vvEPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/625/R-99/012
August 2000
Abiotic In Situ
Technologies for Ground water
Remediation Conference
Proceedings
Dallas, TX
August 31 -September 2, 1999
-------
-------
EPA/625/R-99/012
August 2000
Abiotic In Situ Technologies for
Ground water Remediation Conference
Proceedings
Dallas, Texas
August 31 - September 2, 1999
Technology Transfer and Support Division
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Printed on Recycled Paper
-------
Notice
The preparation of this report has been funded wholly (or in part) by the U.S. Environmental Protection Agency
(U.S.EPA) under Contract No. 68-C7-0011, Work Assignment No. 1-58 issued to Science Applications International
Corporation (SAIC).The views expressed in these Proceedings are those of the individual authors and do not neces-
sarily reflect the views and policies of EPA. This document has been reviewed in accordance with EPA's peer and
administrative review policies and approved for publication. Mention of trade names or commercial products does not
constitute endorsement or recommendaton for use.
-------
Foreword
The U.S. Environmental Protection Agency (U.S. EPA) is charged by Congress with protecting the nation's land, air,
and water resources. Under a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural systems to support and
nurture life. To meet this mandate, EPA's research program is providing data and technical support for solving environ-
mental problems today and building a science knowledge base necessary to manage our ecological resources wisely,
understand how pollutants affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation of techno-
logical and management approaches for reducing risks from threats to human health and the environment. The focus of
the laboratory's research program is on methods for the prevention and control of pollution to air, land, water, and
subsurface resources; protection of water quality in public water systems; remediation of contaminated sites and ground-
water; and prevention and control of indoor air pollution. The goal of this research effort is to catalyze development and
implementation of innovative, cost-effective environmental technologies; develop scientific and engineering information
needed by EPA to support regulatory and policy decisions; and provide technical support and information transfer to
ensure effective implementation of environmental regulations and strategies.
This Proceedings document on Abiotic In-situTechnologies for Groundwater Remediation is a technical resource
guidance document for hydrogeological environmental engineering practitioners.
E. Timothy Oppelt
National Risk Management Research Laboratory
in
-------
Abstract
The U.S. Environmental Protection Agency (EPA) Conference on Abiotic In Situ Technologies for Groundwater
Remediation was held in Dallas, TX, August 31 -September 2,1999. The goal of the meeting was to disseminate current
informaton on abiotic in situ groundwater treatment technologies. Although much information is being provided about
biotic groundwater remediation approaches, less is known about abiotic processes. There is a recognized need to
communicate current design and performance data, technical feasibility, implementation impediments, and cost infor-
mation to regulators, decision makers, and other stakeholders who would benefit from this type of technology transfer.
The development of cost-effective in situ treatment technologies is a major priority for the EPA's Office of Research
and Development, the National Risk Management Research Laboratory, and the Technology Information Office of
EPA's Office of Solid Waste and Emergency Response. EPA is particularly interested in disseminating information to
potential users of these technologies to promote more-effective application to sites that have, historically, been difficult
to remediate.
The conference provided information on treatment technologies in the following areas:
Permeable Reactive Subsurface Barriers (Treatment Walls)
Thermal Enhancement Treatment
VOC Stipping Treatment
Chemical Oxidation Treatment
Geotechnical Methods and Treatment
Reactive Zones Treatment
Technical Practices and Design.Considerations
Economic Considerations
This document includes conference presentations on these topics, as well as case studies. Presenters are listed
on Page 109. These proceedings will be useful to environmental and regulatory personnel at the Federal, State, and
local level; university professors, researchers and students; and private-sector personnel including industry representa-
tives and environmental consultants.
IV
-------
Contents
Notice ii
Foreword iii
Abstract iv
Acknowledgements , , vii
Plenary Session
In Situ Abiotic Technologies for Groundwater Cleanup: Current Trends, Needs, and Expectations 1
Groundwater and In Situ Technologies Under Texas Risk Reduction Program Rule 3
Permeable Reactive Subsurface Barriers (Treatment Walls)
Overview of Abiotic Subsurface Remediation 7
Permeable Reactive Barriers for Remediation of Acid Mine Drainage 10
Long-Term Performance Monitoring of a PRB for Remediation of Chlorinated Solvents & Chromium 12
Long-Term Performance Monitoring of Permeable Reactive Barriers at DOE Sites 14
Demonstration of Permeable Reactive Barriers for Groundwater Remediation:
Status and Preliminary Results of the Fry Canyon Project 17
An Overview of Installation Methods for Permeable Reactive Barriers 18
Session A: Thermal Enhancement Treatment
Mechanisms in In Situ Thermal Remediation . 21
Remediation of Soils and Groundwater Using Steam Enhanced Extraction 22
Field Demonstrations of Thermally Enhanced Extraction 25
In Situ DNAPL Remediation Using Six-Phase Heating™ 26
RF Heating Technology for Soil Remediation 27
Thermal Conduction Heating for In Situ Thermal Desorption 30
Session A: Case Studies for In Situ Technologies
In Situ Destruction of Chlorinated Hydrocarbon Compounds in a Reactive Well Using Pd-Catalysts 31
Case Studies of In situ Chemical Oxidation using the Geo-Cleanse® Process 33
Session A: VOC Stripping Treatment
Interagency DNAPL Consortium 37
Field Pilot Test of In Situ Chemical Oxidation Through Recirculation Using Vertical Wells at the
Portsmouth Gaseous Diffusion Plant 42
In-Well Aeration/Stripping Technology: Overview and Performance Results 50
Air Sparging for Groundwater Remediation of Toluene and Other VOCs: Case Studies 54
Application of VFLUX to Assessment of Soil Venting Performance and Closure 55
In Situ Treatment of Hexavalent Chromium and VOCs Using Recirculating Wells 58
-------
Contents (continued)
Session B: Chemical Oxidation Treatment
An Overview of In Situ Chemical Oxidation Technology Features and Applications 61
Fenton Oxidation, Carbon Regeneration, and Groundwater Remediation 70
In Situ Chemical Oxidation for Groundwater Remediation Near a Waste Injection Well 71
Groundwater Remediation Using the ClenOX® In Situ Chemical Oxidation Process 74
Degradation-Desorption Relationships in the Treatment of Contaminated Soils Using
Modified Fenton's Reagent 78
Session B: Geotechnical Methods and Treatment
Containment Technology and Monitoring 80
Hydraulic Fracturing Overview and Issues 82
In Situ Chemical Treatment Using Hydraulic Fracturing To Emplace Fe° Metal and KMnO4 Reactive Solids.... 85
Session B: Reactive Zones Treatment
In Situ Treatment of Chromium Source Area Using Redox Manipulation 93
In Situ Redox Manipulation for Treatment of Chromate and Trichloroethylene in Groundwater 96
In Situ Remediation of Chromium Contamination of Soil and Groundwater 98
In Situ Reduction of Hexavalent Chromium in Groundwater and Surface Soil Using Acidified
Ferrous Sulfate 99
In Situ Remediation of Hexavalent Chromium in Groundwater: Practical Implementation 101
Technical Practices and Design Considerations
Design of Cosolvent Flooding Solutions for NAPL Remediation 104
In Situ Flushing for Enhanced NAPL Site Remediation: Metrics for Performance Assessment 105
Case Studies and Economic Considerations
Application of Pervaporation for the Removal of VOCs and Recovery of IRA from Surfactant-Based
Soil Remediation Fluids 106
Coupling In Situ Flushing with Other Remediation Technologies 108
Speaker List 109
VI
-------
Acknowledgments
The success of the conference and this document is due largely to the efforts of many individuals. Gratitude goes
to each person who was involved.
Presenters and Exhibitors
A special thanks goes to the authors of the papers and demonstrations that were presented at the conference.
Their efforts in preparing papers made this document possible and led to the overall success of the conference. The
participation of the exhibitors whose displays added to the conference is also appreciated.
Special Thanks
The contributions of the following individuals in the development of the conference are especially appreciated.
Richard Steimle, EPA, Office of Solid Waste and Emergency Response, Washington, D.C.
Steve Schmelling, Ph.D., EPA, Office of Research and Development, Ada, OK.
Jerry Jones, EPA, Office of Research and Development, Ada, OK.
Robert Puls, Ph.D., Office of Research and Development, Ada, OK.
Vincent Malott, Region 6, Dallas, TX.
Kathleen Yager, EPA, Office of Solid Waste and Emergency Response, Washington, D.C.
Technical Direction and Coordination
Douglas Grosse, EPA, ORD, NRMRL, Cincinnati, OH, provided technical direction throughout the development of
the conference and the preparation of the conference proceedings. Science Applications International Corporation
(SAIC) of Reston, VA, handled conference logistics and provided support for many aspects of the conference.
Editorial Review and Document Production
Thomasine Bayless and Jean Dye of EPA's Office of Research and Development, Cincinnati, OH, guided the
compilation and editing of this publication. John McCready provided graphics support.
VII
-------
-------
In Situ Abiotic Technologies for Groundwater Cleanup:
Current Trends, Needs, and Expectations
Walter W.Kovalick, Jr.
Technology Innovation Office
Office of Solid Waste and Emergency Response
USEPA
Washington, DC 20460
703-603-9910
kovalick. waiter® epa.gov
Introduction
The scope of the hazardous waste remediation challenge
in the United States is large and diverse in nature. In addi-
tion to Superfund sites, corrective action under the Re-
source Conservation and Recovery Act and leaking un-
derground storage tanks, and clean ups by the Depart-
ments of Defense and Energy, emerging state voluntary
clean up programs and local Brownfields redevelopment
projects offer many challenges. Groundwater contamina-
tion is a major issue at many sites; it is further compli-
cated by the presence of nonaqueous phase liquids
(NAPLs). Achieving effective and less costly cleanups re-
quires the development and commercialization of improved
remediation technologies. In situ technologies, in particu-
lar, are in great demand because they are usually less
expensive and more acceptable than above-ground op-
tions. This presentation will summarize important
"benchmarking" information for the developers of new
groundwater technologies as well as some of the expec-
tations of the user community.
Technology Benchmarking
The USEPA, other Federal agencies, universities, and the
private sector have made considerable progress in en-
couraging the development and use of in situ technolo-
gies for both soil and groundwater cleanup. Recently, a
growing body of information is becoming available about
the early application of these new technologies as well as
the cost and performance of more established approaches.
More than ever it is crucial that new technology research-
ers be cognizant of the competitive advantages and dis-
advantages of their new approaches and their relative
costs in order to assure the relevance of their work.
Information on in situ and other innovative cleanup tech-
nologies is available in published reports and electronic
databases. Some of the more important information re-
sources are listed below.
The Technology Innovation Office's (TIO's) Clean Up In-
formation (www.clu-in.org or www.epa.gov/tio) web site
contains over 300 EPA and non-EPA documents directed
at the remediation of soil and groundwater. The listing of
other related EPA and non-EPA web sites leads to numer-
ous other sources of such information.
Also sponsored by TIO is EPA REACHIT (REmediation
and Characterization Innovative Technologies), the on-line
directory of 370 remediation and 160 site characteriza-
tion technologies along with current deployment informa-
tion from over 900 Superfund sites. This web-based data-
base, which can searched by media, contaminant, and
technology, is located at http://www.epareachit.org.
• EPA has compiled summary information of over 300
field demonstrations of innovative cleanup technolo-
gies in North America. Of these 103 are in situ soil
projects, and 36 involve in situ groundwater technolo-
gies.The contacts listed in the report can provide more
detailed project data. It is available for viewing or down-
loading at http://www.clu-in.org, and is being updated
in 1999. Notable for its major contribution to this body
of demonstrations are the over 100 projects for
remediation technologies under the EPA SITE
(Superfund Innovative Technology Evaluation) pro-
gram; program-specific information is found at
www.epa.gov/ORD/SITE.
• Under the auspices of the Federal Remediation Tech-
nologies Roundtable, U.S. federal agencies have pre-
pared a series of 140 detailed remediation case stud-
ies, which include some in situ technologies (e.g.
bioremediation, air sparging, pump and treat, perme-
able walls). Fewer case studies have been completed
on in situ groundwater technologies because they have
only recently been implemented, and they typically
take several years or more to complete. Accessible at
http://www.frtr.gov, this web site will include over 200
1
-------
case studies by the year 2000. In addition, the
Roundtable has prepared a guide for gathering cost
and performance data that includes several of the
newer abiotic technologies. Researchers and others
can expect that the parameters outlined in this Guide
will be the minimum data sets expected by many Fed-
eral users.
• In cooperation with EPA, the Ground-Water
Remediation Technologies Analyses Center
(GWRTAC) serves as the focal point for information
transfer on groundwater treatment technologies.
GWRTAC maintains several databases on the devel-
opment, demonstration, and vendors of groundwater
remediation technologies. The Center has also pre-
pared technical reports on 12 in situ methods for treat-
ing groundwater, including air sparging, bioslurping,
electrokinetics, horizontal wells, surfactant/cosolvents,
thermal enhancements, and treatment walls. Informa-
tion is available via the internet at http://
www.gwrtac.org.
Challenges for New Technology
Considerable sophistication exists among the consulting
and user community about these information resources,
so much will be demanded of new technology offerings in
terms of their cost and performance data. In addition, the
recognition that the problem owner is seeking a total "so-
lution", not a technology, is very important for technology
developers. Thus, as in the case of soild remediation so-
lutions with both organic and inorganic contaminants, tech-
nology developers will find it necessary to link their ideas
together with other technologies to offer the problem owner
a "complete" solution. Also, in the area of groundwater
cleanup, the invisible nature of the problem and the need
to measure the progress of the on-going solution means
that in situ process monitoring and process control is cru-
cial. Unlike above ground soil remediation technologies,
whose operations and results are more easily monitored,
in situ processes suffer from the lack of robust procedures
for problem identification, operational monitoring, and com-
pliance measurement.Thus, new solutions for in s/futreat-
ment must inevitably be linked with other measurement
technologies to guarantee their acceptance.
Beyond these requirements, there are several attributes
of previously successful technologies for soil and ground-
water remediation that are appropriate for review as de-
velopers pursuing new abiotic groundwater solutions. Sim-
plicity, ease of explanation to users and stakeholders, and
the "Hippocratic" principle of "doing no harm" are a few of
the characteristics that have marked "breakthrough"
remediation technologies. It is important to keep these
factors in the forefront as developers create new solutions
for the difficult world of DNAPL remediation.
-------
Groundwater and In Situ Technologies Under Texas Risk
Reduction Program Rule
Greg Tipple
Technical Support Section/Remediation Division
Texas Natural Resource Conservation Commission (TNRCC)
Austin, TX 78711-3087
512-239-2465; (fax) 512-239-2469
gtipple @tnrcc.state,tx.us
The purpose of this paper is to explain the groundwater
performance requirements of Remedy Standard A and B
of the Texas Natural Resource Conservation Commission's
(i.e, the Commission's) proposed Texas Risk Reduction
Program (i.e., TRRP) rule. The commission is scheduled
to meet in Austin on Thursday, September 2,1999, to con-
sider whether to adopt TRRP as a new chapter in the
commission's rules.TRRP will, if adopted, establish a rea-
sonable, consistent, risk-based, performance-oriented
process applicable to most waste program areas regu-
lated by the TNRCC with the goals of: appropriately bal-
ancing protection of human health and the environment
with the economic welfare of the citizens of Texas, and of
preserving and restoring the productive use of land. To
obtain a copy of the entire rule package, download it from
theTNRCC's web site at:
http://www.tnrcc.state.tx.us/waste/trrp.htm
When you download this rule package and start to print it
you should be aware that it is over 1000 pages long.
The primary topic of discussion in this paper is the ground-
water performance standards under TRRP for each of the
two remedy standards with particular emphasis placed
on in-situ technologies, such as monitored natural attenu-
ation (MNA). However, in order to provide some context to
this discussion of the groundwater performance standards,
it is desirable to summarize some general aspects of the
TRRP rule as a whole. All of the TNRCC's program areas
that will be covered under TRRP, if adopted, include: pe-
troleum storage tanks, municipal solid wastes, under-
ground injection control, voluntary cleanup programs, in-
dustrial and hazardous wastes, Superfund, spills, and
wastewater treament plants. The primary implementation
date is May 1, 2000; however, there is a delayed imple-
mentation date for the petroleum storage tank program of
September 1,2003. We started the TRRP project in 1995
when we decided that the two risk-based programs that
we had at that time were not consistent and were not work-
ing well together. After this we published two lengthy con-
ceptual documents and an ecological guidance document.
The agency also proposed and withdrew a draft rule in
1998 and then reproposed this draft rule on March 26 of
1999.
There is a grandfathering of affected properties under the
current Risk Reduction Rule. On May 1, 2000, persons
shall comply with the requirements of TRRP except as
modified by §350.2. Before May 1, 2000, persons may
use TRRP once the effective date of the rule has passed.
The effective date of the rule is 20 days after the rule is
filed with the Texas Register. One aspect of the
grandfathering provisions is that a person who has sub-
mitted an initial notification prior to May 1, 2000 to con-
duct a Risk Reduction Standard 1 or 2 response action
under the current Risk Reduction Rule and has submitted
a final report within five years after that date, may request
that the response action be reviewed under the Risk Re-
duction Rule. Applicants will automatically qualify for this
grandfathering if they have received a letter from the
agency acknowledging receipt of the initial notification or
if they submit other forms of documentation by May 1,
2001 demonstrating that proper and timely notification has
been made. Much more detail may be obtained by study-
ing §350.2(m) of TRRP.
Although the entire TRRP process is long and detailed,
this paper concentrates on the groundwater resource clas-
sification performed during the Affected Property Assess-
ment and on the General Requirements and performance
requirements for Remedy Standard A and B.
Determining the classification of the groundwater at an
affected property is a key step in the TRRP procedure.
The classification of groundwater at a site dictates whether
a pollution cleanup or an exposure prevention response
approach will be established for contaminated groundwa-
ter. The present TNRCC's regulations which classify
groundwater include the Risk Reduction Rule classifica-
-------
tion system which represents the approach traditionally
taken by the commission to classify groundwater. This
approach ascertains whether a groundwater-bearing unit
either is a current or potential groundwater supply. An-
other groundwater classification system used by the pe-
troleum storage tank program is something of an excep-
tion since it recognizes four different classifications of
groundwater with greater emphasis placed on the current
use of the resource and the present location of water wells.
The commission has decided to establish an additional
class of groundwater (i.e., class 1) due to increased reli-
ance on exposure prevention approaches in the rule and
the resulting concern that the most valuable groundwater
in the state should be afforded an extra measure of pro-
tection by requiring a pollution cleanup approach. The
commission has established a separate category for class
1 groundwater to ensure the protection of the most valu-
able groundwater in the state and to allow some flexibility
to take exposure prevention response actions in appro-
priate circumstances for class 2 groundwater. Adopted
criteria for class 2 and class 3 groundwater is essentially
equivalent to the definitions for "potential groundwater
supply" and "not a potential groundwater supply", respec-
tively.
Determining groundwater classification only applies to a
"groundwater-bearing unit" defined as a saturated geo-
logic formation, group of formations, or part of a formation
with a hydraulic conductivity equal to or greater than 1 x
10-s centimeters/second.This means that a saturated clay
shale with a dydraulic conductivity less than 1 x 10'5 cen-
timeters/second would not be considered a class 1, 2, or
3 groundwater resource and would not bu subject to the
groundwater response objectives. Next, we've been care-
ful to establish that the groundwater-bearing unit must be
capable of producing the target yield "at a sustainable rate"
(i.e., not a one time occurrence but on a continuing basis)
in order to meet a particular resource classification. More-
over, the yield standard for each class is expressed in
terms of a groundwater yield for a given size of well; how-
ever, a different size well could be used for the test proce-
dure as long as an equivalent yield is used to make the
classification determination.
Three key verbs are used within TRRP to define the sur-
face soil, subsurface soil, and groundwater response ob-
jectives that must be attained in order to meet the perfor-
mance requirements forthe remedy standards.To "remove"
means to take away from the affected property to another
location for storage, processing, or disposal. "Decontami-
nation" is a permanent and irreversible treatment process
which eliminates chemicals of concern (COCs) at con-
centrations above the critical protective concentration lev-
els (PCLs). And to "control" is to apply a physical control,
such as a cap or a slurry wall, or an institutional control,
such as a deed notice or restrictive covenant, which pre-
vents exposure of receptors to COCs.
TRRP is a performance-based rule. By focusing on rem-
edy performance, TRRP should be effective in minimizing
technical and regulatory barriers to the deployment of in-
novative technologies, including in-situ technologies, as
long as these technologies can achieve the required per-
formance. For example, to attain Remedy Standard A, the
person must, within a reasonable time frame given the
particular circumstances of an affected property, remove
and/or decontaminate the surface soil, subsurface soil,
and groundwater protective concentration level
exceedence (PCLE) zones, other environmental media,
hazardous and non-hazardous waste to achieve COC
concentration levels below the residential or commercial/
industrial critical PCLs, as appropriate. Remedial alterna-
tives, including the use of MNA as a decontamination rem-
edy, must be capable of achieving the Remedy Standard
A objectives within a reasonable time frame, given the
particular circumstances at the affected property, and must
be appropriate considering the hydrogeologic character-
istics of the affected property, COC characteristics and
the potential for unprotective exposure conditions to con-
tinue or during the remedial period. Note that MNA, as
well as any other in-situ technology, must qualify as a de-
contamination action, rather than a control measure, to
be used under Remedy Standard A. In other words, natu-
ral attenuation based upon biodegradation would be an
acceptable remedy under Remedy Standard A, while natu-
ral attenuation based upon adsorption, which is non-per-
manent and reversible, would not.
The commission does not plan to specify across-the-board
what is a reasonable time period for the completion of re-
sponse actions at all affected properties in Texas. This
rule is to be applied on a site-specific basis where the
individual program area is compared against the required
performance to establish remedial time frames. Likewise,
the commission is not specifying the details for the nu-
merous groundwater response methods which could be
used, but rather is defining the performance that what-
ever method is used would have to attain. Evaluation by
the commission of the reasonableness of the use of MNA,
or of any in-situ technology, will be based on data con-
tained in the self-implementation notice (SIN) and the re-
sponse action effectiveness reports (RAERs).The RAERs
are to be submitted at least every three years to docu-
ment whether sufficient progress is being made to achieve
the remedy. If a person can demonstrate that an in-situ
technology will achieve the response objectives within a
reasonable time frame, the technology would be an ac-
ceptable remedy. The agency is prepared to reject soil
and groundwater response plans when it is clear from the
beginning that they will not work or will work so slowly that
the response time cannot be considered "reasonable".The
executive director may require a demonstration of the ap-
propriateness of the remedy. If the executive director de-
termines either that insufficient progress is being made
toward attainment of the remedy standard or that the re-
sponse action is inappropriate, then the executive direc-
tor shall require the person to perform an alternative re-
sponse action. Finally, the provision expressed at
§350.31 (h) allows the executive director to require an in-
stitutional control to be recorded if a response action is
-------
either predicted to take or does take in excess of 15 years
to be completed. This process is designed to encourage
the early completion of response actions.
In summary, Remedy Standard A requires the use of pol-
lution cleanup, walk away response actions. COCs must
be removed and/or decontaminated to the critical PCLs
throughout the soil, groundwater, and other environmen-
tal media at an affected property within a reasonable time
period. No physical controls, such as a cap, may be used.
While attaining the groundwater response objectives, the
groundwater PCLE zone is not allowed to expand. Under
Remedy Standard A, the presence of non-aqueous phase
liquids (NAPLs) or a demonstration of technical impracti-
cability cannot be used to justify variance from the basic
requirement to reduce the COC concentrations in all envi-
ronmental media to at or below the critical PCLs. Remedy
Standard A is a self-implementing standard which means
that once a person files a SIN with the commission, the
person may perform the response action without receiv-
ing written approval. Since physical controls are not al-
lowed, there is no need for either post-response action
care or financial assurance to perform post-response ac-
tion care. And finally, there is no institutional control re-
quired for residential land use under Remedy Standard A.
However, an institutional control is required for commer-
cial/industrial land use.
To achieve Remedy Standard B, the person must remove,
decontaminate, and/or control the surface soil, subsurface
soil, and groundwater PCLE zones, other environmental
media, hazardous and non-hazardous waste such that
human and ecological receptors will not be exposed to
COCs in excess of the critical PCLs at the prescribed, or
any approved alternate, points of exposure (POEs). Since
physical control measures are typically used to achieve
Remedy Standard B, such response actions are gener-
ally referred to as "exposure prevention" remedies. As with
Remedy Standard A, the person must demonstrate that
the response action will be completed within a reason-
able time period. There is, however, a substantial differ-
ence between the performance requirements for the two
remedy standards. For Remedy Standard A, a COCs con-
centration must be reduced to the critical PCL throughout
the soil and groundwater PCLE zones. The performance
for Remedy Standard B is more achievable. The soil and
groundwater PCLE zones must be removed, decontami-
nated, and/or controlled such that the critical PCLs are
not exceeded at the POEs. In other words, the concentra-
tion of COCs does not necessarily need to be reduced to
the critical PCLs throughout the PCLE zones so long as
human and ecological receptors will be prevented by physi-
cal controls from being exposed to this material at the pre-
scribed or approved alternate POEs over the long term.
Under Remedy Standard B, with one exception, a person
may use a remedy which involves removal and/or decon-
tamination, removal and/or decontamination with controls,
or controls only. That exception is PCLE zones in class 1
groundwater which must be removed and/or decontami-
nated to the critical groundwater PCL for each COC. The
person is restricted to a pollution cleanup response ac-
tion because the objective is to restore the class 1 ground-
water to its beneficial use. Thus, the required performance
is to reduce the COC concentrations so that they are be-
low the critical PCLs and not merely to contain and pre-
vent the groundwater PCLE zone from spreading. Also,
the TRRP rule takes a pollution prevention approach for
any class 1,2, or 3 groundwater-bearing unit which does
not contain COCs above the critical PCLs. In further con-
trast to Remedy Standard A, MNA or other in-situ tech-
nology, whether it is a decontamination or physical control
process, could be used under Remedy Standard B, pro-
vided it is capable of attaining the groundwater perfor-
mance objectives.
Remedy Standard B is not self-implementing. A person
must receive the executive director's written approval of
an affected property assessment report (APAR) and re-
sponse action plan (RAP) before commencing the re-
sponse action. Institutional controls (i.e., deed notices,
restrictive covenants, and equivalent zoning or govern-
mental ordinances) are required more extensively under
Remedy Standard B than A. Institutional controls are re-
quired for either residential or commercial/industrial land
use, whenever the response action takes in excess of 15
years to complete, whenever a physical control is used,
whenever a modified groundwater response approach is
used (i.e., waste control unit, technical impracticability, or
plume management zone). Whenever physical control
measures are used to address the soil and/or groundwa-
ter PCLE zone(s), the person must perform the post-re-
sponse action specified in the approved RAP. Also, when-
ever a physical control measure(s) is used, the person
must provide financial assurance adequate to perform the
post-response action care.
The general groundwater response objectives for Rem-
edy Standard B are expressed at §350.33(f)(1). Unless
one of the modified groundwater response approaches
described in (f)(2), (f)(3), or (f)(4) is approved for an af-
fected property, these objectives apply to class 1,2, and 3
groundwater. The person must use either an active resto-
ration approach or MNA (or an in-situ technology) to re-
duce the concentration of COCs to the critical PCLs
throughout the groundwater PCLE zone. The person must
also prevent expansion of the groundwater PCLE zone,
prevent migration of COCs to air or surface water above
the PCLs, and prevent human and ecological receptor
exposure to the groundwater PCLE zone.
§350.33(f)(2), pertaining to waste control units, excludes
the groundwater throughout that portion of the groundwa-
ter PCLE zone directly underlying a waste control unit from
the requirement to meet the groundwater restoration re-
quirements. Beyond the perimeter of the waste control unit,
the response objectives must be met. In response to
§350.33(f)(3), a technical impracticability demonstration
can be used for all three classes of groundwater. The per-
son must demonstrate that it is not feasible from a physi-
-------
cal perspective using currently available technologies to
reduce the concentration of COCs throughout all or a por-
tion of the groundwater POLE zone to the critical PCLs
within a reasonable time frame. Where possible, the per-
son must reduce the concentration of COCs to the critical
PCLs for any portion of the groundwater PCLE zone where
it is technically practicable. The person must prevent mi-
gration of COCs from that portion of the groundwater PCLE
zone which satisfies the technical impracticability demon-
stration. Also, the person must prevent COCs at concen-
trations above the critical groundwater PCLs from spread-
ing beyond the existing boundary of the groundwater PCLE
zone. The person must also satisfy the requirements for
NAPLs which are expressed at §350.33(f)(4)(E) in the rule
and summarized later in this paper.
