permeable reactive barrier to remove U is feasible for groundwater
remediation. However, U retained by Fe° may be reoxidized and remobilized when the precipitated U(IV)
on Fe° is exposed to atmospheric oxygen. The effect of dissolved oxygen in water on the rate of re-oxidation
has not been determined in this study, but will be included in future work. As long as a reducing condition
is maintained in the permeable reactive barrier by Fe°, it is likely that the reduced U(IV) is coprecipitated
and eventually cemented with the Fe° corrosion products. Both laboratory and field-scale experiments are
underway to determine the removal efficiency in situ, and to gauge the geochemical influence to such
technology. The long-term performance of Fe° reactive barriers with respect to its efficiency, byproduct
formation, and clogging is still a matter of debate [32]. Results of this work have demonstrated that Fe° is
an effective medium that can be used to remove certain redox-sensitive radionuclides and metals, in addition
to its ability to degrade many chlorinated organic compounds.
ACKNOWLEDGMENTS
Cercona™ Bone-Char materials were supplied by R. Helferich at Cercona of America, Inc. Technical
assistance by Dr. M. Dickey, X. Yin and assistance on SEM and XRD analyses by Y. Roh and S. Y. Lee are
gratefully acknowledged. We thank D. Watson for his helpful comments on the manuscript. This work was
supported by the Subsurface Contaminants Focus Area of Office of Environmental Management, U.S.
Department of Energy. Travel support for L. Liang through the UK Environment Agency is gratefully
acknowledged.
41
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
REFERENCES
1. R.W. Gillham, U.S. Patent 5266213, Nov 30, 1993.
2. S.H. Shoemaker, J.F. Greiner, R.W. Gillham, in "Assessment of Barrier Containment Technologies",
R. R. Rumer, J.K. Mitchell (eds.). (1995) Section 11, 301.
3. R.W. Gillham, S. F. O'Hannesin, Ground Water 32(6) (1994) 958.
4. K.J. Cantrell, D.I. Kaplan, T.W. Wietsma, J. Hazardous Materials 42 (1995) 201.
5. R.M. Powell, R.W. Puls, S.K. Hightower, D.A. Sabatini, Environ. Sci. Technol. 29 (8) (1995) 1913.
6. L. Liang, J. D. Goodlaxson, in 'Emerging Technologies in Hazardous Waste Management Vn, extended
Abstracts for the special symposium, Atlanta, GA' (1995a) 46.
7. T.L. Johnson, M. M. Scherer, P.G. Tratnyek. Environ. Sci. Technol. 30(8) (1996) 2634.
8. American Chemical Society Extended Abstracts of 209th National Meeting, Division of Environmental
Chemistry, Anaheim, CA 35 (1) (1995).
9. Emerging Technologies in Hazardous Waste Management Vn, extended Abstracts for the special
symposium, Atlanta, GA (1995).
10. L.J. Matheson, P. G. Tratnyek, Environm. Sci. Technol 28(12) (1994) 2045.
11. T.M. Vogel, C.S. Griddle, P.L. McCarty, Environ. Sci. Technol. 21(8) (1987) 722.
12. T.M. Sivavec, D.P. Horney, in Extended Abstracts of 209th National Meeting. Anaheim, CA. Division
of Environmental Chemistry. 35(1) (1995) 695.
13. L. Liang, J. D. Goodlaxson, N. E. Korte, J. L. Clausen, D. T. Davenport, in Extended Abstracts of 209th
National Meeting. Anaheim, CA. Division of Environmental Chemistry. 35(1) (1995b) 728.
14. TJ. Campbell, D.R. Bums, A.L. Roberts, J.R. Wells. Environ. Tox. Chem. 16(4) (1997) 625.
15. A.L. Roberts, L.A. Totten, W.A. Arnold, D.R. Bum's, TJ. Campbell, Environ. Sci. Technol. 30(8)
(1996) 2654.
16. C. Grittini, M.Malcomson, Q. Fernando, N. E. Korte, Environ. Sci. Technol. 29(11) (1995) 2898.
17. L. Liang, N.E. Korte, J.D. Goodlaxson, J. Clausen, Q. Fernando, R. Muftikian, Ground Water
Monitoring and Remediation, Winter (1997a) 122.
18. T. M. Sivavec, P.D. Mackenzie, D.P. Horney, S.S. Baghel, Preprint from the 1997 International
Containment Technology Conference, St. Petersburg, FL., Feb 9-12,1997.
19. D.W. Blowes, C.J. Ptacek (1994). 'System for treating contaminated groundwater,' U.S. Patent No.
5,362,394.
20. L. Liang, B. Gu, X. Yin. Separations Technol. 6 (1996) 111-122.
21. Grambow, B.; Smailos, E.; Greckeis, H.; Muller, R.; Hentschel, H. Radiochim. Acta 1996,74,149-154.
22. Wersin, P.; Hochella Jr., M. F.; Persson, P.; Redden, G.; Leckie, J. O.; Harris, D. W. Geochim.
Cosmochim. Acta 1994, 58, 2829-2843.
23. Bostick, W. D.; Jarabek, R. J.; Fiedor, J. N.; Farrell, J.; Helferich, R. Proceedings of the 1997
International Containment Technology Conference and Exhibition. St. Petersburg, FL 1997, 767-
773.
24. Gu, B.; Dowlen, K. E.; Liang, L.; Clausen, J. L. Sep. Technol. 1996, 6, 123-132.
25. Gu, B.; Schmitt, J.; Chen, Z.; Liang, L.; McCarthy, J. F. Environ. Sci. Technol. 1994,28, 38-46.
26. Dai, S.; Metcalf, D. H.; Del Cul, G. D.; Toth, L. M. Inorg. Chem. 1996, 35, 7786-7790.
27. Hsi, C. K. D.; Langmuir, D. Geochim. Cosmochim. Acta 1985,49,1931-1941.
28. Ho, C. H.; Miller, N. H. J. Coll. Interf. Sci. 1986,110,165.
29. Sposito, G. The surface chemistry of soils; Oxford University Press, New York 1984.
30. Baes, C. F.; Mesmer, R. E. The Hydrolysis of Cations; John Wiley & Sons: New York, 1976.
31. Dai, S.; Toth, L. M.; Del Cul, G. D.; Metcalf, D. H. J. Phys. Chem. 1996,100, 220.
32. Liang, L.; Jacobs, G. K.; Gu, B. Coll. Surf. 1997, (in review).
33. B. Gu, L. Liang, M.J. Dickey, X. Yin and P. Cameron. Preprint of extended abstract for RTDF
permeable reactive barriers Action Tarn Meeting, Virginia Beach, VA. Sept. 18-19, 1997.
42
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
Discussion
Wolfgang Wiist asked whether Liang observed a maximum capacity for sorption of uranium. He noted that
he did not see a problem with leaving uranium in an iron reactor because it is very anoxic and uranium will
remain reduced unless conditions become oxidizing and mobilize the uranium. Liang indicated that leaving
uranium in the reactor is a different issue from leaving iron in situ when used for treating organic solvents
because of the environmental implications. She said that the experiments that were conducted examined the
worst-case scenario. In column flow-through experiments, no suspended solids were noted.
Timothy Vogel asked whether nitrate was transformed mainly to ammonia or whether it was converted to
nitrogen gas as well. Liang responded that the nitrate was mainly transformed to nitrite because reduction
to nitrogen takes longer. Wiist added that he did not find nitrite in his experiments. We all agree that the
reactive media determine the by-products of the reaction.
43
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Bioprocesses in Treatment Walls: Bioscreens
Huub H. M. Rijnaarts1
Abstract
The feasibility of intrinsic and enhanced bioremediation approaches for 16 contaminated sites in the
Netherlands are discussed. At least five out of 10 chlorinated solvent sites, natural attenuation can be used
as one of the tools to prevent further dispersion of the plume. At two sites stimulation of the intrinsic
dechlorination processes in a bioactive zone is required, and pilot field tests are currently under way. In three
sulphate-reducing/methanogenic aquifers contaminated with aromatic compounds, natural anaerobic
degradation of the risk-determining benzene was demonstrated not to occur spontaneously by microcosm
studies. Benzene biodegradation could be initiated by feeding small amounts of oxygen (in all samples) or
nitrate (only in one sample). Investigations at two locations contaminated with hexachlorocyclohexanes
(HCHs) indicated significant intrinsic bioattenuation of HCHs under very low field-redox conditions, and
of the HCH-degradation products chlorophenol and benzene under sulphate/iron reducing conditions.
Biostimulated zones remain required to complete the degradation of HCH and the degradation product
monochlorobenzene.
INTRODUCTION
The ideas concerning best approaches for remediation of polluted soil and groundwater are rapidly changing.
A few years ago many engineers and regulators considered contaminated aquifers as biologically inactive
systems that could only be remediated by very drastic and expensive methods like soil vapor extraction or
large scale pump-and-treat. Nowadays, one has become aware that the natural microbial population in soil
and groundwater react actively when confronted with pollution, and thus offer new and better ways to
protect groundwater resources at acceptable costs.
Often, natural biodegradation processes can convert large amounts of contaminants without any external
stimulus. Sometimes, the natural processes can completely degrade or immobilize pollutants during their
down-gradient transport. When sufficient time and space is available, an intrinsic remediation approach can
in principal be completely protective for man and surrounding ecosystems. At other sites, the natural
biodegradation processes need to be enhanced in order to obtain a sufficient risk reduction and groundwater
protection. Such a stimulation can be performed in a bioactive zone or bioscreen. In such a zone, the
autochthonous bacteria are supplied with the appropriate electron donors, acceptors or nutrients. At some
sites, the microbial population is not yet fully adapted to the most optimal biodegradation process. In such
a case bioaugmentation of the bioscreen may be a solution. Intrinsic remediation, when necessary combined
with bioscreens, is becoming the new approach to deal with large and complex contaminated sites.
Natural and enhanced attenuation as a risk based solution for contaminated soil and aquifers is currently
being investigated at various sites in the Netherlands. A number of the 16 cases investigated by TNO will
be discussed below.
1TNO/MEP, P.O. Box 342,7300 AH Apeldoorn, The Netherlands, tel: 31/55-5493-380, fax: 31/55-5493-410, e-mail:
h.h.m.rijnaarts ©mep.tno.nl
44
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
RESULTS AND DISCUSSION
Chlorinated Solvents
At the Rademarkt site (Groningen, The Netherlands) contaminated with perchloroethylene (PCE) and
trichloroethylene (TCE), mixed redox conditions control the intrinsic biodegradation processes. In a
methanogenic/sulfate reducing zone a complete reductive dechlorination reaction via vinyl chloride to
ethene and ethane occurs. However, the transformation rates of vinyl chloride observed in the field (and in
the laboratory) are too slow to prevent migration of this hazardous compound to areas to be protected. The
low amounts of DOC (< 10 mg/1) indicate a lack of sufficient amounts of electron donor naturally present.
Laboratory experiments identified that a mixture of electron-donors is most suitable to enhance the in situ
reductive dechlorination. An in situ pilot test with an anaerobic activated zone designed for complete
reductive dechlorination is planned this fall. In another flow direction in the field, the redox condition
changes from methanogenic/sulfate reducing to oxic conditions. In the reduced part, PCE and TCE are
transformed to cis-l,2-dichloroethylene (DCE) and vinyl chloride (VC), which both disappear after entering
the oxic zone. Probably, these compounds are removed by intrinsic oxidation. Hence, in this flow direction,
a complete sequential degradation Of the chlorinated compounds is achieved.
At another PCE/TCE site in Maassluis, the contaminated aquifer contains high concentrations of organic
carbon (up to 700 mg/1 DOC). Here, complete reductive dechlorination is observed. Most likely, this is a
result of the high amounts of intrinsic electron donors and a well-adapted autochthonous microbial popula-
tion.
The results of the characterization of redox conditions and intrinsic biodegradation at 10 chlorinated solvent
sites will be presented. For at least five of these sites studied, intrinsic remediation is an important part in
an approach to effectively control the risks. At two sites at least, in situ biodegradation needs to be
stimulated in in situ activated zones to protect down stream areas.
Intrinsic chlorinated solvent remediation is further investigated by testing electron donors and measuring
in situ hydrogen pressures in the field and in laboratory microcosms. Thus a further insight into mechanisms
involved will be obtained.
At present, new technologies are being developed by TNO for application in bioscreens and remediating
chlorinated solvent hot spots and chlorinated solvents in low-permeable soils and subsurface layers.
Combination of electro-reclamation techniques and biological methods appear to bring new solutions in the
near future.
Aromatic arid Oil-Related Hydrocarbons
Aromatic compounds are often the risk-controlling compounds at oil and gas production sites, and at sites
of coating and nutrition industries. At three sites in the north part of The Netherlands, deep anaerobic
aquifers contaminated with benzene, toluene, ethylbenzene or xylenes (BTEX) have been investigated.
Under the existing sulfate-reducing conditions, the intrinsic biodegradation of toluene and ethylbenzene
could be demonstrated in the field and in microcosm studies. Benzene biodegradation could not be
evidenced with the field data. Laboratory microcosm studies with five different sediment samples were
performed. For each sample a series of comparable triplicate microcosms were incubated; in total about 300
microcosms were used. The results demonstrate thus far (200 days of incubation) that spontaneous intrinsic
anaerobic benzene biodegradation does not occur in these sediment and groundwater samples.
Microcosm method development studies indicated that special care is required to prevent artifacts: "apparent
intrinsic anaerobic" benzene biodegradation could be demonstrated to originate from low (often not
45
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
measurable) amounts of oxygen introduced into the systems, when inappropriate materials and techniques
were used.
Microcosms were also used to investigate possibilities to stimulate biodegradation of benzene and BTEX
compounds. Especially, addition of nitrate and low amounts of oxygen to the anaerobic samples was studied.
In one out of the five series of sediment-microcosms, nitrate-reducing benzene biodegradation appeared to
occur after lag-times greater than 100 days. In addition, anaerobic benzene degradation could be initiated
in all samples by spiking low amounts of oxygen. Further investigations are underway to elucidate
mechanisms and quantify remediation process parameters. The results are used for designing pilot
demonstration tests to be performed this spring/summer.
At an oil refinery site in the Rotterdam Harbor area, an aerobic reactive-trench bioscreen is tested for
managing a plume of the dissolved fraction of a mineral oil contamination (80% of the compounds belong
to the C6 - C12 fraction). Bench scale experiments are currently performed to establish: i) optimal grain-size
and packing density for the porous media used in the trench; and ii) optimal oxygen supply rates to
sufficiently initiate aliphatic hydrocarbon biodegradation and to minimize clogging with iron(EI)oxides.
Chlorinated Pesticides
HCH isomers are important pollutants introduced by the production of lindane (gamma HCH). At two sites,
intrinsic anaerobic biodegradation of HCH and corresponding degradation-intermediates (benzene and
monochlorobenzene) is currently being investigated. At one site, natural biodegradation processes appear
to completely degrade all compounds except the monochlorobenzene; possibly some biostimulation may
be required at the downstream end of the plume. At the other site, interception of the HCH/chloro-
benzene/benzene plume is required. A bioactivated zone as an alternative to conventional large scale pump-
and-treat is currently being investigated. At present, laboratory process research aimed at developing a
combination of anaerobic-microaerophilic in situ stimulation in such a bioactivated zone is being performed.
The results indicate good prospects for further development and pilot scale demonstration.
Guidelines for Application of Natural Attenuation or Bioscreens
Dutch and Nicole-supported guidelines for assessing the feasibility of natural attenuation approaches for
various contaminant situations are currently under development, making use of already existing protocols
(USA) and taking Dutcri/European specific conditions, knowledge and regulatory constraints into account.
These guidelines are to become tiered decision tools. Li the first steps, a minimum of data is required for
an initial feasibility screening, thus saving costs. At higher tiers, data requirements are designed to address
natural degradation under various redox-situations (including sequential redox conditions) that often occur
in European sedimentary regions.
CONCLUSION
Intrinsic and enhanced in situ bioremediation approaches become more and more accepted as appropriate
cost-effective solutions for aquifers and soils contaminated with organic chemicals. These techniques still
need to be further developed and demonstrated and could also be expanded to control and remediate
mixtures of organic and inorganic (heavy metal) pollutants.
Discussion
Hexachlorocyclohexanes can be degraded using zero-valent iron and without producing benzene as a by-
product. Zero-valent iron may be an appropriate reactive material for these types of compounds as long as
the mass-balance considers monochlorinated benzene.
46
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
Because there is no natural nitrate at the site, anaerobic benzene degradation was stimulated by introducing
nitrate. Even though The Netherlands has a 50 mg/1 nitrate drinking water standard, the introduction of
nitrate did not pose a problem because concentrations in the stimulated system ranged from only 10-20 mg/1.
47
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Novel Catalyses for Reactive Barriers1
Timothy M. Vogel2 ;
The presentation that we will make today is a combination of several different presentations, not necessarily
those planned for today. But the point is simply to talk about different types of material that might be able
to catalyze or to react with compounds of interest in barriers and in wall systems, so we can target a number
of different reactions (Figure 1), and we'll use some information that has been collected in other laboratories
and with participants, including Nada Assaf-Anid at the University of Manhattan, Pedro Alvarez from the
University of Iowa, and Larry Nies at
Purdue University, who all have
contributed in some or the other to the
data that we will present today.
There are basically two different
concepts for the remediation of ground-
water contamination: (1) Methods based
on transfer and concentration of the
contaminants based on activated carbon
—activated carbon even in a wall is not
a reactive wall but an adsorption
wall—and more innovative methods
such as mobilization of contaminants Figure 1. Reactions
with surfactants or in situ absorption barriers—but these methods all produce secondary waste streams that
have to be treated separately. (2) Destructive methods—such as thermal, oxidative, biological, or catalytic
dehalogenation—that are shown in the laboratory to work effectively.
What we'd like to do is to talk about these reactive systems from two points of view: what kind of reactions
we would like to observe, and what reactions we would not like to observe. The example is with the
chlorinated alkenes, where a lot of individuals do not like to see the presence of vinyl chloride. What we'll
talk about is ways to catalyze some of these reactions and to avoid certain by-products.
Our approach is to look at the different problems
that have been associated with literature reports on
catalyzed reactions and to try to identify ways to
answer these questions. Today one of our goals is
to, in a sense, initiate some discussion, because we
are going to talk about some things that have not
necessarily been brought to a scientific, critical
review, nor applied in full-scale processes, so any
comments that you have would be gratefully
received.
We looked at reaction rates (Figure 2), and the
question often asked is, how we can do reactions
faster. In a talk presented yesterday, there was a
Reaction rates
Compounds degraded
Corrosion of material
Metal inactivation
Activation energy, steric
hinderance, diffusion
limits
Electrophilicity,
hydrogenation
Selective reactivity,
protection, pretreatment
Chemical reactivity,
redox changes —
electron sources
Figure 2. Problems and Solutions
1 This paper was prepared from a transcript of Dr. VogeFs presentation.
2RhodiaEco ServicesXATE, 17 rue Perigord, 69330 Meyzieu, France; tel: 33/4-7245-0425; Fax: 33/4-7804-2430- e-
mail: timothy.vogel@rhone-poulenc.com.
48
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
comparison between zero-valent iron and biological reactions, where the rates were much faster with the
zero valent. And then just recently, we had another on the use of bio-screens, so there is a question about
reaction rate that has to be considered with cost—we're not going to talk about cost—we're going to talk
about what we can to increase reaction rates. So either we try to change the activation energy of the system,
or in the case where there is some kind of steric hindrance—where the molecules cannot come in contact
with the reactive surface—or in the same respect diffusion limits between the molecules in the bulk solution
and the site of reactivity. These are ways to address the reaction rates and maybe improve them.
Other questions are related to the compounds that are degraded. Many compounds are degraded by existing
systems, but are degraded at rates that are not acceptable, therefore some people say these compounds are
not degraded by a particular system. For example, some people say iron cannot degrade a certain chlorinated
compound, when in fact it's more the rate that is not acceptable, not whether it can degrade or not degrade.