Plume management zones (PMZs), the final modified
groundwater response approach, are discussed in sub-
stantial detail in §350.33(f)(4). A PMZ may be established
under Remedy Standard B only for PCLE zones in class 2
or 3 groundwater, not class 1 groundwater. The key to a
PMZ is, if approved, the POE to groundwater will be
changed from throughout the groundwater PCLE zone to
an alternate location situated some distance hydraulically
downgradient of the existing PCLE zone. This alternate
groundwater POE is established in accordance with
§350.37(1) or (m) for class 2 or 3 groundwater, respec-
tively. Approval of a PMZ by the executive director is not
automatic. §350.33(f)(4)(A)(i) and (ii) present a substan-
tial list of groundwater and surface water factors which
are to be considered to determine whether such an ap-
proval would be appropriate.
This paragraph looks in a general manner at the estab-
lishment of a PMZ in class 2 groundwater. Not all perti-
nent details are discussed. The analysis starts with a de-
termination of the current length of the residential ground-
water PCLE zone as of the submittal date of the RAP. The
length of the PMZ is equal to the length of the residential
PCLE zone plus a distance "x". The objective is to deter-
mine the distance "x" since that is the acceptable distance
that a PCLE zone could expand. The maximum additional
length (i.e., x) of the PMZ is established as the smallest of
the following applicable distances: (1) up to 500 feet be-
yond the current length of the residential-based ground-
water PCLE zone; (2) a length of up to 0.25 times the
current length of the residential-based groundwater PCLE
zone (i.e., up to 25% additional plume length); (3) to within
two years groundwater travel time of the closest hydrauli-
cally downgradient off-site property (a) for which the owner
has not provided written concurrence to allow the record-
ing of an institutional control or (b) which does not contain
the residential-based PCLE zone and the groundwater has
a reasonably anticipated future beneficial use; (4) the cur-
rent extent of the residential PCLE zone when the resi-
dential groundwater PCLE zone is already within the two-
year travel time setback distance; or (5) the distance to a
surface water POE. As an exception, when the affected
property is subject to zoning or a governmental ordinance
which is equivalent to the deed notice, voluntary cleanup
program certificate of completion, or restrictive covenant
which would have otherwise been required, the criteria of
paragraphs (3) and (4) do not apply.
The final subject of this paper is to summarize how NAPLs
within PMZs must be managed. This also applies to NAPLs
within technical impracticability zones.The wording of this
requirement has been modified and no longer requires
that "NAPLs be removed to the maximum extent practi-
cable". Instead, TRRP requires that a person reduce
NAPLs which contain COCs in excess of PCLs "to the
extent practicable". First, the agency is not requiring NAPLs
to be removed solely because they are in a free phase
form.The NAPL must contain COCs in excess of the PCLs
to be of concern. Second, the rule establishes the follow-
ing criteria which the person and the executive director
will use to determine the required extent oi NAPL removal:
whether readily recoverable NAPLs have been recovered;
whether the NAPLs will generate explosive conditions;
whether the NAPLs will discharge to the ground surface,
to surface waters, to structures, or to other groundwater-
bearing units; whether there will be increased NAPL ex-
tent; and whether critical PCLs will be exceeded at POEs.
The extent of NAPL removal efforts to satisfy the 1o the
extent practicable" requirement will be determined on a
case-by-case basis depending upon the circumstances
at the particular affected property.
Persons who are interested in obtaining future guidance
pertaining to TRRP should periodically check theTNRCC
web site cited in the first paragraph of this paper for new
material. The agency plans a significant effort to prepare
guidance to explain TRRP and to facilitate its implemen-
tation.
-------
Overview of Abiotic Subsurface Remediation
Stephen G. Schmelling
USEPA
National Risk Management Research Laboratory
Subsurface Protection and Remediation Division
Ada, Oklahoma 74821
580-332-8540
schmelling.steve@epa.gov
The poor performance of pump-and-treat systems in the
early to mid 1980s provided an impetus for research to
better understand abiotic subsurface processes and to
apply the results of that research to the development of
better aquifer remediation technologies. Subsurface het-
erogeneity, sorption of contaminants to aquifer solids, and
the presence of nonaqueous phase liquids (NAPLs) were
recognized as significant barriers to successful aquifer
remediation. In addition, inadequate site characterization
severely limited the probability of success. Research on
subsurface processes led to the development of new
remediation technologies, based on applications of sub-
surface in situ abiotic processes, to overcome the fac-
tors limiting remediation success. Many of these new tech-
nologies have proved successful in small-scale field tests,
but only a few have been evaluated at full field scale, and
even fewer are in common use. The next few years are
likely to prove crucial in bringing the results of this re-
search to fruition and moving these technologies to regu-
lar use for site cleanup.
Introduction
Aquifer remediation was begun in earnest in the early to
mid 1980s, largely in response to the new Superfund leg-
islation which required the restoration of groundwater
quality at NPL sites. Early efforts relied almost exclusively
on what came to be known as "pump-and-treat." Pump-
and-treat was based on the idea that the contaminated
groundwater could be extracted from the aquifer, treated,
and either returned to the aquifer, or otherwise disposed
of, while the extracted water was replaced by clean wa-
ter. Most of these early systems were designed using
practices that had proved successful in the development
of ground-water supplies for public, agricultural, or indus-
trial use, but were not optimized for aquifer cleanup. A
study of 19 pump-and-treat systems by the EPA's Office
of Emergency and Remedial Response (OERR) in 1989
concluded that while many of these systems had achieved
mass removal and some degree of plume control as well
as reductions in contaminant concentrations, most of
them were not achieving aquifer restoration and would not
be likely to do so at any time in the foreseeable future.The
causes of poor performance for the systems in the study
included subsurface heterogeneity, sorption of contami-
nants to aquifer solids, the presence of NAPLs, inadequate
site characterization, and poor system design. Among
other results of this study was a realization that new ap-
proaches for aquifer remediation were needed and that
this required a major research effort.
Solving the Problem
As a result of the widespread recognition of the problems
of these early pump-and-treat systems, research was ini-
tiated by the ERA and other federal agencies, along with
the academic community and other organizations, to bet-
ter understand subsurface contaminant transport and fate
processes and to apply that understanding to develop and
evaluate innovative subsurface remediation technologies.
Many of these innovative technologies are the subject of
this conference and are based on application or manipu-
lation of abiotic subsurface processes.There has also been
research to develop better subsurface contaminant trans-
port and fate models that integrate process-level models
and may be used to design and evaluate these new
remediation technologies. At the same time there has been
a parallel research effort to develop better subsurface site-
characterization technologies.
All of the abiotic in situ technologies that we will hear about
at this conference rely on one or a combination of two
basic approaches. They either apply subsurface abiotic
processes to transform or destroy the contaminants in situ
or use abiotic processes to enhance contaminant removal
for subsequent surface treatment or disposal. A wide vari-
ety of approaches have been developed to accomplish
both of these objectives. While recognizing that subsur-
face contaminant problems need to be addressed as a
whole, including the vadose zone, the source area, and
the dissolved plume, it is often fruitful to apply different
approaches to different portions of the problem. For ex-
-------
ample, dealing with the dissolved plume initially may be
sufficient to limit or control risk to down-gradient recep-
tors while allowing time to develop a remedy for the con-
taminant source. Permeable reactive barriers often func-
tion in this mode.
Most of the technologies that we will hear about at this
conference are designed in one way or another to over-
come the barriers to aquifer cleanup that were identified
by the OERR study: aquifer heterogeneity, sorption, and
the presence of NAPLs. They are also looking to do so in
the most cost-effective manner.
As a consequence of subsurface heterogeneity, contami-
nants are often sequestered in low permeability zones that
are interspersed among zones of higher permeability.
Movement of the contaminants out of these low perme-
ability zones is often limited by diffusion rates which tend
to be slow relative to advective transport. There is no way
to completely overcome this problem, but a variety of ap-
proaches can lessen its impact. Abiotic technologies that
address this problem include the addition of energy to heat
the aquifer using steam, hot water, electrical resistance
heating, or radio-frequency heating. Electrokinetics is an-
other alternative for dealing with zones of low hydraulic
conductivity. Permeable reactive barriers can be a cost-
effective way around this problem if the goal is protection
of down-gradient receptors. In addition, better system de-
sign, including the use of wells that induce vertical flow
also address this problem. Heterogeneity is one of the
most prominent features of fractured rock sites which, to
date, remain a major challenge for successful subsurface
remediation.
Sorption of contaminants to aquifer solids reduces the rate
at which they can be removed by extraction technologies.
Technologies which address the issue of sorption include
the addition of thermal or electrical energy to increase the
rate of desorption or the addition of a chemical agent that
will increase contaminant solubility and effectively change
the partition coefficient. Addition of a chemical reactant,
such as an oxidant, may also be used to destroy the sorbed
contaminant. Conversely, particularly for inorganic con-
taminants, sorption may provide a means of sequestering
the contaminant so that its concentration in the adjacent
groundwater is reduced. Some permeable reactive barri-
ers are examples of this approach.
The presence of NAPLs presents some of the most chal-
lenging problems for subsurface remediation. Most abi-
otic technologies that address this problem try to increase
the rate at which the contaminant can be removed by in-
creasing its solubility or its mobility through the addition of
a chemical agent, such as a surfactant, or a physical agent,
such as heat.Thursday morning's session on in situ flush-
ing will provide an extensive review of the status of efforts
to remove NAPLs by increasing solubility or mobility.
The Current Situation
Abiotic in situ subsurface remediation technologies ap-
pear to be at a critical stage in their development and imple-
mentation. Pump-and-treat is now better understood, both
in terms of what it can reasonably be expected to accom-
plish and how to design systems that operate more effec-
tively, and pump-and-treat is still the single most commonly
used ground-water remediation technology. Innovative
technologies that will be discussed at this meeting are
making progress, although most have yet to be widely used
for site cleanup. Included among those in more widespread
use are air sparging combined with soil vacuum extrac-
tion, permeable reactive barriers, and to some extent, ther-
mal technologies. In addition, monitored natural attenua-
tion, which, for organic compounds, relies predominantly
on biodegradative processes, is being implemented at a
number of sites, usually in combination with some other
remedial technology. Other technologies are still at some
sort of demonstration/evaluation stage ranging from field
pilot scale to full scale. Included in this group are in situ
flushing, in situ oxidation/reduction, and the use of reac-
tive zones.
Current Needs and Issues
While a great deal of progress has been made in the last
ten years, there are still a number of needs and issues
that need to be addressed before abiotic in situ remediation
technologies will fulfill their promise. In addition to tech-
nology-specific needs, there are also general issues that
cut across most of the technologies that will be presented
and discussed at this meeting. For those technologies that
have shown success in field pilot-scale demonstrations, a
critical next step is to take the technology to a larger scale
demonstration and to get it in use for an actual site cleanup.
For those technologies that have been implemented at
one or a few sites, the next goal is to make the leap to
becoming accepted practice for site remediation. An im-
portant part of this transition is the acquisition of credible
cost data and the development and evaluation of cost
models that can be used for system design.
No technology is likely to remove 100 percent of the con-
taminants that are present at a site. Consequently, it is
important to determine the benefits of partial mass re-
moval. Another way to phrase this issue would be in terms
of relating mass removal to risk reduction. This is an ac-
tive area of research, but the questions are far from re-
solved.
Better site characterization, while not the subject of this
conference, is a continuing need and intimately connected
with the success of all subsurface remediation technolo-
gies. This need is particularly great for sites with complex
hydrogeology, including fractured rock. Site characteriza-
tion is closely related to the problem of actually measur-
ing the performance of a subsurface remediation technol-
-------
ogy. In many situations, the reported performance of a
technology may be highly dependent on how it is mea-
sured. Needs in this area include both specific measure-
ment techniques and more general agreement on bench-
marks for measuring performance. There is a need for
both pre- and post-cleanup monitoring methods and real-
time methods to measure the ongoing progress of a re-
medial action.
Many sites will require different approaches for different
parts of the site. As a result, there is a need to better un-
derstand how to link technologies together to achieve site
cleanup in the most cost-effective manner.
One conclusion to be drawn from this brief overview is
that, while a great deal of progress has been made in im-
proving the effectiveness of subsurface remediation tech-
nologies, there is still a great deal of work yet to be done
before the promise of innovative approaches is realized.
Disclaimer
This is an extended abstract of the proposed presenta-
tion. It does not necessarily reflect USEPA policy and no
official endorsement should be inferred.
-------
Permeable Reactive Barriers for Remediation of Acid Mine Drainage
David W. Blowes, Shawn G. Benner, and Carol J. Ptacek
Department of Earth Sciences
University of Waterloo
Waterloo, Ontario, N2L 3G1, Canada
519-888-4878
blowes@sciborg.uwaterloo.ca
Permeable reactive barriers are emerging as an alterna-
tive to traditional pump and treat systems for groundwater
remediation. This technique has progressed rapidly from
laboratory bench-scale studies to full-scale implementa-
tion over the past decade. Laboratory studies indicate the
potential for treatment of a large number of inorganic con-
taminants, including As, Cd, Cr, Cu, Hg, Fe, Mn, Mo, Ni,
Pb, Se, Tc, U, V, NO3, PO4, and SO4. Small-scale field
studies have indicated the potential for treatment of Cd,
Cr, Cu, Fe, Ni, Pb, NO3, PO4, and SO4. Solid-phase or-
ganic carbon in the form of municipal compost has been
used to remove dissolved constituents associated with
acid-mine drainage, including SO4, Fe, Ni, Co Cu, Cd and
Zn.
Introduction
Over the past decade, permeable reactive barriers have
been developed and used to treat groundwater contami-
nated by inorganic constituents. Permeable reactive bar-
riers are placed in the path of a migrating plume of con-
taminated groundwater. Reactive materials within the bar-
rier are selected to promote geochemical reactions that
result in the destruction or stabilization of the groundwa-
ter contaminants. Ideally these materials are sufficiently
reactive to treat water for periods of years to decades.
Mixtures selected for the attenuation of inorganic species
must be designed to maintain their permeability as sec-
ondary precipitates accumulate. The barrier design must
also ensure that the contaminant will remain immobilized
within the aquifer, or can be retrieved with the reactive
material following treatment.
Metal Removal by Sulfate Reduction
Biologically mediated reduction of sulfate to sulfide, ac-
companied with the formation of metal sulfides occurs
through the reaction sequence:
2CH20(S)
,->H2S(a,0+2002(ll4)+H20(,)
(aq)
'(s)
where CH2O represents organic carbon and Me2*
represents a divalent metal cation. Biologically mediated
sulfate reduction has been used to treat metal cations
derived from mining activities in permeable reactive
barriers (Blowes et al., 1995,1998; Waybrant et al., 1995,
1998; Benner etal., 1997,1999; McGregor et al., 1999).
Although these systems are designed to promote the
removal of dissolved metals, these barriers also effectively
remove sulfate. In laboratory studies, Waybrant et al. (1995)
observed sulfate removal at rates of 0.14 to 4.23 mg L1
day1 g-1.
Laboratory studies indicate that many metals, including
Ag, Cd, Co, Cu, Fe, Ni, Pb and Zn, are treatable using this
approach (Waybrant et al., 1995, 1998). Column experi-
ments, conducted using a range of organic substrates
demonstrated the potential to remove a range of dissolved
metals at groundwater velocities similar to those observed
at sites of groundwater contamination (Waybrant et al.,
1995).
In August, 1995, a full-scale continuous reactive barrier
was installed in an aquifer downgradient from an inactive
mine tailings impoundment at the Nickel Rim mine site,
Sudbury, Ontario (Benner et al., 1997). The barrier was
installed into a confined bedrock valley by a cut-and-fill
technique. The barrier is 15 m long, 4.3 m deep and 3.7 m
wide. It is composed of a reactive mixture containing mu-
nicipal compost, leaf compost, and wood chips to promote
bacterial sulfate reduction and metal sulfide precipitation
reactions. These organic materials were mixed with pea
gravel to attain a permeability greater than that of the aqui-
fer. One meter wide buffer zones containing coarse sand
were installed on the up and down gradient sides of the
reactive material.
After passing through the permeable barrier, concentra-
tions of SO4 decrease from 2400 - 3800 mg/L to 110 -
1900 mg/L. The concentrations of Fe decrease from 740 -
1000 mg/L to < 1 - 91 mg/L, and alkalinity values increase
from 60 - 220 mg/L (as CaCO3) to 850 - 2700 mg/L (as
10
-------
CaCOJ. A comparison of the equivalents of potential acid-
ity to the equivalents of alkalinity indicate that the water
entering the barrier will generate acidity when exposed to
atmospheric oxygen.The water leaving the barrier has an
excess of alkalinity and will consume acidity when dis-
charged from the aquifer. As a result of the barrier, the
plume water will begin to neutralize the pH of the receiv-
ing surface-water flow system. Concentrations of dissolved
Ni of up to 10 mg/L up gradient of the barrier are decreased
to < 0.1 mg/L within and down gradient of the barrier. Enu-
meration of sulfate-reducing bacteria indicates an abun-
dance of these species within the wall, and elevated num-
bers in the down gradient aquifer (Benner et al., 1997).
In March 1997, a pilot-scale, compost-based reactive bar-
rier was installed at an industrial site in Vancouver, Canada
(McGregor et al., 1999). The barrier was installed in the
path of a plume of groundwater containing Cd, Cu, Ni, Pb
and Zn. The barrier was installed by trenching using bio-
degradable slurry. Dissolved Cu concentrations were ob-
served to decrease from 300 mg/L to <5 mg/L within the
barrier. The concentrations of Cd, Ni, Pb and Zn showed
similar decreases, with effluent concentrations generally
below instrument detection limits.
References
Benner, S.G., Blowes, D.W., Ptacek, C.J., 1997. A full-scale
porous reactive wall for prevention of acid mine drainage.
Groundwater Monitor. Remed. XVII (4), 99-107.
Blowes, D.W., Ptacek, C.J., Bain, J.G., Waybrant, K.R.,
Robertson, W.D., 1995. Treatment of mine drainage water
using in situ permeable reactive walls. Proc. Sudbury '95,
Mining and the Environment, CANMET, Ottawa, ON Vol.
3, 979-987.
Blowes, D.W., Ptacek, C.J., Benner, S.G., Waybrant, K.R.,
Bain, J.G., 1998. Porous reactive walls for the prevention
of acid mine drainage: a review. Min. Pro. Ext. Met. Rev.,
19,25-37.
McGregor, R.G., Blowes, D.W., Ludwig, R., Pringle, E.,
Pomery, M., 1999. Remediation of a heavy metal plume
using a reactive wall. Proc. In Situ and On-Site
Bioremediation Conference, April 19-22, 1999, San Di-
ego, California.
Waybrant, K.R., Blowes, D.W., Ptacek, C.J., 1995. Selec-
tion of reactive mixtures for the prevention of acid mine
drainage using in situ porous reactive walls. In Proc.
Sudbury '95, Mining and the Environment, CANMET, Ot-
tawa, ON Vol. 3, 945-953.
Waybrant, K.R., Blowes, D.W., Ptacek, C.J., 1998. Selec-
tion of reactive mixtures for use in permeable reactive walls
for treatment of mine drainage. Environ. Sci.Technol., 32,
1972-1979.
11
-------
Long-Term Performance Monitoring of a PRB for Remediation of
Chlorinated Solvents and Chromium
Robert W.PuIs
USEPA
National Risk Management Research Laboratory
Ada, Oklahoma 74820
580-436-8543; (fax) 580-436-8703
puls.robert@epa.gov
Timothy Sivavec
General Electric Corporate Remediation
Schenectady, New York
Permeable reactive barriers (PRBs) are an emerging, al-
ternative in situ approach for remediating contaminated
groundwater that combine subsurface fluid flow manage-
ment with a passive chemical treatment zone. In the last
few years, there has been extensive research conducted
to improve our understanding of the mechanisms and ki-
netics of the transformation reactions responsible for the
removal of contaminants from the aqueous phase in such
in situ treatment systems. The few pilot and commercial
installations which have been implemented have proven
that passive permeable reactive barriers can be a cost-
effective and efficient approach to remediate a variety of
different compounds. However, in all of the pilot and com-
mercial installations to date there has been very little data
collected or research focused on the long-term perfor-
mance of these in situ systems, particularly with respect
to the build-up of surface precipitates or bio-fouling.
A detailed analysis of the rate of surface precipitate buildup
in these types of passive in situ systems is critical to un-
derstanding how long these systems will remain effective.
Different types of minerals and surface coatings have been
observed to form under different geochemical conditions
which are dictated by the composition of the permeable
reaction zone and aquifer chemistry. Microbiological ac-
tivity impacts are also important to understand and better
predict how long these systems will remain effective in
the subsurface.
Continuous Wall Emplacement
A permeable in situ subsurface reactive barrier composed
of 100% granular zero-valent iron (ZVI) was installed in
June, 1996, at the U.S. Coast Guard Support center near
Elizabeth City, North Carolina to treat overlapping plumes
of chromate (Cr(VI)) and chlorinated solvent compounds
(trichloroethylene (TCE), cis-ichloroethylene (c-DCE), and
vinyl chloride (VC)). Concentrations in excess of 10 mg/L
Cr and 19 mg/L TCE had been detected in the groundwa-
ter at the site since 1991. The wall was emplaced using a
continuous trenching machine. The continuous trencher
excavates native soil and allows the iron to be emplaced
in one continuous operation. Excavated aquifer materials
were brought to the surface by an excavating belt, and
then conveyed to the side of the machine. An estimated
3.2 m3 of iron-filings were emplaced per linear meter and
about 280 tons of iron was installed. The installation was
completed during a 6 hour period on June 22,1996. The
PRB is 46 m long, 7.3 m deep and 0.6 m wide and ori-
ented perpendicular to groundwater flow.
Wall Emplacement Verification
Wall emplacement was verified using a conductivity probe
manufactured by Kejr Engineering, Inc that was advanced
through the soil/iron interface using a Geoprobe™ . The
tool provides real-time, specific conductance data versus
depth on a portable computer. The radius of influence of
the probe is 2-3 cm. The data was used to identify the
location of the plume, the outlines of where the iron was
emplaced, and the density of packing of the iron filings
within the aquifer. Conductivity values greater than 100
mS/m indicate the presence of the granular iron. Differ-
ences in conductivity between the aquifer sediments and
the iron filings were as much as 2 orders of magnitude,
making this a very useful tool in locating the relatively thin
wall (0.6 m). Data collected thus far indicates a wall thick-
ness of 45 cm, or 15 cm less than the designed thickness
of 60 cm. Other measurements have ranged from 48 to
55 cm.
Monitoring Network
Performance monitoring entails the following: sampling of
10-5 cm dia PVC compliance wells and 15 multi-level sam-
plers (7 to 11 sampling points per sampler) arranged in
three transects perpendicular to the wall for the following
constituents: TCE, cis-DCE, vinyl chloride, ethane, ethene,
acetylene, methane, major anions, metals, Cr(VI), Fe(ll),
total sulfides, dissolved H2, Eh, pH, dissolved oxygen (DO),
specific conductance, alkalinity, and turbidity.
12
-------
Geochemical Indicator Parameters
Ferrous iron concentrations within the wall increase from
background levels within the aquifer of less than 0.5 mg/L
to as much as 14.8 mg/L but are variable with depth, loca-
tion and time. Total iron concentrations are similar to fer-
rous iron values indicating most of the soluble iron is in
the ferrous form. Within 1.5 m downgradient, ferrous iron
concentrations persist as high as 2.2 mg/L. Eh values within
the wall are as low as -600 mv but generally range from -
250 to -550 mv, whereas upgradient Eh ranges from 250
to 450 mv. Upgradient pH ranges from 5.7 to 6.5, whereas
within the wall the pH is generally between 9 and 10.7.
DO values within the aquifer range from 0.2 to 2.0, whereas
within the wall DO is generally less than 0.2. Dissolved
hydrogen increases from background concentrations of
less than 10 nMol to greater than 1000 nMol. These data
are consistent with the effects of the iron corrosion reac-
tion. The following equation shows that the oxidation of
the iron filings would be expected to generate ferrous iron
and dissolved hydrogen, decrease Eh and increase pH:
Fe° + 2H2O-> Fe2++ H2 + 2OH-
Geochemical conditions within 1.5 m downgradient of the
wall show increasing reducing conditions over time, indi-
cating that a "redox front" may be moving downgradient
from the wall and the area affected by reducing conditions
may increase over time.
Wall Performance: Contaminant Removal
The center of the Cr(VI) plume resides between 4.5 and
5.5 m below ground surface while the extent of the plume
ranges from 4 to 7 m below ground surface. Cr(VI) con-
centrations decline from upgradient values as high as 5.1
mg/L to less than detection limits (<0.01 mg/L) within the
first few centimeters of the wall. No chromate is detected
downgradient of the wall.
Under the highly reducing conditions which prevail within
the wall, the reduction of Cr(VI) to Cr(lll) is predicted and
the subsequent formation of an insoluble precipitate is
formed as shown below:
Fe° + CrO42 +4H2O -» (FexCr,.x)(OH)3 +5OET
Geochemical modeling calculations indicate that this pre-
cipitate should form and support Cr(lll) aqueous concen-
trations less than 0.01 mg/L. Cr (III) has been detected on
the surface of the iron in a few cores using x-ray photo-
electron spectroscopy (XPS) analysis. Cr has also been
detected in these same samples using scanning electron
microscopy with energy dispersive x-ray analysis (SEM-
EDS) and appears in association with iron as surface coat-
ings on the iron filings.
The vast majority of the multi-layer sampling ports show
reduction of the chlorinated volatile organic compound
(CVOC) concentrations to less than North Carolina (De-
partment Environmental Quality) regulatory target levels
(5 mg/L.TCE; 70 mg/L, c-DCE; 2 mg/L VC). With the dis-
appearance of the TCE, c-DCE and VC, there has been a
steady increase in detectable ethane, ethene, and acety-
lene. These data indicate that the organic compounds are
degrading via both reductive dechlorination and beta elimi-
nation pathways.
Analysis of Emplaced Iron
Limited mineralogic analyses of cores have been com-
pleted. Due to increased pH within the wall, there is a shift
in equilibrium form a bicarbonate to a carbonate domi-
nated system. Analysis of total inorganic carbon to evalu-
ate precipitation of likely carbonate phases (e.g. siderite
[FeCO3], calcite [CaCO3]), shows increases in inorganic
carbon within the iron compared to the upgradient aquifer
sediments. This is true for both the upgradient and
downgradient interfaces, but differences are greater for
the upgradient cores.
The presence of a large reservoir of iron and favorable pH
and substrate availability conditions may favor the activity
of iron and sulfate reducing bacteria and methanogens.
This enhanced activity may favorably influence zero-va-
lent iron reductive dehalogenation reactions through fa-
vorable impacts to the iron surface or through direct mi-
crobial transformations of the target compounds. However,
this enhancement may come at the expense of faster cor-
rosion leading to faster precipitate buildup and potential
biofouling of the permeable treatment zone. Increases in
microbial biomass have been detected at the upgradient
and downgradient wall interfaces.
The most ubiquitous mineral phases observed thus far
are amorphous iron (oxy)hydroxides, calcite, siderite, and
possibly green rust (general stoichiometry of
[Fe2+4Fe3yOH-)12]2+ • [SO/- • H2O]2').The corrosion layer
is greatest (20-25 um) within the first 5 cm and decreases
significantly (<5 um) within 20 cm of the upgradient inter-
face.These observations have been made with cross-sec-
tional SEM-EDS and additional studies are underway.
Summary
Chromium is removed from the groundwater to less than
detection limits (<0.01 mg/L) and considerably less than
regulatory target levels (0.1 mg/L). This is accomplished
via redox reactions accompanied by precipitation pro-
cesses due to the corrosion of the iron. Likewise there is
reduction in CVOC concentrations to less than regulatory
limits where these compounds are entering and being
treated by the iron wall.
Acknowledgments
The authors wish to thank the following individuals for their
contributions to this project: Dr. David Blowes, Dr. Robert
Gillham, Mr. Tim Bennett, University of Waterloo; Mr. John
Vogan and Ms. Stephanie O'Hannesin, EnviroMetal Tech-
nologies, Inc.; Ms. Cindy Paul and Mr. Frank Beck, USEPA;
Mr. Jim Vardy, Mr. Murray Chappell, Mr. Frank Blaha,
USCG.
13
-------
Long-Term Performance Monitoring of Permeable
Reactive Barriers at DOE Sites
Nic Korte and Liyuan Liang
Oak Ridge National Laboratory
Environmental Sciences Division
Grand Junction, CO 81503
970-248-6210
nek@ornl.gov
A permeable reactive barrierwas installed to destroy chlo-
rinated hydrocarbons at the Department of Energy's (DOE)
Kansas C'rty Plant. In the Feasibility Study, it was predicted
that mineral precipitation would not be a problem for a
minimum of five years. It is hoped, of course, that the bar-
rier will last much longer. Nevertheless, most scientists
believe that barriers will eventually require regeneration
or retrieval. Indeed, where removal of uranium with Fe° is
concerned, as at deployments on the Oak Ridge reserva-
tion, some believe that barriers will have to be removed or
re-oxidation and remobilization will occur. When Fe° is used
in an in situ barrier, corrosion and mineral precipitation
alter the surface composition. Pilot studies have shown
that compositional changes decrease the reactivity of the
iron and mineral precipitation decreases flow through the
barrier [McKenzie et al 1997].
One field test exhibited significant clogging within 6 months
of operation [Korte et al. 1997]. The influent side of the
media was solidly cemented and sampling of the media
revealed the presence of a diverse microbial population.
Clearly, cost-effective monitoring in conjunction with pre-
dictive capability to determine when barrier failure is im-
minent are needed. For these reasons, the DOE, the De-
partment of Defense (DOD), and the EPA (EPA) have
embarked on a joint project to determine cost-effective
long-term monitoring strategies. In this presentation, ob-
servations from field installations are summarized, and a
biogeochemical conceptual model, to be used as a basis
for developing long-term monitoring strategies, is pro-
posed. A brief discussion of hydraulic monitoring concepts
is also included.
Water Quality
Water quality data from Fe° barriers is not always consis-
tent with respect to pH, dissolved oxygen (DO), alkalinity,
and concentrations of cations and anions. At Hill Air Force
Base (AFB), the pH increased from 7.5 to >9. Similarly,
pH was significantly higher within the Fe°than in the influ-
ent in installations at Sunnyvale, CA; Lowry AFB, CO; and
Elizabeth City, NC. In contrast, the pH remained near neu-
tral (from ~ 6 to ~7) at the Portsmouth Gaseous Diffusion
Plant (PORTS) in Ohio [Liang et al. 1997].
At Ports, significant clogging occurred within the Fe°-me-
dium, forcing the system to cease operation after 6 months.
A significant decrease in SO4z-concentration was also
observed with the effluent SO/- between 125 to 175 mg/L
versus an influent concentration of 400 mg/L. The ground-
water at Ports contained iron oxidizers, sulfate reducers,
anaerobic heterotrophs, and iron reducers at levels of 1 to
100/mL. In contrast, after 1 year of operation of the Fe°
barrier at the Borden site, samples did not show biological
growth, mineral precipitation, or alteration of the iron
[Matheson 1994,0'Hannesin and Gillham 1998]. At most
field sites, however, precipitates are detected. Precipitate
formation was observed at the top of the reactors at the
SQL Printed Circuits site in New Jersey, at the Lowry AFB
site in Colorado, and at the Y-12 site in Tennessee. At the
Lowry site, total alkalinity, Ca2+ and SO42' concentrations
rapidly decreased along the reactor flow path [Edwards et
al. 1996]. At Hill AFB, precipitates of iron and calcium car-
bonate accounted for a 14% reduction in porosity within a
few months. A lower annual porosity reduction (0.5% po-
rosity loss) was estimated after 10 months of operation at
the Denver Federal Center site [McMahon 1997].
A Conceptual Model
The variable results from the field indicate that an increased
understanding of both chemical and microbial reactions
should permit more accurate predictions regarding clog-
ging and biogeochemical impacts on Fe° barrier perfor-
mance. In considering a conceptual model, it is assumed,
as a first approximation, that all reactions reach a pseudo-
equilibrium.
The chemical aspects of a conceptual model can be
bounded by the following hypotheses:
14
-------
• An elevated DO in groundwater will be a major factor
contributing to the corrosion and plugging of a reac-
tive barrier due to the precipitation of ferric oxide/hy-
droxide at the up-gradient interface.
• Groundwater with limited buffering capacity will ex-
hibit a high pH in the Fe° barrier, thus suppressing
microbial activity, which, in turn, will result in little risk
of biofouling.
• Groundwater with a high bicarbonate buffering capac-
ity will result in near-neutral pH in the Fe°and greater
risk of fouling/clogging from precipitation of metal sul-
fides and carbonates.
• Natural organic matter will compete for iron and ox-
ide/ hydroxide surfaces; thus inhibiting reductive sur-
face reactions with contaminants.
• When Fe° contacts groundwater, oxidation-reduction,
mineral precipitation, and associated equilibrium re-
actions occur. Mixed oxides (e.g., green rusts) that
incorporate different types and amounts of anions (Ch
, SO42- or CO32- etc.) may form. Transformation of one
mineral to another, such as green rust to ferric
oxyhydroxide or to iron sulfides, also will occur.
Using pH and pe, Fig. 1 demonstrates that different iron
minerals are observed depending on the nature of the in-
PH
0 24 6 8 10 12 14 16
20
16
Figure 1. pH-pe diagram for the system Fe-CO2-H2O, illustrating esti-
mated pH, pe, and the iron mineral precipitates that correspond to the
type of influent and effluent groundwater chemistry for an Fe° reactive
barrier. The solid phases are am-Fe(OH)3(amorphous ferrous hydrox-
ide), and FeCO3 (siderite), Fe(OH)2(ferrous hydroxide), and Fe°. Lines
are calculated based on total carbonate species at 10/3M, ferrous and
ferric iron at 10/SM. (Adapted from Strumm and Morgan 1996).
fluent groundwater and the corresponding effluent from
the Fe° barrier. Thus, the amount of precipitation produced
depends on the characteristics of the influent groundwa-
ter (shown in regions I and II in Fig. 1) as well as the re-
sulting effluents (as region III and IV).
Typical groundwater pH ranges from ~5 to -8. High DO
groundwater is illustrated in region I in Fig. 1 and low DO
in region II. As groundwater passes through the Fe°, DO
will be consumed rapidly, decreasing the redox potential
to about -400 to -600 mv as illustrated in regions III and
IV.
Precipitation as Influenced by Dissolved
Oxygen
Region V in Fig. 1 typifies the condition under which oxi-
dative precipitation of ferric oxides may cause clogging at
the upgradient interface. The aerobic oxidation of Fe° re-
sults in the precipitation of amorphous iron oxide:
Fe°(s) + 1-H20 + -02
Fe(OH)3(s)
When treating water with a high DO and a neutral pH,
orange precipitates typical of oxidized iron have been ob-
served in the Fe°-medium in bench and pilot-scale stud-
ies [Liang et al. 1997, McKenzie et al. 1997, Shoemaker
et al. 1995]. Under neutral pH conditions, the solubility of
ferric hydroxide is low and its precipitation is expected.
Thus, the influent DO determines the amount of ferric ox-
ide produced at the interface to the Fe°-barrier. For every
% mole of oxygen consumed, 1 mole of iron oxide is pro-
duced (Eq. 1). Thus, if a groundwater contains 4 mg/L
(0.125 mM) DO, the molar concentration of ferric hydrox-
ide generated could be 0.1 7 mM (1 8 mg/L). Assuming that
water enters the Fe° media through a cross-sectional area
of 1 ft2 (0.09 m2) at a flow rate of 0.1 gal/min (0.39 L/min),
within 6 months the amount of ferric hydroxide generated
could approach 2 kg and form a 1 .3-cm-thick plug (as-
suming a porosity of 0.5 and a density of 3 g/cm3).
System pH and Alkalinity
A poorly buffered groundwater will show a large pH in-
crease, characterized by region IV in Fig. 1 . Most labora-
tory tests have been conducted in weakly buffered solu-
tions, and the pH observed is typically >9 [Gillham and
O'Hannesin 1994]. Examples from the field include Hill
AFB and a site in Sunnyvale, California, where a high pH
typical of low alkalinity groundwater was observed [Shoe-
maker etal. 1995].
As noted previously, alkalinity will largely determine the
pH of the barrier effluent. Groundwater with a limited buff-
ering capacity will show an increase, and groundwater
containing high alkalinity and/or high SO42' will have a near-
neutral pH as in region III, Fig. 1 . Although carbonate spe-
cies are the major contributors to groundwater alkalinity,
redox and microbial reactions could produce species that
15
-------
neutralize H*. For example, ferric oxide species formed
from Fe° oxidation and sulfide species from SO42' reduc-
tion can increase alkalinity. Microbial SO42'-reduction can
cause fouling and clogging from mineral precipitation as
well as from the growth of the bacteria.
Water quality changes at the PORTS site provided direct
evidence of sulfate-mediated biogeochemical reactions.
The influent SO42- concentration decreased across the Fe°
medium, while pH remained near neutral. The change in
Fe(ll) concentration with time coincided with that of alka-
linity. The cause of the alkalinity increase has been attrib-
uted to the presence of dissolved ferric and ferrous hy-
droxide species, sulfides, and titratable organic compounds
produced by SO 2--reducing bacteria. The simultaneous
decrease of Fe(ll) and alkalinity indicated the precipita-
tion of siderite within the medium.
Hydraulic Monitoring
Hydraulic monitoring is as important as geochemical moni-
toring when evaluating barrier performance. Thus, this
project has also focused on determining the type of hy-
draulic testing that should be performed. Various types of
flow meters have been used in and around barriers. The
colloidal borescope, which uses natural colloid flow within
the barrier as a means of measuring groundwater flow
rate and direction, has been used at several DOE sites.
Directional data with a heat-pulse flow meter at the Kan-
sas City Plant, however, showed little correlation with data
from the colloidal borescope or with the potentiometric
surface.The borescope, while it provides consistent data,
can be difficult to use within a barrier because there are
so many finely divided iron particles floating within the well
bore. Continuous water-level measurements, therefore,
appearto be the most useful hydraulic monitoring method.
Additional testing at a variety of sites should yield specific
recommendations regarding which procedures are appro-
priate.
References
Edwards, R.W., D. Duster, M. Faile, W. Gallant, E.Gibeau,
B. Myller, K. Nevling, and B. O'Brady, 1996. In RTDF Meet-
ing Summary, San Francisco, CA, Aug. 15-16,1996.
Gillham, R.W., and S. F, O'Hannesin, 1994. Groundwater
32(6): 958.
Korte, N., O. R. West, L. Liang, M. J. Pelfrey, and T. C.
Houk. 1997. Federal Facilities Environmental Journal,
Autumn, 1997:105.
Liang, L., O. R. West, N. E. Korte, J. D. Goodlaxson, D. A.
Pickering, J. L. Zutman, F. J. Anderson, C. A. Welch, and
M. J. Pelfrey. 1997. TM-13410. Oak Ridge National Labo-
ratory, Grand Junction office, Grand Junction, Colorado.
Matheson, L. J. 1994. 'Abiotic and Biotic Reductive
Dehalogenation of Halogenated Methanes,' PhD Disser-
tation, Oregon Graduate Institute of Science and Technol-
ogy, Portland, Oregon.
McKenzie, P. D., T. M. Sivavec, and D. P. Horney. 1997.
Extending Hydraulic Lifetime of Iron Wall, International
Containment Technology Conference Proceedings, Feb
9-12,1997 St Petersburg, Florida, 781-787.
McMahon, P. B. 1997. Field Evaluation of a Permeable
Reactive Barrier Containing Zero-Valence Iron at the Den-
ver Federal Center. RTDF Meeting Summary, Virginia
Beach, VA. September 18-19,1997.
O'Hannesin, S. R, and R. W. Gillham. 1998. Long-Term
Performance of an in Situ "Iron Wall" for Remediation of
VOCs. Groundwater, 36 (1): 164.
Shoemaker, S. H., J. F. Greiner, and R. W. Gillham. 1995.
In "Assessment of Barrier Containment Technologies", R.
R. Rumer, J. K. Mitchell (eds), Section 11:301.
16
-------
Demonstration of Permeable Reactive Barriers for Groundwater
Remediation: Status and Preliminary Results of the Fry Canyon Project
Edward M. Feltcorn and Ronald Wilhelm
USEPA
401 M St., SW
Washington, DC 20460
202-564-9422
feltcorn.ed@epamail.epa.gov
David L. Naftz; James A. Davis; Christopher C. Fuller; Geoff W. Freethey; Michael J. Piana
USGS
Stan J. Morrison
Weston, Inc.
An abandoned uranium (U) ore upgrading facility in south-
eastern Utah, known as the Fry Canyon site, was selected
as a long-term demonstration site to assess the perfor-
mance of selected permeable reactive barriers (PRBs)
for the removal of U and other trace elements from ground-
water. PRBs consist of reactive material placed in an aqui-
fer for the purpose of immobilizing, destroying, or render-
ing a contaminant less toxic or mobile. Previous PRB in-
stallations targeted primarily organic contaminants.
Project partners include the USEPA (USEPA), Utah De-
partment of Environmental Quality, and U.S. Geological
Survey. Project funding was provided by the USEPA. Ob-
jectives of the project were: (1) hydrologic and geochemi-
cal site characterization prior to barrier emplacement; (2)
design, installation, and operation of selected materials;
and (3) evaluation of barrier(s) performance for site
remediation and economic viability.
Three PRBs were installed in 1997 using funnel and gate
construction techniques. Reactive materials demonstrated
are zero valent iron (ZVI), amorphous ferric oxyhydroxide
(AFO), and phosphate (PO4). Uranium concentration in
the aquifer ranged from 60 micrograms per liter (mg/L) in
water from a background well to 20,700 mg/L in water
beneath the tailings. Uranium reduction ratios (input U
concentration/PRB U concentration) were calculated for
all water samples collected from each PRB. Initial results
indicate that the ZVI PRB is most efficient in removing U
from groundwater (>99% removal rate); however, U re-
moval efficiencies in the PO4 and AFO PRBs also are
high (>90% removal rate).
17
-------
An Overview of Installation Methods for Permeable Reactive Barriers
Stephanie O'Hannesin
EnviroMetal Technologies Inc.
745 Bridge Street West, Suite 7
Waterloo, Ontario Canada N2V 2G6
519-746-2204; (fax) 519-746-2209
sohannesin @ eti.ca
In situ permeable reactive barriers (PRBs) for groundwa-
ter remediation have been proven to be a cost effective
alternative to conventional groundwater treatment. Vari-
ous configurations and installation methods have been
implemented to install PRBs into the subsurface at vari-
ous depths. Innovative configurations have evolved to re-
flect site specific parameters such as site geology,
hydrogeology and contaminants distribution. Various con-
ventional geotechnical construction methods have been
adapted to accommodate installations involving reactive
materials.
Configurations
Various PRB configurations have been implemented to
ensure complete plume capture and remediation. A con-
tinuous permeable wall configuration distributes the reac-
tive material across the entire path of the contaminated
groundwater. This configuration is least sensitive to com-
plexities in the flow field and does not significantly alter
the natural groundwater flow path.
In a funnel and gate configuration, the contaminated
groundwater is directed by low permeability funnels to-
wards a permeable treatment zone or gate. The funnel
sections could consist of sheet piling (which eliminates
the removal of soil and reduces the soil disposal cost), or
slurry cuttoff walls. Depending on the type of slurry wall,
some portion of the excavate soil may be incorporated
Into the wall, however, soil disposal costs must be taken
Into account for these types of funnels. The funnel and
gate system must be keyed into a lower hydraulic con-
ductivity layer. Taking take into consideration aquifer het-
erogeneity and to minimize groundwater mounding, a low
funnel to gate ratio is preferred. In order to assure com-
plete capture of the plume, the length of a funnel and gate
system is typically 1.2 to 2.5 times the plume width de-
pending on the funnel to gate ratio and the number of gates.
This system has a larger water flux per unit of reactive
material area. The funnel and gate configuration allows
for reactive material to be more easily replaced, however,
field experience indicates funnel and gates are more sus-
ceptible to hydraulic uncertainties which cause by pass of
flow around the system.
A slight adaptation of the typical funnel and gate is the in
situ reactive vessel, where the funnels or collection trench
collect the water and as a result of hydraulic head build up
on the upgradient side, causes the groundwater to flow
through the buried vessel. A GeoSiphon™ is a innovative
configuration developed by the Westinghouse Savannah
River Company which is designed to passively induce
contaminated groundwater flow through a permeable treat-
ment system at an accelerated rate via a siphon, taking
advantage of the natural hydraulic head difference between
two points. Once the siphon is established the passively
induced flow draws the contaminated groundwater towards
and through the treatment cell, treats the groundwater and
is subsequently discharged to a lower hydraulic head at
ground surface or surface water.
The theoretical amount of reactive material needed in a
PRB is independent of treatment system configuration,
since any configuration will have to treat the same mass
flux. An exception may be when a minimum amount of
reactive material is required due to construction limitations.
Construction Methods
The choice of construction method will depend on the depth
of installation, site geology and amount of reactive mate-
rial requiring placement. Recent papers discussing vari-
ous construction methods are given in Gavaskar, et al.
(1998), Day et al. (1999), MSB (1996) and Rummer and
Ryan (1995). Conventional excavation and replacement
methods are typically utilized at shallow depths (less than
35 ft) and are typically less expensive than those imple-
mented deeper into the subsurface.
Shallow Methods
Several construction methods are available to accommo-
date these various configurations at shallow depths. At
depths typically less than 35 ft, methods generally involve
18
-------
excavation and replacement procedures.The least expen-
sive trenching method would be backhoe trenching which
could be implemented if the formation soil does not cave
in. However, the limitation would be the excavation width.
A continuous trenching machine, which is currently lim-
ited to depths of less than 30 ft, allows for simultaneous
excavation and backfilling without an open trench. It al-
lows for very rapid installation, however, it has consider-
ably large equipment and slightly higher mobilization costs.
Several trenching contractors have installed these sys-
tems, with costs ranging from $200 to $400 per linear foot.
Another common installation method is cofferdam or sheet
pile excavation boxes which are formed and braced using
interlocking sheet piling. The sheet piling maintains the
dimensions of the treatment zone during excavation and
backfilling. After backfilling is complete, the sheet piling is
removed and the groundwater is allowed to flow through
the treatment zone. Cost for this installation average about
$80 ft2. Trench boxes, like sheet piling are used to main-
tain trench integrity during excavation and backfilling op-
erations and range about $10 to $20 per ft2. Auger holes
have also been utilized to install treatment zones. Rotat-
ing a continuous flight of hollow stem augers into the re-
quired depth, the reactive material can be placed through
the auger stem as the augers are removed. These treat-
ment zones can be created by overlapping holes or in
well arrays where two or more rows are required.
Deep Methods
For deep installations typically greater than 35 ft, excava-
tion and replacement can be-costly. Costs are often influ-
enced by the need to excavate to considerable depths
through uncontaminated soil before reaching the plume.
Caisson installation involves driving a large circular steel
caisson into the ground and augering our the native mate-
rial. The caisson is then backfilled with a reactive material
and removed. Overlapping or tangential caisson emplaced
treatment zones can be used to create a larger perme-
able treatment zone. However, the overlapping caissons
causes wastage of iron ranging from 10% to as high as
30%. Caisson implementation averages about $200 per
vertical foot.
A mandrel or H-beam is a hollow steel beam with a dis-
posable shoe at the leading edge which is driven into the
ground to create a thin continuous treatment zone. Once
the mandrel reaches the maximum depth of the treatment
zone the reactive material is placed inside the mandrel
and the disposable shoe is removed. This process is then
repeated creating a continuous zone of reactive material.
In previous applications, parallel treatment zones were
created to provide sufficient reactive material and to re-
duce the risk that the contaminants would not come in
contact with the reactive material. This installation method
ranges from $10 to $20 per ft2.
Another installation method that has been proposed at
several sites, but has not been used to date, is ground
freezing which has been implemented in the construction
industry for many years. This involves the use of refrigera-
tion to convert in situ pore water into ice. The ice acts as a
bonding agent which fuses together particles of soil to in-
crease the strength of the mass and make it impervious.
Excavation can be performed safely inside the barrier of
water-tight frozen earth with conventional excavation
equipment.
Deep Construction Methods Employing
Biodegradable Slurry
Other deep installation methods require that the reactive
materials be carried in a biodegradable slurry (bioslurry),
usually guar. These methods have been employed in the
construction industry for years, and are currently being
modified to implement PRBs along with various reactive
materials deep into the subsurface. As part of pre-con-
struction activities, tests to determine that the site water
chemistry is compatible with the reactive material and
bioslurry mixture and to assure that the bioslurry break-
downs over a suitable time period at groundwater tem-
peratures should be undertaken. Additional testing should
also be undertaken to determine the effectiveness of the
reactive material once it has been in contact with the
bioslurry.
A bioslurry trench installation is similar to constructing a
conventional impermeable slurry wall. As the trench is
excavated, bioslurry provides stability to the trench walls,
and the reactive material is placed via a tremmie tube into
the trench. There should be minimal contact between the
reactive material and bioslurry. The method of circulation
of the breaker/enzyme plays a critical role in assuring com-
plete breakdown of the bioslurry. Costs for implementing
this installation method range from $15 to $25 per ft2.
Vibrated beam technology has been used for years to in-
stall thin impermeable slurry walls and recently has been
adapted to inject reactive material and bioslurry. The large
I-beam is driven into the subsurface and as the beam is
vibrated out, a reactive material and bioslurry is pumped
into the formation filling the void created by the beam. This
process is then repeated and several lines in parallel could
provide the required amount of reactive material. This tech-
nique could cost on the order of $10 per ft2.
Deep soil mixing has been suggested where the reactive
material is mixed with biodegradable slurry and pumped
to the mixing augers while they are being advanced slowly
through the soil. Over time the bioslurry breakdowns al-
lowing the groundwater to flow through the reactive mate-
rial and aquifer mixture or treatment zone. High costs are
associated with mobilization and demobilization for deep
soil mixing. Costs typical range from $75 to $120 per cu-
bic yard.
Two other deep installation methods that have been suc-
cessfully demonstrated in the field are jetting and vertical
hydrofracturing. Jetting uses high pressure to inject fine
19
-------
grained reactive material into the natural aquifer forma-
tion. The jetting tool is advanced into the formation to the
desired depth, then the reactive material and bioslurry is
injected through the nozzles as the tool is withdrawn. Ei-
ther columnar zones or diaphragm walls can be created.
Jetting may cost on the order $75 per vertical foot for co-
lumnar walls. Vertical hydrofracturing uses a specialized
tool to orient a vertical fracture and initiate the fracture
process.The tool is placed to the desired depth through a
borehole and the interval for fracturing isolated by pack-
ers. The reactive material and bioslurry in then pumped
under low pressures into the formation to form a thin ver-
tical plume along the line of the induced vertical fracture.
Construction Uncertainties
Many unforeseen conditions have been encountered dur-
ing PRB construction. These issues can pertain to unex-
pected "complexities" in geological formation or complica-
tions with the construction technique itself. These unan-
ticipated issues have included:
• compaction or densif ication of the aquifer material dur-
ing the vibration of equipment such as caisson or sheet
pile;
• compaction effects on the aquifer soils when using
mandrel, I-beam, jetting and vertical hydrofracturing;
• flowing sands with continuous trenching requiring
additional passes of the trencher;
• slurry wall flows into the excavated void created for
the reactive material;
• smearing of clay and sand zones during gate con-
struction, resulting in inadequate flow through the treat-
ment zone;
• slow breakdown of bioslurry used in jetting, deep soil
mixing, vertical hydrofacturing and bioslurry trench;
• sufficient flushing of the bioslurry to prevent high bio-
logical activity zone downgradient;
• insufficient reactive material injection during jetting.
These problems are addressed during PRB construction,
on a site specific basis, but have added time and costs to
system installation. However, these "lessons learned" have
helped improve the cost effectiveness of subsequent PRB
installations, making the technology an even more attrac-
tive remedial alternative.
References
Gavaskar, A.R., Gupta, N., Sass, B.M., Janosy, R.J. and
O'Sullivan, D., 1998. Permeable barriers for groundwater
remediation: Design, construction and monitoring. Batielle
Press, Columbus, OH, pp. 176.
Day, S.R, O'Hannesin, S.F. and Marsden, L, 1999.
Geotechincal techniques for the construction of reactive
barriers. Journal of Hazardous Materials (in press).
MSETechnology Applications, Inc., 1996. Analysis of tech-
nologies for the emplacement and performance assess-
ment of subsurface reactive barriers for DNAPL contain-
ment. MSE Technology Applications, Inc., Butte, MT, pp.
105.
Rummer, R.R.and Ryan, M.E., 1995. Barrier containment
technologies for environmental remediation applications.
John Wiley & Sons, Inc., New York, pp. 170.
20
-------
Mechanisms in In S/fuThermal Remediation
Eva L. Davis
USEPA
PO Box 1198
Ada, OK 748121
580-436-8548
davis.eva @ epa.gov
Many of the physical properties of organic contaminants,
as well as the properties of porous media, limit the recov-
ery of contaminants through techniques such as pump
and treat or soil vapor extraction. Remediation techniques
that rely on the addition of chemicals surfer from the lack
of the ability to mix the chemicals with the contaminants
in the subsurface, and from limitations on recovering the
chemicals. The physical properties of fluids and their in-
teractions with soil will control their movement in the sub-
surface, as well as their recovery. All of these properties
are temperature-dependent. In addition, various types of
liquid and vapor movement can be created in porous me-
dia due to temperature gradients.
Understanding these temperature effects and their rela-
tive magnitudes is important in understanding how in situ
thermal remediation processes can be used to enhance
the recovery of organic contaminants. For volatile organic
compounds, which includes fuels such as gasoline and
diesel and chlorinated solvents, the main mechanism for
enhanced recovery is the increased vapor pressure with
temperature which allows these compounds to be recov-
ered in the vapor phase. Enhanced solubility and de-
creased adsorption at higher temperatures will also con-
tribute to the recovery of these chemicals.
The recovery of semivolatile and (essentially) nonvolatile
oils can be enhanced at higher temperatures through vis-
cosity reduction, increased relative permeability, and de-
creased capillary forces.These changes in physical prop-
erties allow oils to be displaced in the liquid phase when
there saturation is greater than residual saturation, and
also tend to decrease the amount of residual saturation.
Field experience demonstrates that even semivolatile or-
ganic compounds can be distilled at significant rates when
high temperatures and low vapor phase pressures condi-
tions are created in the subsurface.
Ongoing research shows that naturally existing soil mi-
crobes can survive temperatures of 100°C, and poten-
tially can degrade the small amount of contaminants that
remain after a thermal remediation. Also thermophysical
reactions have been found to partially or fully oxidize some
organic contaminants. These biological and
thermophysical process can make contaminants easier
to recover, or destroy them in situ, and contribute to the
cost effectiveness of thermal remediation processes.
21
-------
Remediation of Soils and Groundwater Using Steam Enhanced Extraction
Kent S. Udell
University of California
6179 Etcheverra
Berkeley, CA 94720
510-642-2928
udell@euler.berkeley.edu
Soil and groundwater contamination by non-aqueous
phase chemicals is a problem of international importance.
The concern over these chemicals is due, in part, to their
potential to contaminate groundwater for centuries to come.
Decades of work in the area of in situ remediation has
resulted in few successes. But in situ thermal techniques
have been gaining popularity, both in the U. S. A. and in-
ternationally, due to their robustness. Part of the success
of such techniques is from the fact that once the soil-wa-
ter-NAPL system is heated to the boiling point of water,
equilibrium thermodynamics requires complete NAPL va-
porization. Thus, if coupled with active soil vapor extrac-
tion, the removal rate of the liquid chemical is determined
by the relatively fast heat transfer rate, not the slower mass
transfer rate. In addition, concentrations of chemicals that
tend to partition into the vapor phase can be easily de-
creased to drinking water standards through depressur-
ization, electrical heating or radio frequency heating if the
soils are at the water boiling point.
These mechanisms, plus enhanced NAPL mobility with
increasing temperature, can be easily be exploited by
steam enhanced extraction (SEE). In SEE, steam is in-
jected into the subsurface to heat the target zone of con-
tamination to the boiling point of water while actively pump-
ing water, NAPL and vapors from other wells. This is con-
tinued until the zone is sufficiently heated. Thereafter, the
steam injection pressure and extraction well vacuum are
varied to induce pressure changes in the heated zone.
During periods of decreasing pressure (depressurization),
the temperature decreases to maintain thermodynamic
equilibrium between the steam and the water in the soil,
giving up energy to drive the vaporization of interstitial
water. This in situ boiling occurs in a manner relatively
independent of the permeability thus allowing convective
fluxes from lower permeability regions. Thus, mass trans-
fer limitations are relaxed and recovery rates are increased.
After the pressures in the soil have reached nearly steady
values, the pressures are increased to recharge the soil
with energy through the condensation of steam. This pres-
sure cycling continues until the recovery rates and ex-
tracted fluid concentrations decrease to low values.
Two field examples of this process are briefly discussed.