We'd like to separate the issue of the possibility of a degradation occurring from an unacceptable rate. Thus,
we want to examine the kinds of mechanisms and how they affect these dechlorinations or reductions. And
finally, there are issues about the lifetime of the material, and how it's affected, and how you can engineer
support material and different catalyst systems, such that the lifetime is long enough for the application.
In addition, we will discuss some biological aspects and using organisms coupled with other systems like
iron to deal with problems that result from the applications of catalysts.
The general schematic is a cyclic system (Figure 3)
in which an electron comes from some source, is
transferred to the molecule that is the pollutant, the
pollutant is reduced by that reaction, and the source
is then regenerated (or maybe not) (Assaf-Anid et
al.). There are two ways of looking at this: One is
based on thermodynamics, and thermodynamics
are (for those who are not aware of these systems)
the pressures pushing the system around—the
force. But then, there is also the question about the
kinetics, the rates. You can have a large force, but
if you have a small rate, it will be slow. Take a
hydraulic dam. You can have a lot of water behind
red = e 4- ox
J
oxidized mediator
fast
rats-limiting
reduced mediator
e = red
Figure 3. Simplified electron flow
a hydraulic dam, but if the hole is small, not much water will go through it, so we need to separate
conceptually thermodynamic data from kinetic data. The thermodynamic possibility is necessary, but it
doesn't provide you with a real process that you can apply if the kinetics are unacceptably slow.
If we look at the thermodynamics of different systems, we can see in Table 1 (taken from several literature
studies) that there are measures of different redox couples that are used to determine whether the
thermodynamics of a system is favorable not. This is the redox couple between the oxidized and reduced
phases of these compounds. Some of the data are pretty standard, for example water—the oxygen/water
couple. This provides you with some information about the potential energy—how high the water is behind
the dam—but it does not tell you anything about the rate at which these reactions will go. These are just
potential electron donors in systems, and those in Figure 4 are ones from biological systems based on a 1987
review (Vogel et al., ES&T) concerning these types of catalysts.
Table 1. Standard potentials (E°) of biologically relevant electron donors or reductants
Reductants J . '- ~ -
Vitamin B12
Co(I) tetraphenylporphin
^' fe°"(volte) ' ' - ,'
-0.59 to -0.8
-0.56
49
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Reductants
Ferrodoxin (reduced)
H2
Cr(m)
NADH + H*
Cytochrome P450
(unactivated)
Glutathione (reduced)
Cytochrome P450
(activated)
Fe (II) deuteroporphin IX
Ubiquinone (reduced)
Cytochrome c (+2)
Fe(H)
H2O
E° (volts)
-0.43
-0.42
-0.41
-0.32
-0.30
-0.23
-0.17
0.00
0.10
0.22
0.77
0.82
What is of interest is the effect that the complex has on the metal. If you talk about iron, or oxidized kon,
and there is some work with iron porphyrins, the important aspect is that the term, "iron reactive barriers,"
for example, is uninformative because it's necessary to consider whether it's zero valent kon; whether the
iron is complexed in some form or the other; or what the iron is doing.
Table 2 shows redox couples between Fe3* and Fe2* in different complexes or the different ligands. Redox
energy varies for the same kon couple, based upon the way that it's complexed by the organic: the potential
for reduction or oxidation in these systems varies with the ligand. There is also some cobalt data in the table.
A metal is not in an isolated situation, but has certain complexes associated with it that may have an effect
on the thermodynamics.
Table 2. Standard reduction potential of Fe3* and Co3* to the respective bivalent ions
Ligand
1,10 phenanthroline
3 bipyridin
water
glycine
6NH3
6(CN)~
cytochrome c
/neso-tetrakis (N-methyl-pyridil) porphin
EDTA4-
rubredoxin
E°Fe(+3/+2),V
+1.20
+0.95
+0.77
-+0.4
+0.37
+0.36
+0.25
+0.17
+0.13
-0.05
E0Co(+3/+2),V*
+0.31
+1.88
+0.15
+0.11
-0.80
+0.37
50
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
Ligand ^ - , >' '<,.?'-, •"" . -"x -
iron protoporphyrin
8-hydroxyquinoline
^E°¥e(+3/+2),V '..
-0.12
-0.15
E°C6(+3/+2),V
The redox potential of your chlorinated or non-
chlorinated organic that you want to reduce needs
to be examined in order to determine (based upon
your electron donor and your electron acceptor)
whether you have a potential energy production.
And again, borrowing from a review (Vogel et al,
1987, ES&T), Figure 4 shows data where, by
classic chemical methods, you can determine the
energy that would be derived, and the thermo-
dynamic potentials for coupling different reactions,
electron donors, and acceptors.
For example, one reductive dechlorination is from
tetrachloroethylene (PCE) to trichloroethylene
(TCE). That is one-half of the couple, and if you
coupled it then with vitamin B12, for example,
which is a cobalt-containing porphyrin, you would
have a favorable reaction. The distance would be
the amount of energy, the potential energy from
that reaction is about one volt. In other words, it is
a thermodynamically favorable reaction to use
vitamin B12 to dechlorinate PCE to TCE. This is
just a simple figure that has two couples, an
electronic donor and an acceptor. These types of
tables and calculations are useful because they
allow you to determine what the thermodynamic
potentials are in your system, assuming you calcu-
late this accurately for the concentrations in your
actual system and not for standard-state
calculations that is typical in the literature. Then
you can determine in your system whether you
Half-reaction reduction potentials for reducing and oxidizing agents
Glutaihione reduced to oxidized |»--.i.&!)-
c. ...I .:.-,;^...i..'r*£.;WV.
Figure 4. Thermodynamic potentials
have a gap here, and whether you have the potential to cause the reaction that you want. Note that as you
get less chlorine on some of these chlorinated aliphatics, the redox potential decreases.
In summary, this concept of coupling these different kinds of complexes with reductive dechlorination or
with reductions of molecules has been studied and discussed extensively in the literature. This is all in
published review articles, but it's enough to give you an idea. Table 3 shows some of the chlorinated
compounds that have been observed to degrade and the complexes that have been reported to be responsible.
The diversity of potential catalysts for chlorinated compounds is relatively high, and from a thermodynamic
point of view, there is considerable interest.
Table 3
. Reduction of halogenated aliphatic compounds by transition metal cor
Compound , * ., :
Methanes
Chloromethane
Jfto&nd** .-"" --'?: .- ;/
Alkylated co-complex
Reductant
Co(I) chelates
nplexes
51
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Compound
Dichloromethane
Trichloromethane
Tetrachloromethane
Bromomethane
Dibromomethane
Tribromomethane
Ethanes
Chloroethane
1 ,1-Dichloroethane
1,1,1 -Trichloroethane
Hexachloroethane
iromoethane
. , 1 -Dibromoethane
. ,2-Dibromoethane
*ropanes
-Chloropropane
, 1 -Dichloropropane
-Bromopropane
,2-Dibromo-3-
hloropropane
Products
Methane
Cl-alkylated 812
Methane
Dichloromethane
Dichloromethane
Cl-alkylated Ei2
Methane
Chloroform
Cl-alkylated B12
Methane
Alkylated co-complex
Methane, ethene
Br2-alkylated B,2
Alkylated co-complex
Ethane, ethanol
Ethane, ethanol, ethene,
chloroethane
L , 1 -Dichloroethane
1 , 1 -Dichloroethane
Tetrachloroethene
Tetrachloroethene
ithane
Ethane, ethanol
ithene
ithene
Alkylated co-complex
Propane, propanol, propene
Alkylated co-complex
Propene, allyl chloride
Reductant
Cr(H) SO4
B12-Co(m)
(methylcobalamine)
Cr(n)SO4
Fe(II)P
Fe(II)P
B12-Co(m)
Cr(E)S04
Fe(II)P
Bi2-Co(ffl)
Cr(n)SO4
Co(I)-complex
Fe(U)P
B12-Co(m)
Co(I)-complex
Cr(H)SO4
Cr(H)SO4
Fe(II)
Fe(II)P
Fe(H)P
Cr(II)SO4
Ni(I)
Cr(II)SO4
FedD
Fe(H)P
Co(I)-complex
Cr(II)S04
Co(I)-complex
Cr(II)SO4
Note: 'Ihe homolytic cleavage of cobalt-carbon bonds requires 15-30 kcal/mol and can occur at relatively low
temperatures.
Figure 5 shows an example where a carbon tetrachloride (CT) is dechlorinated by cobalt in solution to
chloroform (CF) over a period of about eight hours. When that cobalt was placed into a porphyrin (Figure
6), you see that the reaction rate changes drastically and you have dechlorination within about an hour
(Figure 7 is an example of a porphyrin).
52
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 11)
2-
Ov
0
cr
c?
10
12
Figure 5. Carbon tetrachloride (CT) dechlorinated to
Carbon fluoride (CF).
Figure 6. Carbon tetrachloride dechlorinated to
carbon fluoride in the presence of vitamin B12
. PORPHYBIN COSE
Direct link bstwesa A and B In conins
Methene
bridge in
porpfayrins
A, B, C, D = pyrrole
Ms = metal
The same kind of process is observed with a large range of
metals, thus, the practical question becomes where can the
metals be placed: on what kind of surface, or in what kind of
complex in which good thermodynamics and a high kinetic rate
are possible. In other words, what we want is a lot of water
behind the dam and a large hole in the dam, so that the reaction
flows.
Another example is with zinc (Figure 8): by itself; with carbon
tetrachloride again; and zinc with vitamin B12, which goes
much faster. So using that concept of thermodynamics and that
aspect of kinetics, you can engineer systems where your
catalyst is potentially active from an energetic point of view,
and where the system has physical characteristics and chemical
characteristics such that the kinetics will be quite rapid.
Another proposed use of iron is with a microorganism for the
conversion of nitrates to N2 to avoid ammonia (Alvarez et a/.).
Figure 8 is a schematic of the process: an organism that would
use hydrogen gas and would be responsible for this reaction.
The thought was that it would provide some synergy between
the iron and the microorganism. This synergy would allow for
some benefits in terms of the hydrogen production, in terms of
possible effects on the iron, and also possibly reducing the
competition of nitrate with other compounds like chlorinated solvents on the iron (Figure 9). In addition,
it might cause a higher reactivity for certain
compounds that don't react very rapidly with iron,
like dichloromethane. That's the basis for the
synergy between the two. Figure 10 shows a mass
balance in different tests: iron powder, steel wool,
iron with this organism, Pseudomonds,
Pseudomonas and steel wool, and with hydrogen.
The "Unrecovered NO3~" in the figure is the lack of
the mass balance. With iron powder alone, mostly
ammonia was produced. With steel wool alone, there
Tetrspyrrole
Figure 7. Porphyrin Core
NO3-
Figure 8. Nitrate removal can be enhanced by
combining Fe° with hydrogeonotrphic denitrifiers.
Bacteria will use cathodic H2 as electron donor to
produce harmless N2 instead of objectionable NH4+.
53
-------
Treatment Walls
NATO/CCMS Pilot Study Phase HI
e° corrosion rapidly induces anoxic conditions.
Bacteria may remove passivating H2 layer from Fe°
surface, increasing Fe° reactivity. Use of H2
increases e'flow from Fe° (cathodic depolarization).
Removal of inhibitory products by Fe° enhances
bacterial participation in cleanup process (e.g., NO2"
reduction by Fe°; metals precipitation due to high pH
caused by Fe° corrosion).
Bacteria could remove pollutants that Fe° cannot
(e.g., dichloromethane).
was considerable un-reacted nitrate and ammonia.
And in the presence of the microorganism,
considerable amounts of the N2O were measured in
these mass balances.
Ftgura 9. Basis of Synergism for Fe° + Bacteria
Mass Balance Results (after 8-12 days)
• UnmctKlNOy
• Rtduod by F<(0) to NH4*
• Unracowrad NO,-
* AMbnlbted by bacttita
" D«nHrHI«d by btctmrit to N.O
Fe°Powd.r*
P. danltrfHcins
P. donltriflcans
Steel Wool &
H2 & P. denltriflcans
Figure 10. Mass balances
Discussion
In response to a question from Resat Apak, Vogel indicated that some metal chelates used as catalysts are
inside membranes. They are trapped in a system that allows the transfer of the chlorinated compound, but
keeps the solubilized porphyrins in membrane systems. Vogel indicated that there are several techniques
documented in the literature to re-oxidize the cobalt(H)EDTA to cobalt(DI)EDTA, including the use of
electron donors, solid catalysts, and chemical reductants.
Wolfgang Wiist asked whether any column experiments were done before conducting the field work,
because his experience has shown that if catalysts are used, iron hydroxides or other precipitates inhibit the
diffusion (and thus reactivity) of EDTA. Schiith said that he was currently running laboratory experiments
in which several thousand pore volumes have been exchanged with groundwater. He has not observed much
precipitation in these experiments because the pH was lowered, not raised. However, microbial growth did
pose a problem for mass transfer due to diffusion.
Liyuan Liang asked whether and how much hydrogen was introduced into the system. Vogel indicated that
hydrogen had been introduced in the laboratory, and the amount was calculated by stoichiometry. In the
column and laboratory experiments, hydrogen gas was bubbled through the water to create hydrogen-
saturated water. However, Vogel cautioned that this approach may not work in the field.
Paul Bardos observed that many of the projects that were discussed in the treatment wall session involved
highly engineered solutions in terms of emplacing walls and improving the reactive matrix. He asked if the
advantages of these improvements were actually worth the increased costs, given that a trench with iron
filings treats groundwater reasonably well. Vogel responded that iron filings cannot treat all contaminants,
nor is there any other existing PRB suited for all contaminants and subsurface environments. Co-
contaminants can cause problems, and some contaminants are less reactive than others. The contaminants
that cannot be treated by iron walls must be determined, and inexpensive reactive materials must be found
to treat them.
Noting that the cost of iron filings for an iron wall accounts for a significant proportion of the construction
costs, Bob Siegrist asked how the use of a more expensive catalyzed metal such as palladium could be cost-
effective when compared to pump-and-treat methods. Vogel responded that although palladium costs more
54
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 1)1
than iron, the reactivity of palladium is also greater, so it could still be more cost effective because you do
not need as much material. Wtist noted that his tests have indicated that the cost of using 0.1 percent
palladium (the optimum amount) was approximately the same as iron, but the reactivity of the palladium
was much greater. He expressed concern, however, that the palladium could be mobilized under certain
conditions and contaminate the groundwater.
Harry Whittaker noted that there has been a lot of discussion thus far on reductive dehalogenation barriers
and asked whether oxidative dehalogenation barriers are being developed. Vogel indicated that techniques
for oxidative dehalogenation barriers have been patented recently, including systems involving Fenton's
agent.
55
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Funnel-and-Gate Systems for In Situ Treatment of Contaminated
Groundwater at Former Manufactured Gas Plant Sites
Hermann Schad1 and Peter Grathwohl2
Introduction
Groundwater contamination by polycyclic and monocyclic aromatic hydrocarbons (PAH and BTEX) is
typical for many former manufactured gas plant sites. Most of the PAHs are very persistent in the subsurface
environment—they are still present in high concentrations many decades after the contamination occurred,
and they cannot be removed from the subsurface within a reasonable time period by pump-and-treat. This
persistence above all is caused by slow dissolution kinetics of the compounds from non-aqueous phase
liquids, slow diffusion of the contaminants from low permeability zones (that have accumulated the
pollutants over decades) or resistant adsorption of the contaminants by the aquifer material.
Fast remediation of such contaminations is only possible by excavating the contaminated soil. This,
however, can only be applied at shallow sites and is still expensive because of the high disposal or treatment
costs of the contaminated material. Also the location and extent of the subsurface contamination has to be
known in detail, which is not feasible at large and abandoned industrial sites. Thus, new alternative
approaches for in situ groundwater remediation have to be developed. Since the goal of remediation in
general is the protection of groundwater resources downgradient from a contaminated area, in situ treatment
may be focused on the plume rather than on the source. This can be achieved using the concept of in situ
passive treatment systems, which in general are built perpendicular to the flow direction (e.g., Teutsch et
al, 1996). The technical implementation is either in the form of a continuous permeable wall or as a funnel-
and-gate system. In both cases no active pumping is required to move the groundwater through the treatment
zone. Within the reactive zone the pollutants are either degraded, sorbed or precipitated through biotic or
abiotic processes.
In this paper, we present an example for a funnel-and-gate system, which is currently being planned for the
long-term remediation of the former manufactured gas plant site of the city of Karlsruhe in southern
Germany.
Site Characterization
The site is located within the municipal area of the city of Karlsruhe and covers approximately 100,000 m2.
Site investigation programs were completed in 1996 (Trischler and Partner, 1996) with the recommendation
for funnel-and-gate as the most favorable remediation technique for that site. The latter was based on a
feasibility study (IMES, 1996), in which the general and site specific advantages of passive in situ treatment
were discussed and evaluated technically and economically.
The geological and hydrogeological situation at the site may be classified as typical for the upper Rhine
valley. The aquifer has a mean thickness of about 12m and consists of mostly sandy gravel, which is
1 IMES GmbH, Kocherhof 4, 88239 Wangen, Germany, tel: +497528 97130, Fax: +49752897131, e-mail:
hermann.schad.imes@t-online.de
2 Applied Geology, University of Tobingen, Sigwartstr. 10, 72076 Tubingen, Germany, tel: +49 7071 297 5429, Fax:
+497071 5059, e-mail: grathwohl@uni-tuebingen.de
56
-------
Treatment Walls
NATO/CCNIS Pilot Study Phase HI
funnel-and-gate system
KA.H-Concentration >• 1OO (J.g/1
PAH-Concentration > 1O
Figure 1. Site map showing the PAH plume and the location of the funnel-and-gate system
underlain by a clay layer at a depth of 15 m below surface. Groundwater drainage is dominated by the Rhine
river northwest of the site. The mean groundwater flow direction therefore is from southeast to northwest
(Figure 1). The contamination of the site is dominated by PAHs with acenaphthene being the highest
concentrated compound of the plume extending about 400 m down gradient from the site. Several
infiltration hot spots of non-aqueous phase (NAPLs) liquids were located within the saturated zone at the
site. PAH concentrations of the soil at many places exceed 10 g/kg. Altogether approximately 55,000 m3 or
100,000 tons of soil (within the saturated zone) are highly contaminated.
The hydraulic conductivity of the aquifer was estimated to 1.4 x 10"3 m/s and the hydraulic gradient is
approximately 0.1%. The groundwater flow rate from.the contaminated site under natural conditions is about
2.7 1/s.
Based on the environmental regulations to be applied, the groundwater contamination at the site has to be
remediated. For several remediation techniques the investment and running costs were estimated prior to
the final evaluation of the site investigation results (Table 1).
Table 1. Summary of estimated costs in DM for 50 years operation
for different remediation techniques (Trischler and Partner, 1996)
Technique
Thermal Treatment
Containment
Pump-and-Treat
In Situ Immobilization
Funnel-and-Gate
Investment
Costs
34,900,000
11,600,000
5,600,000
15,600,000
7,500,000
Running Costs,
200,000
4,000,060
8,300,000
600,000
4,100,000
Unforeseeable
Costs
3,500,000
3,100,000
1,400,000
1 3,200,000
1,700,000
Total Costs
38,600,000
18,700,000
15,300,000
9,400,000
13,300,000
With a feasibility study (IMES, 1996) it was shown, that a funnel-and-gate can be installed at the site. The
main components of such a system will be a slurry wall of approximately 240 m in length and two gate
buildings with a total volume of about 400 m3. The principal questions with regard to the system design are
related to the decontamination process within the gates and to ^the hydraulics of the entire system.
57
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Decontamination within the gates will be based on adsorption of the PAHs on activated carbon. In order to
determine the design parameters for the gates the processes relevant for operating the gates have to be
identified and parameterized. In order to answer the questions related to the hydraulic behavior of the
funnel-and-gate a numerical model will be necessary. Based on these experimental and numerical
investigations the technical construction details have to be designed.
In the following sections the key questions related to sorption, hydraulic behavior and technical construction
will be discussed based on the results of the feasibility study.