The first is the Gas Pad at Lawrence Livermore National
Laboratory. The subsurface geology was of alluvial ori-
gin, with layers of high permeability interspersed with clays
and clayey silts. The horizontal correlation of the high per-
meability layers was moderate except for a thick gravel
layer found below the water table from about 35 m (140
ft.) to 42 m (150 ft). This gravel layer was found in each
well surveyed. The gasoline was detected as a second
phase in an area about 50 m (150 ft.) in radius. The verti-
cal distribution ranged from 17 m (50 ft) to about 45 m
(137 ft.), with significant spreading due to major ground-
water elevation fluctuations. Indeed, separate phase gaso-
line was found in the deep water-bearing zone nearly 8 m
below the water table. This situation rendered others tech-
nologies such as ambient temperature vacuum extrac-
tion to be impracticable, and groundwater extraction to
be time-prohibitive. There were two major flow zones at
this site: the deep gravel layer and an upper unsaturated
zone consisting of intermingled sands, silts, clays and
gravels. A contaminated clay layer of variable thickness
separated the two permeable zones.
Six injection wells were placed around the perimeter of
the area contaminated by separate phase gasoline; three
recovery wells were installed near the center of the plume.
The injection wells were constructed to allow separate
steam injection into the upper and/or lower permeable
regions. The recovery wells were completed over the en-
tire height of the contaminated zone. To monitor the sub-
surface temperatures, 11 temperature observation wells
were placed throughout the site.
The above ground equipment included a boiler fired by
natural gas, a flat-plate condenser cooled by water recir-
culated though a cooling tower, cooling heat-exchanger
for the liquids pumped from the subsurface, separation
tanks, water treatment facilities, and air treatment. Auto-
motive internal combustion engines were used to oxidize
the recovered hydrocarbons in the air exiting the con-
denser and to serve the purpose of the vacuum pump.
22
-------
During the two months preceding steam injection, 3-phase
electrical heating was conducted during off-hours. A total
of 202 Mw-hrs of electrical energy was dissipated in pre-
dominantly clay-rich zones.
During the first 37 days of operation, steam was injected
first into the lower gravel layer below the water table for
two weeks.The total electrical energy input was surpassed
with steam after the first three days of injection. Steam
broke through at the recovery wells after about 10 days of
injection. The steam was injected on a variable schedule
thereafter into the upper zone and/or lower zone to target
lower temperature regions. At the end of 37 days, opera-
tions ceased for 75 days. Steam injection resumed for a
second pass after 9 days of groundwater extraction and
vacuum pumping.The steam injection schedule was peri-
odic with periods of steam pressurization and de-pres-
surization. Steam was injected a total of 21 days during
this 46 days of gas and groundwater extraction.
The recovery rates observed during the second pass were
highest before steam injection began as the extraction
system recovered high-concentration gases evolved dur-
ing 45 days of down-time conductive heating. Thereafter,
periodic peaks in the recovery rates were observed cor-
responding to times when steam injection ceased and
depressurization began. The 2nd pass was terminated due
to permitting restrictions on the boiler while recovery rates
remained high and may have continued with additional
cycles of steam injection. During the next 80 days, soil
samples were collected, the steam boiler was taken from
the site, and additional electrical heating wells were in-
stalled. The treatment system was shut down during this
period.
The soil concentrations found after the second pass
showed that, in general, gasoline was effectively removed
from the hot steam-bearing zones and concentrations were
significantly reduced in the low permeability zone sepa-
rating the upper and lower permeable units. An estimate
of 2840 liters (750 gallons) of gasoline remained in the
soil after the end of the second pass based on the soil
concentrations after the second pass.
The recovery rates during the final phase were significant
at the beginning of pumping, but the rates fell with time,
with magnitudes much less than those observed during
the second pass. Some improvement in the recovery rates
were achieved by converting upper zone injection wells
into vapor recovery well. However, little could be done to
increase recovery rates during the final few days of op-
eration - including electrical heating. Regardless, an addi-
tional 3800 liters (1000 gallons) of gasoline were recov-
ered in this final phase before the recovery rates dropped
to small values.
Subsequent gas and water sampling is showing decreas-
ing concentrations of gasoline components with time with-
out additional pumping, implying that there is no separate
phase gasoline remaining at this site. Ethybenzene, tolu-
ene, and xylene concentrations in the water have dropped
to below regulation limits. Benzene concentrations con-
tinue to decrease. Hydrocarbon-degrading biological ac-
tivity was found in the zones subjected to steam tempera-
tures, indicating that the application of steam did not leave
the site sterile. The culture make-up however, has been
dramatically altered. No further treatment at that site was
required.
The second demonstration of interest is Site 5 at Alameda
Point (formerly Naval Air Station Alameda) in California.
This site is the location of an underground waste solvent
tank that has leaked a mixture of solvents into fill soils
only 3 m. (10 feet) below the ground surface. This site
presented special challenges for SEE due to the shallow
depth and complex NAPL chemistry.
The aboveground equipment included a boiler fired by
propane, an air-cooled condenser, a cooling tower pro-
viding the cooling for the liquids pumped from the subsur-
face, separation tanks, and carbon for water treatment and
air treatment. Sealed ring vacuum pump placed down-
stream of the air-cooled condenser provided up to 20
inches of Hg vacuum at the well.
Six injection wells were placed around the perimeter of
the area contaminated by nonaqueous phase solvents;
one recovery well was installed near the center of the
plume.The injection wells were constructed to allow sepa-
rate steam injection into the upper vadose zone and/or
lower saturated regions. The recovery well was completed
over the entire height of the contaminated zone. Two hun-
dred and seventeen theromocouples were placed through-
out the site to monitor subsurface temperature distribu-
tions.
Steam was injected for 70 days during this demonstra-
tion. After temperatures began to increase near the ex-
traction well, separate phase solvents began to flow from
the subsurface. After a period of about 15 days of sepa-
rate phase solvent pumping, the separate phase flow
stopped and recovery was dominated by vaporization. After
18 more days of steady flow, vaporization rates dropped
to low values and cyclic steam injection began. Recovery
rates increased again to high values during the first cycle,
then decreased sequentially with each additional cycle.
Nearly 2000 liters (528 gallons) of NAPL were collected in
drums and sent out for disposal during the demonstration.
TheTCE was removed primarily in the vapor phase (-200
kg), with significant mass also measured in the extracted
water (22 kg) and NAPL (18 kg).
Near the end of the demonstration, TCE and other sol-
vents concentrations in the extracted water decreased to
background values. Salt concentrations and oxygen 18
isotopic values in the extracted water also indicated that
the water came from outside of the demonstration area.
Since the goal of this demonstration was source removal
to support natural attenuation as the final remedy, the
demonstration was considered to be highly successful.
23
-------
These projects and others currently underway have shown
the effectiveness of SEE to remove NAPLs from soils and
their soluble fractions from groundwater. The scientific
basis of the process is quite well-defined and the in situ
application is surprisingly robust. The keys to successful
implementation were found to be appropriate monitoring
to control the locations and rates of steam injection, and
planning for the various phases of recovery including start-
up, high NAPL rates, high vaporization rates and cyclic
operation. This planning includes modeling to ensure
proper sizing of all treatment equipment, design for flex-
ibility of operation, and measurement locations to provide
confidence in monitored variables.
24
-------
Field Demonstrations of Thermally Enhanced Extraction
Lloyd D. Stewart
Praxis Environmental Technologies
1440 Rollins Rd.
Burlingame, CA 94010
650-548-9288
praxis @ praxis-enviro.com
Thermally Enhanced Extraction (TEE) is an innovative
combination of groundwater extraction, soil vapor extrac-
tion, steam injection, hot air injection, and control tech-
niques for accelerated and complete remediation. Results
from two field demonstrations of TEE are described. The
two TEE applications were removal of a dense nonaque-
ous phase liquid (DNAPL) pooled atop an aquitard in the
saturated zone and stripping of volatile compounds from
a residual hydrocarbon NAPL in a smear zone across the
water table.
The DNAPL site contains two former trenches used to
dispose of residuals from a solvent recovery unit. The dis-
posed liquids seeped through the uppermost aquifer and
accumulated in a geologic channel forming a pool in the
saturated zone about 50 feet below the surface. The
DNAPL consists of over 70% trichloroethene. Prior to the
TEE demonstration, roughly 26,200 gallons of DNAPL had
been recovered with a pump-and-treat system.
The TEE process included eleven days of thermal injec-
tion followed by twelve days of ambient air injection.
Groundwater extraction and soil vapor extraction were
performed during all phases. A bank of DNAPL preceded
the thermal front and began to appear in extraction wells
after one day of injection. Thermal breakthrough occurred
in the extraction wells after three days of injection. Over
500 gallons of DNAPL were recovered from the demon-
stration area during the first five days of TEE. The TEE
process removed an estimated total of 908 gallons of
DNAPL in a 3-week period.
The EPA Superfund Innovative Technology Evaluation
Program collected and analyzed soil samples from 14 lo-
cations immediately after the demonstration. Post-dem-
onstration analyses of 42 soil samples revealed the aver-
age contaminant concentrations in the channel soil were
reduced by 96%. The soil preliminary remediation goals
were met in all 23 samples collected from the gravel chan-
nel and 13 of 19 samples collected from the underlying
sand lens. This brief period of TEE treatment reduced the
average soil concentrations in the underlying aquitard over
50%. This demonstration was the first field application of
steam injection to remediate DNAPL in the saturated zone.
A second demonstration sponsored by the Strategic En-
vironmental Research and Development Program
(SERDP) was performed to increase understanding of
steam injection and evaluate its effectiveness for cleanup
of hydrocarbon NAPL. The field test was performed in a
10-by-15-foot cell enclosed by steel sheet-pile walls ex-
tending into a low permeability clay layer 30 feet below
ground surface. Constituents of the NAPL were primarily
weathered fuels and oils and chlorinated solvents.
Results of the steam injection test revealed very effective
distillation of volatile contaminants from the residual NAPL.
Compounds of concern showed reductions in soil con-
centrations greater than 95% in steam-swept layers. The
final groundwater concentrations in the test cell were close
to cleanup levels with only two target compounds exceed-
ing the drinking water standard. As predicted by theory,
the bulk NAPL was not effectively mobilized by the steam
because of its high viscosity and low residual saturation;
yet the low residual saturation allowed good contact be-
tween the steam and NAPL for distillation. Extracted va-
por concentrations were well below saturated values for
the elevated temperature suggesting the presence of wa-
ter in the system inhibited distillation.
25
-------
In Situ DNAPL Remediation Using Six-Phase Heating™
Brett Trowbridge
Current Environmental Solutions
PO Box 50387
Bellevue.WA 98008
949-388-5808
www.cesiweb.com
The in situ cleanup of dense nonaqueous phase liquids
(DNAPL) remains one of the remediation industry's tough-
est challenges. Standard remediation technologies require
years of continued application to produce even marginal
results. A new technology, Six-Phase Heatinga (SPH), has
successfully remediated DNAPLs in demonstrations for
the Environics Directorate of the US Air Force at the Do-
ver AFB and for the Army Corps of Engineers at Fort
Richardson. SPH has now been successfully implemented
in full-scale remediations at several commercial and DOD
sites including a recently completed large DNAPL site that
is ready for site closure.
SPH is a polyphase electrical technology that uses in situ
electrical resistive heating and steam stripping to achieve
subsurface remediation. Developed by Battelle Memorial
Institute at the Pacific Northwest National Laboratories
(PNNL) for the US Department of Energy to enhance the
removal of volatile contaminants from low-permeability
soils, SPH has subsequently proven capable of
remediating DNAPLs from saturated zones. SPH is now
commercially available through Current Environmental
Solutions (CES).
SPH transformers convert regular three-phase electricity
into six separate phases.These electrical phases are then
delivered throughout the treatment volume by steel pipe
electrodes that are inserted using standard drilling tech-
niques. SPH can quickly increase subsurface tempera-
tures to the boiling point of water and is equally effective
in all soil types, including fractured rock, under both va-
dose and saturated conditions.
The technology was originally developed for the treatment
of low permeability soils. Since the original technology
development, the treatment applications have been sig-
nificantly expanded. SPH is now commercially available
for the following treatment applications:
• Low permeability soils
• DNAPL and LNAPL Remediation
• Cold Regions Treatment
• Heavy Hydrocarbon Mobilization
• Rapid Clean Up
• Enhanced Biological Remediation
• Aquifer Heating
This presentation will also cover two case studies of the
application of the SPH technology. The first case study
will cover the successful remediation of a DNAPL plume
under a former manufacturing building.
The second case study will cover the implementation of
the SPH technology for the rapid remediation of a site with
a pending property transfer." The site was a former dry
cleaner and was contaminated with PCE and daughter
products. The site was cleaned up to MCLs in 90 days of
SPH operations.
26
-------
RF Heating Technology for Soil Remediation
Stephen L. Price and Raymond S. KasevJch
KAI Technologies
199 Constitution Ave, Bldg 1
Portsmouth, NH 03801
603-431-2266
price @ xdd-llc.com
Results from two sites, Ft. Wainwright, Alaska and an ac-
tive retail petroleum facility in the Midwest demonstrate
two applications of radio frequency (RF) antenna heating
for safely accelerating site closures. At Ft. Wainwright lo-
cated in a cold climate, preliminary results suggest that
RF enhanced bioremediation during one year of an ongo-
ing pilot test. The combination of RF with soil vapor ex-
traction (SVE) at a Midwest United States retail petroleum
facility and convenience store resulted in a six-month in-
stead of a two to three year cleanup using only SVE.
Introduction
Remediation of soil and groundwater by established tech-
nologies is often costly and lengthy. RF heating applied by
antennas offers a safe, cost-effective complement for a
wide variety of remedial technologies. RF has application
for accelerating remediation at sites contaminated with a
variety of volatile and semi-volatile compounds. The RF
heating process is not limited to 100°C temperatures.
RF heating for soil remediation by antenna technology
brings controlled heating to the subsurface, increasing the
rate of removal of contaminants from the soil.The antenna
delivery system permits RF energy to be safely focused
where needed, steered and directed (employing the inter-
actions of multiple antennas), and delivered at the required
depth. The antenna applicators may also be deployed with
flexibility because casings can be oriented vertically, hori-
zontally, or on an angle. RF delivered via antenna applica-
tors is inherently safe because any high voltage points
are shielded by the transmission line, soil, metal ground
plane (if necessary), or the RF generating equipment.
Periodic monitoring permits unrestricted site access by
ensuring that above ground electromagnetic field levels
are below OSHA (Occupational Safety and Health Admin-
istration) levels. Operation is at one of the ISM (Industrial,
Scientific, and Medicine) frequencies allotted by the FCC
(Federal Communications Commission). RF of course
does not remove contaminants; it is a complement to other
remedial technologies that increases their removal rates.
Heating causes changes in the physical, chemical, and
biological properties of soils, groundwater, and contami-
nants, making them more amenable to remediation efforts,
thereby reducing remedial time frames and costs. RF has
been applied with SVE, groundwater aeration (air
sparging), product recovery, and bioremediation.
Two uses of RF are reviewed: the RF enhanced
bioremediation pilot test at Ft. Wainwright, AK and the
commercial cleanup in the Midwest United States at an
active retail petroleum facility.
RF Heating Pilot Test at Fort Wainwright,
Alaska
Cold soil temperatures limit the achievable biodegrada-
tion and volatilization rates of contaminants in soils in
Alaska's interior. RF energy is being used at Ft. Wainwright
to enhance two remedial technologies: bioremediation
through bioventing and SVE during an extended pilot test
of over a year in duration.
The pilot test is being conducted near the former fuel
pump islands, where high concentrations of gasoline and
diesel range organic compounds were measured from
about 3 m below surface to several feet below the water
table. Thus far, bioventing has been implemented for over
a year, in which, a 12.2-m by 12.2-m by 3-m thick volume
of soil was heated and maintained at a target temperature
range of between 15°C and 35°C. An identical unheated
(control) target area is located about 36.6 m from the RF
heated area for comparison and analyses purposes. The
ambient temperature of these soils is less than 6°C dur-
ing the entire year.
The commercial RF heating system designed and de-
ployed for heating the vadose zone consists of four inde-
pendent 2.4-kilowatt (kW) continuous wave solid-state RF
generators. The generators operate at 27.12 MHz. Each
RF generator supplies energy via 2.2-cm transmission line
to a 3-m long, 8.9-cm diameter antenna. The antennas
are located in 15.7-cm diameter vertical fiberglass cas-
ings.The four antennas are separated by 6.1 -m, deployed
27
-------
in a square formation at wells RF#1, RF#2, RF#3, and
RF#4, and are centered in the vadose zone at 3.3 m BG.
Preliminary results from the ongoing pilot test at a Fort
Wainwright source area suggest that biodegradation rates
can be improved substantially by RF heating. From May
19,1998, the beginning of focused vadose zone heating,
through March 11,1999,4,439 kWh were supplied by the
RF heating system to raise the source area to the target
temperature range of 15°C to 35°C. After heating the soils
an additional 35,453 additional kilowatt-hours were re-
quired to maintain the within the target temperature range.
The input of this RF energy has resulted in optimal soil
temperatures for bioremediation. Biodegradation rates
observed between the RF heated and unheated control
areas increased by a factor of 6.
The RF system has operated remotely and essentially
unattended since system startup and debugging. The
biophase of the pilot test is ongoing and will evaluate the
effect of adding nutrients to the soil on the biodegradation
rate starting this summer. Column studies conducted at
the University of Alaska in Anchorage on site soil suggest
that biodegradation rates increase substantially when nu-
trients are added to RF heated soil. Following the biophase
an 2-month long SVE phase will commence where the
soils will be heated and maintained at70°Cto 90°C.When
the pilot test is complete, a value engineering analysis will
determine the potential for enhanced biodegradation and
SVE as cost-effective remediation strategies at the site.
Commercial Application of RF Enhanced
SVE at a Retail Petroleum Facility
An active Midwestern United States retail petroleum facil-
ity and convenience store is a site where RF antenna heat-
ing technology was applied to enhance SVE and achieve
closure in substantially less time than conventional SVE.
Contamination at this site was due to leaking underground
storage tanks, pumps, and pipes and because of this, a
dissolved phase plume was located approximately 9 m
below grade. The site soils are glacial deposits consisting
of coarse sands. In six months, from January through June
of 1998, Dahl & Associates, a commercial licensee of KAI's
RF heating technology accomplished what normally would
have taken two to three years to achieve at ambient soil
temperatures; site closure.
The RF system consisted of a trailer mounted computer
controlled remotely operated 10-kilowatt 27.12 MHz RF
gene-ator and a 3 m long antenna applicator. The applica-
tor was designed to heat vadose or capillary zone soils,
but it was also used to heat the capillary zone by adjust-
ing its length. Two at grade vertical wells were used to
deploy the antenna applicator. The antenna applicator was
raised and lowered in the wells so that it could focus RF
energy into the capillary fringe or vadose zone during the
course of the remediation. In addition to the RF wells there
were seven air extraction wells and three air sparging wells
at grade level. All transmission line and piping for the SVE
and air sparging system were underground. A trailer
mounted thermal oxidizer was used to destroy the organic
compounds as they were extracted.
The successful cleanup of the retail petroleum facility
marked the first commercial application of RF antenna
heating with SVE and air sparging. During six months of
operation the RF system raised soil temperature by ap-
proximately 30°C at 3 m from the RF wells and achieved
temperatures in excess of 100°C at the RF wells.The SVE
system during this same period recovered 2,305 kg (853
gallons) of gasoline range organics (GRO), an amount
sufficient to achieve site closure. Based on its experience
with similar sites in Minnesota, DAHL estimates it would
normally take two to three years of SVE operation to close
the site. This is over four times longer than when RF was
applied with SVE. This result is also supported by the
USEPA's Hyperventilate software which calculated a fac-
tor of four increase in the recovery rate of GRO when soil
temperatures are maintained at an average 37.8°C (see
Figure 1 below)
100
a
•a
w
3
13
-------
Conclusion
KAI's RF antenna heating technology has been shown to
significantly improve the performance of bioremediation
and SVE thereby reducing time on site and the cost of
remediation. Preliminary results from the ongoing pilot test
and column studies at Ft. Wainwright, AK indicate biodeg-
radation rate increases of two to ten fold are possible with
the application of RF and the addition of nutrients. During
its first commercial application, at a Midwest United States
retail petroleum facility, RF antenna heating increased SVE
off gas contaminant concentrations substantially and
helped complete the cleanup of the site in six months in-
stead of the usual two to three years required at ambient
soil temperatures. RF antenna heating has another ben-
efit an ability to co-exist safely with other site usage. This
was most clearly demonstrated during the remediation of
the retail petroleum facility and convenience store.
References
Hyperventilate V2.0 for Widows/Plus, USEPA ID#
IS120016, Dec. 12,1996
Laboratory Column Study Final Report West
Quartermaster's Fueling System, Ft. Wainwright, AK, Pre-
pared for Dept. of the Army, U.S. Army Engineer District,
AK under Contract DACA85-95-D-0015 by CH2MHHI,
University of Alaska School of Engineering. Anchorage,
AK, and Geosphere, Inc., Jan. 1999
Marley, M.C., Price, S.L., Kasevich, R.S., PariklrJ.M.,
Droste, E.X., Acomb, L, Fosbrook,C., Wallace, W., Horn,
R., "Radio Frequency Heating System for Enhanced
Bioremediation: Pilot Test Results, Fort Wainwright,
Alaska," ASCE 10" Annual Conference on Cold Regions
Engineering, August 16-19,1999.
29
-------
Thermal Conduction Heating for In Situ Thermal Desorption
George L. Stegemeier
GLS Engineering
5819 Queensloch Dr.
Houston, TX 77096
713-245-7785
gstegmeier® shell.com
Harold J. Vinegar
Shell Exploration and Production Technical Co.
In situ Thermal Desorption is a soil remediation process
in which heat and vacuum are applied simultaneously to
subsurface soils, either with surface heater blankets or
with an array of vertical heater/vacuum wells. Near the
high temperature (~900°C) heat sources, radiation heat
transport dominates; however, at greater distances into
the soil, thermal conduction accounts for the majority of
heat flow.
As soil is heated, volatile contaminants in the soil are va-
porized by a number of mechanisms, including: (1) evapo-
ration into the air stream, (2) steam distillation into the
water vapor stream, (3) boiling, (4) oxidation, and (5) py-
rolysis. The vaporized water, contaminants, and natural
organic compounds, are drawn by the vacuum in a counter-
current direction to the heat flow into the vacuum source
at the blankets or wells.
Compared to fluid injection processes, conductive heat
injection is very uniform in its vertical and horizontal move-
ment. Furthermore, transport of the vaporized contami-
nants is improved by the creation of permeability, which
results from drying and shrinking of the soil. Flow paths
are created even in tight silt and clay layers, allowing es-
cape and capture of the vaporized contaminants.The com-
bined effectiveness of both heat and vapor flow yields
nearly 100% sweep efficiency.
In addition, very high displacement efficiency (approach-
ing 100%) can be reached in the heated soil. This occurs
because the soil can be heated to high temperatures
(>500°C) for prolonged times (many days). Field project
experience has confirmed that a combination of high boil-
ing temperature and long times result in extremely high
overall removal efficiency of even the high boiling point
contaminants from the soil.
In practice, most of the contaminants are destroyed in the
soil before reaching the surface. Contaminants that have
not been destroyed in situ are removed from the produced
vapor stream at the surface with an air pollution control
(ARC) system.
30
-------
In Situ Destruction of Chlorinated Hydrocarbon Compounds in a
Reactive Well Using Pd-Catalysts
Martin Reinhard
Department of Civil Engineering
Stanford University
Terman Engineering Center 5-5
Santa Theresa/Morris Way
Stanford, CA 94305
650-0723-0308
reinhard @ ce.stanford.edu
Walt W. McNab, Jr. and Roberto Ruiz
Environmental Restoration Division
Lawrence Livermore National Laboratory
Livermore, CA 94550
Chlorinated solvents such as trichloroethylene (TCE),
perchloroethylene (PCE), carbon tetrachloride, and chlo-
roform are among the most commonly encountered
groundwater contaminants. Conventional methods to
remediate chlorinated solvent sites are expensive and thus
there is a significant need for more cost-effective tech-
nologies. Many chlorinated hydrocarbon compounds are
rapidly dechlorinated (within minutes) in the presence of
dissolved H2 and small quantities of Pd catalysts
(Kovenklioglu, etal., 1992, Schreierand Reinhard, 1995;
Siantar et al., 1996, Schuth and Reinhard, 1998, Lowry
and Reinhard, 1999) without the production of intermedi-
ate transformation products often observed in biological
systems. The rapid dechlorination rates permit the design
of reactors that can be placed within the well bore of a
treatment well for in situ groundwater remediation. This
technology can be applied in aerobic aquifers, even though
the process is reductive, because dechlorination of chlo-
rinated ethylenes has been observed to be faster than the
reduction of dissolved O2.
Here we report on a prototype reactor for well bore treat-
ment of chlorinated hydrocarbons that uses the H2/Pd
approach.The particular location at the LLNL site required
a novel treatment approach because tritium co-contami-
nation of the aquifer complicated the task of treating
groundwater at the surface with conventional methods.The
site is underlain by unconsolidated alluvial sediments, with
the depth to groundwater approximately 30 m below the
surface.The aerobic groundwater environment precluded
natural reductive dehalogenation as a means of
remediation, while the depth to groundwater rendered the
placement of a permeable iron wall impractical. Given these
constraints, the decision was made to test the remediation
of the chlorinated hydrocarbons by an hyPd reactor placed
within a flow-through treatment well screened across two
permeable sandy zones, separated by an aquitard. Both
zones are characterized by similar contaminant chemical
profiles: TCE, PCE, 1,1-DCE, chloroform, carbon tetra-
chloride, and 1,2-DCA, with a total contaminant concen-
tration of 4 to 5 mg/L. Tritium was also present in both
zones at a concentration of approximately 8,000
picocuries/L.
The reactor design is shown on Figure 1. Contaminated
influent water is drawn into the well from the lower sand,
hydrogenated, treated for chlorinated hydrocarbons by
exposure to the Pd catalyst, and then discharged into the
upper sand where tritium self-remediates through radio-
active decay. The reactor has operated on a continuous
basis for almost one year. The daily operating schedule
has entailed 10 hours of operation, with flow reversal in
the catalyst columns after 5 hours, followed by catalyst
regeneration by flushing the reactor with aerobic water
and subsequent overnight aeration.
Performance data, which have remained essentially con-
stant over the operating history, are summarized on Table
1. Average contaminant removal is better than 98% ex-
cept for chloroform (91 %) and 1,2-DCA. The inability of
the process to remove 1,2-DCA was as expected based
on laboratory results (Lowry and Reinhard, 1999). Cost
estimates suggest that a one year catalyst life-time is
needed for the H./Pd technology to be competitive with
granulated activated carbon (GAG) adsorption, not con-
sidering intangible benefits such as a small footprint and
in situ destruction.
31
-------
Tablo 1. Ranges of contaminant concentrations (mg/L) measured at sampling points within the treatment system.
Compound
TOE
PCE
Chloroform
1,1-DCE
1,2-DCA
Carbon tetrachloride
C/S-1.2-DCE
Vinyl chloride
Influent
3612-3777
366-370
167-235
130-180
26-28
18-21
0.6-0.7
<0.4
1st Catalyst
column effluent
<0.4-0.9
<0.4
11-38
<0.4
20-26
<0.4
<0.4-1
<0.4
2nd Catalyst
column effluent
<0.4-0.8
<0.4
6.5-36
<0.4
19-27
<0.4
<0.4-0.9
<0.4
System removal
efficiency
>99%
>99%
-91%
>99%
~0
>98%
-
-
References
Kovenklioglu, S., Z. Cao, D. Shah, R. J. Farrauto, and E. N.
Balko, JAIChE, 38(7), 1003 (1992).
Lowry G.V. and M. Reinhard, Environ. Sci. Technol., 33(11),
1905, (1999).
McNab, W.W., Jr., and R. Ruiz, Chemosphere, 37(5), 925
(1998).
Schreier, C.G., and M. Reinhard, Chemosphere, 31 (6),
3475 (1995).Schuth, C., and M. Reinhard, Appl. Cat. B,
18(3-4), 215 (1998).
Siantar, D.R, C.G. Schreier, C.S. Chou, and M. Reinhard,
Water Research, 30(10), 2315 (1996).
Acknowledgments
Funding was provided by the Deutsche
Forschungsgemeinschaft (DFG), by the Office of Research
and Development, USEPA, under agreement R-815738-
01 through the Western Region Hazardous Substance
Research Center and R-825421. Work also performed
under the auspices of the U.S. Department of Energy by
Lawrence Livermore National Laboratory under Contract
W-7405-Eng-48.
8-inch well bore
Packed bed
catalyst
columns
Pneumatic packer
Pump
Contaminated influent
Figure 1. H/Pd Reactor
T.D. = 140 ft
32
-------
Case Studies of In Situ Chemical
Oxidation using the Geo-Cleanse® Process
J. Daniel Bryant and James T. Wilson
Geo-Cleanse International, Inc.