Sorption/Retardation
The adsorption of hydrophobic organic contaminants from the aqueous phase generally increases with
decreasing solubility of the compound (or increasing octanol/water partition coefficient, K^J and increasing
organic carbon content of the aquifer solids. Natural materials with high organic carbon content such as
coals or bituminous shales cause significant retardation of organic contaminants from groundwater. Such
materials with high sorption capacities may be used for passive removal of strongly hydrophobic
contaminants in groundwater (such as PAHs), which are the key problem at many former manufactured gas
plant sites). Much more efficient is adsorption onto activated carbon, which is already a well-established
technology for ex situ treatment of drinking water, polluted groundwater or waste water. For successful use
in a permeable wall or in a funnel-and-gate system permeability and sorptive properties of the adsorptive
wall material (adsorbent) must be optimized. Both the permeability and the sorption rates depend on the
grain size of the adsorbent, the permeability increasing with increasing grain size, and the sorption rates
decreasing with increasing grain size squared.
In general, an economically efficient in situ groundwater treatment system that relies on adsorption of the
contaminants must fulfill the following requirements:
1. Long regeneration cycles (3-10 years). Thus, high sorption capacities or adsorption combined with
biological or abiotic degradation of less sorbing compounds are required.
2. Relatively high permeability in comparison to that of the aquifer in order to prevent steep hydraulic
gradients.
3. Fast sorption kinetics in order to achieve high retardation factors even at high groundwater flow
velocities (short contact times in funnel-and-gate applications).
4. No decrease in permeability or "chemo-biofouling" of the adsorbent due to competitive adsorption of
dissolved organic matter or the growth of a biofilm that may plug adsorbent pores (although
biodegradation of contaminants would be highly desirable for a sorptive plus reactive wall).
The retardation of contaminants, such as in a sorptive permeable wall, can be calculated based on the
sorption coefficient, K^:
n
where p denotes the bulk density and n the porosity of the filter (in packed beds both values lie within a
narrow range). Rd may be interpreted as the number of pore volumes that can be displaced before break-
through of the contaminant occurs. Kd denotes the ratio between the sorbed and aqueous concentrations of
the contaminant. If sorption is nonlinear (which is often the case for strongly hydrophobic contaminants and
strong adsorbents such as granular activated carbon, GAC), Kd can be calculated in a first approximation
from the commonly-used Freundlich type adsorption isotherm:
58
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
V - s - V
K.J - - A.,.
d C... F'
\ln-l
where Ca Cw and KFr denote the equilibrium solid and aqueous concentrations of the solute and the
Freundlich sorption coefficient, respectively. The Freundlich exponent 1/n is usually smaller than 1,
resulting in decreasing Kd values with increasing concentration of the contaminant.
Activated Carbon Adsorption
Adsorption capacities for organic compounds using activated carbons are much higher than in naturally
occurring organic sorbents. Figure 2 shows literature data on adsorption of a variety of organic compounds
onto activated carbon (F300). Distribution coefficients (Kj) were calculated based on the Freundlich sorption
isotherms (KFl, 1/n) for an aqueous concentration of 10% of the pure compound's water solubility. For
compounds with high water solubilities (S > 1 mg/1), an almost linear inverse relationship between Kd and
S is observed (solid line in Figure 2). Since bulk density and porosity in a packed bed of granular activated
carbon are both close to 0.5, Kdm a GAC passive treatment wall is approximately equal to the retardation
factor Rd.
Sorption Kinetics and Permeability
The data shown in Figure 2 are valid for equilibrium sorption, which is not always applicable since the mean
residence times (contact times) in a permeable sorptive wall may be too short for complete equilibration.
Sorptive uptake of solutes by a porous adsorbent particle is diffusion limited. This results in Kd values that
increase with the square root of time for short-term sorptive uptake (less than 50% of sorption equilibrium
reached). Fast equilibration would be achieved for small particles with high surface to volume ratios. Small
particles, however, result in low permeabilities. In a funnel-and-gate system the permeability of the gate has
to be as least as high as in the aquifer. In a first approximation the permeability (K) increases with the grain
radius of the adsorbent particle squared and the intergranular porosity. Since the porosity in loose packed
beds will generally lie within a relatively narrow range, between 0.4 and 0.6, the main factor influencing
the permeability is the size of the adsorbent particles. For typical groundwater flow velocities and a
thickness of the sorptive wall of about 1-2 m, the mean residence time of contaminated groundwater will
range from less than a day to a few days. If the mean residence time is too short for equilibration, the Kd
values will be much smaller than expected for equilibrium conditions, which would in turn result in an early
breakthrough of the contaminant through the sorptive wall.
Column Tests with Contaminated Groundwater From the Site
A natural bituminous shale sample and three different GACs were used in laboratory column tests in order
to investigate the transport of PAHs within the gates. Breakthrough of PAHs in the bituminous shale column
was already observed after approximately 10 displaced pore volumes, indicating that the equilibrium
sorption capacity was not reached (by far). In the GAO columns, no breakthrough was observed, and PAHs
were concentrated in the first few centimeters of the packed bed (Table 2), as determined by methanol
extraction of thin GAO layers after the column test.
The results obtained for the GAO indicate that sorption at these relatively slow flow velocities is reasonably
close to equilibrium (typical flow velocities in drinking water treatment are much higher) and the retardation
factors to be expected are higher than 3,000.
59
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Table 2. Column tests
concentration a
Adsorbent (GAG)
Origin
Grain Size [mm]
BET-Surface Area mz/g
Column Length [cm]
Porosity (intergranular)
Run Time (days)
Pore Volumes Displaced
Flow Velocity [m/day]
Mean Residence Time [h]
PAH Penetration Depth
for removal of PAH
pproximately 2500 u
F100
Chemviron
0.5-3
770
14
0.52
23
885
5.3
0.6
<1.5cm
s (16 EPA PAHs) from contaminated groundwater (PAH
e/1, predominantly naphthalene and acenaphthene)
D15/1
Lurgi
1.25
1,102
12
0.68
23
731
4.0
0.76
<1.5cm
C40/4 (pellets)
Carbo Tech
4.2
1,281
13
0.61
23
814
4.6
0.68
<4cm
Bit. Shale
Lias & (Aalen)
2-4
16
21
0.48
112
1,097
2.1
2.6
breakthrough
Batch Experiments on Sorption of PAHs onto GAC
Batch experiments were performed in order determine the equilibrium adsorption isotherms and the sorption
kinetics of selected PAHs (acenaphthene, fluorene, phenanthrene, fluoranthene, and benz(a)anthracene) for
three different GACs and a O18-modified silica gel used as reference sorbent (Table 3).
Table 3. Properties of GACs Used hi Batch Experiments.
Property/GAC
Origin
Manufacturer
BET Surface Area
[m«/g]
Total Pore Volume
[cm3/g]
Average Pore
Radius [A]
Radius fmml
F100
Bituminous Coal
Chemviron
770
0.422
10.95
0.8-1.0
C40/4
Bituminous Coal
Carbo Tech
1280
0.630
9.87
1.0-2.0
TE143
Coconut
Pica
993
0.483
9.71
0.4-0.5
C18
Silica Gel
1ST
F,,, = 0.1 9
- ' •
-.
20-35um
KJ and KFr, determined in the equilibrium sorption experiments (Table 4) agree reasonably well with
literature data shown in Figure 2. The Kd values measured in the sorption kinetic experiments were much
lower than expected from the equilibrium values (Table 4). The sorption kinetic data agreed very well with
the numerical solution of Pick's second law for intraparticle diffusion (Figure 3). The apparent diffusion
coefficients were between 1 x 10'12 and 1 x 10'13 m2 s"1.
Table 4. Freundlich coefficients (KFa 1/n), equilibrium Kd and Kd values measured in the batch kinetic
experiments after 24 h. 3 days and 7 days (KH *).
F100
LoqKp,
-1/n
Log Kd at 10% S
Log Kd* 24 h
Loq Kd* 3 davs
Ace
6.2
0.59
5.2
3.4
4.0
Fin
6.1
0.81
5.6
3.7
4.3
Phe
5.9
0.56
5.0
4.1
4.4
C40/4
Ace
6.0
0.87
5.6
3.2
3.6
Fin
5.9
0.99
5.8
3.2
3.7
Phe
5.7
0.70
5.1
3.3
3.7
TE143
Ace
5.2
0.92
5.0
3.3
4.5
Fin
5.6
0.67
4.9
3.3
4.4
Phe
5.7
0.56
4.7
3.4
4.2
C1fl Silica
Ace
4.8
0.81
4.3
4.6
4.6
Fin
5.0
0.82
4.6
4.9
4.9
Phe
5.2
0.79
4.8
5.1
5.1
60
-------
treatment Walls
NATO/CCMS Pilot Study Phase ffl
IF. 100
Log Kd* 7 day5 | 4.0
| 4.5 | 4.7
C40/4
3.9 1 3.9 |S.9 .
, TE 143
4.8 1 4.8 1 4:6
C1fl Silica I
4.7 1 5.0 1 5.3 I
Hydraulic Aspects of Funnel-and-Gate Systems
There is a basic difference between the hydraulic design of a funnel-and-gate system and a continuous
permeable wall. For the continuous permeable wall, the only essential hydraulic design criterion is the
hydraulic conductivity of the wall material, which should be more permeable than the aquifer material. In
the case of a funnel-and-gate system, the hydraulic design depends on a number of factors, including the
number, position, and size of the gates, the length and angle of the funnel, and the hydraulic conductivity
of the reactive material within the gates. Starr and Cherry (1994) provided the basic functional relationships
for one-gate systems. From a practical perspective, the major design criterion is the width of the capture
zone (treatment zone) achieved for a given length of the funnel and the gate(s). The width of the capture
zone can be approximated from the total flow rate through the gate(s). In Table 5, the numerical results
obtained from the feasibility study (DVIES, 1996) are summarized. It is obvious that the number of gates
(equally distributed over the length of the funnel) is of much greater influence for the width of the capture
zone than the total length of the gates.
Table 5. Dependence of the total flow rate through the gate(s) on the number
of gates and the total length of the gate(s)
Number of gates
1
1
2
2
3
3
.Total length of gates [m]
12
36
12
24
18
24
Total flow rate [1/s]
1.04
1.55
1.58
2.01
2.09
2.22
Another aspect that needs to be considered in the design of funnel-and-gate systems is the heterogeneity of
, the aquifer. The location of the capture zone can be affected by low and high permeability zones in the
subsurface. Figure 4 shows results from the 3-dimensional stochastic groundwater flow model. The three
examples (one homogeneous case and two stochastic realizations) show that the size of the capture zone
does not change even though the parameters vary considerably. However, the position of the capture zone
(little elongated box) is different for every case. The stochastic model study showed that the length of the
funnel had to be increased by a 20% safety factor in order to compensate the uncertainty related to the
heterogeneity of the aquifer.
It is also interesting to note that a tenfold increase of the conductivity of the gate only resulted in a less than
10% increase of the flow rate through the gates for a given funnel-and-gate configuration. This finding was
found to be true for gate permeabilities exceeding the aquifer permeability.
Technical Construction of the Funnel-and-Gate System at the Site
For constructing the'fuhnel several different possibilities, e.g., slurry wall or sheet pile wall etc. are readily
available. These techniques are state-of-art and have been applied in the construction business many times.
The innovative part of the funnel-and-gate system is the construction of .the gate(s) and the connection
between the funnel and the gate(s). Since it is anticipated that the sorptive activated carbon fill of the gate(s)
has to be exchanged from time to time, the gate(s) will be reinforced concrete buildings, which can be
61
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
tmg/kg:mg/l]
at 10% S
1000000
100000
10000
KF,
[g/kg:mg/l]
1000
100
10
Bwi
B«ru
O
Btnzo
PRITOfk
co(ghl)p«r
•
3(«)pyi»n.
O
O
bjfluoroi
fliioroint
fl»n»
m
O
ithww
-N
O
luoroanth
^•Riinj-n
\
Dlmth
O
o
HW
1ft
tnapMlw
V
>Ni|
rlbtnztn*
Toll
D
J2.tr,
«
Mhilim
. Ethylb
XPwchl
»M rf\
BenzsiM
cMoronM
nm-Dlchlo
u
&
0
HU*IM
rarathtnt
• Trichlor
\
^\
o»th»n«
0 <
O
O
XtlWM
Phenol
•
\
\
o
0.0001 0.001 0.01 0.1 1 10 100 1000 10000
S [mg/L]
Fig. 2: Freundlieh sorption coefficients KFr (open symbols) as reported in the literature (American
Water Works Association, 1990; Sontheimer et al., 1985) for GAG (F300) and distribution co-
efficients (Ka: filled, labeled squares) were calculated for a concentration of 10% of the water
solubility S [mg/L] of the respective compounds.
100000
10000
1000
100
t
0
-JU
Phe
Rn
Ace
10
Time[h]
100
1000
Fig. 3: Sorption kinetics of Acenaphthene (Ace), Fluorene (Fin), Phenanthrene (Phe) and
Fluoranthene (Fth) onto GAG (F100). The solid line was calculated using a numerical model
(Crank-Nicolson implicit/explicit scheme) for sorptive uptake of Fin in a bath of limited volume.
Dashed, horizontal lines denote the equilibrium Kd (Ace, Fin, Phe).
62
-------
Treatment Walls
NATO/CCMS Pilot Study Phase IB
inel
1 Gate (12 m)
HomogowuK-FWd
-I I I I I I I I . I I
3M «0 00 650 7» mo
x[m]
emptied, m Figure 5 the conceptual
design of the gate construction is
illustrated In a horizontal (A) and two
vertical sections (B) and (C). The pit
will be excavated after installing a
stiffened sheet pile caisson. The
concrete building will be constructed
either in place of will be prefabricated.
Simultaneous to the construction of the
gate(s), a gravel pack will be filled
between the gate building and the sheet
piles at the inflow and outflow sides of
the gate(s). After completion of the gate,
construction the sheet piles will be
pulled out, allowing the contaminated
groundwater to flow through.
With the upper two figures ("1" and
"2") in Figure 6, two conceptually dif-
ferent possibilities of how contaminated
water may flow through the gates are
demonstrated: (1) horizontal flow and
(2) vertical flow. Vertical flow would be
favorable from sorption considerations.
It can be expected that the PAH-
concentrations will be different at
different depths. Mixing due to vertical
flow within the gravel pack will level
out the concentration profile before the
groundwater enters the gate. Horizontal flow would be favorable from hydraulic considerations. Owing to
the aquifer geometry, the cross-sectional flow area is much larger in the case of horizontal flow, leading to
a smaller hydraulic pressure build-up upstream of the gates.
x[ml
Fig. 4: Stochastic modeling of streamlines for a funnel and gate system planned at an forme
manufactured gas plant site in Southern Germany
The lower two figures in Figure 6 give
some details about how the funnel (e.g.,
slurry wall) may be connected to the
gates.
Conclusions and Further Plans
Granular activated carbon appears to be
applicable for in situ removal of PAHs
occurring in the groundwater at the
former manufactured gas plant site in
Karlsruhe. For the preliminary design of
a funnel-and-gate system at the site with
two gates and about 400 m3 total gate
volume the sorption capacity of the
activated carbon fill can be expected in
the order of 8 to 12 years.
outer gravel pack-
inner gravel pad
clay seal
V. sheet piles
outline of the gate
stiffening
' vertical section
pit stiffening
pit stiffening
suction pipes
Figure 5. Conceptual horizontal and vertical sections of the gate
construction pit
63
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Numerical flow modeling showed that the PAH-plume can be captured by a two-gate system with a total
gate length of 24 m. From the modeling results it also obvious that multi-gate systems are advantageous over
single or two-gate systems.
Heterogeneity structures with regard to hydraulic conductivity may have a major influence on the position
of the capture zone relative to the location of the funnel-and-gate system.
In order to work out the final design of the funnel-and-gate system, additional experimental and numerical
work will be necessary, including the performance of sorption column experiments at the site simulating
the conditions of in situ activated carbon filtration, laboratory experiments to optimize the choice of the type
of activated carbon to be used with regard to sorptive and hydraulic considerations, numerical stochastic
modeling in order to determine the length and location of the funnel and also to determine the number and
geometry of the gates.
These investigations will be carried out during the first half of 1998 and will provide the knowledge required
for the design and engineering of the funnel-and-gate system. The construction of the funnel-and-gate
system is expected to be carried out in 1999.
' grmvel pack toner
close-up of the connection between funnel and gate
- slurry wall (J) •— slurry wall
ClMESOmbH, I'
Figure 6. Design possibilities for the gates (1,2) and for the
connection between funnel and gates (3,4)
References
American Water Works Association (1990): Water Quality and Treatment, 4th ed.
IMES GmbH (1996). Machbarkeitsstudie Funnel-and-Gate am Standort Gaswerk-Ost in Karisrahe -
unpublished report.
Sontheimer, H., Frick, B.R., Fettig, J., Horner, G., Hubele, C., Zimmer, G. (1985). Adsorptionsverfahren
zur Wasserreinigung.- DVGW-Forschungstelle am Engler Bunte Institut der Universitat Karlsruhe
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
Teutsch, a, Grathwohl, P., Schad, H., Werner, P. (1996). In situ Reaktionswande - ein neuer Ansatz zur
passiven Sanierung von Boden- und Grundwasserverunreinigungen. Grundwasser, 1,12-20.
Trischler und Partner (1996). Sanierung ehemaliges Gaswerk-Ost in Karlsruhe - unpublished report.
Discussion
Noting ttiat carbon is an excellent medium for bacterial growth, Shad was asked how to prevent bacterial
slime from clogging the system. He said that the system is anaerobic and must be maintained under
anaerobic conditions. He has performed laboratory column tests at 20-25 °C and did not observe bacterial
growth. He also plans to perform sorption column tests under in situ conditions, using the same water, to
further examine the potential for bioclogging.
Wolfgang Wiist asked about the sum of the dissolved organic carbon in the system. Schad was not sure, but
said that there is highly organic material (including humic acids) present. Dissolved organic carbon will be
the main parameter examined during upcoming column tests. Schad will be assessing co-sorption of
contaminants and variations in sorption rates.
65
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Reactive Treatment Zones: Concepts and a Case History
Stephan A. Jefferis1 and Graham H. Norris2
ABSTRACT
The concept of the reactive treatment zone whereby pollutants are attenuated as they move along a pathway
in the ground has enabled a re-thinking of many of the concepts of containment. In particular it allows the
control of the flux from a contaminated area by controlling the contaminant concentration in the pathway(s)
as well as or instead of using a low permeability barrier. This paper outlines the basic concepts of the
reactive treatment zone and the use of permeable and low permeability reactive systems. It then gives a case
history of the design, installation and operation of a reactive treatment zone using an in situ reaction
chamber.
1. INTRODUCTION
The use of containment systems is now widely accepted for municipal and toxic waste landfill sites but up
to now it has been more reluctantly accepted for the remediation of contaminated land. This reluctance
perhaps stems from a desire to use process orientated treatment technologies as a permanent means of
dealing with contaminated land problems. However, as the effectiveness of some in situ treatment tech-
nologies remains in doubt and the costs of treatment in general remain relatively high, interest in contain-
ment continues to develop and particularly in combining treatment with containment and the prospects for
using intrinsic processes as low cost longer term remedial options (Bardos and van Veen 1996).
1.1 Terminology
In the UK, it is common practice to refer to barrier systems that impede groundwater flow as passive
containment systems, and the term "active containment" has been coined for those that reduce pollutant flux
by treating the contaminant(s) in a flow path between a source and a receptor. However, these terms are not
universally accepted. For example, in Canada and the USA, "active containment" is often used to describe
processes where there is an energy input, such as pump-and-treat systems, and the term "permeable barrier"
is used for the UK's "active containment." In this paper the more general term "reactive treatment zone" will
be used for in-ground systems involving reactions that tend to reduce contaminant concentrations in a
pathway, and "passive containment" for barriers designed to impede flow.
2. REACTIVE TREATMENT ZONES
The essential feature of a reactive treatment zone is that contaminants can be controlled in a pathway
without necessarily preventing the flow of the carrier fluid (almost always water but gases are also amenable
to such treatment).
Reactive treatment zones can be distinguished from passive (low permeability) containments by considering
the controls on the flux, which for any contaminant may be written as:
1 Colder Associates (UK) Ltd, 54-70 Moorbridge Road, Maidenhead, Berkshire, SL6 8BN, England, +44-1628-771731,
e-mail: sjefferis@golder.com
2 Nortel Ltd, Oakleigh Road South, New Southgate, London Nl 1 1HB, England +44-181-945-3556, graham.norris®
nt.com
66
-------
Treatment Wads
NATO/CCMS Piiot Study Phase Ul
Flux = Concentration x Permeability x Hydraulic gradient x Row area
For a traditional passive barrier, the flux is controlled by ensuring that the permeability of the system is low
and hence that the flow through the barrier is low. However, as a result the flow area will be high as the
barrier must surround the source and the concentration(s) of the contaminant species and the hydraulic
gradient (unless modified by pumping) also may be high.
An alternative way to control the flux is to reduce the concentration of the contaminant(s). If a sufficient
reduction in concentration can be achieved then permeability ceases to be the controlling parameter and it
may be high or low as appropriate to the application. With a high permeability system the hydraulic gradient
will be low as the carrier fluid is allowed to escape and the area for flow may be less than that of a contain-
ment as it may be deliberately focused with an engineered system or naturally higher permeability regions
may be exploited—if they can be sufficiently well defined: This may simplify monitoring, and where
necessary, in situ control. If space is available the input concentration may be varied by selecting the position
of the reactive zone in relation to the contaminant source or in-ground intrinsic remediation may be
exploited upstream or downstream of the reactive treatment zone. It should be noted that reactive treatment
zones often will be used as pathway control mechanisms to prevent contaminants from a source reaching
a receptor rather than as source clean-up technologies (at least in the short term).
2.1 Reactive Zone Configurations
In principle a reactive treatment zone or zones should completely enclose identified pathways from the
pollutant sources to the receptors. In practice this may require that the zone(s) cover a substantial area.
Furthermore the flow through any element of a zone may be uncertain as a result of factors including
seasonal variations in groundwater level and flow. It follows that large volumes of reactive material may
be necessary to ensure sufficient residence time throughout the zone under all operating conditions.
Optimization of the deployment of the reactant is therefore an important consideration. Possible
configurations include:
• The use of injection wells to introduce chemicals into the natural groundwater flow.
• The use of passive wells containing, for example, replaceable canisters of treatment materials that
dissolve into the natural groundwater flow.
• Treatment zones in the region of the source that reduce pollutant concentrations to levels at which natural
intrinsic remediation can operate in zones down gradient of the source.
• Systems that focus the flow either in plan or elevation or both.
It should also be possible to monitor and, if appropriate, control the processes within the treatment zone. The
requirements for monitoring and control can present problems for in situ treatments or intrinsic remediation
in an undefined area, given the heterogeneity typically associated with ground conditions and the large
volumes of soil that may require treatment. Martin and Bardos (1996) describe examples of in situ treatment
zones where treatment remains in the ground but under conditions of better definition. The gate of a Funnel
and Gate™ system is an example of such a system where monitoring can be particularly straightforward and
control systems can be incorporated in the works if appropriate. ,
2.2 The Permeability of Reactive Treatment Zones
Reactions in a treatment zone may influence its permeability to the carrier fluid just as they may influence
the permeability of a passive containment. It follows that chemical, biochemical and physical phenomena
within contaminant control systems must be recognized as interrelated. Contaminant control systems should
be considered as a continuum bounded by the two extreme situations:
67
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
• Permeable reactive treatment zones. The best known example is the use of zero valent iron to
dechlorinate solvents in groundwater (Gillham and O'Hannesin, 1992).
• Low permeability active containments, achieved, for example, by exploiting the chemistry of, or adding
species to, low permeability barriers such as a soil-bentonite (Evans, 1991) or cement-bentonite (Jefferis,
1996).
A reactive treatment zone must include some process that removes, destroys precipitates or otherwise
attenuates the contaminants migrating from the source(s) towards the receptor(s). Processes that may be
employed alone or in combination include: chemical reaction, physical separation, biological degradation,
andsorption. • ......
Many of these processes will rely on an interaction between the flowing groundwater and a solid phase in
the treatment zone and although the permeability of the zone is not the controlling parameter it may be
necessary to consider it. For example, a moderate to high permeability may be necessary if problems such
as ponding within a containment are to be avoided. Designing and maintaining such a permeability may pose
some problems as the reaction rate in the zone may be dependent on the surface area of a solid phase within
it. As a result it may be necessary to use finely divided (and hence low permeability) materials to limit the
required residence time of the contaminants in the zone. Also precipitates or biological slimes may form
within the zone. Thus careful consideration must be given to the initial permeability of the treatment zone
and to processes that may lead to any change over time (an increase as well as a decrease in permeability can
be damaging if the water level or residence time in the reactive zone is sensitive to its permeability). It is
perhaps ironic that a low permeability barrier may be compromised by an increase in its permeability
whereas a reactive treatment zone may be compromised by a decrease. The problems are reciprocal and if
one type of containment could be perfectly achieved there would be no need for the other.
2.3 Low Permeability Reactive Zones
Reactive treatment zones may be designed to be of low permeability, and the physical effect of the barrier
may be enhanced by processes such as chemical reaction in the region of the barrier. For example, heavy
metals may be precipitated by the high pH of cement-bentonite walls though credit is seldom given for this
in the design of such containments (except for the containment of radionuclides in purpose-designed
repositories).
A significant amount of work has been reported on the effect of contained chemicals on the physical
properties of barrier materials such as clay liners and cement or soil-bentonite walls. The prime focus of
much of this work has been the assessment of physical damage to the cut-off material, for example, its
permeability that may be caused by contained chemicals. Much less work has been reported on the beneficial
effects of chemicals, though an examination of the literature shows that some work focused on the damage
has actually demonstrated beneficial effects.
Interactions in low permeability systems are potentially much more complex than in high permeability
systems as it is necessary explicitly to consider the effects of chemical, biochemical and mineralogical
changes on the permeability and not just design to minimize their impacts. At the laboratory scale it has been
demonstrated that the chemical history of a containment material and its permeability can be strongly
interlinked. For example, leaching a cement-bentonite material by sustained permeation with water, so
reducing its pH, will usefully reduce its permeability and also inhibit expansive cracking and damage if it
is subsequently exposed to sulfates. High pH, if required, can be re-established by permeation with lime and
this can further reduce the permeability but it will re-establish sensitivity to sulfates (Jefferis 1996).
68
-------
Treatment Walls
NATO/CCMS Pilot Study Phase \\\
That such interaction can occur is not surprising but it does raise questions about the interpretation of
laboratory-investigations of chemicalidamagei'to barriers; using samples that haye not" been exposed to the
chemical history that occurs in situ (including, and very importantly, periods of leaching by water). Also it
has been demonstrated that some chemical damage can be reversed by re-introduction of appropriate
chemical species. It follows that there is considerable potential to modify the behavior of barriers once in
the ground. This could be to repair them, extend their life, protect .them from damaging species or modify
or re-charge their .chemical activity. It is therefore quite simplistic to consider cement-bentonites and indeed
most passive" barrier-systems, as having a unique permeability. They can be "physically; chemically or
biochemically active and if we can learn to exploit this activity new barriers concepts will be forthcoming.
3. CASE HISTORY OF THE APPLICATION OF A PERMEABLE REACTIVE ZONE
In addition to the problems of design of reactive treatment zones with appropriate chemical reactions there
can be significant practical problems (and also opportunities) in fitting such systems to the circumstances
of a site and this is illustrated by the following case history.
At an industrial site in Belfast, Northern Ireland, currently being used for the manufacture of electronic
components, historic spillages of chlorinated solvents had led to an intense though localized contaminant
source and plume. A site investigation identified the following contamination in the groundwater:
• Trichloroethene: • , , •-•
In situ permeability tests and pumping tests conducted to evaluate potential remedial strategies showed that
conventional pump and treat remediation would require numerous extraction wells to address the thin
saturated zone and this, together with the presence of a free product DNAPL source, meant that such a
system would have to be maintained for many years. Similarly there would be major problems with a soil
vapor extraction system. ,
3.1 Reductive Dechlorination
One of the best known and researched reactive treatments is the use. of iron filings to reductively
dechlorinate chlorinated solvents (Gillham and O'Hannesin, 1994). The process essentially leads to the
development of an extremely reducing environment, allowing reactions occurring on the surface of the iron
to strip halogens from the dissolved organic compounds as they flow through the zone. The reaction is
abiotic and chlorinated solvents are converted to chloride ions and various hydrocarbons.
69
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
The process does not degrade the resulting hydrocarbons but as non-chlorinated species they have much
higher permissible regulatory control concentrations. Also they are much more readily bio-degraded by
intrinsic remediation processes occurring in the ground. The iron filings reactive containment is thus a
conversion process that reduces toxicity. This demonstrates a general principle that a reactive treatment zone
need not degrade organic contaminants to basic species such as carbon dioxide and water, but need only
convert them to species that are acceptable lexicologically and environmentally in the local setting.
The iron filings technology seemed to have potential for the Belfast site and samples of the site groundwater
obtained from sampling wells were shipped to EnviroMetal Technologies, Inc., Guelph, Ontario, for
treatability studies. The results of these studies are reported in Thomas et al. (1995). Briefly, a series of
laboratory column studies were undertaken to determine the degradation kinetics of the species of interest
and to identify whether there were any constraints imposed by the natural geochemistry of the site
groundwater. The results showed that TCE, PCE and TCA were completely degraded from initial
concentrations in the test groundwater samples of 300,000,170 and 200 ug/1 respectively. The half life of
TCE in the site water was estimated to be 1.2 hours in the presence of iron filings.
In contrast small concentrations of cis-dichloroethene (DCE) and vinyl chloride (VC) built up in the water
as a result of the dechlorination of the TCE, etc., with VC reaching a peak at the end of the column of 700
u.g/1 in the initial studies. The reason for this is that although the dechlorination does not appear to be a
stepwise process (i.e., tetrachloroethene to tri- and then dichloroethene, vinyl chloride, and ultimately to
chlorine-free hydrocarbons), and it seems that the majority of the solvent is dechlorinated in a single step.
A small amount may be released from the surface of the iron prior to full degradation. This leads to a
temporary increase in species, such as DCE and VC, as these species are more slowly dechlorinated by iron
filings. Considerable care was necessary in the design to ensure that the DCE and VC would be satisfactorily
degraded. Otherwise, the net effect of the treatment zone could be to increase the concentrations of these
species, which would be most unsatisfactory, particularly as vinyl chloride has markedly lower regulatory
control limits than many other chlorinated solvents.
Further column studies were then undertaken at lower flow rates, the increased residence time resulting in
much earlier degradation of the parent compounds thus enabling estimation of the degradation rate of vinyl
chloride. From these laboratory trials, it was concluded that a residence time in contact with the iron filings
of about 12 hours was required for the site water to achieve removal of all chlorinated species to below
regulatory limits. This result, together with the results of a groundwater modeling exercise, enabled the
calculation of the required reactive zone path length from the estimated groundwater flow rate and the
required residence time of the contaminated groundwater in contact with the iron filings.
3.2 Design of the Reactive System
A preliminary review of the site situation concluded that the optimum strategy for deployment of the iron
filings was in the gate of a Funnel and Gate™ system. However, the site geology and site circumstances
placed a number of constraints on the design:
• The contaminant source extended to within a few meters of the site boundary, outside which there was
a public road. The reactive treatment zone therefore needed to be very compact.
• The solvent source was underlain by a thin layer of clay, which had inhibited its migration to greater
depths. If this layer were penetrated by the gate the free product solvents would sink and pollute a lower
aquifer stratum.
• The groundwater perched on the thin clay layer was shallow and showed seasonal variations in depth.
It would be difficult to maintain any significant depth of horizontal flow in a gate seated on this layer.
70
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
• A perched water table also existed in the fill covering the surface of the site. In wet seasons this water,
if allowed to enter the gate, could dominate flow through it and unacceptably reduce the residence time.
It therefore had to be excluded.
• Because of the proximity of buildings and the risk of damage by vibration and also the cost, sheet piles
could not be used to form either the funnel or the gate of the system.
• If a slurry trench procedure were used to form the system then it was imperative that the iron filings were
not inundated and thus blocked by slurry from the wall construction. It would therefore be necessary to
contain the iron or to construct the slurry wall before placing the iron filings. Prior formation of the cut-
off wall could allow it to slump during excavation for the iron gate and thus a poor seal between the wall
and the iron filings.
• The residence time of groundwater within the reactive treatment zone is critical. A zone that is too thin
may lead to the escape of lower chlorinated species, such as vinyl chloride, which can pose a greater
lexicological or environmental threat. Extensive groundwater modeling studies were undertaken to
ensure that the design residence time could be achieved.
• The clean-up was being undertaken voluntarily by the owner of the site.
After consideration and rejection of many reactive treatment zone designs, the in situ reactor configuration
shown in Figure 1 was developed as best fitting the constraints imposed by the site. In place of previously
used horizontal flow reactive treatment zones the flow was arranged to be vertical in a 12 m tall by 1.2 m
diameter steel reactor shell which was filled with iron filings as shown in the figure. This design enabled
the reactor to be placed between the contamination
and the site boundary. This could not have been
achieved with a horizontal flow regime as the design
calculations had shown that the flow path length
needed to be at least 5 m plus entry and exist zones
to collect and disperse the flow. The vertical flow
direction within the reactor also ensures that the full
depth of the iron filings will be saturated whatever
the seasonal variation in groundwater level.
The reactor was placed in an enlargement in a
cement-bentonite cut-off wall (Figure 1), which was
used to funnel the flow to the reactor. This wall was
toed into an aquiclude layer at 10 m and the enlarge-
ment was taken to a depth of slightly over 12 m to
accommodate the reactor shell. The cut-off and
enlargement penetrated through the clay layer on
which the chlorinated solvents were retained.
However, as the cut-off material was designed to
have a permeability of <10'9 m/s minimal downward
migration of solvents will occur.
Because of the relatively low permeability and
heterogeneity of the adjacent soil, it was decided that
the flow to the reactor should be collected via an
upstream high permeability cplledtor and that
downstream of the reactor there should be a similar
Gravel
, fitted
collector
1
J
*•
aminated area
In-situ
reactor
- 4
&
'si
diswbutoe
Clean area
12m
_<3round
level
Cut-off wall
Figure 1. Reactor configuration, elevation,
and plan view
71
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
distributor. The collector and distributor were formed from gravel filled piles taken down to the top surface
of the thin clay layer. A gravel backfilled trench, excavated under a degradable polymer slurry, could have
been used to form the collector and distributor but there were concerns about the effect, on the iron filings,
of any polymer remaining in the trench and so augured pile holes were used.
Finally, the internal geometry of the reactor was arranged so that the pipe connections to the gravel filled
collector and distributor piles could be made from within the clean environment of the reactor shell without
the need for any hand excavation or for anyone to enter the excavations all of which were undertaken either
with a backhoe under cement-bentonite slurry or with a piling auger. During the works airborne solvent
concentrations were monitored and found to be undetectable.
3.3 Performance of the Reactor
The performance of the reactor has been and continues to be very satisfactory though there are two marginal
complicating factors:
• The flow rate through the reactor was very low during the first months of operation and. is likely to
remain quite modest (this is as expected from the hydrological investigation). It follows that the internal
pore water chemistry is rather sensitive to any disturbance of the flow regime
• Sampling from within the relatively modest volume of the reactor has proved to be a problem if
traditional procedures are used. Only small volumes of water can be withdrawn if the local chemistry
within the reactor is not to be disturbed. Particular techniques are being developed to obtain small but
representative samples from the reactor whilst avoiding the need to purge the sampling tubes.
The results of five campaigns of sampling within the reactor and in the upstream collector and downstream
distributor are shown in Figures 2,3, and 4. The results relate to a period of 19 months from installation of
the reactor.
Figure 2 shows the data for TCE as a
three dimensional plot of concentration
as a function of position in the reactor
system and time. There are five sampling
points within the reactor shell. Sampling
point R5 is at the entry point to the iron
filings bed from the collector and Rl at
the discharge point to the distributor.
The other three sampling points, R2
through.,R4, are, distributed at 1.5 m
intervals through the 6 m deep bed.
It can be seen that the of TCE within the
reactor drops very rapidly and is at very
low levels by the time it has reached the
18-Apr-96
17-JUI-9S
24-OCt-98
21-Jan-97
12-Aug-97
Figure 2. TCE concentrations in the reactor zone
first sampling point (R4) at 1.5m into the reactor bed. Li August 1997 the average TCE concentration within
the reactor bed-was less than 4 micrograms/liter. Figure 2 also shows that there is still a significant, though
reducing, TCE concentration in the downstream distributor and the reactor discharge point (Rl). This is as
a result.of the TCE originally present in the downstream area some of which may have diffused back into
the reactor exit zone or have been drawn back into the reactor by the sampling process. However, it seems
most likely that the actual reason for its presence is the advection of TCE contaminated groundwater through
.72
-------
Treatment Waffs
NATO/CCMS Pilot Study Phase HI
the distributor zone as a result of groundwater flow in the downstream area. It should not be forgotten that
there may be a downstream flow vector parallel to as well through a gate.
Figure 3 shows the chlorine balance in the reactor zone. For this it is necessary to consider both the chlorine
in the inorganic chloride ion and in the TCE. The TCE chlorine drops rapidly with distance into the reactor
and only increases again in the distributor area. The total chlorine shown in the figure is the sum of the TCE
and chloride chlorine and it can be seen that it is relatively consistent throughout the reactor (the high total
chlorine on 21 August 1997 was as a result of a high chloride ion concentration, the reason for it is not
known but could be as a result of some slight seepage of de-icing salts used in winter time.
300
TCE
..»..18-Apr-96
..+~i7-Jul-96
.. + ..24-Oct-96
.••+•• 21 -Jan-97
• -m--12-Aug-97
Total
—»— 18-Apr-96
— »- 17-JUI-96
-+—24-Oct-96
12-Aug-97
Collector
Distributor
Figure 3. Chlorine balance in the reactor zone
On each sampling occasion TCA was at undetectable levels. Vinyl chloride was found at a single sampling
point within the reactor at 4 months (at a level of 0.4 mg/1) but not in the collector or distributor and has not
been detected on any subsequent occasion.
With regard to the remaining downstream contamination it is interesting to note, that at times of low
seasonal flow, there is potential for the recycling groundwater from wells downstream of the reactor to the
upstream collector (by pumping) to treat the downstream area should treatment of this contamination be
required. Also the reactor will work with the flow in either direction. Thus, if, for some reason, the flow
should be reversed naturally then the site would be protected from the inflow of any contaminants spilled
or remaining outside the reactor system.
4. CONCLUSIONS
The reactive treatment zone concept offers not only a new tool .for the control of contamination but also
forces a re-thinking of low permeability barrier systems. Prior to the development of the concept,
contaminant/barrier interaction had generally been considered only as a possibly damaging process that
could increase the permeability of a barrier. Now barrier interactions should be seen differently. They may
be deliberately engineered to be beneficial.
73
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Reactive treatment zones may be most economically employed as pathway control procedures but will
contribute to the slow clean-up of a source. The pathway control they offer is their essential feature and it
maybe argued that they should not be designed to achieve more than slow source clean-up as this minimizes
the rate at which any treatment chemicals, energy or other process inputs are required i.e. it minimizes the
process demands and operating costs whilst maximizing the opportunity for natural, intrinsic, remediation.
At the present time a significant limitation to the use of reactive treatment zones is the rather narrow range
of contaminants that can be treated and the complications introduced by mixed contaminants etc. However,
work is in progress and the technology has strong parallels with intrinsic remediation (e.g., intrinsic
bioremediation), and it is likely that the reactive treatment zone concept will find much application in
stimulating intrinsic degradation or attenuation.