4 Mark Road, Suite C
Keniworth, New Jersey 07033
908-206-1250; (fax) 908-206-1251
geocleanse@earthlink.net
www.geocleanse.com
Introduction
The Geo-Cleanse Process is a patented, in situ chemical
oxidation technology to deliver hydrogen peroxide and fer-
rous iron catalyst (Fenton's reagent) to the subsurface.
The Geo-Cleanse Process is a pressurized system ca-
pable of oxidizing many organic contaminants in both soil
and groundwater. The Geo-Cleanse Process has in the
past six years been applied at over 50 sites in 17 states
and in Canada, to address common organic contaminants
ranging from petroleum hydrocarbons (BTEX), chlorinated
solvents (PCE andTCE) including DNAPLs, nitroaromatic
compounds (RDX, TNT, DNT), and manufactured gas plant
PAH contamination. Cleanup objectives as low as non-
detectable levels (<1 mg/L) in groundwater.
The purpose of this presentation is to briefly review sev-
eral recent case studies of Geo-Cleanse Process appli-
cations at several Department of Defense and private in-
dustrial facilities. Seven sites, all treated within the past
year, are described. Four of the seven sites are impacted
by chlorinated aliphatic solvents (e.g., PCE), one is im-
pacted by chlorinatedaromatic solvents (trichlorobenzene),
one is impacted by jet fuel, and one is a former manufac-
tured gas plant impacted by extremely high levels of poly-
cyclic aromatic hydrocarbons (PAHs). These case stud-
ies represent a wide range of geological and
hydrogeological conditions and a variety of organic con-
taminants. Two of the sites are currently in treatment, and
thus only the pre-treatment site conditions and the work
plan are summarized.
Kings Bay Naval Submarine Base, Georgia
PCE and its natural degradation products (TCE, cis-1,2-
DCE, vinyl chloride) were detected in groundwater under-
lying the Old Camden County Landfill, now located within
the boundaries of Kings Bay Naval Submarine Base. The
groundwater plume was migrating off base towards an
adjacent residential neighborhood.The maximum ground-
water VOC concentration was 9,074 mg/L of total chlori-
nated aliphatic hydrocarbons (sum of PCE, TCE, cis-1,2-
DCE, and vinyl chloride concentrations), primarily PCE
(8,500 mg/L). The PCE concentration exceeded 1% of the
water solubility, implying the presence of a residual DNAPL
phase of unknown mass. The treatment goal was to
achieve a total chlorinated aliphatic hydrocarbon concen-
tration of <100 mg/L, a level considered sufficient to allow
natural attenuation to reduce VOCs to MCLs to prevent
migration of the plume beyond base boundaries. The site
geology is characterized as medium to fine sands, and
the targeted treatment zone was an interval of relatively
high conductivity located from approximately 30 to 40 ft
below grade. An oval-shaped source area approximately
120 ft long x 40 ft wide was targeted for treatment.
A full-scale Geo-Cleanse Treatment was conducted at the
site in cooperation with Bechtel Environmental, Inc. and
the Naval Facilities Engineering Command, Southern Di-
vision. A total of 23 specially designed injectors were in-
stalled in an array encompassing the target treatment area.
A total of 12,045 gallons of 50% hydrogen peroxide was
injected in 26 days during two treatment phases in No-
vember 1998 and February 1999. Groundwater samples
were collected before, during and after treatment. The re-
medial objective of <100 mg/L was achieved within the
treatment area following the first treatment phase. The
second treatment phase targeted the area primarily
downgradient of the primary treatment area, in order to
achieve further reductions. The results indicate destruc-
tion of >99% of the dissolved chlorinated aliphatic hydro-
carbons in groundwater, with a reduction to 9 mg/L from
9,074 mg/L within the primary source area. All monitoring
wells within or adjacent to the treatment area achieved
the cleanup objective of <100 mg/L and have maintained
compliance to date, (up to 9 months after treatment). In
addition, monitoring wells and recovery wells located
downgradient also exhibited significant reductions.
33
-------
As a result of the Geo-Cleanse Treatment Program, the
Navy was recognized and given an award for environmen-
tal awareness by the State of Georgia Department of Natu-
ral Resources, and a long-term pump-and-treat system to
control plume migration was removed from service. Based
upon these very positive results, the scope of work was
expanded in June 1999 to include additional downgradient
and upgradient source areas in order to achieve even
greater reductions outside of the initial treatment area.The
additional treatment was conducted in June and July 1999.
Preliminary data indicate further reductions and success-
ful treatment of the expanded area.
Pensacola Naval Air Station, Florida
TCE and its natural degradation products (cis-1,2-DCE
and vinyl chloride) and aromatic hydrocarbons (primarily
chlorobenzene isomers) were detected in a groundwater
plume originating under a former sludge drying bed at the
industrial Wastewater Treatment Plant at Pensacola Na-
val Air Station. Site characterization data delineated a
groundwater plume migrating northeast towards Pensacola
Bay. The cleanup goal was to eliminate source area con-
tamination in order to enhance natural attenuation of the
downgradient plume and thereby prevent subsurface dis-
charge to Pensacola Bay. The maximum concentration of
total VOCs dissolved in groundwater was 4,896 mg/L,
composed primarily of TCE (3,600 mg/L). The site geol-
ogy is characterized as medium to fine sands with ground-
water at approximately 4 ft below grade. The primary tar-
geted treatment zone was an interval of relatively high
conductivity located from approximately 35 to 45 ft below
grade.
A full-scale Geo-Cleanse Treatment was conducted at the
site in cooperation with the Naval Facilities Engineering
Command, Southern Division and the Naval Facilities
Engineering Service Center. A total of 15 specially de-
signed injectors were installed in an array encompassing
the target treatment area. Six of the injectors were installed
at shallower depths within the source area to target con-
tamination at the suspected source. The remaining nine
injectors were installed within the 35 to 45 ft depth inter-
val.
A total of 10,160 gallons of 50% hydrogen peroxide was
injected in 13 days during two treatment phases in De-
cember 1998 and in May 1999. Groundwater samples were
collected before, during and after each treatment.The re-
sults indicate destruction of >96% of the dissolved chlori-
nated aliphatic hydrocarbons in groundwater, with a re-
duction of total VOCs to non-detect (<1 mg/L) from 4,896
mg/L in the most contaminated monitoring well. These
reductions have been maintained in all compliance points
to date (from a sampling round conducted 30 days follow-
ing injection).
Fort Dix 4400 Area Spill Site, New Jersey
Fort Dix is located in the Pine Barrens region of central
New Jersey, considered an environmentally sensitive re-
gion because it is the recharge area for the Kirkwood-
Cohansey Aquifer, the primary source for domestic wells
in the area. The shallow aquifer at the site is character-
ized as medium to fine sand extending to a depth of ap-
proximately 40 ft below grade. Motor pool operations at
the Fort Dix facility resulted in a discharge of chlorinated
solvents to the shallow groundwater. The maximum his-
torical concentration of total VOCs was 132 mg/L, prima-
rily PCE (65 mg/L). The maximum concentration immedi-
ately prior to treatment was 24.6 mg/L, primarily PCE (11.2
mg/L). Due to the environmental sensitivity of the site, an
extremely rapid and effective solution was desired, with a
cleanup objective of 1 mg/L.
A full-scale Geo-Cleanse Treatment Program was con-
ducted at the site in cooperation with ICF Kaiser Engi-
neers, Inc. (now part of the IT Group). Twenty-four spe-
cially designed injectors were installed in two discrete lev-
els at the site to encompass the vertical and horizontal
distribution of contaminants. A total of 5,708 gallons of
50% hydrogen peroxide was injected over the course of a
7-day field treatment program in February 1999. The re-
sults indicate that the maximum dissolved concentration
of total VOCs was reduced to 1.0 mg/L from 24.6 mg/L
(96%), thus achieving the remedial goal.These levels were
maintained for a groundwater sampling event 14 days fol-
lowing treatment, and additional quarterly monitoring is
planned.
Fort Stewart, Georgia
Fueling operations resulted in jet fuel contamination of soil
and groundwater at Wright Army Airfield, Fort Stewart,
Georgia. The geology of the site is marked by a distinctive
stiff clay- and silt-rich sand interbedded with red and grey
mottled clay layers, interpreted as an ancient soil (paleosol)
weathering horizon. The paleosol is underlain by fine to
medium quartz sand extending at least 25 ft below grade.
Soil investigations identified two distinct layers of contami-
nation, at approximately 13 and 17 ft below grade, indi-
cated by stained soils, petroleum odors, and globules of
free product floating on water samples. Soil sampling indi-
cated maximum total petroleum hydrocarbon contamina-
tion of 906 mg/kg, and maximum dissolved concentrations
of fuel components (sum of BTEX, 2-methylnaphthalene,
and naphthalene) of 7,331 mg/L.
A Geo-Cleanse Pilot Treatment was conducted in collabo-
ration with Science Applications International Corporation
(SAIC) and the Fort Stewart Directorate of Public Works,
Environmental Branch, to demonstrate effectiveness of the
Geo-Cleanse Process on jet fuel at the Fort Stewart site.
Six specially designed injectors were installed surround-
ing a monitoring well. Four of the injectors targeted the
shallower contaminated interval (13 ft below grade) and
two injectors targeted the deeper zone (17 ft below grade).
A total of 2,537 gallons of 50% hydrogen peroxide was
injected over a five-day period in September 1998.
Because the pre- and post-injection soil sampling loca-
tions do not correspond exactly, the site was partitioned
34
-------
into two areas. Pre-injection soil results for Area 1 ranged
from 1.4 mg/kg to 906 mg/kg, and one post-injection
sample yielded 4.9 mg/kg. Pre-injection soil results for Area
2 ranged from 65.8 mg/kg to 837 mg/kg, and one post-
injection sample yielded 779 mg/kg. Groundwater samples
also indicate significant reductions in dissolved fuel com-
ponents (sum of BTEX, 2-methylnaphthalene, and naph-
thalene). The central monitoring well exhibited a reduction
to 80 mg/L from 839 mg/L (90% destruction). On a site-
wide average comparing three pre- and post-treatment
samples, dissolved fuel components were reduced to 906
mg/L from 4,051 mg/L (78% destruction). The objective of
demonstrating effectiveness of the Geo-Cleanse Process
on jet fuel contamination at the Fort Stewart site, in a very
heterogeneous, dense, clay-rich paleosol horizon, was
successfully achieved.
Former Industrial Site, Ohio
Operations at this former industrial site used
trichlorobenzene as a solvent during synthesis of an or-
ganic chemical product. A DNAPL phase composed pri-
marily of 1,2,4-trichlorobenzene was identified in a sand
and gravel aquifer. The DNAPL collected in discrete pools
on an underlying stiff clay layer, under the razed building
foundation, at a depth of approximately 20 to 22 feet be-
low grade. One pool was targeted for a pilot treatment with
the Geo-Cleanse Process. The volume of DNAPL present
was difficult to estimate, but as much as 4.5 feet of DNAPL
accumulated in a recovery well within the treatment area.
Other organic contaminants reported at the site included
other isomers of trichlorobenzene, isomers of dichloroben-
zene, naphthalene and bis(2-ethylhexyl)phthalate. Addi-
tionally, ammonia was reported at levels up to 590 mg/L
within the treatment area. Ammonia is a highly basic com-
pound that is not effectively destroyed by Fenton's reagent.
Groundwater at the site was found to have a pH of 7.2 to
8.2, with some samples yielding a pH as high as 11. The
ammonia and very basic groundwater pH presented chal-
lenges because of the buffering capacity of the ground-
water (Fenton's reagent is most effective at mildly acidic
groundwater pH conditions) and radical trapping proper-
ties of the non-targeted compounds.
The goal of the Geo-Cleanse Pilot Treatment were to re-
duce DNAPL mass within the targeted treatment area to
demonstrate effectiveness of the Geo-Cleanse Treatment
Program at the site. Six specially designed injectors and
two specially designed vent wells were installed within the
treatment area. The vent wells were installed with screens
extending into the vadose zone soil and a blower pump
was attached, in order to draw any ammonia vapors that
might be entrained with gases (oxygen and carbon diox-
ide) resulting from the Geo-Cleanse Process. The pump
effluent was passed through a water trap and carbon filter
to capture any ammonia vapors. A total of 3,016 gallons
of 50% hydrogen peroxide solution was injected over the
course of five days in August 1998.
Dissolved 1,2,4-trichlorobenzene levels decreased over-
all across the site as a result of the Geo-Cleanse Pilot
Treatment. The volume of DNAPL within the treatment area
was reduced dramatically. DNAPL was eliminated during
the treatment in three monitoring wells that previously
accumulated DNAPL, including PMW-14 (formerly 3.78
feet collected), PMW-1 (formerly 0.04 feet collected) and
PRW-3 (formerly 1.60 feet collected). Two other monitor-
ing wells located adjacent to the treatment area experi-
enced reductions in dissolved 1,2,4-trichlorobenzene lev-
els and no change in DNAPL accumulation thickness. An
additional observation was that, after treatment, only the
wells installed into the underlying clay accumulated
DNAPL, suggesting that the residual DNAPL remaining
after treatment was confined to the upper portion of the
underlying clay layer rather thanjn the targeted sand and
gravel interval.
Letterkenney Army Depot (DA Area Bed-
rock), Pennsylvania
Letterkenney Army Depot was established as an ammu-
nition storage facility in 1942, with additional missions in-
cluding overhaul and maintenance of wheeled and tracked
vehicles, issue and shipment of chemicals, maintenance
of missile systems, and ammunition maintenance and
demilitarization. Operations resulted in disposal of indus-
trial solvents and other wastes within the Disposal Area
(DA), and the DA was placed on the National Priorities
List in July 1987. A Geo-Cleanse Pilot Treatment was con-
ducted within the DA in order to evaluate effectiveness of
the Geo-Cleanse Process within a limestone bedrock for-
mation. This application was innovative because of the
relatively high alkalinity and pH of the groundwater and
expected reactivity of the native limestone bedrock at the
site towards acidic reagents utilized during the Geo-
Cleanse Process. Contaminated areas in the unconsoli-
dated soil overburden was excavated in 1993-94. How-
ever, groundwater contamination within the underlying frac-
tured bedrock aquifer was not addressed. The bedrock
aquifer at the site was the subject of intensive investiga-
tion supported by the U.S. Army Corps of Engineers. Total
VOC concentrations in groundwater ranged from 18 to as
high as 3,000,000 mg/L, with TCE and its natural degra-
dation product cis-1,2-DCE as the primary contaminants
present within the pilot treatment area.
A Geo-Cleanse Pilot Treatment was conducted at the site
in collaboration with Roy F. Weston, Inc. and the U.S. Army
Corps of Engineers. A total of 12,635 gallons of hydrogen
peroxide were injected over 4 days in June 1999. Final
analytical results are not yet available. However, field pa-
rameters indicate a successful treatment. The bedrock
aquifer was able to sustain very high overall injection rates,
resulting in an expected very wide radius of influence.
Groundwater samples for VOC analysis have not yet been
collected. However, hydrogen peroxide is persistent in the
monitoring wells and injectors within the area 25 days af-
ter completing the injection.
35
-------
Former Manufactured Gas Plant, Wisconsin
The site of a former manufactured gas plant in southeast-
em Wisconsin is slated for redevelopment by the year 2001.
Redevelopment cannot be completed until remaining
groundwater and saturated soil contamination is
remediated. Unsaturated soils at the site were excavated
in 1995, but residual contamination within the saturated
zone remains very high. BTEX and PAH compounds are
the primary contaminants present, with concentrations as
high as 59,288 mg/kg of BTEX and 404,060 mg/kg of PAH
sorbed to saturated zone soils within the targeted pilot
treatment area. Groundwater is at approximately 5 ft be-
low grade and limestone bedrock is encountered at ap-
proximately 15 ft below grade. The soils within the satu-
rated zone are generally described as silty sand and gravel
with some cobbles.
Bench-scale testing indicated that the Geo-Cleanse Pro-
cess can effectively oxidize the PAH and BTEX compounds
present in the soil at the site. Accordingly, a Geo-Cleanse
Pilot Treatment was conducted to determine if the Geo-
Cleanse Process offers a viable remedial alternative. Ap-
proximately 7,200 gallons of 50% hydrogen peroxide so-
lution were injected in 11 days in July 1999. Final analyti-
cal data are not yet available. However, field indications,
primarily carbon dioxide production as a result of hydro-
carbon oxidation, indicate a very efficient chemical oxida-
tion reaction.
Summary
While no treatment technology is appropriate for every
site, in situ chemical oxidation can provide a viable tech-
nology to achieve cleanup objectives in a cost-effective
and timely manner. Case studies are presented in this
summary for seven sites treated by the Geo-Cleanse Pro-
cess. The contaminants at the sites include chlorinated
solvents, polycyclic aromatic hydrocarbons, and aromatic
hydrocarbons, in a wide range of geological and
hydrogeological settings. The case studies demonstrate
that the Geo-Cleanse Process is a robust technology for
in situ chemical oxidation that is applicable to a wide vari-
ety of contaminants and site conditions.
36
-------
Interagency DNAPL Consortium
Tom Early and Paul DeVane
Oak Ridge National Laboratory
PO Box 2008
Oak Ridge, TN 37831
423-576-2103
eot@ornl.gov
Federal agencies have been using shrinking environmen-
tal research and development resources to address envi-
ronmental contamination caused by dense non-aqueous
phase liquids. When technologies do appear promising,
there is a lengthy period before they are accepted by site
owners, stakeholders, and the regulatory community. Usu-
ally, there is no direct comparison between technologies
to determine cost and performance data.
In an attempt to better use our resources, and address
the DNAPL problem, several agencies have joined forces.
The Interagency DNAPL Consortium is working together
at Cape Canaveral Air Station, Florida to demonstrate three
promising DNAPL source remediation technologies. We
feel there are several benefits to this partnership.
Benefits
• Multi-agency cooperation
• Cost shared demonstrations
• Side-by-side, real-time technology demonstrations
• Accelerated technology deployment
• Share/develop expertise among agencies with com-
mon problems
• Comparative cost-data for technologies
• Shared agency solutions for common problems
Problems
Dense Non-Aqueous Phase Liquids (DNAPLs) pose seri-
ous, long-term groundwater contamination problems due
to their toxicity; limited solubility in groundwater; arid sig-
nificant migration potential in soil gas, groundwater, and/
or as separate phase liquids. DNAPL chemicals, particu-
larly chlorinated solvents, are among the most common
of environmental contamination problems in the United
States as well as for most industrialized countries. There
are thousands of DNAPL-contaminated sites in the United
States, often at contaminant volumes that are difficult to
detect, but in quantities that can represent significant
sources of groundwater contamination. Many agency and
private-sector sites have DNAPL contamination problems,
including federal, state and local government agencies.
The Office of Management and Budget estimates that the
federal government alone will spend billions of dollars for
environmental clean-up of DNAPL contamination prob-
lems.
While various DNAPL remediation, characterization and
monitoring technologies have been demonstrated in the
past, it is difficult, if not impossible, to make meaningful
comparisons of either performance or cost among these
technologies because of the variable conditions at the
demonstration sites. As a result, "problem holders" and
regulatory officials have been reluctant to deploy these
technologies for site clean up. In order to expedite the
regulatory acceptance and use of these innovative reme-
dial technologies, comparative cost and performance data
must be collected.
Solutions
An important step in reducing technology risk and increas-
ing user and regulatory acceptance of DNAPL remediation,
characterization and monitoring technologies involves
conducting concurrent, "side-by-side" field demonstrations.
These side-by-side demonstrations result in comparative
cost and performance data collected under the same field
conditions. Through appropriate documentation, the re-
sulting cost and performance data can be evaluated for
site-specific applications. Side-by-side demonstrations
help to fill an important "gap" in the process of technology
development and deployment and will accelerate technol-
ogy privatization.
In 1998, a multiagency consortium was organized by the
United States Department of Energy/Office of Environmen-
tal Management (DOE/EM) and the Department of De-
fense (DOD) through the Air Force Research Laboratory
(AFRL) in cooperation with the 45th Space Wing, the Na-
tional Aeronautics and Space Administration (NASA) and
the United States EPA (EPA) to demonstrate innovative
DNAPL remediation and characterization technologies at
a NASA remediation site on Cape Canaveral Air Station,
37
-------
Cape Canaveral, FL. This Interagency DNAPL Consor-
tium (IDC) was formed to:
• address a serious, widespread and shared environ-
mental problem adversely affecting many U.S. federal
agencies (e.g., DOE, EPA, DOD, NASA, Department
of Interior, Department of Agriculture);
• cost-share the demonstration and comparison of these
remediation and monitoring system technologies;
• accelerate both the demonstration and deployment of
DNAPL remediation, characterization and monitoring
technologies for the purpose of reducing the perceived
technology risk associated with these technologies;
• increase regulatory and user acceptance of these
technologies by providing documented, cost and per-
formance data; and
• provide increased opportunities to test new sensors
designed to support in situ remediation of DNAPL con-
tamination problems in addition to ex situ treatment
and disposal.
In order to conduct this side-by-side demonstration, an
IDC Core Management Team was organized. The IDC
consists of representatives from DOE, NASA, USAF, DOD
and EPA. The Team is a collaborative decision-making body
that draws upon the strengths of each agency to solve
problems associated with the project. The Team utilizes a
Technical Advisory Group (TAG) for support in making
decisions that concern individual evaluation of remediation
systems. The IDC TAG is comprised of experts from in-
dustry, academia and federal agencies. With the support
of the TAG, the Team selected three of the most promising
remediation technologies (Figures 1 -3) for deployment and
evaluation at Launch Complex 34. Florida State University's
Institute for the Team for International Cooperative Envi-
ronmental Research provides day to day field project
management through a cooperative agreement with the
Department of Energy.
Remediation Technologies
Thermal Treatment for Mass Removal -
Electrical Resistance Heating
The Six Phase Soil Heating technology (Figure 1) removes
contaminants from soil and groundwater by passing an
electrical current through the soil matrix. The passage of
current generates heat due to electrical resistance within
the soil. This is the same process used in any electrically
heated device (e.g., clothes iron, heater, stove). Heat is
generated throughout the soil in the remediation area and
the temperature of the soil is increased to the boiling point
of water. Soil moisture becomes steam that is captured by
vapor recovery wells for removal. Soil contaminants are
vaporized concurrently and are captured for ex situ treat-
ment.
Benefits
• Heat is generated uniformly throughout the treatment
volume. While low permeability lenses reduce the per-
formance of other technologies that rely on the verti-
cal movement of a fluid or vapor though the soil ma-
trix, soil heterogeneity or low permeability does not
adversely effect Six Phase Soil Heating. In fact, low
permeability soils tend to carry greater current than
do sandy soils, thus, become hotter, and boil constitu-
ents faster.
Anaerobic dechlorination of solvents will add conduc-
tive chloride ions to "hot spots", likewise attracting
current for faster remediation of the impacted regions
of the site.
The boiling of soil moisture in clay lenses forms steam
to "sweep out" volatile organic compounds. This steam
stripping process effectively increases the permeability
of clay soils.
Because Six Phase Soil Heating treats all soils in the
treatment volume, there are no untreated regions from
which contaminants could diffuse later and cause re-
bound. Rebound has not been observed at any Six
Phase Soil Heating site.
The presence of perched water does not reduce the
effectiveness of Six Phase Soil Heating.
In Situ Chemical Oxidation - Potassium
Permanganata (KMnOJ
In situ oxidation using potassium permanganate (Figure
2) is a potentially fast and low cost solution for the de-
struction of chlorinated ethylenes (TCE, PCE, etc), BTEX
(benzene, toluene, ethylbenzene, and xylene) and simple
polycyclic aromatic hydrocarbons. In particular, potassium
permanganate reacts effectively with the double bonds in
chlorinated ethylenes such as trichloroethylene, perchlo-
roethylene, dichloroethylene isomers, and vinyl chloride.
It is effective for the remediation of DNAPL, adsorbed
phase and dissolved phase contaminants and produces
innocuous breakdown products such as carbon dioxide,
chloride ions and manganese dioxide. The permanganate
solution typically is applied at concentrations of one to three
percent solution via injection wells. This solution is easily
handled, mixed and injected and is non-toxic and
non-hazardous.
Bench scale laboratory tests of potassium permanganate
with trichloroethylene have resulted in up to a 90% reduc-
tion of trichloroethylene in four hours of treatment. The
effectiveness of the in situ injection of permanganate is a
function of the reaction kinetics, the transport and contact
between potassium permanganate and the contaminant,
as well as competitive reactions with other oxidizable spe-
cies (e.g., iron, natural organics).The effective use of this
remedial technology requires an engineered approach for
maximizing the contact between potassium permangan-
ate and the target contaminant. As with many technolo-
gies, low permeability and heterogeneity of soils present
a challenge and require a carefully designed application
system.
38
-------
Figure 1. Six Phase Soil Heating.
Figure 2. In Situ Chemical Oxidation with Permanganate.
39
-------
Benefits:
• Chemically oxidizes a wide range of organic com-
pounds to innocuous end-products over a wide pH
range
• Visible (purple) solution makes it easy to track the in-
jection influence or the degree of treatment
• Chemically stable in water (very slow
auto-degradation)—stays in solution until it is reacted
• No off-gas treatment required
Thermal Treatment for Mass Removal
Steam Stripping
Thermal remediation by steam injection and recovery uses
Dynamic Underground Stripping, Steam Enhanced Ex-
traction, Hydrous Pyrolysis/Oxidation, and Electrical Re-
sistance Tomography. Combining these technologies the
Dynamic Underground Stripping System uses boilers to
generate steam which is then pumped into injection wells
that surround the contaminants. The steam front volatil-
izes and mobilizes the contaminants as it pushes the re-
sulting steam front toward a central network extraction well
where it is vacuumed to the surface. Direct electrical heat-
ing of soils, clay and fine-grained sediments causes
trapped water and contaminants to vaporize and forces
them into steam zones where vacuum extraction removes
them. Electrical Resistance Tomography is used as a pro-
cess control method to measure electric resistance and
temperatures in the subsurface that allow for real-time
control of the heating process.
Benefits
• Faster clean-up, potential closure within months to
years, not decades
• Removes source contaminants effectively
• Treats contamination both above and below the water
table, with no practical depth limitation
Sensor Technology Evaluations
In addition to DNAPL remediation technology demonstra-
tions, the project provides the opportunity to evaluate in-
novative characterization technologies for locating DNAPL,
in situ lithologic mapping, in situ vadose zone and satu-
rated zone sampling and in situ hydraulic conductivity mea-
surements. These technologies were deployed using the
DOE and EPA Site Characterization and Analysis Pen-
etrometer System (SCAPS) trucks. In addition to sensor
technology evaluation, the SCAPS trucks have been used
for data collection essential to conceptual model design
and strategic location of critical lithologic units, sediment
sampling and monitoring well placement. The following
cone penetrometer (CPT) based sensors and sampling
tools have been deployed at the site:
Figures. In Situ Thermal Remediation (Steam Injection)
40
-------
• Raman spectroscopy: used for direct detection of
DNAPL in the subsurface
• GeoVIS: soil video imaging system used for visual
characterization of critical stratigraphic units and vi-
sual detection of DNAPL
• Cone Permeameter™: in situ permeability measure-
ments
• Cone Sipper: multiple depth discrete soil gas and
groundwater sampling
• FLUTE: Hydrophobic Flexible Membrane is a sam-
pling device that can provide detailed delineation of
DNAPL in a borehole
• Precision Injection/Extraction (PIX) Probe: charac-
terization method for determining the presence or
absence of depth discrete DNAPL
Agency Contacts
Skip Chamberlain
U.S. Department of Energy
(Germantown, Maryland) (301) 903-7248
grover.chamberlain@em.doe.gov
(or)
Jim Wright
U.S. Department of Energy
(Aiken, South Carolina)
(803) 725-5608
jamesb.wright@srs.gov
Major Paul B. DeVane
U.S. Air Force Research Laboratory
(Tyndall AFB, Florida)
(850) 283-6288
paul.devane@mlq.afrl.af.mil
Thomas Holdsworth
USEPA
(Cincinnatti, Ohio)
(513)569-7675
noldsworth.thomas~epamail.epa.gov
Ed Worth
U.S. Air Force, 45th Space Wing
(Patrick AFB, Florida)
(407) 853-0965
edwin.worth @ pafb.af.mil
Jackie Quinn
National Aeronautics and Space Administration
(Kennedy Space Center, Florida)
(407) 867-4265
Jaqueline.Quinn-1 @ ksc.nasa.gov
Interagency DNAPL Consortium home page: www.getf.org/
dnaplguest
41
-------
Field Pilot Test of In Situ Chemical Oxidation Through Recirculation Using
Vertical Wells at the Portsmouth Gaseous Diffusion Plant
K. S. Lowe and F. G. Gardner
Oak Ridge National Laboratory
4830 W. 28th Ave.
Denver, CO 80212
303-966-3430
kslowe@ornl.gov
R. L. Siegrist
Oak Ridge National Laboratory
Colorado School of Mines
T. C. Houk
Becthel Jacobs Company, LLC
Piketon, Ohio
Introduction
In s/ft/chemical oxidation is an emerging remediation tech-
nique in which chemical oxidants are delivered to the sub-
surface to rapidly degrade organic contaminants (Siegrist
1998, US EPA 1998). For the past 6 years, engineers and
scientists at Oak Ridge National Laboratory (ORNL) have
been developing this technology for in situ degradation of
organic chemicals such as trichloroethene (TCE),
perchloroethene (PCE), naphthalene, and pyrene (Gates
and Siegrist 1995; West et al. 1997); Siegrist et al. 1998a
and 1998b). Laboratory-scale experiments performed to
date have demonstrated that permanganate (MnO •), if
applied at sufficient loadings, can effectively oxidize TCE
and PCE. The following describes the overall chemical
reaction forthe oxidation of TCE using potassium perman-
ganate (eqn.1):
2KMnO4 +C2HC13 ->• 2CO2 +2MnO2 +2KC1 + HC1
Permanganate is commonly used in waste water treat-
ment to oxidize organic compounds and is commercially
available in two forms, KMnO4 and NaMnO4. While the
behavior and oxidation potential of the two compounds is
similar, the aqueous solubility of NaMnO4 is much higher
(-50% vs. -6%). The relative stability of MnO4" makes it
attractive as an in situ chemical oxidant since it can be
delivered and dispersed in soil and groundwater and has
the ability to migrate by diffusion into fine-grained zones
where residual contaminants may exist.