The Funnel and Gate™ concept is a powerful adjunct to the technology as it can allow the treatment zone
to be at some distance from the source, if this is required by surface or buried structures, geological
conditions or for process optimization (e.g., exploiting natural dilution or attenuation in the soil), etc. Sites
need not be capped to prevent rainwater ingress, indeed it can be an advantage in driving groundwater flow
to a treatment zone.
The concept of an in situ reactor adds further flexibility to the design of reactive treatment zones, allowing
more precise control of the reactive zone and also recharging or recovery and replacement of the active
material should this be required. Furthermore several reactors may be linked in series to treat mixed
contaminants.
Although most reactive treatment systems that have been advanced to date have been for the control of
groundwater pollution there is no reason why other systems such as reactive caps should not be developed
to control, for example, the escape of semi-volatile or volatile organics from petroleum impacted sites or gas
or odor from landfills. Peat or humic based materials could be particularly cost effective in this respect (and
of course for the removal of organics from impacted water). Bio-oxidation zones to control methane
migration through unsaturated soils offer some exciting prospects but there could be difficulties with
excessive local heating.
Low-permeability reactive treatment zones are a more specialist application and may require site capping
or groundwater pumping (and treatment) to avoid ponding. Reactive treatment zone concepts will be useful
where long term secure containment is necessary as chemical interactions can provide additional security
and lifetime to a physical containment. Furthermore the concepts may be useful in developing in situ repair
strategies for barriers that have suffered damage and in analyzing the effects of reaction on the permeability
of these systems (Jefferis, 1996).
5. ACKNOWLEDGMENTS
Colder Associates gratefully acknowledges the proactive technical and financial support given by Nortel Ltd
who funded the reactive treatment project referred to in the case study.
The support of the UK Department of the Environment, Contaminated Land and Liabilities Branch, is also
gratefully acknowledged. They have funded Golder Associates to undertake a research project EPG 1/6/16
on active containment concepts. This paper includes review work carried out for the Department of the
Environment. However, the views expressed in this paper are those of the authors only and do not
necessarily reflect the views of the Department of the Environment. Reference to any individual or
organization or mention of any proprietary name or product in this paper does not confer any endorsement
by the authors nor the Department of the Environment.
74
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
The award, of an ERDF Northern Ireland Single Programme (1994 - 1999) Environmental Services and
Protection Sub-Programme grant of 50% of qualifying costs associated with the Nortel reactive barrier
project is also gratefully acknowledged.
6. REFERENCES
Bardos, R.P. and L van Veen. (1996) Longer term or extensive treatment technologies. Land Contamination
& Reclamation^ 4(1), pp. 19-36.
Evans, J.C., Y. Sambasivam and Z.R. Zarlinski. (1991) Attenuating materials in composite liners, Waste
Containment Systems: Construction, Regulation and Performance, ASCE Geotechnical Special
Publication, No 26.
Gillham, R.W. and S.F. O'Hannesin. (1994) Enhanced degradation of halogenated aliphatics by zero-valent
iron, Groundwater, 32(6), pp. 958-987.
Gillham, R.W. and S.F. O'Hannesin. (1992) Metal catalysed abiotic degradation of halogenated organic
compounds, IAH Conference "Modern trends in Hydrogeology". Hamilton Ontario, Canada, pp. 94-
103.
Jefferis, S.A. (1996) Contaminant - barrier interaction: friend or foe? Mineralogical Society Conference,
Chemical Containment of Wastes in the Geosphere.
Martin, I. and R.P. Bardos. (1996) A review of full scale treatment technologies for the remediation of
contaminated soil. (Final report for the Royal Commission on Environmental Pollution, October 1995).
EPP Publications, 52 Kings Road, Richmond Surrey.
Starr, R.C. and J.A. Cherry. (1994) In situ remediation of contaminated groundwater: The funnel-and-gate
system, Groundwater, 32, pp. 465-476.
Thomas, A.O., D.M. Drury, G. Norris, S.F. O'Hannesin, and J.L. Vogan. (1995) The in situ treatment of
trichloroethene contaminated groundwater using a reactive barrier - results of laboratory feasibility
studies and preliminary design considerations. Contaminated Soil '95, W.J. van den Brink, R. Bosnian
& F. Arendt (eds).
Discussion
There was some discussion concerning the placement of the reactor vessel relative to the collector and
distributor, and whether there would be any cost savings in installing the reactor at the same time as the cut-
off wall, collector, and distributor. Jefferis replied that there are circumstances where there would be no cost
savings for simultaneous construction. He said that sometimes it would be cheaper to build a pipe through
the cut-off wall rather than place the reactor within it.
The reactor was designed to have a retention time of approximately 1 day. It is difficult to determine the
exact flow through of the reactor because it was operated in the summer when the through-flow was too low
to measure. The highest flow occurs hi the winter, and the reactor was designed for a maximum flow through
of 5 m3/day. Other chemical parameters measured at the site were pH, PCE, DCE, vinyl chloride, and some
inorganics.
75
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Harald Burraeier questioned whether the groundwater was pumped through the reactor. Jefferis indicated
that it is a gravity-flow reactor, but if necessary, groundwater could be pumped through to treat water, or to
pump down in the winter during peak flow to augment the low flow during the summer.
Wiist commented that the system seemed to work well, but wondered whether the hydraulic design was
adequate to clean up other contaminants besides TCE. Jefferis noted that no other contaminants were
present, other than cw-DCE, which was found only in the distributor (not in the flow through). He added
that the higher concentrations shown in Figure 2 were found in the distributor located downstream of the
reactor. He said that the reactor has unused capacity, and he would like to pump back the flow through; he
acknowledged his disappointment in how slowly the concentration was dropping. Geologically, it is a very
heterogeneous, complex site. There may be some cross-flow bringing in contamination directly from the
plume.
76
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
Horizontal Treatment Barriers of Fracture-Emplaced Iron and Permanganate Particles
' Robert L. Siegrist1, Kathryn S. Lowe?, Lawrence W. Murdoch3,
Traci L. Case1, Douglas A. Pickering2, and Thomas C. Houk4
1.0 Abstract
In many contaminated sites, vertical leaching and/or volatilization of organic compounds can lead to high
exposures and unacceptable human health or environmental risk. One remediation alternative for these sites
is to emplace horizontal treatment barriers to reduce the source concentration and mass as well as intercept
and treat any mobile organics in the subsurface. Development and demonstration activities have been
conducted to evaluate the feasibility of creating horizontal treatment barriers by employing hydraulic
fracturing for emplacement of chemically reactive solids. In this work, iron metal particles and a new
permanganate solid media were both tested in horizontal barriers to achieve interception and degradation
of chlorocarbons like trichloroethene (TCE). Laboratory experiments were completed to develop suitable
delivery fluids and to determine reaction rates and efficiencies. The iron metal particles could be suspended
and delivered using a typical guar gum gel while the permanganate solids required a new mineral-based
carrier fluid. Degradation of TCE at -10 to 80% aqueous solubility with iron metal particles exhibited a half-
life in the range of 1 to 2 hr while that of the permanganate solid was much faster with a half-life of only
several min or less. A full-scale field test was subsequently conducted at an old land treatment site in the
midwestern USA. Two test cells were installed with horizontal barriers comprised of reactive fractures
emplaced at -1.2,2.4, and 3.6-m depth in silty clay soils using hydraulic fracturing methods. After 3,10 and
15 mon of emplacement, continuous cores were collected across the emplaced fracture zones and
morphology was observed, geochemical properties were measured, and degradation tests were completed
with TCE contaminated groundwater (TCE concentrations at -5-50% solubility). These analyses revealed
strong short-range trends in redox potential and pH consistent with the anticipated behavior of the two
different reactive media emplaced. Highly oxidizing zones were present in and around the permanganate
barriers. TCE degradation efficiencies during 2 hr of contact of >99% in 10-cm thick soil zones after -3 mon
of emplacement with even broader effects at -10 and 15 mon. The treatment barrier comprised of 200 urn
iron-metal particles exhibited highly reducing conditions within the iron-metal fracture but no marked
effects in the soil zone surrounding it. Degradation of -35% was achieved after 24-48 hr of contact but only
in the iron-metal itself. The iron exhibited some corrosion, but this did not appear to affect its reactivity. The
results of the work to date are very positive regarding the performance for both types of horizontal barriers
and the estimated costs for a typical application are comparably low at only $30/m3. Further work is ongoing
to enable design and deployment of this technology.
2.0 Introduction
Petrochemicals (e.g., benzene) and chlorocarbons (e.g., TCE) are common and problematic contaminants
of concern (COCs) at federal facilities and industrial sites across the U.S. and abroad (USEPA, 1992; API,
1995; DOE, 1996). These contaminants can be released to the environment through leaks in storage tanks
and transfer lines, spills during transportation, and the land treatment of wastes. They are often present in
source areas and in soil and groundwater plumes as dissolved or sorbed phase constituents as well as
nonaqueous phase liquids (NAPLs). When these COCs are present in silt and clay media either as massive
'Colorado School of Mines, Environmental Science and Engineering Division, Golden, CO. USA 80401-1887. Ph. 303-
273-3490.Fax.303-273-3490.Email.rsiegris@mines.edu.
2 Oak Ridge National Laboratory, Environmental Sciences Division and Life Sciences Division, Grand Junction, CO.
3 FRx, Inc., Cincinnati, OH and Clemson University, Clemson, SC.
4 Lockheed Martin Energy Systems, Inc., Piketon, OH.
77
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
surface deposits or as interbedded lenses in otherwise permeable formations, there are major challenges with
assessment of their behavior and implementation of effective in situ remediation technologies. Despite a low-
bulk permeability of silts and clays (e.g., K^ < 10'7 cm/s), organic compounds can contaminate fine-grained
deposits by preferentially moving into and through naturally occurring pore and fracture networks and
partitioning between the nonaqueous, gas, aqueous, and sorbed phases. In the unsaturated zone, organics can
continually volatilize into the soil air or leach into percolating water while in the saturated zone they can
slowly dissolve into advecting groundwater. Exposures and unacceptable risk can result from ingestion of
contaminated drinking water or surface soil media, or inhalation of vapors in ambient air or emitted during
showering events. In recognition of the severity of the problem and the need for effective in situ treatment
methods, organic compounds in fine-grained deposits was recently ranked as one of the top environmental
restoration needs across the DOE Complex (DOE, 1996). Similarly, within the petroleum industry, nearly
40% of the underground storage tanks in the world are located on clay soils and leaking underground tanks
and remediation of organic contamination in these settings remains a major challenge (API, 1995).
In situ remediation by conventional methods such as mass recovery by soil vapor extraction (SVE) or in situ
destruction by biodegradation or chemical oxidation are often ineffective for NAPL compounds at sites with
silt and clay media. Poor accessibility to the contaminants and difficulty in delivery of treatment reagents
throughout these deposits have made rapid and extensive in situ remediation difficult. Some aggressive
techniques have been developed involving subsurface disruption by soil mixing or alternative driving forces
based on electrokinetics (e.g., Siegrist et al., 1995). In seeking less intensive methods that could be used over
larger areas, soil fracturing techniques have been developed whereby hydraulic or pneumatic fluids can be
injected to create permeable horizontal layers. This can increase the overall permeability of the subsurface
and improve contaminant recovery as existing channels or pathways can be expanded or new channels or
pathways can be created (U.S. EPA, 1993; Murdoch et al, 1997). Hydraulic methods normally employ an
agent or proppant (e.g., sand) to fill and support the fracture opening that was created and thereby prevent
fracture closure during natural healing processes in unconsolidated deposits. If the fractures can be spaced
throughout the deposit, they potentially could be used to (1) enhance the recovery of organic compounds,
(2) deliver and distribute treatment fluids into the deposit and accomplish destruction in place, or (3) place
reactive solids as an integral part of the emplaced fracture that could serve as a horizontal treatment barrier.
Two reactive media that appeared appropriate for use in horizontal treatment barriers included: (1) iron
metal particles and (2) permanganate crystals. The use of zero valent iron metal as a reactive barrier media
has been developed and deployed in vertical orientations for several years (e.g., O'Hannesin and Gillham,
1998). The reductive dechlorination reaction for TCE is shown in equation 1. The reaction involves single
and multiple electron transfers following pseudo-first order kinetics (at least at lower TCE concentrations)
with half-lives on the order of 30-60 min as normalized to solution to solid surface area. The pH of the
reacting system tends to rise, but appears to stabilize in the pH 9-10 range due to iron hydroxide
precipitation.
2Fe°+
3H2O -
H2 + Cl' + 3OH'+ 2Fe+
[1]
The use of potassium permanganate for oxidative degradation of organic pollutants in contaminated land
has evolved over the past few years (e.g., Gates et al, 1995). The oxidative destruction of TCE is given in
equation 2. The reaction can include destruction by direct electron transfer or free radical advanced
oxidation. The reaction follows pseudo-first order kinetics and is very rapid with half-lives on the order of
sec to min. The pH of the reacting system can decline to very low values depending on the buffering capacity
of the system.
2KMnO4
- 2CO + 2MnO2 + 2KC1 + HC1
[2]
78
-------
Treatment Walls
NATO/CCMS Pilot Study Phase TO
3.0 Horizontal Treatment Barriers
3.1 Development and Demonstration Activities.
While there has been extensive research and development followed by increasing application of vertical
treatment walls, research and development of horizontal treatment barriers has only recently been
accomplished (e.g., Murdoch et al. 1997, Siegrist et al. 1997). During the past three years, these authors have
been studying horizontal barriers comprised of either iron metal or permanganate solids delivered via
hydraulic fracturing methods. This work has included laboratory and field testing to evaluate their feasibility
and performance for in situ degradation of TCE in fine-grained media deposits (Figure 1). The results and
implications of this work are highlighted
in this paper while details regarding the
methods and results can be found else-
where (e.g., Murdoch et al., 1997;
Siegrist et al., 1997; Case and Siegrist
1997).
The laboratory and field research
conducted by these authors has demon-
strated that conventional hydraulic
fracturing equipment and methods could
be employed to emplace reactive solids
in silty clay soils at depths of 1.2-5 m
with diameters of 6-9 m. Fractures
containing 200 um iron particles as the
proppant were successfully emplaced at
depths of -1.8,2.4, and 3.6 m. Based on
morphology and geochemical
measurements along profiles transecting
the fractures, the iron proppant remained
reactive even after 10-15 months of
emplacement, but there was little effect on the enveloping soil matrix (Figure 2). Based on batch degradation
tests with samples of the iron from the fracture or soil media surrounding it and initial TCE groundwater
concentrations ranging from -50-500 mg/1 (equivalent to -400-4000 mg/kg), the degradation achieved was
equal to only -35% during reaction periods of 24- or 48-hr. While the iron metal in the fracture did show
signs of surface corrosion, its degradation efficiency was comparable to that of unused iron even after nearly
a year of emplacement in the moist silty clay subsurface.
Fractures containing a new permanganate solid mixture as the fracturing fluid and proppant were
successfully emplaced at depths of -2.1,2.7, and 3.4 m. Morphology and geochemical measurements along
representative profiles indicated that this process created highly reactive fractures enveloped by zones of
reactive soil 30-60 cm thick (Figure 3). Dissolved TCE with initial concentrations of from -50-500 mg/1
(equivalent to -400-4,000 mg/kg) was completely degraded (100% efficiency) in batch tests using either
material from the fractures or the reactive soil from above and below the fractures. Complete degradation
was observed in as little as 2 hr. The degradation potential as a function of time and space, was controlled
by the mass of permanganate ion that was present in the fracture or the fraction of it that had migrated into
the soil matrix and had not been previously consumed.
Figure 1. Illustration of a horizontal treatment barrier concept with
hydraulic fracturing for reactive media emplacement (Note: the
barriers shown contain reactive media that has the ability to
dissolve and effect soil-solution biogeochemistry changes).
79
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
3.2 Discussion and Implications.
The viability of reactive fractures
emplaced by hydraulic fracturing
methods as horizontal barrier systems
requires consideration of the horizontal
continuity, degradation capacity, and
longevity. The results obtained by these
authors enable a very interesting contrast
between reactive fractures created with
surface reactive particles (i.e., iron metal
particles) versus those created with a
reactive grout that dissolves and yields a
wider reactive zone (i.e., permanganate
grout). Both types of reactive media
were successfully handled and emplaced
by conventional hydraulic fracturing
equipment and methods. The geometry
of the fractures was similar to that of
conventional sand-filled fractures
emplaced at the same site. Thus, there
was no unusual behavior associated with
the different fracturing proppants (/.«.,
iron particles in guar gum versus
permanganate particles in mineral carrier
versus sand in guar gum).
Redox (mV)
horizontal fracture location
pH
Figure 2. Features of iron metal horizontal barriers 10 months after
initial emplacement in a silty clay soil (after Siegrist etal., 1997).
(a) Media redox potential and pH above and below the horizontal
treatment fracture, (b) TCE degradation potential above and below
the horizontal treatment fracture as measured in batch tests using
3-5 g of media in 40 ml of groundwater.
GW1 =480mg/LTCE
Of)
10
*•"•*
1 '
2 o
§ o
\
'0 k
. « hr ™, (b) GW2
[] 48 hr rxn • • _••
&••"*' sp% 100
Reduction (%)
Co
Since the iron metal fractures are
discrete reactive sheets, it is not likely
that these fractures would be self-healing
within or between fractures or effective
beyond the boundary of initial
emplacement. This challenges the fracture emplacement to be continuous and uniform horizontally with no
breaches through it. Also, degradation of the TCE or related compounds within the deposit must rely on their
being mobilized to the fracture where they can contact the metal surface. Since the TCE reaction rates with
iron metal are comparatively slow, any mobile contaminants must remain in contact with the iron metal for
a reasonably long period of time (e.g., on the order of 18-24 hr). If flow through the fracture is controlled
by unsaturated conductivity in the surrounding soil media and assuming the conductivity is about 10"6 cm/s,
then the retention time in a fracture of 5 mm thickness could be on the order of a day or more depending on
the effective porosity in the iron. Thus, adequate retention time can be reasonably expected. As a treatment
barrier, the reactivity of the iron surface would need to exist for an extended period (e.g., 2-5 years or more).
Analysis of the micromorphology of the surface of the unused and used iron metal revealed some corrosion
of the iron surface after emplacement for ~10 months. The effect was limited however to only a fraction of
the available iron surface and this corrosion had no apparent effect on TCE degradation. Whether this would
hold true for longer emplacement periods is currently unknown.
With permanganate fractures, the dissolution of the solid permanganate particles will yield MnO4" ions that
will enter the soil solution and migrate away from the original location of emplacement by advection and/or
diffusion. This will yield an oxidizing zone that can treat TCE within the affected zone enveloping the
fracture as well as intercept and degrade mobile contaminants that enter the oxidation zone. This behavior
suggests that the permanganate fractures will be somewhat self-healing internally and between horizontally
80
-------
Treatment Walls
NATO/CCMS Pilot Study Phase ffl
500
10 X)
7.00 9.00 A 11
00
Radox (mV)
horizontal fracture location
(a) Media redox potential and pH above and below the horizontal
treatment fracture.
Figure 3. Features of permanganate horizontal barriers as
observed 10 months after initial emplacement (after Siegrist et al.,
1997).
(b) TCE degradation potential above and below the horizontal treatment
fracture as measured in batch tests using 3-5 g of media in 40 ml of
groundwater.
GW1=480mg/LTCE
54mg/LTCE
proximal fracture emplacements. With
regard to longevity, the permanganate
solids present in a 5 mm thick fracture
contains about 0.4 g KMnO4 per cm2 or
fracture horizontal area Based on
complete oxidation and a stoichiometric
dose of -2.5 wt./wt., each cm2 of
fracture can treat -0.16 g of TCE, which
is equivalent to roughly 161 of percolate
with a concentration of 10 mg/1 of TCE.
At a deep percolation flux of 1 cm/day,
this potential is equivalent to about 50
years of life. Realistically though, it is
anticipated that advective loss of oxidant
out of the treatment region or the oxidant
demand of natural organic matter will
markedly diminish this life. Based on
direct observation in this study, the
oxidation capacity within and around the
permanganate fracture was very high
even after 10-15 months of
emplacement. If alkenes, such as TCE,
were migrating downward in percolating
groundwater at reasonably high levels
(e.g., 10-100 mg/1), the capacity of the
fractures seemingly would be high
enough to intercept and treat this mobile
TCE.
The cost of a horizontal treatment barrier
comprised of reactive fractures such as
tested in this study were estimated.
When compared to standard hydraulic
fracturing and SVE, the major cost differences would be added costs for the reactive media (~$1.6/lb for
permanganate and ~$0.38/lb for iron metal) as compared to standard fracturing sand (~$0.10/lb) but lower
operational costs. The costs for media per fracture amounts to roughly $100 for sand, $1,000 for iron, and
$ 1,500 for permanganate. This increases the installation cost for the iron and the permanganate fractures.
However with lower resource consumption (e.g., power), less sampling and analysis (e.g., no off-gas),
reduced labor requirements, and no off-gas treatment costs, the operational costs would be substantially
lower for the horizontal barriers as compared to fracture-enhanced vapor extraction. For a 2.2 ha site and
a 5-m deep zone of contamination, the costs for emplacement of horizontal treatment barriers was estimated
to be in the range of $30/m3.
In summary, horizontal treatment barrier systems appear viable for TCE contaminated sites with silt and clay
deposits. Further work is necessary and appropriate, however, to fully develop this horizontal treatment
barrier approach and provide needed design and performance data for a range of full-scale applications. For
example, information is still needed on the depth and interval of emplacement as well as reactivity and
degradation capacity over time. The behavior of horizontal barrier systems under conditions of forced
advection also needs to be evaluated to understand the benefits/costs of recirculation approaches to in situ
treatment and source area mass reduction as opposed to more passive barriers for simple interception and
treatment.
Reduction (%)
Reduction (%)
81
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
4.0 References
American Petroleum Institute. 1995. Petroleum Contaminated Low Permeability Soil: Hydrocarbon
Distribution Processes, Exposure Pathways and In Situ Remediation Technologies. Health and
Environmental Sciences Dept. Publication No. 4631. Sept. 1995.
Case, T. 1997. Reactive Permanganate Grout for Horizontal Permeable Barriers and In Situ Treatment of
Groundwater. M.S. Thesis. Colorado School of Mines, Golden, CO.
Case, T.L. and R.L. Siegrist. 1997. Oxidation Kinetics of Trichloroethylene in Groundwater by Perman-
ganate under Varying Matrix Conditions. J. Water Environment Research. In review.
DOE. 1996. In Situ Remediation ofDNAPL Compounds in Low Permeability Media: Transport/Fate, In
Situ Control Technologies, and Risk Reduction. Oak Ridge National Laboratory Report, ORNL/TM-
13305, for the U.S. Department of Energy. August, 1996.
Gates, D.D., R.L. Siegrist, S.R. Cline. 1995. Chemical Oxidation of Volatile and Semi-Volatile Organic
Compounds in Soil. Proc., Air and Waste Management Association Conf. June.
Murdoch, L., B. Slack, B. Siegrist, S. Vesper, and T. Meiggs. 1997. Hydraulic Fracturing Advances. Civil
Engineering. May 1997. pp. 10A-12A.
O'Hannesin, S.F. and R.W. Gillham. 1998. Long-term Performance of an In Situ "Iron Wall" for
Remediation of VOCs. Ground Water. 36(1): 164-170.
Siegrist, R.L., O.R. West, J.S. Gierke, et al. 1995. In Situ Mixed Region Vapor Stripping of Low
Permeability Media. 2. Full Scale Field Experiments. Env. Sci. and Technol. 29(9):2198-2207.
Siegrist, R.L., K.S. Lowe, L.C. Murdoch, D.A. Pickering, T.L. Case. 1997. In Situ Oxidation by Fracture
Emplaced Reactive Solids. J. Environmental Engineering. In review.
U.S. EPA. 1992. Dense Nonaqeuous Phase Liquids-r-A Workshop Summary. EPA/600/R-92/030. Office
of Research and Development, Washington, D.C. 20460.
U.S. EPA. 1993. Hydraulic Fracturing Technology—Technology Evaluation Report. EPA/540/R-93/505.
Office of Research and Development, Cincinnati, OH.
Discussion
Noting the rapid release rate for permanganate, Resat Apak asked whether permanganate sheets can be used
only in low-permeability aquifers. Siegrist acknowledged that his work involves the use of permanganate
grout in low-permeability zones; however, there is work being done with controlled releases of perman-
ganate. With a slower release rate, perhaps the technology can be tailored to higher permeability zones.
Apak also whether Siegrist has observed clogging due to the formation of manganese hydroxide. Siegrist
noted that the product of the reaction between manganese and TCE is manganese oxide precipitate, which
ranges in size from 1-2 |im. The formation of the solids is somewhat related to the concentration of reactants.
Siegrist has observed limited formation of solids at the test site, but has not seen an accumulation at the
interface. Clogging has not been a problem with low-permeability zones, but may be a problem with higher
permeability zones.
82
-------
treatment Walls
NATO/CCMS Pilot Study Phase m
Paul Bardos questioned whether the capacity of the permanganate is reduced due to oxidation of organic
matter in the soil. Siegrist agreed that permanganate is an oxidant that reacts with TCE and natural organic
matter in the soil. However, permanganate does not seem.to be consumed as much by the organic matter.
Siegrist noted that residual permanganate remains in zones with organic material after 15 months; however,
the soil matrix should be assessed for the potential to consume the permanganate.
The cost difference between using permanganate and iron is not significant. The cost per fracture of using
sand alone is approximately US$100. The cost per fracture is US$1,000 for iron and US$1,500 for
permanganate. The retention time in the iron-filled fracture is 1-5 days, during .which there is 30-40 percent
degradation. Siegrist noted that the cost estimate of US$30/m3 that he cited during his presentation applies
to both iron and permanganate. Both materials appear to be viable, although they have different
characteristics. .
83
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
In Situ Remediation Research in a Complexly Contaminated Aquifer:
The SAFIRA Test Site at Bitterfeld, Germany
H. Weiss1, F.-D. Kopinke, P. Popp, and L. Wunsche
The experience of the last 20 years has shown that hydraulic soil and groundwater remediation schemes are
often very ineffective. This is especially true of large-scale pollution, where the source of contamination
either cannot be precisely located or is problematical to remove. Owing to the frequently low solubility of
hydrophobic organic substances (for example), the existence of a separate residual phase, the usually uneven
subterranean flow, and the sometimes slow return diffusion of pollutants that have penetrated the rock over
a period of years or even decades, solely hydraulic remediation can only achieve limited clean-up. The
consequences are very prolonged (and thus expensive) pump-and treat remedies, whose completion is
usually impossible to predict.
Against the background of the now generally acknowledged problems of pump-and-treat activities, intensive
work has been carried out in recent years on developing low-price in situ remediation measures. The
currently most advanced remediation technology working in the flow is that of "in situ reaction/adsorption
walls" [1], which have already been tried out for simple pollutant mixtures (e.g., LCKW, BTEX, and PAK)
at a number of model sites. Nevertheless, there is still a great deal of development work to be done with
respect to the complex pollutant mixtures commonly encountered above all in the vicinity of abandoned
chemical plants.
As this problem is extraordinarily important in connection with the large-scale ecological remediation
schemes in eastern Germany, the SAFIRA2 project has been set up to examine and further develop the usage
of in situ reaction walls at a contaminated model field location.
In order to create the necessary basis for this model project, a preliminary study was carried out by UFZ in
cooperation with the Universities of Dresden, Halle, Leipzig, Stuttgart, and Tubingen. During the course of
this preliminary study, the hydrogeological and hydrogeochemical conditions at the planned field location
were investigated and various technologies for the design of the in situ pilot plant were examined and tested.
Choosing the Model Location
The Bitterfeld region was selected as the model location for investigations into developing powerful in situ
technologies for the remediation of complexly contaminated groundwater.
The soil and water environmental compartments in the Bitterfeld/Wolfen district have suffered sustained
damage as a result of over a century of lignite-mining and chemical industry. Whereas relevant soil pollution
is mainly confined to industrial locations (plant sites) and landfills, the persistent penetration of the
groundwater by pollutants has resulted in contamination attaining a regional scale. Consequently, an area
of about. 25 km2 with an estimated volume of some 200 million m3 is now partly highly polluted and must
1 UFZ- Centre for Environmental Research Leipzig-Halle, Department Industrial and Mining Landscapes, P.O. Box 2,
D - 04301 Leipzig, Germany
2 "Sionierungs-Forschung/n .Regional kontaminierten Aquiferen" - Remediation Research in Regionally Contaminated
Aquifers
84
-------
Treatment Walls
NATO/CCMS Pilot Study Phase M
be regarded as an independent source of contamination [2]. This pollution is characterized by the extensive
distribution of halogenized hydrocarbons, especially chlorinated aliphatics and chlorinated aromatics.
The development of the hydraulic conditions and thus regional pollutant transport is determined by the
draining techniques used for lignite-mining and its subsequent abandonment, as well as the imminent
flooding of Goitsche opencast mine. Since 1990, the hydraulic and conditions have been described and
evaluated in several studies (e.g., [2,3,4]; for further reading see [5]), so that there is an extraordinarily high
level of knowledge concerning the area in question. The containment and remediation concepts drafted since
1994 [2-4] feature both "active" and "passive" hydraulic containment measures to minimize groundwater
mobility. Successive measures to contain landfills (e.g., encapsulation) are also being discussed. "Active"
measures for groundwater remediation on a regional scale are not under consideration owing to the volumes
involved and the ad infinitum emission of pollutants from the landfills.
3/97 1/96
2/96
800
'•70.0
90.0
50.0
40.0
-30.0
Kf 2,0 - 22 E-4 m/s
gravel/sand
lower terrace
, woicnsw Cold Stage
- (WO '"
micaceous sand
marine-littoral-fluvial
upper Oligocene
Upper Middle Qllgocone
Figure 1. Geological section (E-W) at the experimental site in the Bitterfeld area
Given this background, Bitterfeld appears especially appropriate as the SAFIRA model location for
developing methods and technologies for minimally invasive "passive" decontamination techniques for
complexly polluted groundwater within a genuine scenario, as well as demonstrating their suitability in the
field. The results of the feasibility study for the experimental site are compiled in [6]. Figure 1 contains a
geological section of the area under investigation.
Pollution
Groundwater
Contamination of the groundwater with inorganic pollutants (e.g., heavy metals, arsenic, etc.) has proved
minor. The only noteworthy feature is the high levels of sulphate (up to 1,000 mg/1) and chloride (1,300
mg/1).
In order to characterize the organic pollution of the groundwater at the various water levels, several samples
were examined from different levels. The "main components" of organic contamination were found to be
85
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
chlorobenzene, 1,2-dichlorobenzene, 1,4-dichlorobenzene, benzene, trichloroethene, cis-l,2-dichloroethene
and trans-1,2-dichloroethene. Mass-spectrometry screening also detected other substances, albeit in very low
concentrations.
Although high organic pollution with halogenated hydrocarbons was confirmed in both the upper and lower
aquifers, contamination considerably varied between the two in terms of both quality and quantity. The lower
aquifer is dominated by aliphatic chlorinated hydrocarbons (trans-1,2-DCE, cis-l,2-DCE and TCE), while
the main component of the upper aquifer is chlorobenzene. During depth-oriented sampling, contamination
was found to display significant stratification: whereas the sample from a depth of 7-8 m only contains very
low amounts of pollutants, samples from depths between 16 and 20 m were above all found to contain high
levels of chlorobenzene (8-51 mg/1), as well as substantial concentrations of dichlorobenzenes (up to 1 mg/1).
Solid samples from the cores
In addition to the water samples, core material was also examined. The core samples from bore Saf Bit 2/96
were subjected to headspace-screening over the entire profile between 0 and 48.5 m. The relative sensitivi-
ties, which are proportional to the concentration of the substances in the sample, are summarized in Figure
2, showing clearly the influence of the lignite seam on the contaminant distribution.
One striking aspect is that the aliphatic halogen hydrocarbons and benzene chiefly occur in the strata
between 19.5 m and 24.5 m, while chlorobenzene and the dichlorobenzenes are found in higher concen-
trations in the strata between 12 m and 22.65 m. The main components are trans-1,2-dichloroethene, cis-1,2-
dichloroethene, chlorobenzene, trichloroethene, and benzene. Bromobenzenes and perchloroethene were also
identified by mass spectrometry.
Investigations into the adsorption and desorption of the pollutants in coal samples revealed that the sorption
capacity of the coal has not yet been exhausted. Hence the coal can act as both a pollutant sink (high
concentrations in the groundwater) and a pollutant source (lower concentrations of the pollutants in the
groundwater).
Laboratory Experiments on Pollutant Decomposition
Bioremediation
The basic requirements for deciding whether microbiological in situ remediation techniques can in principle
be used, selecting the most suitable technique and assessing the prospects of success are as follows:
• Determining the microbial density.
• Determining the pollutant breakdown capacity of the autochthonous microbial biocenoses under
in situ conditions.
• Examining ways of boosting the breakdown capacities of autochthonous microbial biocenoses
in situ using technically and financially feasible measures.
• Determining the microbial density.
• Determining the pollutant breakdown capacity of the autochthonous microbial biocenoses under
in situ conditions.
86
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
relative abundance
0 SOOOO 0 100000 0
t-Dichloro-
ethene
c-Dichloro-
ethene
Trichloro-
ethene
Benzene Chloro- 1,4-Dichloro-
benzene benzene
Figure 2. Contaminant profile, core 2/96
Examining ways of boosting the breakdown capacities of autochthonous microbial biocenoses
in situ using technically and financially feasible measures.
In addition to the pollutant level, the aquifers at the location are characterized as microbial habitats by certain
combinations of microbe-ecologically relevant abiotic milieu factors. The quaternary aquifer has a lack of
oxygen. Sulphate and nitrate are present as potential electron acceptors for anaerobic processes. The pH
levels are around 7 and the temperatures about 13°C, only varying slightly along the depth profile.
The groundwater and aquifers are inhabited by bacteria down to the maximum drilling depth (50.5 m below
the surface). Nitrate, iron, and manganese reducing bacteria predominate in the autochthonous biocetioses.
87
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
As expected, a lack of yeasts and mycelium-forming fungi was found in most habitats; a tiny abundance
(^xlO2 g"1) was only detected in the coal-bearing strata, individual horizons of the quaternary aquifers (6,
16-17 and 19-21 m) and the tertiary aquifer at 48 m.
The pollutants found to be the main contaminants in the quaternary aquifer were selected for investigations
into microbiological breakdown. These were monochloro-benzene (MCB, 30 mg/1), 1,2-dichlorobenzene
(1,2-DCB, 10 mg/1) and 1,4-dichlorobenzene (1,4-DCB, 10 mg/1). Benzene (100 mg/1) was also included
in the investigation as the expected product of the chemically catalyzed dehalogenation of chlorobenzenes.
The three primary pollutants MCB, 1,2- and 1,4-DCB (as well as benzene) were removed from the auto-
chthonous bacteriocenoses under simulated in situ conditions within technologically relevant periods of
times. However, the extent and speed of removal significantly differed depending on the pollutants'
chemical structure, the metabolism type of the ecophysiological groups involved, and the availability of
electron acceptors and additional carbon substrates (Figure 3).
Of the three chloroaromatics, 1,2-DCB easily proved to be the hardest compound to break down; in fact
under anaerobic conditions no decomposition whatsoever was established without the availability of
additional carbon substrates. After adding non-resident carbon substrates (a mixture of acetate and lactate),
the 1,2-DCB was however almost completely removed after just 20 days under nitrate, sulphate and iron
reducing conditions. When using these approaches, the other pollutants were no longer detectable either by
this time.
Benzene
MCB
1,2-DCB
1,4-DCB
10 20 30 40 0
time(d]
10 20 30 40
tima[d]
10 20 30 40 0
tim»[d]
10 20 30 40
b'mo[d]
10 20 30 40 0
thm[d}
10 20 30 40 0
ttme[dj
10 20 30 4
tlme[d]
10 20 30 40 0
timo[d]
10 20 30 40 0
Vm»[d]
10 20 30 40
ttm»[dj
Figure 3. Decomposition of chlorobenzenes and benzene
by autochthonous bacteria groups in the boundary layer
of the quaternary aquifer and coal (19-23 m beneath the
surface)
Figure 3. a) Without the addition of substrates (A
aerobic; B-E anaerobic; B without additional
electron acceptors; C nitrate; D sulfate; E
hydrated iron oxide)
b) With the addition of substrates (lactate,
acetate, yeast extract, ammonium, and
phosphate); A-E
Under aerobic conditions, 1,2-DCB was in all experiments only broken down very slowly (if at all). The
residual concentrations found after an incubation time of 40 days were usually 40 - 50 %, and in just one
case about 30% of the initial concentration - thus on the whole they were only slightly below the residual
concentration of 50 % defined as the evaluation limit for biological removal.
88
-------
Treatment Waffs
NATO/CCMS Pilot Study Phase ffl
Assuming sufficient quantities of terminal electron acceptors and carbon substrates were available, the
pollutants, were, somewhat surprisingly, broken down relatively quickly in the anaerobic environment under
nitrate-, sulphate-, and iron-reducing conditions. ,•..,,.-.".-
The findings of the screening programme indicate that in principle the microbiological in situ remediation
of the contaminated aquifers under the anaerobic conditions prevailing at the location is possible. Further-
more, the in situ removal capacities can be increased by stimulating autochthonous anaerobic groups via the
addition of electron acceptors and usable carbon substrates,
Abiotic pollutant removal
Physical and chemical methods for pollutant reduction in the groundwater are grouped together within the
SAFIRA Project under the term abiotic techniques. Starting from the familiar process of dehalogenating
aliphatic chlorinated hydrocarbons with Fe° walls, new methods were sought that above all enable the
removal of chloroaromatics and are suitable as in situ techniques. Initial approaches are seen in electro-
chemical techniques, sonochemical methods and bimetal systems. In addition, the usage of activated carbon
adsorption was tested, which could in particular be interesting as an in situ method if the residence times
were extended by microbial inhabitation. The goal of the preliminary study was to assess the possibilities
of such methods and to find new approaches. In the following, the findings of electrochemical removal are
presented by way of example; it should be pointed out that other approaches also yielded very promising
results.
A distinction was made between two different techniques for electrochemical dechlorination: systems using
palladium as a catalyst for hydrogen activation and those without palladium. The former represent the
current state of research, whereas the latter have not previously been described for chlorobenzene
degradation.
Pd was either precipitated electrochemically on the carbon cathode or mixed with the cathode material as
a carrier catalyst (e.g., 5% Pd on activated carbon). The main difference between the two methods is that
in the first case the Pd is on the electrode potential, whereas in the second it is not conductively linked to
the electrode.
In both methods, water electrolysis produces molecular hydrogen, which is activated at the surface of the
Pd. The concentration of chlorobenzene in the electrolytes (Bitterfeld groundwater) in these experiments
was 50-150 ppm. In many cases, chlorobenzene was completely removed. Figure 4 shows a typical course
of concentration over time for a mixture of mono-, di-, and trichlorobenzenes. The constancy of the chloro-
benzene concentrations was measured up to 450 hours in order to record sorption and other losses. After
starting water electrolysis, the chlorobenzenes were mostly removed within about 50 hours. The regular
arrows indicate the renewed addition of the chlorobenzene mixture; the jagged arrows mark the beginning
of electrolysis. This sequence can be repeated several times without a drop in activity. The only reaction
products found were benzene and chloride (>90% of the converted chlorobenzenes). Similar results were
obtained when Pd on different carriers (activated carbon, soot) was freely suspended as a powder in the
groundwater.