To continue development of in situ chemical oxidation for
widespread use and commercial viability, techniques for
delivering chemical oxidants in adequate amounts to the
subsurface are being developed (Case, 1998; DOE 1996a;
Gates and Siegrist 1995; Lowe et al. 1998; Siegrist et al.
1994,1995,1998a, 1998b, 1999; Struse 1999; West et al.
1995). Initially, an in situ chemical oxidation through recir-
culation (ISCOR) treatability study was conducted at the
U.S. Department of Energy Portsmouth Gaseous Diffu-
sion Plant (PORTS), Piketon, Ohio, involving injection and
recirculation of an oxidant solution (~2wt.% KMnOJ
through paired horizontal wells and selected vertical wells
(West et al. 1997). Although oxidant delivery/recirculatron
was impacted by heterogeneity within the demonstration
area, where the oxidant was able to permeate the aquifer,
significant and sustained reductions in TCE were mea-
sured in soil and groundwater. The goal of this field pilot
test was to test an alternative oxidant delivery approach
using recirculation via vertical wells in a 5-spot pattern to
develop a sufficient understanding of ISCOR to facilitate
completion of the Corrective Measures Study process. The
test design was comprised of a 5-spot vertical well pat-
tern with the central well serving as the injection point (Fig-
ure 1).
The four corner extraction wells were placed in a square
grid surrounding the injection point on a fixed radius of 45
ft. Several monitoring wells were also installed throughout
the region to monitor changes in contaminant concentra-
tions, oxidant concentrations, water quality parameters,
and water levels during operation. The system was de-
signed to accommodate recirculation flow rates of 2 to 20
gpm supplemented with NaMnO4 to maintain delivery of
the oxidant concentration at -250 mg NaMnO4 per L of
re-injected groundwater.
42
-------
Mnord
Conceptual Only
Not to Scale
Figure 1. Conceptual diagram of vertical well injection/recirculation system.
The advantages of the use of liquid NaMnO4 and a verti-
cal well recirculation approach include: (1) better control
of oxidant and contaminant migration within the treatment
zone when compared to single well injections, (2) the in-
troduction of higher volumes of oxidant solutions in a given
time because existing soil pore water is extracted concur-
rent to oxidant injection, (3) easier oxidant handling, and
(4) potentially lower overall cost for treating larger volumes.
Methods
The field pilot test was conducted within the source area
of the 5-Unit Investigative Area, immediately south of the
X-770 Building, at PORTS. The geology under the pilot
test cell area is comprised of 23-26 ft of Minford sediments
underlain by 3 to 8 ft of Gallia. Both units are unconsoli-
dated, Quaternary age members of the Teays formation.
The Gallia lithology is typically comprised of a saturated
silty-sandy gravel unit of higher permeability overlying a
drier lithified unit of coarse angular gravel in a silty clay
matrix. The upper more permeable member of the Gallia
varies in thickness from 0.5 to 2.5 ft. The Gallia is thicker
and more permeable across the southern two thirds of
the test cell area. In general, 5-Unit Area groundwater flows
horizontally from north to south and the hydraulic gradient
is very low (0.001) because of the flat valley floor, the
presence of thicker and more permeable Gallia deposits,
and the proximity of the east-west-trending groundwater
divide present along the central part of the facility.
The field pilot test required the installation, development,
and hydraulic testing of wells, baseline soil and ground-
water characterization, and assembly of the oxidant deliv-
ery system. One 6-in diameter injection well, four 4-in di-
ameter extraction wells and nine 2-in diameter monitoring
wells to monitor the pilot test operations were installed
(Figure 2). Soil and groundwater characteristics, physical
and chemical, were obtained throughout the area prior to
recirculation. Hydraulic tests (pressure injection test and
single well tests) were performed to select the recircula-
tion pumping and injection rate and to determine the dis-
tribution of permeability within the test area.
A conservative tracer (bromide) was added at the initia-
tion of injection/recirculation to enable detection of the in-
jected water within the treatment area. Approximately 500
gals of 1800 mg/L bromide solution were injected at 18
gpm into the central injection well with equal extraction
(4.5 gal/min) at the four corner extraction wells. Concur-
rently, groundwater samples were collected from each
monitoring location and extraction well and analyzed for
bromide using an ion selective electrode at 15 to 30 minute
intervals until movement of the bromide past the monitor-
ing location was detected. This information was used to
determine preferential flow paths, to measure the extent
of hydraulic control in the treatment area, and to define
the rate and extent of anticipated NaMnO4 transport.
The oxidant delivery system was assembled along the
northern side of the test cell. The major components of
the system comprise flow and pressure control, oxidant
dosing and delivery, and sample ports. System flow rates,
pressures, water levels, oxidant feed concentrations, move-
ment/reaction of the oxidant through the Gallia aquifer
43
-------
MW14 MVI06 moi
<£> ®
MW10
®
HW13
MW04
.EW03
®
MWM
EHrtW
Q Existing Extraction Well
® Existing Monitoring Well
@ Existing Injection Well
9 Soil Boring Location
Figure 2. Field pilot test site.
during recirculation, and contaminant concentrations were
monitored daily. Additional analyses for cations, anions,
and water quality parameters were collected during recir-
culation of the first pore volume, but terminated after de-
livery of the oxidant throughout the test area due to inter-
ferences in the analysis method.
During injection and recirculation, oxidant solution was
injected into the Gallia through the central well and ex-
tracted at equal rates from the 4 perimeter wells located
at 45 ft from the central injection well. Groundwater was
extracted from the four perimeter wells at a maximum com-
bined rate of 18 gpm (4.5 gpm from each extraction well).
Groundwater from the four extraction wells was pumped
through a manifold system into a single line where the
NaMnO. solution was added with an oxidant-resistant
chemical metering pump.The metering pump was adjusted
to feed the concentrated stock solution at a rate sufficient
to maintain 250 mg/L of NaMnO4 in the injection water.
The oxidant-laden water was then pumped into a 120 gal
holding tank providing approximately 7 minutes of resi-
dence time for the oxidant with the contaminated ground-
water. This ensured that only treated water was re-injected
into the aquifer. Injection of the extracted water was per-
missible as long as the extracted water had been "treated".
A 250 mg/L NaMnO4 solution was recirculated at 18 gpm
for the 1st pore volume (approximately 82,000 gals or 3
days). Then the recirculated water with residual NaMnO
was supplemented with approximately 100 mg/L of
NaMnO4 solution to maintain a delivery of -200 to 250
mg/L of oxidant into the region at 18 gpm for the 2nd and
3rd pore volumes. Approximately 78 gals of 40wt% NaMnO4
(162.4 kg NaMnO.) were delivered throughout the test
region within 10 days. Approximately 240,660 gals of
treated groundwater were recirculated. The recirculation
flow rates remained steady throughout the test duration
with ~1 gpm reduction in total recirculation flow rate over
the course of the 10-day test (Table 1). This gradual but
minor decline in the total recirculated flow rate is attrib-
uted to injection or extraction well fouling, matrix plugging,
pump performance, or clogging/plugging within the sys-
44
-------
Table 1. Summary of Process Operation Parameters.
Location Ave. Flow Rate (gpm) Ave. Pressure (psi)
Table 2. Single Well Aquifer Testing Results.
X770-IW01
X770-EW01
X770-EW02
X770-EW03
X770-EW04
17.37
4.17
4.39
4.36
4.59
14.5
65.4
146
146
150
tern lines (at in-line valves, gauges, etc). Pressures within
the system and at the injection well gradually increased
over time. The injection pressure at the well head nearly
doubled, from 10 psi to 18.5 psi, but remained within safe
operating ranges based on optimum flow out of the se-
lected well screen, potential fracturing of the formation,
and potential damage to the well construction. It is likely
that the gradual increase in system pressures and de-
cline in extraction well flow rates is in part, a result of in-
jection well performance or potential redistribution of fine
particles in the formation near the well.
Performance assessment was based on TCE reduction
as determined by characterizing contaminants in the aque-
ous phase before and after treatment. If VOC mass was
substantially reduced by oxidation, then the TCE aque-
ous levels in the extraction and monitoring wells should
also be substantially lower. Groundwater was recirculated
through the pilot test area until sufficient oxidant had been
delivered or breakthrough of the residual oxidant at the
injected concentration (indicating that the NaMnO4 was
not being consumed in the aquifer) was observed. The
TCE concentrations in the test area were monitored dur-
ing recirculation and for 24 hrs after recirculation was ter-
minated. Finally, post-treatment groundwater concentra-
tions were monitored weekly for approximately 1 month
and once at 2 months after recirculation to determine the
potential aquifer rebound effects.
Results
A pressure injection test conducted at the injection well
(Bouwer and Rice 1976 and Horner, 1951) indicated that
the injection well could sustain recirculation rates up to 26
gpm without exceeding optimum pressures at the well
head. Observation of water level fluctuations during the
pressure injection test also indicated significant variability
in hydraulic conductivity across the 64-ft by 64-ft test area.
In relative terms, greater water level increases indicate
regions of higher permeability while smaller water level
increases indicate regions of lower permeability. Single
well tests confirmed the variability and provided a relative
measure of permeability with values ranging from approxi-
mately 26 to >300 ft /d (Table 2).
Bromide tracer test results confirmed the hydraulic tests
and indicated that the Gallia is very heterogeneous
throughout the test region. The bromide curves from the
northern wells tend to flatten out indicating a more hetero-
Well Number
X770-EW02
X770-EW03
X770-EW04
X770-MW01
X770-MW03
X770-MW05
X770-MW07
X770-MW08
X770-MW10
X770-MW12
X770-MW14
Ave Pre-lnj
ft/day
26.5
86.5
44
148
297.5
691.5*
202.5
121
145
—
601.5*
Ave Post-lnj
ft/day
26.7
87.3
54.3
94
271.7
532.3
187.7
143
113.7
366.3
337
geneous flow system and greater dispersion while the
curves to the west and south indicate a more pronounced
(i.e., sharp) bromide front suggesting a more homoge-
neous flow system with less dispersion (Figure 3). Addi-
tionally, bromide rapidly moved (within two hours) in a
southeastern direction as delineated by well X770-MW03,
but was not initially detected in X770-EW02, the south-
eastern most well, until 10 hours after injection with the
curve peak at -22 hours after injection. Water levels, mea-
sured continuously by a data acquisition system, indicate
that the system reached hydraulic equilibrium within a few
hours after recirculation was initiated and remained con-
stant throughout the test.
The bromide tracer test also provided insight on the pref-
erential flow paths within the region. Based on preliminary
calculations (assuming laminar plug flow and homogenous
subsurface conditions) and model predictions (WinTrans)
it was estimated that it would take approximately 3 days
for the tracer to reach each of the extraction wells located
45 ft away. However, initial detection of the bromide in the
extraction wells was within 5 to 1.0 hours. The tracer was
detected at the northeastern well (X770-EW01), but bro-
mide concentrations did not exceed 5 rhg/L (near the mini-
mum detection limit of the field probe). Thus, the majority
of the water produced from X770-EW01 is presumed to
be from outside the test region as indicated by the slow
delivery of bromide and oxidant to the well. In relative terms,
the region north and east of X770-EW01 appears to be
more permeable than the region between the injection well
and X770-EW01, including X770-MW10. Finally, the rapid
arrival of the bromide tracer at the western extraction wells
and the slow arrival at northeastern well illustrates how
preferential flow zones control movement of fluids through
the Gallia.
Groundwater samples collected prior to system operation
indicated TCE concentrations ranged from 133 mg/L to
2148 mg/L with higher concentrations detected in the
northern and western monitoring wells. Soil samples
ranged from 3 to 4527 mg/kg, typically with increasing
concentrations with depth. Oxidant injection using vertical
wells in a 5-spot pattern was capable of providing suffi-
cient hydraulic control to deliver oxidant throughout the
45
-------
permeable zones of the Gallia within the pilot test area
within 3 days or less. The pre-test TCE concentrations
throughout the region were reduced to <10 mg/L through-
out all but the lower permeable eastern edge of the test
region within 3 days indicating an apparent reduction in
contaminant levels of -92% (Figure 4). Continued recir-
culation for 10 days, provided uniform oxidant delivery
throughout the entire area excluding the northeast extrac-
tion well (X770-EW01). As of two hours after recirculation
was terminated, TCE concentrations were reduced to be-
low detection limits (5 mg/L) at all but one location, (from
-1500 mg/L to 650 mg/L at X770-EW01), indicating an
apparent -97% reduction in TCE. When the TCE concen-
trations are compared to delivered oxidant concentrations
(Figure 5) at the same time intervals, the correlation of
oxidant delivery with TCE reduction is apparent. The slower
arrival of oxidant to the northeast portion of the region
(specifically X770-EW01 and X770-MW10) is also notice-
able and attributed to the lower permeability of this region.
It is important to note that while the permeability of the
Gallia varied by over an order of magnitude in the region
(Table 2), that oxidant was effectively delivered through-
out the area in a relatively short time frame (i.e., 3 days).
To evaluate the short term effects of the residual oxidant
and TCE within the region, groundwater samples were
collected at 1,2, 4, 6, 9, 21, and 27 hrs after system shut
down. Additionally, groundwater samples were collected
weekly for one month after recirculation. Although low lev-
els of TCE were present along the upgradient edge of the
1400
1200
> 1000
aoo
ooo
•400 ,
200
ft
* ,1 *
-r-f
" I >'•
I I
4.0 6.0
Elapsed Time (hrs)
10.0
0.0 10.0 20.0 30.0 40.0 SO.O 60.0 70.0
Rgure3. Bromide curves.
TCE
Concentra-
tions ppb
pre-injection
3 days
10 days
Rgure 4. TCE distribution within the pilot test area.
46
-------
NaMnO4
Concentra-
tions ppm
3 days
10 days
B/V-01
^ MW-10
EW-02
EWQi
MW-10
EVWK
Figure 5. NaMnO4 distribution within the pilot test area.
test area (14 mg/L at X770-MW08,39 mg/L at X770-EW04
and up to 738 mg/L at X770-EW01) when the system was
shut down, the TCE concentrations continued to decline
after recirculation was terminated. This decline in TCE
concentrations is due to the oxidation of organic com-
pounds by the oxidant remaining in the pore spaces. TCE
remained near or below detection limits (<20 mg/L) for
approximately two weeks after recirculation at which point
the TCE concentrations along the upgradient northern
edge (X770-EW01, X770-MW08, and X770-EW04) and
lower permeable eastern edge (X770-MW10) of the test
area began to increase (Table 3). The higher concentra-
tions detected at these locations can be attributed to the
migration of untreated upgradient plume groundwater into
the test region (northern edge) as well as TCE diffusion
from the less permeable zones of the test area (eastern
edge). It is important to note that near the injection well
(15 ft radius monitoring wells) where the delivered oxidant
mass was the highest, TCE concentrations remained be-
low detection limits one month after recirculation except
at the most upgradient location (35 mg/L at X770-MW01).
One month after system shut down, the oxidant concen-
trations within the region had declined to less than 1 mg/L
while a gradual increase in TCE concentrations was ob-
served. However, the apparent TCE mass reduction within
the mobile groundwater fraction of the test cell remained
at ~83%.The oxidant consumption overtime is as expected
and presumed to be due to a combination of factors in-
cluding oxidation of the natural organic material present
Table 3. TCE Concentrations (in mg/L) throughout the Pilot Test Area at Selected Intervals (see Rgure 2 for well locations).
X770-MW01
X770-MW03
X770-MW05
X770-MW06
X770-MW07
X770-MW08
X770-MW10
X770-MW12
X770-MW14
X770-EW01
X770-EW02
X770-EW03
X770-EW04
X770-IW01
Pre-test
baseline
(average)
820
462
904
724
895
1433
1420
1270
1591
1605
1164
2148
2040
133
3 days after
injection (one
pore volume)
ND
ND
2
ND
7
8
72
3
8
1043
20
37
40
41
End of inj/rec.
(10 dor 3 pore
volumes)
ND
3
4
ND
ND ,
14
2
ND
9
738
9
8
39
78
2 hours
after
shutdown
ND
ND
4
ND
1
ND
48
5
28
340
ND
11
3
ND
2 weeks
after
shutdown
8
ND
ND
ND
ND
70
553
ND
ND
1421
7
ND
132
7
4 weeks
after
shutdown
35
ND
ND
ND
ND
205
610
ND
10
1461
6
ND
355
18
47
-------
and anyTCE that was diffusing from the finer grained less
permeable zones in the Gallia or the overlying Minford
and advecting from the upgradient contamination into the
test area.
Efforts to examine reaction intermediates and products
(e.g., chlorinated organic acids or partially degraded
chlorocarbons) as well as system toxicity have been initi-
ated and revealed no adverse effects. For example, using
a Microtox Model 500 analyzer (AZUR Environmental,
Carlsbad, CA), groundwater samples were analyzed for
acute toxicity to a luminescent bacteria (Vibrio fischeri)
when exposed to 91% of the total sample concentration.
No toxicity was measured for background or post-recircu-
lation samples.
Single well tests were repeated at each monitoring loca-
tion at the end of the pilot test to quantify any potential
formation plugging, possibly due to MnO2 generation and
the potential redistribution of fines within the matrix. Post
injection/recirculation hydraulic conductivity values ranged
from 26 to 530 ft/d and were not significantly different from
the pre-test values (Table 2).
Conclusions
Oxidant injection using vertical wells in a 5-spot pattern
was capable of providing sufficient hydraulic control to
deliver oxidant throughout the permeable zones of the
Gallia within the pilot test area within 3 days. Pre-test
baseline TCE concentrations were reduced to <10 mg/L,
throughout all but the lower permeable eastern edge of
the test region within 3 days indicating an apparent reduc-
tion in contaminant levels of -92% within 3 days and -97%
reduction at 10 days (2 hours after the end of the test).
This oxidant delivery technique is applicable to relatively
permeable, saturated subsurface media contaminated with
dissolved and sorbed phase contaminants and potentially
ganglia of non-aqueous phase liquids (NAPLs). An un-
derlying aquitard is required in order to prevent spread of
contamination during the injection phase. It is noted that
application to situations with large masses of NAPLs (e.g.,
pools) may require substantially higher oxidant loadings
and potentially modified hydraulic control approaches.
The costs of the 5-spot ISCOR will vary depending on the
scale of the application and the performance goals re-
quired. A major cost savings with the recirculation deliv-
ery method is avoidance of costs associated with treat-
ment and discharge of extracted groundwater, which in
this pilot test was -240,000 gal. The estimated cost per
gallon of treated groundwater during the pilot test at a DOE
facility was ~$1.70/gal, including well installation costs but
excluding PORTS site support (e.g., health and safety,
health physics, construction engineering, waste manage-
ment, and project management oversight). While it is im-
portant to recognize that specific site conditions will greatly
influence the cost, the most significant costs associated
with the 5-spot ISCOR approach are well installation and
the oxidant. The cost of the injection wells and extraction
wells will be dependent on the size, depth, location, and
materials as well as the drilling subcontractor. The cost of
the oxidant will be dependent on the oxidant concentra-
tion delivered to the subsurface over some period of time
based on performance goals and initial contaminant con-
centrations.
Acknowledgements
This work was sponsored by the US DOE Portsmouth Site
Office, Piketon, Ohio. Dee Perkins, US DOE, and Dave
Taylor, Bechtel Jacobs Company LLC are acknowledged
for their oversight and guidance. Mark Mumby, Doug
Pickering, Bob Schlosser, and John Zutman are acknowl-
edged for their assistance in completing the field pilot test.
Considerable leveraging was also gained from the US DOE
Office of Science and Technology through the Subsurface
Contaminants Focus Area in situ chemical oxidation stud-
ies.
References
Bouwer, H. and R.C. Rice, 1976. A Slug Test Method For
Determining Hydraulic Conductivity of Unconfined Aqui-
fers with Completely or Partially Penetrating Wells, Water
Resources Research, 12(3):423-428.
Case, T. 1997. Reactive permanganate grouts for horizontal
permeable barriers and in situ treatment of groundwater.
M.S. thesis, Environmental Science & Engineering Divi-
sion, Colorado School of Mines, Golden, CO.
DOE. 1996a. In situ Remediation ofDNAPL Compounds
in Low Permeability Media: Transport/Fate, In situ Control
Technologies, and Risk Reduction. Joint project report
containing 16 focus papers authored by national experts.
Oak Ridge National Laboratory Report, ORNL/TM-13305,
for the U.S. Department of Energy, Office of Technology
Development. August, 1996.
Gates, D.D. and R.L Siegrist. 1995. In Situ Chemical Oxi-
dation of Trichloroethylene Using Hydrogen Peroxide.
ASCE Journal of Environmental Engineering 121 (9): 639-
644.
Horner, R.D. 1951. Pressure build-up in wells. Proceed-
ings of the Third World Petroleum Congress, Leiden, Hol-
land, E. J. Bill, editor. Sect 11:503-521.
Lowe, K. S., R. L. Siegrist, F. G. Gardner, D. A. Pickering
and T. C. Houk. 1998. In Situ Chemical Oxidation Recircu-
lation Pilot Test at the 5-Unit Investigative Area using Ver-
tical Wells. DOE/OR/11 -3011 &D1. Prepared by Oak Ridge
National Laboratory for Bechtel Jacobs Company LLC,
Piketon Ohio.
Siegrist, R. L., M. I. Morris, O. R. West, D. A. Pickering, et
al. 1994. X-231B Technology Demonstration for In situ
Treatment of Contaminated Soil: Field Evaluation of Mixed
Region Vapor Stripping, Chemical Oxidation, andSolidifi-
48
-------
cation Processes. ORNL/TM-12261. Oak Ridge National
Laboratory, Oak Ridge, Tennessee.
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. Environ. Science & Tech-
nology. 29(9):2198-2207.
Siegrist, R. L. 1998. "In Situ Chemical Oxidation: Technol-
ogy Features and Applications." Invited presentation at
the Conference on Advances in Innovative Groundwater
Remediation Technologies. December 15, 1998. USEPA
Technology Innovation Office.
Siegrist, R. L., K. S. Lowe, L. C. Murdoch, D. A. Pickering,
and T. L. Case. 1998a. In Situ Oxidation by Fracture
Emplaced Reactive Solids. J. Environmental Engineering.
125(5):429-440.
Siegrist, R. L., K. S. Lowe, L. C. Murdoch, W. W. Slack, and
T. C. Houk. 1998b. X-231A Demonstration of In situ
Remediation of DNAPL Compounds in Low Permeability
Media by Soil Fracturing with Thermally Enhanced Mass
Recovery or Reactive Barrier Destruction. ORNL/TM-
13534. Oak Ridge National Laboratory, Oak Ridge, Ten-
nessee.
Siegrist, R. L., K. S. Lowe, M. Urynowica, etal. 1999. Per-
meation and Dispersal of Reactive Fluids for In situ Treat-
ment of DNAPL Compounds in Low Permeability Media:
Field Studies. ORNL/TM-13597. In publication. Oak Ridge
National Laboratory, Oak Ridge, Tennessee.
Struse, A. M. 1999. Mass Transport of Potassium Perman-
ganate in Low Permeable Media and Matrix Interactions.
M. S. Thesis, Environmental Science & Engineering Divi-
sion, Colorado School of Mines, Golden, Colorado.
USEPA. 1998. In Situ Remediation Technology: In Situ
Chemical Oxidation. EPA 542-R-98-008. Office of Solid
Waste and Emergency Response. Washington, D.C.
West, O.R., R.L. Siegrist, J. S. Gierke, et al. 1995. In Situ
Mixed Region Vapor Stripping of Low Permeability Media.
1. Laboratory Experiments. Environ. Science & Technol-
ogy.29(9):2191-2197.
West, O.R., S. R. Cline, et al. 1997. A Full-scale Demon-
stration of In situ Chemical Oxidation through Recircula-
tion at the X-701B Site: Field Operations and TCE Degra-
dation. ORNL/TM-13556. Oak Ridge National Laboratory,
Oak Ridge, Tennessee.
49
-------
In-Well Aeration/Stripping Technology: Oveview and Performance Results
Tom McKeon
Project Performance Corp.
16935 SE 39th St.
Bellevue, WA 98008
425-643-4634
tmcvkeon @ ppc.com
Joseph Johnson and Carl Bach
The Boeing Company
In-well stripping is an in situ remediation technology for
VOCs in groundwater that can be applied as an alterna-
tive to pump/treat systems. This presentation provides a
summary of in-well stripping technology along with per-
formance data from several sites.
Summary of Technology
In-well stripping technology relies on pressurized air to
circulate and clean water flowing through a well. A pres-
surized air delivery line is placed in the well to deliver a
stream of air bubbles into the well. The rising column of
bubbles acts as an air-lift pump pushing the combined
stream of air/water up the casing while drawing contami-
nated water in through the extraction screen. As the air
bubbles and water move up through the casing, volatile
contaminants transfer from an aqueous dissolved phase
to a vapor phase in the air bubbles. A vacuum line is ap-
plied at the well head at a point above the packer to re-
cover contaminant vapors for treatment. A typical layout
of a stripping well is shown in Figure 1.
The casing is perforated above and below the packer with
an upper screen (the recharge screen) to allow the ground-
water to flow back into the aquifer. Modeling studies have
predicted and experimental studies have verified that the
reinfiltrating water completes a toroidal circulation pattern
within the aquifer.This permits a large portion of the water
to be drawn back in and reprocessed. The water discharg-
ing from the treatment well contains elevated levels of dis-
solved oxygen (typically at or near saturation). Some of
the important criteria for proper system design and imple-
mentation include the hydrogeologic circulation patterns,
mass transfer efficiency of the in-well stripping, and
geochemical stability calculations.
Case Studies
Performance data are available from operating in-well strip-
ping systems at a variety of sites. An example data set is
presented for a single project (site background, geologic
and hydrogeologic conditions, and concentration reduc-
tion profiles). Additional summary performance data for
other sites is also presented.
The case study presented is for an industrial site in Se-
attle, Washington. The property was utilized for manufac-
turing operations and groundwater beneath the site is con-
taminated with TCE. Operations were discontinued in the
late 1970's and contaminated soil was discovered in the
mid1980's.
AIR
INJECTION
T WATER TABLE
VOC-CONTAMINATED
WATER
UNCONTAMINATED WATER/AIR
WATER-AIR MIXTURE
CONTAMINATED WATER/VAPOR
Figure 1. Example In-well Stripping System
50
-------
The VOCs present in groundwater at the site include TCE
and several degradation byproducts. The observed distri-
bution and relative concentration ratios indicates that sig-
nificant degradation of TCE is occurring. Based on site
conditions and plume distribution, the half life for TCE is
estimated to be 0.4 years (assuming a first-order decay
process). The maximum TCE concentration detected in
water was at the solubility limit. Other important geochemi-
cal considerations at this site include the presence of dis-
solved iron and the equilibrium of calcite (CaCO3) with
respect to the groundwater.
The treatment system is comprised of two Density Driven
Convection (DDC) treatment wells and several perfor-
mance monitoring wells. Monitoring activity includes char-
acterization of VOCs in the inlet and outlet at each treat-
ment well as well as sampling from 12 monitoring wells.
All monitoring has been conducted on a quarterly basis
for the duration of system operation.
The site geology consists predominantly of unconsolidated
sand and silt to a depth of about 35 ft. The sediments are
relatively homogenous in an areal extent. However, dis-
continuous silt layers are encountered indicating signifi-
cant heterogeneity over the vertical profile. The bottom
portion of the aquifer (-35 to 45 ft bgs) grades from an
interbedded silt layer to a competent clay unit.