Scaling Up '
By now, the main site-specific investigations have been completed. The location has been found to be
suitable and the fundamental mechanisms for removing the "pollutant cocktail" are known. The investiga-
tions still underway are mainly designed to refine the data basis for hydraulic and matter transport modeling,
and to test and optimize additional removal techniques. Moreover, the technical principles for structural
planning are being worked out. ,
89
-------
Treatment Walls
NATO/CCMS Pilot Study Phase HI
The next stage is to scale up the experiments carried out so far in the laboratory. For this purpose a mobile
decontamination unit has been designed as a "window in the aquifer". Groundwater from a depth of about
20 m is pumped into a storage tank without coming into contact with oxygen from the air. This polluted
water will then be used to charge five possible test columns with the physico-chemical conditions of the
aquifer being preserved (T = 13°C, pH = 7, eH = -250 mV, x = 2.5 mS/cm, AOX • 100 mg/1).
3000000 I
2000000-
1000000
Chlorobenzene
Dichlorobenzene
Trichlorobenzene
400 500
time [h]
600
700
800
Equilibration time (voltage off)
Electrolysis (5V DC)
Addition of stock solution
Figure 4. Reductive dechlorination of mono-, di- and trichlorobenzenes (c0 = 100 ppm MCB, 30 ppm DCB and
10 ppm TCB) on Pd (5 wt-% on 15 mg activated carbon) with hydrogen produced by electrolysis (cathode:
150 mg electrographlte, U = 5 V, I = 2 mA, electrolyte: 150 ml groundwater).
The methods tested successfully in the laboratory and in the mobile decontamination unit have to prove their
chemical and hydrological long term stability and will be optimized in a pilot plant. Several vertical,
experimental columns (up to 3 m in diameter) will be installed to a depth of 20 m directly into the aquifer.
Numerous sampling and process controlling facilities and a variable design of the reaction columns will
enable the analyses of relevant chemical and hydraulic processes during operation and a competitive
technology development under real-world conditions. Operation of the pilot plant will start by the end of this
year. Subsequently an extension of the pilot plant for a horizontal groundwater flow is intended.
References
[1] Teutsch, G, Grathwohl, P., Schad, H. and Werner, P. (1996): In wta-Reaktionswande - ein neuer Ansatz
zurpassiven Sanierung von Boden- und Grundwasserverunreinigungen.- Grundwasser 1/96,12-20.
[2] Peter, H., GroBmann, J. and Schulz-Terfloth, G. (1995): Rahmensanierungskonzept des GroBprojektes
"Bitterfeld/Wolfen".- Li: LUHR, H.-P. (Hrsg.): Grund-wassersanierung 1995.- IWS Schriftenreihe, 23:
123-138; Berlin, Erich Schmidt-Verlag
[3] Schulz-Terfloth, G. and Walkow, F. (1996): MaBnahmen zur Sanierung des Grundwassers unter
Beriicksichtigung der wasserwirtschaftlichen und bergbaulichen Situation im GroBprojekt Bitterfeld-
Wolfen.- In: LOHR, H.-P. (Hrsg.): Grund-wassersanierung 1996.- IWS Schriftenreihe, 27: 307-320;
Berlin, Erich Schmidt-Verlag
[4] Landratsamt Bitterfeld / GFE GmbH: Grundwassermonitoring 1992-1996.
90
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
[5] GroBmann, J. and Liihr, H.-P. (1994): Sanierungsrahmenkonzept fur GroBprojekt Bitterfeld-Wolfen.
[6] WeiB, H., Teutsch, G. and Daus, B. (Hrsg.) (1998): Sanierungsforschung in regional kontaminierten
Aquiferen (SAFIRA) - Bericht zur Machbarkeitstudie fur den Modellstandort Bitterfeld - UFZ-Bericht
27/1997, ISSN 0948 94252; Januar 1998; Leipzig
Discussion
Kahraman Unlii asked whether the removal of iron chloride is a concern. Weiss noted that lowering chloride
concentrations is not feasible due to the presence of underground brine.
Weiss was asked if he considered using iron to produce hydrogen because it is cheaper than producing it
electrolytically. Weiss indicated the oxygen that is produced in the electrolytical reaction can be used for
aerobic reactions.
One of Weiss's photographs showed gas bubbles being produced at the water surface during the reaction,
which were identified as methane and nitrogen.
91
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Summary and Conclusions
Harald Burmeier
Construction of Treatment Walls
The costs of pilot studies and field studies should be considered when computing the costs of implementing
a technology. The costs presented for the funnel-and-gate technology only reflect the materials (iron) and
construction. Field demonstrations are not conducted very frequently anymore; in most cases, laboratory
studies suffice. At approximately 90 percent of the North American sites for which treatment walls, TCE
or PCB is the primary contaminant; with few exceptions, the half-lives fall within a narrow range (30
minutes to 1 hour). Therefore, it becomes difficult to justify field tests as cost-effective.
Eberhard Beitinger recently performed field tests with two columns of activated carbon at a site with several
types of contaminants. The columns were installed on site to model actual conditions. The results of the
column tests were very encouraging; the performance time of the wall would be higher than had been
expected. The minimum time designed for the wall was 30 years for the first removal of reactive materials.
A 50-m continuous wall will be installed with impermeable "wings."
Although residence times to degrade chlorinated solvents are generally known, geological and hydro-
geological characterization is needed for any treatment approach. Furthermore, information on the plume's
location and where it is moving is also necessary. If this information is well understood, design costs can
be lowered.
Harry Whittaker doubted that chromium can be removed permanently with an iron treatment wall because
iron becomes very finely divided ferric oxide in the environment, especially in wet soils. He suggested that
there is a continuous long-term cycle of Fe2+, Fe3+, Fe°, and Cr3* and Cr6+. Once all of the iron is converted
to Fe3*-, the chromium that was retained by the treatment wall will be released. Wolfgang Wiist added that
he has seen this occur in column tests.
Bob Puls said that the state environmental agency in North Carolina had to be convinced that chromate
would not be released eventually from the permeable reactive barrier (PRB) at Elizabeth City. Originally,
natural attenuation of the chromate was examined. There was a natural reduction of chromate observed
(primarily due to the high organic carbon content of the surface soils), and the chromate was immobilized
in a chromium and iron hydroxide phase. Researchers had to prove to the regulatory authority that the iron
would not reoxidize the Cr3* to Cr6*. The only oxidant in natural soil strong enough to oxidize the chromium
is manganese oxide. As a result, they analyzed the manganese oxide content of the soil and groundwater to
prove that whatever immobilization method was occurring would not reverse itself. The PRB was not
considered a permanent remedy for the chromate plume, and they are looking into eliminating the chromate
source.
Harald Burmeier said that long-term aspects of installing treatment walls must be considered during the
feasibility study. For example, the construction of the wall should be approached so that the reactive media
can be removed and replaced as necessary. Although it may be relatively inexpensive to construct a
treatment wall, the need to remove it 30 years in the future must be considered in cost estimates.
Paul Bardos asked if there is any true distinction between permeable reactive barriers and treatment walls.
Burmeier explained that permeable reactive barriers are a type of treatment wall that uses a material that is
permeable and reactive to treat the groundwater. Treatment walls can also be partially permeable and treat
92
-------
Treatment Walls
NATO/CCMS Pilot Study Phase 111
groundwater using adsorptive, biological, or electrochemical processes. In addition to treatment wall
constructions, treatment barriers can be constructed as piles, drainage trenches, and funnel and gate systems.
There was additional discussion on the issue of patents for treatment walls. To the extent that EnviroMetal
Technologies has a worldwide patent on the emplacement of iron in the ground for remediation of
chlorinated organic solvents, there may be cases where researchers are not adhering to the patent restrictions.
Hermann Schad indicated that the funnel-and-gate patent applies to situations where groundwater flow is
diverted toward the treatment wall. This design is protected in Germany and several other countries in the
European Union. Schad noted that funnel-and-gate systems are advantageous when intending to use a
material to react with or absorb a contaminant. The continuous walls, however, have hydraulic advantages.
In response to a question regarding treatment walls that do not work as designed, Burmeier acknowledged
that the examples presented in the session emphasized successes rather than failures. However, he
emphasized that reactive materials must be tested prior to emplacement. He added that when building a wall
over 10 to 20 m deep, technologies must be matched to the permeability of the subsurface as well. Gillham
was aware of at least three cases in the United States where regulators required a contingency for the
installation of a reactive barrier. He added that in any case, an alternative treatment method is almost always
proposed along with the reactive barrier.
Walt Kovalick emphasized that the cost savings of using PRBs instead of pump-and-treat systems is signifi-
cant. He suggested that 10 to 20 percent of the pump-and-treat systems currently active in the United States
could benefit from the installation of PRBs. The issue of contingencies was insignificant from a cost
standpoint because the cost of a typical wall (US$400 thousand) is significantly less than a pump-and-treat
system (US$12-14 million). Therefore, the cost of research and development is also justified.
Reactive Materials, Full-Scale Projects, and General Conclusions
Harald Burmeier summarized the topics discussed and the questions on treatment walls that still remain to
be answered. He pointed out that the Special Session presentations have examined the reaction, absorption,
and degradation processes of treatment walls, but noted that combinations of processes are also being
considered.
Several issues need closer examination, such as how to estimate and influence the long-term performance
of treatment wall; determination of by-products of organic contaminant degradation; the cost of materials
(e.g., the range in costs and corresponding effectiveness of zero-valent irons); calculation of mass balances;
the determination of what reactive substances can be placed in the groundwater without risk; and how to
influence biological processes in combination with treatment walls.
The presentations showed that zero-valent iron can be used to degrade chlorinated solvents and adsorb
uranium, but the nature and toxicity of the adsorbed uranium still must be determined. Clearer explanations
are needed from scientists on how to evaluate degradation and sorption processes. Optimization of reactive
materials using bench-scale tests have frequently been replaced by laboratory tests followed by trial-and-
error field testing. Further research on bioscreens is needed to determine long-term effects on remediation
as well as possible limiting factors such as bioclogging, redox limitations, and concentrations or
combinations of contaminants. There are questions beyond technological solution, that cross into regulatory,
legal, and public policy, such as what concentration of a contaminant is harmless? Can reactive materials
be emplaced in the groundwater, and at what concentrations? What is the long-term implication of in situ
chemical engineering, including any production of toxic by-products?
93
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
From the standpoint of constructing treatment walls, good definitions of terms such as "permeable reactive
barriers" and "treatment walls" must be developed. Although it is now possible to excavate and install deep
walls, the cost effectiveness of such constructions should be evaluated in order to access the market.
In conclusion, treatment wall technology has been proven to work; however, there are some limitations to
its use. Several questions still must be answered before their use can become widespread.
94
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
ABOUT THE AUTHORS
Harald Burmeier is a civil engineer with 20 years of experience in waste and hazardous waste management,
contaminated site investigations, feasibility studies, and health and safety issues. A professor of engineering
on the faculty of Applied Studies and Research of the University of Northeast Lower Saxony, Prof. Burmeier
maintains an active consulting practice throughout Germany on project management for waste and hazardous
waste remediation, issues related to worker health and safety, and emissions control. He is the author of a
number of books and articles on hazardous waste investigation, management, and remediation. Prior to
assuming his present position last year, Prof. Burmeier held a series of senior technical and management
positions in the German construction industry and trade associations. Prof. Burmeier holds a M.S. in Civil
Engineering from the Technical University of Hannover, Germany.
Gerard Evers is a senior engineer and head of the Department of Technique and Development at Soletanche
Bachy (Nanterre, France), the world's largest specialized contractor in foundation techniques and
underground construction. Mr. Evers holds a diploma in civil engineering from the Technical High School
of The Netherlands.
Robert Gillham has been a professor in the Department of Earth Sciences at the University of Waterloo
since the early 1970s. He was Director of the Waterloo Centre for Groundwater Research from 1987 to 1992,
was Chair of the Earth Sciences Department from 1993 to 1997, and currently holds the NSERC/Motorola/
ETI Industrial Research Chair in Groundwater Remediation. In 1997 Dr. Gillham was elected a Fellow of
the Royal Society of Canada. Dr. Gillham's primary research areas concern contaminant transport in
groundwater systems and groundwater remediation.
Liyuan Liang received her B.S. in civil engineering from Northeastern University and her M.S. and Ph.D.
in Environmental Engineering from the California Institute of Technology. Dr. Liang is Senior Lecturer in
the Department of Earth Sciences, University of Wales (UK), where she teaches aquatic geochemistry and
environmental remediation technologies, conducts research on physico-chemical processes, and directs the
Virtual Institute for Environmental Research. Prior to her current appointment, she was at Oak Ridge
National Laboratory (USA), where she managed a group of researchers investigating in situ phsico-chemical
treatment technologies for groundwater remediation and on the mechanistic understanding of geochemical
processes in the environment. She also conducted research on colloidal transport in groundwater, the kinetics
of reductive dechloriantion of organic solvents using zero-valent metals, and on the peroxidation of organic
contaminated soils. While at Oak Ridge, she was a member of the Steering Committee of the U.S. Permeable
Reactive Barriers Action Team. Her primary research interests include in situ treatment technologies for
mixed-wastes (organic solvents and radionuclides); laboratory and field research on particle dynamics in
aquatic environments (including coagulation, dispersion, and transport of colloidal particles); oxidative
precipitation and mineral dissolution; and fate of metals and organic contaminants in the environment. Dr.
Liang has published extensively in refereed journals.
Stephan Jefferis (M.A., M.Eng., M.Sc., Ph.D., C.Eng., M.I.C.E., C.Geo.l, F.G.S.) has qualifications in
natural sciences, chemical engineering, civil engineering, and law. He spent 20 years, with the University
of London (UK) where his research group developed a class of cement-bentonite cut-off materials that has
been used at many contaminated sites in Europe. Dr. Jefferis is presently an associate of Golder Associates,
where his role is world-wide and typically involves the investigation and resolution of unusual geotechnical,
geoenvironmental and materials problems often associated with aggressive chemical and microbiological
processes in the ground. He is also a visiting professor in the Department of Civil Engineering, Imperial
College, London.
95
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Robert W. Puls has a B.S. in Soil Science and Natural Resources from the University of Wisconsin, a
Masters degree in Forest Resources from the University of Washington, and a Ph.D. degree in Soil and
Water Science from the University of Arizona. He has worked as a Research Soil Scientist at the R.S. Ken-
Environmental Research Center, of the National Risk Management Research Laboratory of U.S. EPA in
Ada, Oklahoma since 1987. Prior to that he worked for DOE in Richland, Washington on the High Level
Nuclear Waste Repository Research Program. He has more than 19 years experience working in the
environmental science field. Dr. Puls is the current Co-Chair of the Permeable Reactive Barrier Action
Team, one of seven Action Teams comprising EPA's Remedial Technology Development Forum (RTDF).
He has been actively involved in research related to permeable reactive barriers since 1991. His recent
publications have covered a range of topics including groundwater sampling, colloidal transport in
groundwater, organic-metal-mineral interactions, the development and evaluation of in situ remediation
techniques for soils and groundwater, and metal and metalloid sorption-desorption reactions governing
subsurface contaminant transport and transformation processes. He has authored more than 50 publications
on the above topics and given numerous presentations at national and international scientific meetings. Dr.
Puls has served on several scientific advisory committees for USEPA, the USGS, DOE and the National
Research Council.
Huub H. M. Rynaarts earned his M.Sc. degree in soil chemistry and soil microbiology and his Ph.D. from
the University of Wageningen Agricultural University (The Netherlands), where he studied bacterial
transport in porous media and biodegradation of hazardous organic compounds. Since 1994, Dr. Rijnaarts
has been a Senior Researcher with The Netherlands Organization for Applied Scientific Research (TNO)
Institute of Environmental Sciences, Energy Research and Process Innovation, where he has been involved
in developing new in situ soil and groundwater bioremediation techniques, including biological barriers, and
natural attenuation.
Hermann Schad holds an M.S. and Ph.D. in geology from the University of Oregon (USA) and the
Unviersity of Tubingen (Germany), respectively. His Ph.D. thesis was on the "Variabliity of Hydraulic
Parameters in Non-Uniform Porous Media: Experiments and Stochastic Modelling at Different Scales." Dr.
Schad presently is the Waste Management Director of I.M.E.S. (Innovative Mess-, Erkundunes- und
Sanierungstechnologien GmbH) in Wangen, Germany.
Robert L. Siegrist earned his B.S. and M.S. in Civil Engineering and his Ph.D. in Environmental
Engineering at the University of Wisconsin. During 20 years of experience, he has held research and
teaching appointments with the Colorado School of Mines, Oak Ridge National Laboratory (ORNL), the
University of Wisconsin, and the Agricultural University of Norway. Since January 1995, Dr. Siegrist has
been affiliated with Colorado School of Mines (USA) as a research associate professor in the Environmental
Science and Engineering Division while also holding an adjunct faculty participant position as a senior staff
member in the Environmental Sciences Division at ORNL. His research has focused on characterization,
assessment, and in situ remediation technologies for contaminated land. The technology related work has
included in situ treatment processes (chemical oxidation, redox reactive barriers, bioremediation) as well
as subsurface manipulation methods (vertical and horizontal recovery and recirculation wells, lance
permeation, deep soil mixing, hydraulic fracturing). In related work, Dr. Siegrist has continued research into
the hydrodynamic and purification processes impacting land treatment and disposal of wastes. Since 1990,
Dr. Siegrist has been leading a program of research concerning in situ chemical ozidation involving peroxide
and permanganate systems including fundamental studies of oxidation reaction chemistry and kinetics,
contaminant mass transfer, oxidant delivery systems, and field evaluations through pilot- and full-scale
technology demonstrations. His research has been sponsored by the U.S. Department of Energy, U.S. EPA,
the U.S. Department of Defense, the National Science Foundation, and private industry. He has published
his results in over 35 refereed articles and more than 100 conference proceedings and reports. He is an active
member of several national societies and currently servces on national committees, including the
Groundwater, Hazardous Waste, and McKee Medal Committees of the Water Environment Federation, he
96
-------
Treatment Walls
NATO/CCMS Pilot Study Phase UI
is a registered professional engineer and has served as an advisor and technical expert for state and federal
agencies in the United States, Canada, Norway, and Sweden. He is currently a Fellow with NATO/CCMS.
Timothy M. Vogel received his B.S. degrees in biological oceanography and geology and his M.S. in
environmental engineering from the University of Washington (USA), and his Ph.D. in environmental
engineering (microbiology) from Stanford University (USA). Following his graduate studies, Dr. Vogel
taught at Michigan State University and the University of Michigan. Between 1992 and 1997, he held
positions of research and management with Rhone-Pbulenc Industrialisation and, since 1998, he has been
the research and development project manager for Rhodia Eco Services in Meyzieu, France. His current
research activities include the acquisition, evaluation, and development of novel technologies for the
remediation of polluted soil and groundwater. Reactive barriers are among his most intense research
subjects, based on his previous scientific research involving the development .of reductive dechlorination
systems—such as vitamin B12—and on the biodegradatibn of petroleum hydrocarbons.
Holger Weiss is Head of the Department of Industrial and Mining Landscapes at the Center of
Environmental Research, Leipzig-Halle (Germany), where he coordinates interdisciplinary and international
research on contaminated and degraded landscapes, minimization of environmental impacts of mineral
exploitation and development, landscape reclamation, and ecosystem stability in industrial and mining areas.
He is conducting research on in situ technologies for complexly contaminanted aquifers (active and passive;
physico-chemical; microbiological), on the environmental significance and human and ecological affects
from industrial residues, and on regional emissions and effects of chore-organic contaminants. Dr. Weiss
holds a Ph.D. in geology from the Technical University of Clausthal, Lower Saxony (Germany), and he has
published extensively in the peer-reviewed literature in the United States and Europe.
Wolfgang Wust was born in Nordlingen (Bavaria, Germany). After high school he got a scholarship from
the state Bavaria and decided to study Gedokologie (Environmental Sciences) at the University of Bayreuth
arid Lancaster (UK) with the focus on environmental chemistry and modeling of pollutant transport and
degradation. In 1990 he received his diploma on the transport of heavy metals from the street into sewer
systems in the working group of Professor R. Herrmann. He continued his training in the field of
hydrogeology and groundwater hydraulics at the chair of groudwater hydraulics (Prof. Kobus) and was
involved in different research projects related to groundwater remediation in the VEGAS facility at the
University of Stuttgart. Since three years he is focussing on the chemical processes defining the scope and
the limitations of permeable reactive iron walls (working group of Professor Dahmke) and will soon finish
his Ph.D. on this topic.
97
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
NATIONAL CONTACTS
98
-------
Treatment Wafls
NATO/CCMS Pilot Study Phase iU
Directors
Stephen C. James (Co-Director)
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
26 Martin Luther King Drive
Cincinnati, Ohio 45268
United States
tel: 513-569-7877
fax: 513-569-7680
e-mail: james.steve@epamail.epa.gov
Walter W. Kovalick, Jr. (Co-Director)
Technology Innovation Office
U.S. Environmental Protection Agency
401 M Street, SW (5102G)
Washington, DC 20460
United States
tel: 703-603-9910
fax: 703-603-9135
e-mail: kovalick.walter@epamail.epa.gov
Co-Pilot Directors
Volker Franzius
Umweltbundesamt
Bismarckplatz 1
D-14193 Berlin
Germany
tel: 49/30-8903-2496
fax: 49/30-8903-2285 or -2103
H. Johan van Veen
The Netherlands Integrated Soil Research
Programme
P.O. Box 37
NL-6700 AA Wageningen
The Netherlands
tel: 31/317-484-170
fax: 31/317-485-051
e-mail: anneke.v.d.heuvel @ spbo.beng.wau.nl
Country Representatives
Nora Auer
Federal Ministry of Environment, Youth and
Family Affairs
Dept. m/3
Stubenbastei 5
A-1010 Vienna
Austria
tel: 43/1-515-22-3449
fax: 43/1-513-1679-1008
e-mail: Nora.Auer@bmu.gv.at
Jacqueline Miller
Brussels University
Avenue Jeanne 44
1050 Brussels
Belgium
tel: 32/2-650-3183
fax: 32/2-650-3189
e-mail: jmiller@resulb.ulb.ac.be
Harry Whittaker
Emergencies Engineering Division
Environment Canada
3439 River Road
Ottawa, Ontario, K1A OH3
Canada
tel: 613/991-1841
fax: 613/991-1673
e-mail: harry.whittaker@etc.ec.gc.ca
Jan Svoma
Aquatest a.s.
Geologicka 4
152 00 Prague 5
Czech Republic
tel: 420/2-581-83-80
fax: 420/2-581-77-58
e-mail: aquatest@aquatest.cz
99
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Inge-Marie SkovgSrd
Contaminated Land Division
Danish Environmental Protection Agency
29 Strandgade
DK-1401 Copenhagen K
Denmark
tel: 45/3-266-0100 - direct 45/32660397
fax: 45/3-296-1656
e-mail: ims@mst.dk
Ari SeppSnen
Ministry of Environment
P.O. Box 399
00121 Helsinki
Finland
tel: +358/9-199-197-15
fax: +358/9-199-196-30
Rene" Goubier
Polluted Sites Team
ADEME
B.P.406
49004 Angers Cedex 01
France
tel: 33/241-204-120
fax: 33/241-872-350
Antonios Kontopoulos*
National Technical University of Athens
GR-157 80 Zogrrfos
Athens
Greece
Pdl Varga
National Authority for the Environment
F6U.44
H-1011 Budapest
Hungary
tel: 36/1-457-3530
fax: 36/1-201-4282
e-mail: vargap@kik.ktm.hu
*Due to the death of Prof. Kontopoulos,
communications with the Greek delegation to the Pilot
Study may be directed to:
Manolis Papadopoulos, tel: +30-1-772 2219; fax:
+30-1-772 2218, e-mail papadop@metal.ntua.gr
Matthew Crowe
Environmental Management and Planning
Division
Environmental Protection Agency
P.O. Box 3000
Johnstown Castle Estate
County Wexford
Ireland
tel: +353 53 60600
fax: +353 53 60699
e-mail: m.crowe@epa.ie
Takeshi Nishio , ; .
Soil and Agricultural Chemicals Division
Environment Agency, Water Quality Bureau
Japan Environment Agency
1-2-2, Kasumigaseki, Chiyoda-Ku
Tokyo 100
Japan
tel:+81-3-3580-3173
fax: +81/3-3593-1438
e-mail: takeshi_nishio@ eanet.go.jp
Raymond Salter
Ministry for the Environment
84 Boullcott Street
P.O. Box 10362
Wellington
New Zealand
tel: 64/4-917-4000
fax: 64/4-917-7523
e-mail: rs@mfe.govt.nz
Bj0m Bj0rnstad
Norwegian Pollution Control Authority
P.O. Box 8100 Dep
N-0032 Oslo
Norway
tel: 47/22-257-3664
fax: 47/22-267-6706
e-mail: bjorn.bjornstad@sftospost.md.dep.
telemax.no
Ewa Marchwinska
Institute for Ecology of Industrial Areas
6 Kossutha Street
40-833 Katowice
Poland
tel: 48/32-1546-031
fax.: 48/32 -1541-717
e-mail: ietu@ietu.katowice.pl
100
-------
Treatment Waffs
NATO/CCMS Pilot Study Phase («
Marco Estrela
Institute de Soldadura e Qualidade
Centra de Tecnologias Ambientais
Estrada Nacional 249-Km 3-Leiao (Tagus Park)
Apartado 119 - 2781 Oeiras Codex
Portugal
tel:+351/1-422-8100
fax: +351/1-422-8129
e-mail: maestrela@isq.pt
Branko Druzina
Institute of Public Health
Trubarjeva 2-Post Box 260
6100 Ljubljana
Slovenia
tel: 386/61-313-276
fax: 386/61-323-955 --...-.
e-mail: branko.druzina@ivz.sigov.mail.si
Ingrid Hasselsten
Swedish Environmental Protection Agency
Blekholmsterrassen 36 •
S-106 48 Stockholm
Sweden
tel: 46/8-698-1179
fax: 46/8-698-1222
e-mail: inh@environ.se
Bernhard Hammer :
BUWAL
Federal Department of the Interior
3003 Bern
Switzerland
tel: 41/31-322-9307
fax: 41/31-382-1546
Resat Apak
Istanbul University
Avcilar Campus, Avcilar 34850
Istanbul
Turkey
tel: 90/212-5911-998
fax: 90/212-5911-997
e-mail: rapak@istanbul.edu.tr
Kahraman Unlti
Depratment of Environmental Engineering
Middle East Technical University
Inonu Bulvari
06531 Ankara
Turkey
tel: 90-312-210-1000
fax:90-312-210-1260
e-mail: kunlu@rorqual.cc.metu.edu.tr
Ian D. Martin
Environment Agency
Olton Court
10 Warwick Road
Olton, West Midlands
United Kingdom
tel: 44/121-711-2324
fax: 44/121-711-5830
e-mail: ianmartin@environment-agency.goy.uk
101
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
PARTICIPANTS
Phase m Pilot Study Meeting
Vienna, Austria
February 23-27,1998
102
-------
Treatment Walls
NATO/CCMS Pilot Study Phase jlj
Resat Apak
Istanbul University
Avcilar Campus, Avcilar 34850
Istanbul
Turkey
tel: 90/212-5911-998
fax: 90/212-5911-997
e-mail: rapak@istanbul.edu.tr
Nora Auer
Federal Ministry of Environment, Youth and
Family Affairs
Dept. ffl/3
Stubenbastei 5
A-1010 Vienna
Austria
tel: 43/1-515-22-3449
fax: 43/1-513-1679-1008
e-mail: Nora.Auer@bmu.gv.at
Erik Backhand
Eko Tec
Nasuddsvagen 10 - Box 34
932 21 Skelleftehamn
Sweden
tel: 46/910-333-66
fax: 46/910-333-75
Paul Bardos
R3 Environmental Technologies Ltd
P.O. Box 58
Ware- Hertfordshire SG12 9UJ
United Kingdom
tel: 44/1920-484-571
fax: 44/1920-485-607
e-mail: p-bardos@r3-bardos.demon.co.uk
N. Jay Bassin
Environmental Management Support, Inc.
8601 Georgia Avenue, Suite 500
Silver Spring, Maryland 20910
United States
tel: 301-589-5318
fax: 301-589-8487
e-mail: jbassin@emsus.com
Paul M. Beam
U.S. Department of Energy
19901 Germantown Road
Germantown, MD 20874-1290
United States
tel: 301-903-8133
fax: 301-903-3877
e-mail: paul.beam@em.doe.gov
Eberhard Beitinger
WCI Umwelttechnik GmbH
Sophie Charlotten - Str.33
14059 Berlin
Germany
tel: 49/30-3260-9481
fax: 49/30-321-9472
Bj0rn Bj0mstad
Norwegian Pollution Control Authority
P.O. Box 8100 Dep
N-0032Oslo
Norway
tel: 47/22-257-3664
fax: 47/22-267-6706
e-mail: bjorn.bjornstad@sftospost.md.dep.
telemax.no
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
Diane Dopkin
Environmental Management Support, Inc.
8601 Georgia Avenue, Suite 500
Silver Spring, Maryland 20910
United States
tel: 301-589-5318
fax: 301-5,89-8487
e-mail: ddopkin@emsus.com
103
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Branko Druzina
Institute of Public Health
Trubarjeva 2-Post Box 260
6100 Ljubljana
Slovenia
tel: 386/61-313-276
fax: 386/61-323-955
e-mail: branko.druzina@ivz.sigov.mail.si
Erol Ergag
Istanbul University
Dept. of Chemistry
Avcilar Campus, Avcilar 34850
Istanbul
Turkey
tel: 90/212-5911-998
fax: 90/212-5911-997
Marco Estrela
Institute de Soldadura e Qualidade
Centre de Tecnologias Ambientais
Estrada Nacional 249-Km 3-Leiao (Tagus Park)
Apartado 119 - 2781 Oeiras Codex
Portugal
tel: +351/1-422-8100
fax:+351/1-422-8129
e-mail: maestrela@isq.pt
Gerard Evers
Soletanche
6RueDeWattford
F92000 Nanterre
France
tel: 33/14-7764-262
fax: 33/14-9069-734
e-mail: gerard.evers @ soletanche-bachy.com
Volker Franzius
Umweltbundesamt
Bismarckplatz 1
D-14193 Berlin
Germany
tel: 49/30-8903-2496
fax: 49/30-8903-2285 or -2103
Inger Asp Fuglsang
Contaminated Land Division
Danish Environmental Protection Agency
29Strandgade
DK-1401 Copenhagen K
Denmark
tel: 45/32-66-01-00
fax: 45/32-66-04-79
e-mail: iaf@mst.dk
Robert Gillham
University of Waterloo
Department of Earth Sciences :
Waterloo, Ontario N2L3G1
Canada
tel: 519-888-4658
fax: 519-746-7484
e-mail: rwgillha® sciborg.uwaterloo.ca
Rene Goubier
Polluted Sites Team
ADEME
B.P.406
49004 Angers Cedex 01
France
tel: 33/241-204-120
fax: 33/241-872-350
Diana Halikia
National Technical University of Athens
GR-157 80 Zografos
Athens
Greece
tel: 30/1-722-2167
fax: 30/1-722-2168
e-mail: labmet@metal.ntua.gr
Bemhard Hammer
BUWAL
Federal Department of the Interior
3003 Bern
Switzerland
tel: 41/31-322-9307
fax: 41/31-382-1546
104
-------
Treatment Watts
NATO/CCMS Pilot Study Phase IU
Catherine Harvey
Environment Agency
Steel House
UTothill Street
London
United Kingdom
tel: 44/171-664-6793
fax: 44-171-664-6795
e-mail: diane.williamson@environment-
agency.gov.uk (office e-mail)
Ingrid Hasselsten
Swedish Environmental Protection Agency
Blekholmsterrassen 36
S-106 48 Stockholm
Sweden
tel: 46/8-698-1179
fax: 46/8-698-1222
e-mail: inh@environ.se
Christian Holzer
Department of Waste Treatment and Remediation
of Abandoned Sites
Federal Ministry of Environment, Youth, and
Family Affairs (Dept. m/3)
Stubenbastei 5
A-1010 Vienna
Austria
tel: 43/1-515 22-3429
fax: 43/1-513 16 79-1127
e-mail: christian.holzer@bmu.gv.at
Stephen C. James
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
26 Martin Luther King Drive
Cincinnati, Ohio 45268
United States
tel: 513-569-7877
fax: 513-569-7680
e-mail: james.steve@epamail.epa.gov
Stephan Jefferis
Golder Associates (UK) Ltd.
54-70 Moorbridge Road
Maidenhead, Berkshire
SL6 8BN England
United Kingdom
tel: 44/1628-771-731
fax: 44/1628-770-699
e-mail: sjefferis@golder.com
Harald Kasamas
CARACAS - European Union
Breitenfurterstr. 97
A-l 120 Vienna
Austria
tel: 43/1-804 93 192
fax: 43/1-804 93 194
e-mail: 101355.1520@compuserve.com
Vladimir Kinkor
SEPA s.r.o.
P.O. Box 47
Bezeck£79
169 00 Prague 6
Czech Republic
tel: 420/602-347-679
fax: 420/602-5721-1255
Antonios Kontopoulos
National Technical University of Athens
Athens
Greece
Walter W.Kovalick, Jr.
Technology Jjnnovation Office
U.S. Environmental Protection Agency
401 M Street, SW (5102G)
Washington, DC 20460
United States
tel: 703-603-9910
fax: 703-603-9135
e-mail: kovalick.waiter@ epamail.epa.gov
Tomas Lederer
Aquatest a.s.
Geologika 4
152 00 Prague
Czech Republic
tel: 420/2-581-8995
fax: 420/2-581-8175
e-mail: lederer@aquatest.cz
Liyuan Liang
Department of Earth Sciences
University of Wales, Cardiff
P.O. Box 914
Cardiff GF1 34E
United Kingdom
tel: 44/1-222-874-579
fax: 44/1-222-874-326
e-mail: liyuan@cardiff.ac.uk
105
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
Ian D. Martin
Environment Agency
Olton Court
10 Warwick Road
Olton, West Midlands
United Kingdom
tel: 44/121-711-2324
fax: 44/121-711-5830
e-mail: ianmartin @ environment-agency.gov.uk
Igor Marvan
Grace Dearborn Inc.
3451 Erindale Station Road
P.O. Box 3060, Station A
Mississauga, Ontario L5A 4B6
Canada
tel: 905-272-7435
fax:905-272-7456
Jacqueline Miller
Brussels University
Avenue Jeanne 44
1050 Brussels
Belgium
tel: 32/2-650-3183
fax: 32/2-650-3189
e-mail: jmiller@resulb.ulb.ac.be
Walter Mondt
Ecorem n.v.
Zwartzustersvest 22
B-2800 Mechelen
Belgium
tel: 32-15-21 17 35
fax: 32-15-21 65 98
e-mail: Ecorem@glo.be
Carlos de Miguel Perales
ICADE
Alberto Aguilera, 23
28015 Madrid
Spain
tel: 34/1-586-0455
fax: 34/1-586-0402
Robert Puls
U.S. Environmental Protection Agency
919 Kerr Research Drive
P.O. Box 1198
Ada, Oklahoma 74820
United States
tel: 580-436-8543
fax: 580-436-8703
e-mail: puls.robert@epamail.epa.gov
H.H.M. Rijnaarts
TNO/MEP
P.O. Box 342
7300 AH Apeldoorn
The Netherlands
tel: 31/55-5493-380
fax:, 31/55-5493-410
e-mail: h.h.m.rijnaarts@mep.tno.nl
Hermann Schad
IMES GmbH
Kocherhof 4
88239 Wangen
Germany
tel: 49/7528-971-30
fax: 49/7528-97131
e-mail: hermann.schad.imes@t-online.de
Mathias Schluep
BMG Engineering AG
ffangstrasse 11
8952 Schlieren
Switzerland
tel: 41/1-730-6622
fax: 41/1-730-6622
Christoph Schiith
Eberhard-Karls-Universitat Tubingen
Geologisches Institut
Sigwarstr. 10
72076 Tubingen
Germany
tel: 49/7071-29-75041
fax: 49/7071-5059
e-mail: christoph.schueth@uni-tuebingen.de
106
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
An Seppanen
Ministry of Environment
P.O. Box 399
00121 Helsinki
Finland
tel:+358/9-199-197-15
fax:+358/9-199-196-30
Robert Siegrist
Colorado School of Mines
Environmental Science and Engineering Division1
1500 Illinois Avenue
Golden, Colorado 80401-1887
United States
tel: 303-273-3490
fax:303-273-3413
email: rsiegris@mines.edu
Inge-Marie Skovgard
Contaminated Land Division
Danish Environmental Protection Agency
29Strandgade
DK-1401 Copenhagen K
Denmark
tel: 45/3-266-0100 - direct 45/32660397
fax: 45/3-296-1656
e-mail: ims@mst.dk
Michael Smith
68 Bridgewater Road
Berkhamsted, Herts, HP4 1JB
United Kingdom
tel: 44/1442-871-500
fax: 44/1442-870-152
e-mail: michael.a.smith@btintemet.com
Marek Stanzel
KAP s.r.o.
SkokanskaSO
169 00 Prague 6
Czech Republic
tel: 420/2-2431-3630
fax: 420/2-5721-1255
e-mail: kappraha@login.cz
Kai Steffens
PROBIOTEC GmbH
Schillingsstrabe 333
D 52355 Diiren-Gurzenich
Germany
tel: 49/2421-69090; >
fax: 49/2421-690961
e-mail: info@probiotec.ac-euregio.de
Rainer Stegmann
Technische Universitat Hamburg-Harburg
Harburger Schlobstrabe 37
D-21079 Hamburg
Germany
tel: 49/40-7718-3254
fax: 49/40-7718-2375
e-mail: stegmann@tuharburg.d400de
Jan Svoma
Aquatest a.s.
Geologicka4
152 00 Prague 5
Czech Republic
tel: 420/2-581-83-80
fax: 420/2-581-77-58
e-mail: aquatest@aquatest.cz
Gerhard Teutsch
Eberhard-Karls Universitat - Tubingen
Geologisches Institut
Sigwartstr. 10
72076 Tubingen
Germany
tel: 49/7071-29-76468
fax: 49/7071-5059
Kahraman Unlii
Depratment of Environmental Engineering
Middle East Technical University
Inonu Bulvari
06531 Ankara
Turkey
tel: 90-312-210-1000
fax: 90-312-210-1260
e-mail: kunlu@rorqual.cc.metu.edu.tr
107
-------
Treatment Walls
NATO/CCMS Pilot Study Phase III
H. Johan van Veen
The Netherlands Integrated Soil Research
Programme
P.O. Box 37
NL-6700 AA Wageningen
The Netherlands
tel: 31/317-484-170
fax: 31/317-485-051
e-mail: anneke.v.d.heuvel@spbo.beng.wau.nl
Pal Varga
National Authority for the Environment
F6u.44
H-1011 Budapest
Hungary
tel: 36/1-457-3530
fax: 36/1-201-4282
e-mail: vargap@kik.ktm.hu
Timothy Vogel
ATE/Rhodia Eco Services
17 rue Pe"rigord
69330 Meyzieu
France
tel: 33/4-7245-0425
Fax: 33/4-7804-2430
e-mail: timothy, vogel @rhone-poulenc.com
Holger Weiss
UF2 - Umweltforschungszentrum
Leipzig-Halle GmbH
Permoserstr. 15
04318 Leipzig
Germany
tel: 49/341-235-2060
fax: 49/341-235-2126
Harry Whittaker
Emergencies Engineering Division
Environment Canada
3439 River Road
Ottawa, Ontario, K1A OH3
Canada
tel: 613/991-1841
fax: 613/991-1673
e-mail: harry.whittaker@etc.ec.gc.ca
Wolfgang Wiist
Institut fur Wasserbau, Lehrstuhl fur Hydraulik
und Grundwasser
University of Stuttgart
Pfaffenwaldring 61
70550 Stuttgart
Germany
tel: 49/711-685-4714
fax: 49/711-685-7020
e-mail: ww@iws.uni-stuttgart.de
108
-------
|