The hydraulic conductivity of the aquifer has been esti-
mated at 10'2 cm/sec (28 ft/day) based on an aquifer pump-
ing test. One of the treatment wells is completed into the
interbedded silt unit and the well yield is an order of mag-
nitude less than the well completed in the aquifer material
(2 gpm versus 20 gpm). The hydraulic gradient ranges
from about 0.004 to 0.007 ft/ft.
Since the site was suspected to contain a TCE NAPL, the
treatment well was configured to collect DNAPL in a res-
ervoir at the bottom from which it can be pumped to the
surface and removed as a separate phase liquid. Upon
startup of the system, separate free-phase NAPL was
recovered at a rate of approximately 1 liter/day. Recovery
of NAPL continued for about six weeks of operation and
declined to zero over the operating time. Total NAPL re-
covery was approximately 26 liters. The physical mecha-
nism inducing the DNAPL migration to the well is thought
to be the cyclical pumping of the recirculating well and the
constant vibratory energy into the aquifer formation. This
energy can cause the NAPL ganglia to coalesce and mi-
grate towards the recirculating well. When the well was
not operating, no TCE NAPL accumulated in the well.
Site Cleanup Standards
The site and all surrounding areas are industrial proper-
ties. Groundwater at the site and in the local area is not a
potable resource and water supply wells are not allowed
in the area under State regulations. The risk assessment
completed in the RI/FS identified the reasonable maxi-
mum exposure for groundwater as discharge to surface
water in the nearest river (approximately 4000 feet from
the plume and 2000 from the property boundary).
Based on the reasonable maximum exposure for ground-
water, cleanup levels for groundwater established in the
RI/FS have been based on applicable surface water qual-
ity criteria. The specific standards for protection of benefi-
cial use of surface water are from EPAs Water Quality
Criteria Guidance Documents. The cleanup goals are:
• Trichloroethene—2 mg/L based on the Lowest Ob-
served Effects Level (LOEL) for marine organisms
(Federal Register Notice 45 FR 79341).
• 1,2 Dichloroethene—224 mg/L based on the LOEL
for marine organisms (Federal Register Notice 45 FR
79332).
• Vinyl Chloride —0.525 mg/L based on the ambient
criteria for human-health fish consumption (Federal
Register Notice 45 FR 79341).
Performance Data
The monitoring data collected over the first 18 months of
system operation are presented in Figure 2. These three
wells represent the central area of the TCE plume and
cover the expected area of influence from the treatment
wells. The concentration reductions observed in the per-
formance monitoring wells are significant. Three of the
performance monitoring wells have had TCE concentra-
tion reductions of greater than 99%. Based on the observed
concentration reductions (99+% reduction at about 60 feet
away), the radius of the treatment zone is estimated to be
more than 2 times the distance from the inlet screen to
the water table. The mass removal for the system has been
monitored based on NAPL recovery and off-gas VOC con-
centrations. Mass removal during the first 6 months of
system operation is estimated to have been 400 Ibs.
TCE concentration in the one treatment well where NAPL
was recovered has been reduced by about 80%. The TCE
concentration in the other treatment well has been quite
variable with concentrations generally increasing as the
higher VOC concentrations are drawn up from the
interbedded silt zone. TCE is the primary VOC present in
this central area of the plume. The other VOCs present
(i.e., degradation products cis-1,2 dichloroethene and vi-
nyl chloride) have generally shown significant concentra-
tion reductions but the results are more variable.
A portion of the VOC plume (consisting primarily of the
degradation products cis 1,2 DCE and vinyl chloride) ex-
tends beyond the recirculation zone established by the
treatment wells. The longer-term plan for site remediation
and closure is to operate the treatment system for source
control and mass removal to the extent that continued
operation is practical and effective (continued concentra-
tion reductions are observed). After a point of diminishing
51
-------
-O - MW-9 - (99.9% Reduction at 50 feet down gradient)
p MW-10 — (99.9% Reduction at 60 feet cross gradient)
MW-8 — (99.3% Reduction at 40 feet up gradient)
S ampling Date
Figure 2. Monitoring data-TCE concentration deductions.
returns is reached, the active remediation will be stopped
and trans'rtioned into monitored natural attenuation.
Performance data are also available from other sites from
other industries. One example is a light industrial facility
located in the city of Coeurd'Alene. Groundwater beneath
the site is contaminated with TCE as a result of past waste
disposal practices at several different industries in and
around Coeur d'Alene.The layout of site monitoring wells
includes one well at the outlet of the treatment well
(screened over the water table), one well 19 feet away
(screened from 5 to 15 feet below the water table) and
another 89 feet away (screened from 10 to 20 feet below
the water table). The inlet screen is placed at a depth of
30 feet below the water table and the recharge screera is
placed at the water table. The system operates at a water
pumping rate of 35 gpm. Groundwater at the site is en-
countered at about 190 feet below ground surface.
The maximum groundwater concentration of TCE identi-
fied at the site was 1510 mg/L in a sample taken durmg
drilling. A sample taken just before the system began op-
erating (October 22 1996) had a concentration of 9€0
mg/L.
52
-------
The monitoring data collected over the first 12 months of
system operation are presented in Figure 3. The concen-
tration reductions observed in nearby monitoring wells are
quite dramatic. The closest monitoring well (MW-1, 19
feet away) has shown TCE concentrations reduced by 98
percent in the initial operating period. The most distant
monitoring well (MW-3, 89 feet away), has shown con-
centration reductions of 96 percent in over 12 months and
the downward trend in concentration appears to be con-
tinuing. Based on the observed concentrations reductions
(96% reduction at about 89 feet away), the radius of treat-
ment zone is more than 3 times the distance from the inlet
screen to water table.
M)
en
a
a
o
U
1200
1000
800
600
400
200
_£_ MWlUpgradientWeU
-MW3 89"feetaway
Oct-96 Dec-96 Feb-97 Apr-97 Jun-97 Aug-97 Oct-97 Dec-97 Feb-98 Apr-98
Sampling Date
Figure 3. Monitoring data.
53
-------
Air Sparging for Groundwater Remediation of Toluene and
Other VOCs: Case Studies
Alan Moore
ENSR Consulting and Engineering
35 Nagog Park
Acton, MA 01720
978-635-9500; (fax) 978-636-9180
amoore@ensr.com
www.ensr.com
Air Sparging (also called In situ Air Sparging, IAS) is an in
situ technique being used more and more commonly for
the In situ treatment of groundwater in the saturated zone
contaminated with volatile organic compounds. This pre-
sentation will:
• describe AS and how it works,
• explore the types of site conditions that AS is appli-
cable to,
• discuss the information needed to design a appropri-
ate AS system,
• the advantages of AS over other remedial methods,
• the operating results from three AS systems.
Implementation of AS consists of injecting air below the
water table to strip (transfer) volatile organic compounds
(VOC) from the dissolved and adsorbed phases to the
vapor phase.The injected air now containing the VOC rises
up to the vadose zone and eventually the atmosphere. In
many cases, soil vapor extraction (SVE) is used in con-
junction with AS in order to collect and treat these VOC-
containing soil vapors before they are emitted to the at-
mosphere. Gas transfer of oxygen also occurs from the
injected air to the groundwater. In this way, AS can oxy-
genate groundwater, increasing the redox potential and
the dissolved oxygen content. This dissolved oxygen can
stimulate aerobic biodegradation of biodegradable con-
tamination in the saturated and vadose zone. These pro-
cesses of biosparging and bioventing will not be ad-
dressed, since this conference deals with the abiotic pro-
cesses of AS.
The types of conditions where AS is most applicable in-
clude:
• Sandy or permeable silty sand soils;
• Homogeneous, non-stratified soils;
• VOC only contaminated sites.
The most critical aspect affecting successful AS is the
presence of heterogeneous soil that permits short-circuit-
ing and preferential flow, thereby not remediating all ar-
eas and depths. Additional information on AS can be ob-
tained from many sources, including the In situ Air Sparging
Engineer Manual of the US Army Corps of Engineers (EM-
1110-1-4005).
The three AS sites to be discussed have the following
characteristics:
• each about 1 acre in size;
• contaminated with toluene and/or volatile ketones;
• relatively homogeneous fine sand soils
AS has already remediated one of sites while the other
two sites have been undergoing air sparging for one and
two years respectively.
• Site data to be presented evaluated will include:
• System design features
• Groundwater concentrations and plume size changes
over time;
• Increase in dissolved oxygen from dissolution of
sparge air;
• Prevention of groundwater migration through use of a
"curtain" of AS points;
• Differences in removal rates for compounds with dif-
ferent Henry's Law constants;
• Typical problems with AS and innovative solutions.
54
-------
Application of VFLUX to Assessment of Soil Venting
Performance and Closure
Dominic DiGiulio
USEPA
National Risk Management Research Laboratory
919 Kerr Research Drive
Ada, OK 74820
580-436-8605; (fax) 580-436-5614
digiulio.dominic @ epamail.epa.gov
Dr. Varadhan Ravi
Dynamac Corporation
Ada, OK
Remediation Zone Paradigm
Evaluation of mass flux to and from groundwater is vital in
integrating venting application with progress in ground-
water remediation. Soil remediation goals must reflect the
realities of ground-water remediation. As illustrated in Fig-
ure 1, subsurface remediation are partitioned into three
distinct zones for performance evaluation purposes. Zone
1 is bounded on the upper end by the soil surface and on
the lower end by the seasonally high capillary fringe. It
consists of consistently unsaturated media (above region
of water table fluctuation) where mass flux to and from the
capillary fringe occurs through infiltration and diffusion,
and mass flux from the capillary fringe to the vadose zone
occurs through diffusion. Exclusion of water table fluctua-
tion allows considerable simplification of unsaturated flow
and transport in this zone.
Zone 2 consists of media periodically de-saturated due to
water-table fluctuation or dewatering. Often, it will consist
of a highly contaminated "smear" zone containing residual
NAPL where venting is combined with dewatering to re-
move contaminant mass from a localized region. Zone 2
is bounded on the upper end by the seasonally high cap-
illary fringe and on the lower end by the maximum depth
of the capillary fringe during ground-water dewatering.The
base of zone 2 is located between ground-water extrac-
tion wells to ensure minimum dewatering to the depth of
interest. The maximum depth of zone 2 will be dictated by
the vertical profile of contamination and feasibility of de-
watering to the depth of interest. If a smear zone exists in
highly permeable media where dewatering is infeasible or
requires large pumping rates, zone 2 can be thought of as
the region where another source control technology such
as sparging is used. If dewatering is not to be used in
conjunction with venting, then zone 1 is directly linked to
zone 3.
Zone 3 is bounded on the upper end by the capillary fringe
during dewatering and on the lower end by the maximum
depth of ground-water remediation. It represents media
that remains saturated during venting. Ground-water con-
centrations in zone 3 vary temporally and determine com-
pliance in zones 1 and 2. Zone 2 is in compliance when
ground-water concentrations are less than or equal to
ground-water concentrations in zone 3. This ensures that
remediation of groundwater within a smear zone will be
attempted to levels consistent with deeper contamination.
This avoids reduction of ground-water concentrations in
zone 2 to levels lower than what would occur through ver-
tical recontamination from zone 3.
In the case of a light non-aqueous phase spill where
ground-water concentrations within a "smear" zone are at
much higher levels than beneath the smear zone, low
concentrations in zone 3 forces aggressive dewatering and
venting application in zone 2. In the case of a dense non-
aqueous phase spill where ground-water concentrations
may be at high levels deep within an aquifer, remediation
within the dewatered region or zone 2 proceeds only to
the degree at which it is consistent with remediation in
deeper regions or zone 3. Zone 1 is in compliance when
mass flux to zone 2 through vapor diffusion and infiltration
is less than the mass flux from zone 2 to zonel through
vapor diffusion. This ensures aggressive venting applica-
tion in the vadose zone when groundwater has very low
levels of contamination, and less aggressive venting ap-
plication when vapor diffusion will result in recontamina-
tion of cleansed soils. Venting performance for zones 1
and 2 thus is dynamically linked to the performance of
ground-water remedial efforts in zone 3. Substantial
progress in remediating groundwater translates to increas-
ingly stringent soil venting performance standards, while
lack of progress in remediating groundwater translates into
55
-------
&&&&5??*1* * J
Figure 1. Schematic of three remediation zones.
less stringent soil remediation requirements. Regardless
of the strategy chosen to close an SVE site, there must be
a link between ground-water and soils remediation. Un-
fortunately, this aspect of closure is missing at most sites.
Modeling Approach
Mathematical models that simulate soil-water movement
to groundwater vary in complexity from simple and con-
servative algebraic water-balance equations that neglect
degradation and volatilization to more process descrip-
tive finite-difference and finite-element numerical codes
incorporating Richards' equation. Vadose zone modeling
should ideally: (1) be rigorous enough to incorporate ma-
jor fate and transport processes, (2) have input param-
eters that can be readily collected prior to and during vent-
ing operation, and (3) be sufficiently user-friendly to en-
able use by consultants and regulators. Unfortunately, si-
multaneous attainment of all of these goals is unrealistic.
Lack of data to support spatial discretization of soil prop-
erties (e.g., capillary pressure parameters, hydraulic con-
ductivity, porosity, bulk density, moisture content, total or-
ganic carbon content) and contaminant distribution, how-
ever, commonly limits the use of sophisticated mathemati-
cal models. One approach to this problem is to use ficti-
tious but "reasonable" input for numerical two- or three-
dimensional modeling. Another approach, which is per-
haps more practical and pursued here, is to start with the
data at hand and use fairly simplistic one-dimensional
analytical modeling with the assumption of constant infil-
tration to gain insight into the potential magnitude of sol-
ute transport to groundwater. More sophisticated numeri-
cal modeling can then be utilized if additional supporting
data becomes available. Regardless of sophistication, all
modeling is a simplification of reality containing consider-
able uncertainty in simulated results. Uncertainty analy-
sis is critical in assessment of model output. It is for this
reason that a venting closure strategy or any other va-
dose zone problem decision matrix should include provi-
sions for consideration of other factors. For the assess-
ment of venting performance and closure, these factors
include assessment of site characterization, design, moni-
toring, and rate-limited vapor transport.
To meet the needs of mass flux assessment, an analyti-
cal model termed VFLUX denoting vertical flux simula-
tion, has been developed. VFLUX is similar to the well
known model, VLEACH commonly used in the Superfund
program for vadose zone solute transport assessment ex-
cept that it allows for the presence of non-aqueous phase
liquid residual in soils, degradation, and a time-dependent
boundary condition at the water table interface. The time-
56
-------
dependent boundary condition is the centerpiece of the
mass flux assessment because it dynamically links per-
formance of ground-water remediation to vennnting clo-
sure. VFLUX a significant improvement over VLEACH
because of these increased capabilities and exact method
of solution. The governing equation with associated initial
and boundary conditions of VLEACH was solved numeri-
cally, even though it could have easily been solved ana-
lytically. The determination of mass flux requires calcula-
tion of a spatial derivative at the watertable interface, which
is prone to large error when spatial discretization is coarse
and a sharp concentration gradient exists between grid
blocks. The analytical solution of VFLUX provides exact
values of mass flux at any location. Also, in order to calcu-
late concentration profiles at a specified time, the trans-
port equation has to be solved numerically at each time
step starting from the initial distribution. Analytical meth-
ods allow direct solution at the time of interest. In addition,
analytical solutions are not subject to convergence and
stability problems arising from the use of coarse spatial
grids and large time steps. Thus, there is less chance that
an inexperienced user will generate inaccurate output due
to poor specification of numerical simulation criteria (e.g.,
grid size).
57
-------
In-SituTreatment of Hexavalent Chromium and VOCs Using
Recirculating Wells
T. McKeon, C.J. English, and S.R. Peterson
Project Performance Corp.
1639 SE 39th St.
Bellevue, WA 98008
425-643-4634
tmcvkeon @ ppc.com
Project Performance Corporation (PPC) has implemented
a pilot test of the Enhanced Density-Driven Convection
(EDDC) technology at the Boomsnub/BOC Gases
Superfund Site in Vancouver, Washington. The pilot test
work was funded by EPA Region 10. At this site, the EDDC
technology is being used for treatment of a plume con-
taining halogenated volatile organic compounds (VOCs),
as well as hexavalent chromium [Cr(VI)]. In-well air strip-
ping is being used to treat the VOCs and a chemical re-
ductant, sodium hydrosulfite, is being used to treat Cr(VI)
by reducing it to trivalent chromium [Cr(III)].The following
abstract provides a short project summary and prelimi-
nary evaluation and interpretation of results. The following
sections provide general background on characteristics
of the site and EDDC system and a discussion of results
of groundwater sampling and analysis.
Bench Scale Testing
Some of the initial work on the project was bench scale
testing of several different chromium removal processes
that could be implemented with a recirculating well. The
tests included batch and column tests of different reactive
media to remove the chromium in an infiltration gallery
constructed at the well. The different approaches tested
included the use of zero valent iron and various zeolites
treated with iron.The results of the column tests indicated
that the chromate could be effectively removed from solu-
tion. However, the chromate removal effectiveness for the
reactive media (both zero valent iron and iron-treated zeo-
lites) decreased significantly as the number of pore vol-
ume passed through the column increased. The reactive
capacities of these media were less than 10-3 g chromate/
gm of reactive media.These relatively low chromate load-
ing capacities were consistent with column experiments
conducted by Envirometal Technologies, Inc. in testing
conducted to evaluate the effectiveness of a permeable
reactive wall at the site. Based on the relatively low load-
ing capacity, the change-out interval for a reactive media
would be to frequent (on the order of weeks to months) for
system operation to be cost-effective compared with other
treatment options.
Based on the results of the bench scale testing, it was
decided that continuous injection of a reducing agent was
required to effectively reduce the hexavalent chromium
for a pilot scale or full scale system. Various sulfur based
reducing agents were considered and sodium hydrosulfite
was selected for use in the pilot test.
Site Characteristics
Site stratigraphy consists of a sandy, alluvial aquifer un-
derlain by a silt/clay aquitard. The hydraulic conductivity
of the alluvial aquifer is approximately 1 x 1 Q~z cm/sec and
the hydraulic gradient is approximately 0.003. The depth
to groundwater is approximately 10 ft and the aquifer is
approximately 85 ft thick. The plume is highly stratified,
with contamination present at the base of the aquifer. Ver-
tical plume characterization was conducted using direct
push (Geoprobe) sampling at 10 feet intervals.The princi-
pal VOCs present in groundwater at the site are trichloro-
ethylene (TCE) and trichlorofluoromethane (Freon-11).
Maximum concentrations of these contaminants at the
location of the pilot test are approximately 5,000 mg/L for
TCE and 3,000 ug for Freon-11 .The maximum Cr(VI) con-
centration detected at the pilot test location is 14,000 trig/
L in a discrete Geoprobe sample. Upgradient monitoring
wells have typically indicated Cr(VI) concentrations in the
range of 4,000 mg/L.
System Description
The EDDC well was operated with a pumping rate of ap-
proximately 12 gallons per minute (gpm).The air injected
into the EDDC well is provided by a blower located in an
equipment trailer adjacent to the EDDC well. The blower
is capable of providing a maximum air flow of approxi-
mately 65 standard cubic feet per minute (scfm), for a
maximum air-water ratio of 40:1 at the design pumping
58
-------
rate of 12 gpm. The air returned from the EDDC well flows
through a moisture separator to remove entrained liquid,
is heated to increase its temperature to approximately 95eF,
and is then treated using granular activated carbon (GAG)
to remove VOCs. The treated air then flows to the blower
and is reinjected into the EDDC well. A small amount of
carbon dioxide (approximately 5 standard cubic feet per
hour) is added to the injection air to prevent calcite foul-
ing. The moisture separator, air heater, GAG adsorbers,
and carbon dioxide supply are all located in the equip-
ment trailer.
Hexavalent chromium is being treated by injecting a solu-
tion of sodium hydrosulfite into the treatment zone in the
EDDC well. This reductant solution is stored in a 300-gal-
lon polyethylene tank adjacent to the equipment trailer and
is pumped to the EDDC well using a metering pump lo-
cated in the equipment trailer. The rate of reductant addi-
tion has been varied during the test, but has generally
been in the range of 1.5 to 3.0 liters per hour.
Sodium bromide was added to the reductant solution to
provide a conservative, nondegradable tracer in the water
being recharged from the EDDC well. The purpose of the
tracer was to verify that recirculation was occurring and to
determine the travel time from the EDDC well to monitor-
ing wells and piezometers. The concentration of bromide
in the effluent from the EDDC well depends on the con-
centration in the reductant solution, as well as the pump-
ing rate of the well and the rate of addition of reductant.
The latter two parameters varied during the first part of
the test as operating conditions were being optimized.
Thus, the concentration of bromide in the effluent was not
constant. The concentration of bromide in the reductant
solution was initially selected to maintain a bromide con-
centration of at least 10 mg/L in the well effluent in order
to provide a high confidence of detection.
Grpundwater samples have been periodically collected
from the EDDC well influent and effluent and from the
monitoring wells. Baseline samples were collected at or
before the time of startup. The wells were then sampled at
approximately 1, 2, 3, 4, 6, 8, 12, 14, 16 and 18 weeks
after start of operation.
Analysis of Results
The following provides a discussion of results of ground-
water sampling that has been performed to date. The dis-
cussion is organized by major analyte, i.e., bromide, chro-
mium, TCE, and Freon-11. To assist in interpretation of
results, a three-dimensional, analytical groundwater flow
model was used to predict groundwater flow paths and
travel times. This model was used with the site and oper-
ating characteristics described above. The interpretation
and analysis of the performance monitoring data have
been developed by staff from Project Performance Cor-
poration and have been reviewed by EPA's project team
for the site.
Bromide tracer and other data clearly
indicate that groundwater recirculation is
occurring.
The number of groundwater monitoring points and sam-
pling intervals are necessarily limited in terms of fully de-
scribing the development of the circulation cell. However,
detection of bromide in the EDDC well influent and in
monitoring wells was consistent with arrival time predic-
tions from groundwater modeling. Changes in concentra-
tion of other tracers in the groundwater (i.e., sulfate from
the reducing agent) are consistent with the arrival of the
bromide tracer.The groundwater monitoring data describ-
ing the bromide arrival and ultimate distribution in the aqui-
fer are also consistent with establishment of a circulation
zone. This is evidenced, in part, by the results of samples
from an upgradient deep monitoring point which are in-
dicative of convergence of upgradient flow to the base of
the circulation zone.
The use of a chemical reductant is effec-
tive in treating Cr(VI) to nondetectable
levels in the EDDC well.
Operations to date have shown nondetectable levels of
Cr(VI) in the EDDC well effluent, both with and without
significant excess reductant. The Cr(VI) is reduced to Cr(lll)
which is still present in the treated water. The rate of chro-
mium removal in the treatment well and monitoring wells
was measured as reductions of 76% in the influent water,
75% reduction in a monitoring well 20 feet away and 45%
in a monitoring well 45 feet away (total chrome). Hexavalent
chrome removal rates were essentially the same as the
total chrome. The rate of chromium removal in monitoring
wells located a distance away from the treatment well was
less than originally predicted.The specific processes caus-
ing the less than expected performance in Cr(VI) removal
are uncertain. Possibilities include a solid phase Cr(VI)
present on the aquifer matrix (partitioning to the soil), het-
erogeneity in the aquifer media resulting in silt lenses where
Cr(VI) removal is diffusion limited or other unknown fac-
tors.
Concentrations of Cr(VI) in the aquifer may
be controlled by the presence of a solid
phase.
This phenomenon is evidenced by the apparent rebound
in Cr(VI) concentrations at monitoring wells after arrival of
treated effluent. The presence of a significant inventory of
solid- phase Cr(VI) in the aquifer could have major im-
pacts on the effectiveness of groundwater remediation
systems. Even a low K value can significantly increase
the number of pore volume flushes needed to remove
contamination. Even with a low Kd value of 1, approximately
15 pore volume flushes would be needed to reduce dis-
solved-phase concentrations by 90%. Clearly, the pres-
ence of a solid phase will significantly increase the time
needed for aquifer remediation. Site-specific partitioning
59
-------
values have not been obtained to verify the presence of a
solid phase at the site.
Use of a chemical reductant for Cr(VI)
treatment may reduce the time needed for
remediation by allowing in situ as well as
in-well treatment.
Data collected during the period when significant excess
reductant was being added to the EDDC well suggest that
discharge of excess reductant resulted in in situ treatment
of Cr(VI). If a solid-phase Cr(VI) is present, excess reduc-
ing agent to treat Cr(VI) in situ could reduce the time
needed for cleanup. The use of excess reducing agent
would need to be carefully weighed against the resulting
increased sulfate concentrations.
Operation of the EDDC system has re-
sulted in significant VOC removal.
The intake screen for the EDDC well was located in the
zone having the highest concentrations of VOCs (the base
of the aquifer). The TCE and Freon-11 concentrations in
this influent water were reduced by 93% and 94%, re-
spectively, during the first 12 weeks of operation. Com-
parison of upgradient and downgradient concentrations
in groundwater indicates a VOC removal rate of approxi-
mately 99% as upgradient water passes through the treat-
ment zone established by the circulation cell. The average
TCE concentration reduction for all cross gradient and
downgradient monitoring wells was 70% and this average
percent reduction value includes two monitoring locations
which were initially at low concentrations to start with (due
to the highly stratified plume). The average TCE concen-
tration reduction for the wells with higher TGE levels near
the base of the plume was 86%.
The rate of reduction of TCE concentra-
tions at the bottom of the aquifer may be
affected by solid-phase TCE, which acts as
a continuing source of contamination.
Piezometer PZ-40 and the EDDC well are both screened
at the bottom of the aquifer where the highest levels of
VOC contamination are found. The concentration of TCE
in the sample from PZ-40 after 14 weeks is still quite high,
while EDDC influent samples have shown a TCE concen-
tration reduction of 93% over the first 14 weeks of opera-
tion. Groundwater modeling results and Br tracer data in-
dicate slow groundwater flushing in the zone monitored
by PZ-40 and a high rate of flushing in the EDDC intake
zone. The effect of the rate of flushing on concentration
reductions suggests additional contamination present in
the solid phase. In the zone of slow flushing, dissolution/
desorption of TCE from the solid phase would be expected
to maintain high dissolved-phase TCE concentrations for
longer periods of time. Conversely, silt lenses may exist
within the treatment zone that extend the necessary treat-
ment time frames. An understanding of the site lithology
and chemical partitioning to soil is important design infor-
mation in predicting the performance of the EDDC sys-
tem as well as any other remedial approach for ground-
water.
60
-------
An Overview of In Situ Chemical Oxidation Technology Features and Applications
Robert L. Siegrist and Michael A. Urynowicz
Colorado School of Mines
Environmental Science and Engineering Division
Golden, CO 80401-1887
303-273-3490; (fax) 303-273-3413
rsiegris @ mines.edu
Olivia R. West
Oak Ridge National Laboratory
Environmental Sciences Division
Oak Ridge, TN 37831-6036
Introduction
Subsurface contamination by toxic organic chemicals of
concern (COCs) is a widespread problem in soil and
groundwater at industrial and military sites in the U.S. and
abroad (Riley et al. 1992; NRG 1994; Siegrist and Van Ee
1994; ATSDR 1997). To mitigate current or future risks,
remediation approaches increasingly employ engineered
in situ technologies as well as natural attenuation pro-
cesses (NRC 1994; 1997; Kovalick 1998). In situ treat-
ment of the source or heart of a soil and groundwater plume
is being accomplished using mass transfer and recovery
methods (e.g., soil vapor extraction, air sparging, surfac-
tant/cosolvent flushing) and in place destruction methods
(e.g., bioremediation, oxidation/reduction), sometimes
aided by enabling techniques (e.g., soil fracturing, soil
heating). For treatment of the distal regions of groundwa-
ter plumes, natural biogeochemical attenuation and per-
meable reactive barriers are two strategies that are evolv-
ing (USEPA1997; NRC 1997; NAT01998; Kovalick 1998).
This paper provides an overview of chemical oxidation and
its application for in situ treatment of contaminated sites.
During the 1990's in situ chemical oxidation has emerged
as a promising treatment method (Siegrist 1998, USEPA
1998). Many toxic organics are amenable to rapid and
complete chemical destruction or to partial degradation
as an aid to subsequent bioremediation. The chemistry of
peroxide, permanganate, and ozone are highlighted along
with the results of lab- and field-scale applications. An
approach to design and implementation is outlined as well
as key issues and considerations that must be accounted
for. Further details will be included in a forthcoming guid-
ance document that is under preparation by the authors
and due to be published in 1999.
Chemical Oxidation of Target Organic
Contaminants
Development Background
Research and development into in situ chemical oxidation
has occurred over the past 10 years including laboratory
studies, field pilot demonstrations, and full-scale applica-
tions. As of this writing, a wide variety of organic contami-
nants in soil and groundwater media have been success-
fully oxidized by peroxide, permanganate and ozone oxi-
dants (Table 1).The key features of each oxidant system
as applied in situ are highlighted in Table 2. Optional deliv-
ery methods are illustrated in Figure 1 while example full-
scale applications of each oxidant are summarized in
Tables 3 to 5.
Early studies were primarily focused on hydrogen perox-
ide (H O2) or Fenton's Reagent (H2O2 plus Fe*2) and most
often for ex situ treatment of individual organics in water
(Barbeni et al. 1987; Bowers et al. 1989; Watts and Smith
1991; Venkatadri and Peters 1993), Subsequently, re-
search began to explore peroxide and Fenton's reagent
oxidation as applied in soil environments (Watts et al. 1990;
Watts and Smith 1991; Watts et al. 1991; Tyre et al. 1991;
Ravikumar and Gurol 1994; Gates and Siegrist 1993; 1995;
Watts etal. 1997).
Research was also initiated with alternative oxidants such
as ozone (Bellamy et al. 1991; Nelson and Brown 1994;
Marvin et al. 1998) and potassium permanganate (KMnO4)
(Vella et al. 1990; Vella and Veranda 1994; Gates et al.
1995; Schnarr et al. 1998; West et al. 1998; Siegrist et al.
1998a,b; 1999, Struse 1999). Field demonstrations and
61
-------
Tabto 1. Features of peroxide, permanganate and ozone oxidants as used for in situ remediation1
Features Peroxide (Fenton's) Permanganate
Ozone
Reagent Characteristics:
Form-
Point of generation -
Quantities available -
Oxidation In Situ:
Delivery Methods -
Dose concentrations-
Single/multiple dosing -
Amendments -
Subsurface transport -
Rate reaction/transport -
Companion t school. -
Oxidation Effectiveness:
Susceptible organlcs -
Difficult to treat organlcs -
Oxidation of NAPLs
Reaction products -
System Effects on Oxidation:
Effect of MOM -
Effect of pH-
Effect of temperature -
Effect of Ionic strength -
Oxidation Effects on System:
pH-
Temperature -
Metal mobility -
Permeability loss -
Liquid
Offsite, shipped onsite
Small to large
GW wells, Soil lances
5 to 50 wt. % H2O2
Multiple is common
Fe*2and acid
Advection
High or very high
None required
BTEX, PAHs, Phenols,
alkenes
Some alkanes, PCBs
Direct oxidation possible
Organic acids, salts, O,,
C02
Substantial gas evolution
Demand for oxidant
Most effective in acidic pH
Reduced rate at lower temp.
Limited effects
Lowered if inadeq. buffering
Minor to high increase
Potential for redox metals
Potential for reduction due
to gas evolution and colloids
Liquid or solid
Offsite, shipped onsite
Small to large
GW wells, soil lances, fractur.
0.02 to 4.0 wt.% MnO4
Single and multiple
None
Advection and diffusion
Moderate to high
None required
BTEX, PAHs, alkenes
Alkanes, PCBs
Direct oxidation possible
Organic acids, salts, MnO,,
C02
Minimal gas evolution
Demand for oxidant
Effective over pH 3.5 to 12
Reduced rate at lower temp.
Limited effects
Lowered if inadeq. buffering
None to minor increase
Potential for redox/exch.
metals
Potential for reduction due to
MnO, colloid genesis
Gas
Onsite during use
Small to large
GW sparge wells
Variable
Multiple
Often ozone in air
Advection
Very high
Soil vapor extraction
BTEX, PAHs, phenols,
Alkenes
Alkanes, PCBs
Direct oxidation possible
Organic acids, salts, O.,
C02
Minimal gas evolution
Demand for oxidant
Most effective in acidic pH
Reduced rate at lower temp.
Limited effects
Lowered if inadeq. buffering
Minor to high increase
Potential for redox metals
Potential for reduction due
to gas evolution and colloids
The information presented is for illustrative purposes only.
Permeation with
H2O2orKMnO4
Fracture sheets
with Fe° or MnO,
Mixing with H 2O2
orKMnCX
Sparging with
ozone
Flushing with
NaMnO,
Treatment walls of
redox solids
Figure 1. Example applications of in situ chemical oxidation systems that have been deployed.
62
-------
Table 2. Representative list of organics successfully treated by chemical oxidants
Organic Contaminant Media treated Oxidant
References
Trichloroethylene
Tetrachloroethylene
Carbon tetrachloride and t-1 ,2-DCE
Pentachlorophenol
2,4-dichlorophenol, dinitro-o-cresol
Trifluralin, hexadecane, dieldrin
Naphthalene, phenanthrene, pyrene
Octachlorodibenzo(p)dioxin
Motor oil/diesel fuel
PAHs and PGP
BTEXandTPH
Water (spiked)
Silica sand (spiked)
Silty clay soil (spiked)
Sand & clay soils (spiked)
Groundwater (spiked)
Groundwater (field site)
Groundwater (field site)
Silty clay soil (field site)
Water (spiked)
Silica sand (spiked)
Sand, clay soils (spiked)
Groundwater (field site)
Groundwater (spiked)
Groundwater (field site)
water (spiked)
Silica sand (spiked)
Natural soil (spiked)
Water (spiked)
Soil (spiked)
Clay, sandy soils (spiked)
Soil (spiked)
Soil (field site)
Soil and GW (field site)
Soil and GW (field site)
HA
HA
HA
H.O,orKMnO4
KMnO4
KMnO4
NaMnO,
KMnO4
HA
HA
HO orKMnO.
KMnO,
KMnO4
Ozone
HA
HA
HA
HA
HA
H2OzorKMnO4
HA
HA
Ozone
Ozone
Bellamy etal. 1991
Gurol and Ravikumar 1992
Gates and Diegrist 1995
Gates, Siegrist and Cline 1995
Case 1997, Van and Schwartz 1996
West et al. 1 998, Schnarr et al. 1 998
Lowe etal. 1999
Siegrist et al. 1999
Bellamy etal. 1991
Leugnetal. 1992
Gates, Siegrist and Cline 1995
Schnarr etal. 1998
Van and Schwartz 1996
Dreiling et al. 1998
Bellamy etal. 1991
Gurol and Ravikumar 1 992
Watts etal. 1990
Bowers etal. 1989
Tyre etal. 1991
Gates, Siegrist and Cline 1995
Watts, et al. 1991
Watts 1992
Marvin et al. 1 998
USEPA 1998
Table 3. Example applications of in situ treatment using peroxide (after USEPA 1998; Siegrist 1998)
Location (date)
Delivery
Ohio (1993)
Deep soil mixing
Colorado (1996)
Injectors into GW
Media and COCs
Silty clay soil with
TCE and VOCs.
Groundwater with BTEX.
Application method and results
H2O2+ compressed air injected during deep soil mixing to 15 ft. depth in
3 10-ft. diam. mixing zones, Up to 100 mg/kg mass reduced by 70%
including 50% due to oxidation.
H2O2 + chelated Fe injected via 8 to 14 lances and 7 trenches over 100
ft. x 100 ft. area. Four cycles at 4 to 6 days each. FTEX reduced from 25
Massachusetts (1996)
Injectors into GW
Alabama (1997)
Injectors into GW
South Carolina (1997)
Injectors into GW
TCA and VC in groundwater.
Soil with high levels of TCE,
DCE, and BTEX.
Deep GW Zone with PCE
and TCE DNAPLs in sandy
clay aquifer.
mg/L to <0.09 mg/L. property sold.
H2O2 + Fe + acid via 2 points over 3-days within 30 ft. D.W. TCA reduced
from 40.6 to 0.4 mg/L, VC 0.40 to 0.08 or NO mg/L.
H2O2+ FeSO4via 255 injectors into 8 to 26 ft. bgs zone of clay backfill in
2 acre waste lagoons. 20 days treatment time. 72000 Ibs. of NAPLs
treated down to soil screening levels.
H2O2 + FeSO4 via 4 injectors into zone at 140 ft. bgs beneath old waste
basin. 6-day treatment time. Treatment achieved 94% reduction in COCs
with GW near MCLs. GWTCE reduced from 21 to 0.07 mg/L; PCE from
119 to 0.65 mg/L.
63
-------
Table 4. Example applications of In situ treatment using permanganate (after USEPA 1998; Siegrist 1998)
Location (date)
Delivery
Ohio (1997)
Horizontal well
recirculation
Media and COCs
Groundwater with TCE
DNAPLs in a thin sandy
aquifer.
Application method and results
KMnO4 (2 to 4 wt.% feed) delivered by horizontal recirculation wells 200
ft. long and 100 ft. apart at 30 ft. bgs to treat 106 L zone of groundwater
over 30 days. TCE reduced from 820 mg/L to MCL in 13 of 17 wells.
-300 kg of TCE destroyed. Some MnO2 particles generated. Aquifer
heterogeneities noted.
Kansas (19S6)
Deep soil mixing
Ohio (1998)
Vertical well
Ohio (1996)
Hydraulic fracturing
TCE and DCE in soil and
groundwater to 47ft.
depth.
TCE in silty sand and gravel
groundwater zone at 30 ft.
bgs.
VOCs in silty clay soil from
ground level to 18 ft. bgs.
KMnO4 (3.1 to 4.9 wt.%) delivered by deep soil mixing (8 ft. augers) to 47 ft.
bgs during 4 days. TCE reduced from 800 mg/kg by 82% in the vadose zone
and 69% in the saturated zone (>8 ft. bgs). MnO/ depleted. Microbes
persisted. Comparison tests with mixed region vapor stripping yielded 69%
reduction and bioaugmentation were 38% reduction.
NaMnO4 (250 mg/L) delivered by 5-spot vertical well recirculation system (ctr.
well and 4 perimeter wells at 45 ft. spacing) for 3 pore volumes over 10 days.
TCE reduced from 2.0 mg/L to MCL. Oxidant gradually depleted in 30 d and
no Microtox toxicity. No permeability loss in formation.
KMnO4 grout delivered by hydraulic fracturing to create multi-layered redox
zones. Emplaced over 4 days but sustained oxidative zone for more than 15
mon. Dissolved TCE reduced from equiv. of 4000 mg/kg by 99% during 1 hr
of contact.
Table 5. Example field applications of in situ treatment using ozone (after USEPA 1998; Siegrist 1998)
Location (date)
Delivery
Media and COCs
Application method and results
Colorado (1997)
GW wells
Kansas (1997)
Injectors Into GW
Califormia (1998)
Injectors Into GW
Soil and GW with BTEX
andTPH.
PCE in GW.
Soil and GW with PAHs
and PCP.
Former gas station site. Sand/gravel to 43 ft. bgs with GW at 28 ft. 3 wells to 50
ft. depth cycling air/ozone with water recirculaton. 12 cycles per day. SVE also
continued. TPH in soil from 90 to 2380 mg/kg and BTEX at 7.8 to 36.5 mg/kg.
TPH in GW at 490 mg/L to NAPL. After 6 mon, GW below MCLs. No soils data.
System shut down.
Old drycleaners site. GW at 14 to 16 ft. bgs in terrace deposits. One sparge point
at 3 scfm at 35 ft. bgs. SVE wells in vadose zone. PCE in top 15 ft. of aquifer at
0.03 to 0.60 mg/L. Reduced 91% within 10 ft. of well. Comparisons with air only
indicated 66 to 87% reductions.
Wood treater site 300 ft. by 300 ft. in area. Stratified sands and clays. 4 multilevel
ozone injectors at up to 10 cfm. SVE wells in the vadose zone. After 1 mon,
PAHs at 1800 mg/kg reduced by 67 to 99% and PCP at 3300 mg/kg reduced 39
to 98%.
full-scale applications have evaluated alternative methods
of oxidant delivery including permeation by vertical lances
(Jerome et at. 1997), flushing by vertical and horizontal
groundwater wells (Lowe et at. 1999; Schnarr et at.
1998;West et at. 1998), and reactive zone emplacement
by hydraulic fracturing (Murdoch et at. 1997; Siegrist et at.
1998a,b; Siegrist et al. 1999).
Laboratory studies, field trials, and full-scale applications
have generated considerable insight into the process prin-
cipals and application of in situ chemical oxidation. In gen-
eral, the oxidants have been shown to be capable of achiev-
ing high treatment efficiencies (e.g., >90%) for unsatur-
ated aliphatic (e.g., trichloroethylene (TCE)) and aromatic
compounds (e.g., benzene), with very fast reaction rates
(90% destruction in minutes). Field applications have dem-
onstrated very high reductions in the mass of contami-
nants, but only where adequate oxidant was able to be
delivered and contacted with the target organics. These
field applications have been particularly valuable in that
they have clearly affirmed the control that field-scale re-
action and transport processes exert on design and per-
formance of in situ chemical oxidation.
Oxidants and Reaction Chemistry
Studies of the peroxide degradation of toxic organics in
soil and groundwater initially focused on petrochemicals
(e.g., naphthalene, phenanthrene, pyrene and phenols)
but later work encompassed chlorinated solvents (e.g.,
64
-------
TCE, tetrachloroethfylene (PCE)) (Watts et al. 1990; Watts
and Smith 1991; Watts et al. 1991; Tyre et al. 1991;
Ravikumar and Gurol 1994; Gates and Siegrist 1993; 1995;
Watts et al. 1997). Oxidation using HZO2 in the presence
of native or supplemental Fe+2 produces Fenton's reagent,
which yields free radicals (OH') that can rapidly degrade
a variety of organic compounds (Table 1).
However, the application of peroxide to soil and ground-
water systems involves a variety of competing reactions
as follows:
H2O2+Fe+2 -*OH-a+Fe+3+OH* (1)
H2O2+Fe+3-»HO2*+H++Fe+2 (2)
OH* + Fe+2 ->• OH' + Fe+3 (3)
(4)
HO2*+Fe+3 -> O2 + H+ + Fe"1"2
H2O2 +OHj -» H2O +HO2*
Rj + Fe+3 -*• Fe+2 + products
(5)
(6)
(7)
Hydrogen peroxide can also autodecompose in aqueous
solutions with accelerated rates upon contact with min-
eral surfaces as well as carbonate and bicarbonate (Hoigne
and Bador 1983) (eqn. 8).
H2O2
(8)
The simplified stoichiometric reaction for peroxide degra-
dation of TCE is shown in eqn. 9.
3H2O2+C2HC13->2CO2+2H2O-I-3HC1 (9)
Fenton's Reagent oxidation is most effective under very
acidic pH (e.g., pH 2 to 4) and becomes ineffective under
moderate to strongly alkaline conditions and/or where free
radical scavengers are present (e.g., CO3'2).The reaction
is strongly exothermic and can evolve substantial gas and
heat. The oxidative reactions are extremely rapid and fol-
low second-order kinetics (see following section).
For application in situ, there are three processes that have
been patented based on reaction chemistry and/or mode
of delivery: CleanOX, GeoCleanse, and ISOTEC meth-
ods. While the specifics of an application to a given site
will be very site dependent, in situ chemical oxidation with
peroxides has typically included H2O2 concentrations in
the range of 5 to 50 wt.% and where native iron has been
lacking or unavailable, ferrous sulfate is often added at
mM levels. In some cases acetic or sulfuric acids are also
added to reduce the pH to a more favorable acidic range.
Delivery methods have included common groundwater
wells or specialized injectors. In many cases, multiple
doses or application cycles are used to facilitate more
uniform delivery of reagents and efficiency of treatment.
Compared to peroxide, permanganate oxidation of soil and
groundwater has more recently been studied for in situ
treatment of chlorinated solvents (e.g., TCE, PCE) and
petrochemicals (e.g., naphthalene, phenanthrene, pyrene
and phenols) (Vella et al. 1990; Leung et al. 1992; Vella
and Veranda 1994; Gates et al. 1995; Van and Schwartz
1996; Schnarr et al. 1998; West et al. 1998a,b; Siegrist et
al. 1998a,b; Lowe et al. 1999; Siegrist et al. 1999; Struse
1999). The reaction stoichiometry and kinetics in natural
systems are quite complex and are as yet not yet fully
understood. Permanganate (typically as KMnO4 but also
available in Na, Ca, or Mg salts) can participate in several
reactions as determined to a large degree by system pH.
For example, between a pH of 3.5 and 12, permanganate
ion reacts slowly, but observably, to form manganese di-
oxide (eqn. 10). Above a pH of about 12, manganate ions
(Mn (VI)) may be formed (eqn. 11). Hydroxyl radicals may
also be formed in alkaline solutions (eqn.12). In slightly
acidic solutions, the permanganate ion can decompose
slowly to form manganese dioxide with a release of oxy-
gen (eqn.13). Below a pH of about 3.5, Mn(ll) cations are
formed (eqn.14). Under acidic conditions, the permanga-
nate ion can then oxidize the Mn(ll) to form manganese
dioxide (eqn. 15).
MnO4~ +2H2O+3e" -»MnO2(s)+4OH-
MnO4" +H2O -» MaO4~2
MnO4" + H2O -> MnO;2 + OH* + H+
4MhO4~ +4H+ -»3O2(g)+2H2O+MhO2(s)
MnO4" +8H+ +5e' -> Ma"1"2 +2H2O
; + 3Mn+2 + 2H2O -» 5MnO2 (s) +4H+
(10)
01)
(12)
(13)
(14)
(15)
The stoichiometric reaction for the complete destruction
of TCE by KMnO4 is given in equation 16.
2KMnO4 + C2HC13 -> 2CO2 + 2MnO2 + 2KC1 + HC1 (16)
Under neutral or acidic pH, oxidation is speculated to oc-
cur through the formation of a cyclic ester with further re-
action yielding organic acids and aldehydes as well as
CO2 and MnO2(s) (eqn.16) (Arndt 1981; Leung et al. 1992;
Van and Schwartz 1996). Halogenated substitution with
Cr may facilitate C-C cleavage during oxidation, although
the rate of reaction slows with increasing Ch substitution
(Yan and Schwartz 1996). For example PCE degradation
is slower than TCE. The reaction appears to be 2nd-order
with a rate constant of -0.6 L morV1 (in clean groundwa-
ter). Solution pH between 4 and 8 has little or no effect on
rate, but temperature does effect the rate as described by
the Arhenius equation (Case 1997). In alkaline solutions,
hydroxyl radicals may also be formed which can contrib-
ute to oxidative destruction (eqn.12). The reaction can in-
clude destruction by direct electron transfer or free radical
advanced oxidation. The pH of the reacting system can
decline to strongly acidic conditions (e.g., pH 2 to 3) de-
pending on the buffering capacity of the system. Key re-
action products can include intermediate organic acids
along with production of manganese oxide solids and chlo-
rides.
65
-------
Ozone gas has been injected into the subsurface along
with air sparging to remediate organics in groundwater
zones (Bellamy et al. 1991, Nelson and Brown 1994,
Marvin et al. 1998). Ozone can oxidize contaminants di-
rectly or form OH1 radicals, strong nonspecific oxidants.
Contaminants oxidizable by ozone include aromatics,
PAHs, chloroethenes, pesticides and aliphatic hydrocar-
bons. The simplified stoichiometric reaction of ozone with
TCE in water is given by eqn. 17.
O3 + H2O+C2HC13 -» 2CO2 + 3HC1 (17)
The oxidation reaction can be described as a pseudo-1st
order reaction with extremely fast kinetics. Due to ozone's
high reactivity and instability, O is produced onsite by elec-
trical generators. In addition, the high reactivity of ozone
and free radicals requires relatively closely spaced deliv-
ery points (e.g., air sparging wells). In some settings, scav-
engers for OH'can reduce the reaction efficiency. Gas and
heat can be evolved and in fine-grained sediments, par-
ticulates can conceivably be generated. Decomposition
of the ozone can lead to beneficial oxygenation and
biostimulation.
Kinetics of Reaction and Delivery Methods
Pseudo 1sl- or2nd-order kinetic models often describe the
kinetics of reaction of oxidants with target organics of con-
cern. For example during the past year, kinetic studies
have been completed for treatment of TCE over a wide
range of concentrations (0.5 to 800 mg/L) in simulated
and site groundwaters using permanganate solutions or
solids at stoichiometric dosages in the range of 5x to 10x.
The reaction order and kinetic parameters have been ex-
amined using pseudo 1st and 2nd-order kinetic models fit
to the data:
1st -order d[C1]/dt=-k1[C1]
2nd -orderd[C1]/dt = -k2[C1][C2]
Pseudol51 -order d[C1]/dt = -k'[C1]
*_l t_ fVI T f
(18)
(19)
(20)and(21)
where, d[C,]/dt = rate of change in concentration of the
target compound (MLT'1), k, = 1 st-order rate constant (T
)i Fci] = concentration of the target compound (ML3), k
= 2nd-order rate constant (L3M'1T-1), [CJ = concentration of
the oxidant (ML3), and k'= pseudo 1st-order rate constant
(T'1). With respect to TCE degradation by KMnO4 (in the
absence of natural organic matter (NOM)), the reaction is
clearly 2nd-order (eqn. 19). A recent study by the authors
with a range of TCE and KMnO4 conditions (0.6 to 6.3
mM) yielded an average k,, of 0.9 L mol'V1 with a %R.E.
of only 12%.
The kinetics of oxidation of a given target organic chemi-
cal are also affected by matrix conditions, most notably,
temperature and the concentration of other oxidant de-
manding substances such as natural organic matter
(NOM). Temperature effects can be described by the
Arhenius equation (Case 1997). However, the effects of
NOM (or minerals) on the oxidant demand rate and extent
is not clearly understood. Limited research suggests that
the rate of oxidant consumption is comparable or slower
than that of most target chemicals and that only a fraction
of the total NOM is susceptible to oxidation. It is clear how-
ever that the rate and extent of demand must be accounted
for or a kinetic model, such as equation 19, will grossly
over predict the rate of destruction of a target like TCE.
Moreover, if the NOM demand is too high, it will deplete
the oxidant and cause the reaction with the target organic
to cease altogether.
Another factor affecting the kinetics of destruction is the
phase of the target organic contaminant of concern (i.e.,
whether the organic is in the dissolved, sorbed, or non-
aqueous liquid phase). Most research has been conducted
with dissolved phase organics. Limited research with
Fenton's reagent (Tyre et al. 1991, Li et al. 1997, Watts et
al. 1997) and permanganate (unpublished CSM work),
suggests that sorption is not rate limiting under the usual
high oxidant doses and energetic reaction conditions. In
addition, research on the oxidative destruction of nonaque-
ous phase liquids is limited but preliminary results by
Urynowicz at CSM revealed that the rate of pure phase
dissolution and degradation is accelerated by permanga-
nate in the bulk solution (unpublished CSM work).
Given the relatively indiscriminate and rapid rate of reac-
tion of the oxidants with reduced substances for all oxi-
dants, as well as autodecomposition of peroxide, the
method of delivery and distribution throughout a subsur-
face region of interest is of paramount importance. Oxi-
dant delivery systems often employ vertical or horizontal
wells with forced advection to rapidly move the oxidant
away from the initial point of entry into the subsurface (Fig-
ure 2, Tables 3 to 5). Alternatively, oxidant delivery sys-
tems can employ vertical lance injection or sparging sys-
tems that enable high density delivery to minimize trans-
port distances and enhance contact with target chemi-
cals. In contrast to peroxide and ozone oxidants, perman-
ganate is less prone to decomposition and is more stable
and as a result, it can migrate by diffusive processes al-
beit at slow rates of transport (Struse 1999).
Design and Implementation
The standard-of-practice for the design and implementa-
tion of in situ chemical oxidation technologies is still evolv-
ing. While there have been numerous laboratory studies
and an increasing number of field-scale trials and full-scale
projects, there are still gaps in the current knowledge base
and performance deficiencies have been observed. Engi-
neering of in situ oxidation technologies must therefore
be done carefully with due attention to both reaction chem-
istry as well as delivery and transport processes. Figure 2
provides the conceptual design process that these au-
thors have developed and utilized during the past three
years. The design and implementation process should rely
66
-------
Reac
processes
Figure 2. Process design approach for in situ chemical oxidation.
on an integrated effort involving screening level charac-
terization tests and reaction and transport modeling, com-
bined with treatability studies at the lab and field scale.
Past experience and consideration of the current state-of-
knowledge suggest there are some key issues that must
be carefully considered during design process. These in-
clude: (1) oxidant suitability to degrade the target chemi-
cals at the rate and to the level required under given envi-
ronmental matrix conditions, (2) rate and extent of natural
oxidant demand, (3) viability of and method for oxidant
delivery throughout the subsurface region of interest, (4)
potential for oxidation-induced adverse secondary effects
(e.g. toxic byproduct formation, gas evolution, impurities
in the oxidant, particle genesis and permeability loss,
mobilization of metals), and (5) compatibility of oxidation
with other technologies (e.g., natural attenuation) and post-
treatment land use. The relevance of these issues and the
need for their accurate and complete delineation during
system design is highly dependent on the site specific
conditions and context of the application being contem-
plated.
Conclusions
In situ chemical oxidation is rapidly emerging as a viable
remediation technology for mass reduction in source ar-
eas as well as for plume treatment. The oxidants most
commonly employed to date include peroxide, perman-
ganate and ozone systems, with subsurface delivery to
groundwater by vertical or horizontal wells and sparge
points and to soil by lance injectors and hydraulic fractur-
ing. The potential benefits of in situ oxidation include the
rapid and extensive reactions with various COCs, appli-
cable to many biorecaciltrant organics and subsurface
environments, can be tailored to a site from locally avail-
able components and resources, and can facilitate prop-
erty transfers and Brownfields development projects. Some
potential limitations exist including, requirement for han-
dling of large quantities of hazardous chemicals to be in-
troduced due to the oxidant demand of the target organ-
ics and the unproductive oxidant consumption of the for-
mation, some COCs are resistant to oxidation, and there
is a potential for process-induced detrimental effects in-
cluding gas evolution, permeability loss, and mobilizing
redox sensitive and exchangeable sorbed metals. Full-
scale deployment is accelerating, but care must be taken
to avoid poor performance and unforeseen adverse ef-
fects. Matching the oxidant and delivery system to the
COCs and site conditions is the key to achieving perfor-
mance goals. Further development work is ongoing in
many areas.
Acknowledgments
Sponsorship of the work upon which this paper is based
was provided in part through the Subsurface Contaminants
Focus Area of the DOE Office of Science & Technology.
References
Arndt, D. 1981. Manganese Compounds as Oxidizing
Agents in Organic Chemistry. Open Court Pub. Co.,
LaSalle, IL.
ATSDR. 1997. Agency forToxic Substances and Disease
Registry (ATSDR). 1997 Priority List of Hazardous Sub-
stances. U.S. Dept. of Health and Human Services. Nov.
1997.
Barbeni, M., Nfinero, C., Pelizzetti, E., Borgarello, E., and
Serpone, N. 1987. Chemical Degradation of Chlorophenols
with Fenton's Reagent. Chemosphere, 16, pp. 2225-37.
Bellamy, W.D., Hickman, P.A., and Ziemba, N. 1991 .Treat-
ment of VOC-contaminated Groundwater by Hydrogen
67
-------
Peroxide and Ozone Oxidation. Res. Journal WPCF, 63,
pp. 120-28.
Bowers, A.R., Gaddipati, P., Eckenfelder, W.W. and
Monsen, R.M. 1989. Treatment of Toxic or Refractory
Wastewaters with Hydrogen Peroxide. Wat. Sci. Tech., 21,
pp. 477-86.
Case.T. L. 1997. Reactive Permanganate Grouts for Hori-
zontal Permeable Barriers and In s/YuTreatment of Ground-
water. M.S. Thesis, Colorado School of Mines, Golden,
CO.
Cline, S.R., O.R. West, N.E. Korte, F.G. Gardner, R.L.
Siegrist, and J.L Baker. 1997. KMnO4 Chemical Oxida-
tion and Deep Soil Mixing for Soil Treatment. Geotechnical
News. December, pp. 25-28.
Dreiling, D.N., LG. Henning, R.D. Jurgens, and D.L. Ballard.
1998. Multi-site Comparison of Chlorinated Solvent
Remediation Using Innovative Technology. Proc. First Int.
Conf. on Remediation of Chlorinated and Recalcitrant
Compounds, May, 1998, Monterey, CA.
Gates, D.D., and Siegrist, R.L. 1993. Laboratory Evalua-
tion of Chemical Oxidation Using HydrogenPeroxide. Re-
port from The X-231B Project for In Situ Treatment by
Physicochemical Processes Coupled with Soil Mixing. Oak
Ridge National Laboratory. ORNL/TM-12259.
Gates, D.D., and Siegrist, R.L. 1995. In Situ Chemical
Oxidation of Trichloroethylene Using Hydrogen Peroxide.
ASCE Journal of Environ. Engineering. 121, pp. 639-44.
Gates, D.D., Siegrist, R.L., and Cline, S.R. 1995. Chemi-
cal Oxidation of Contaminants in Clay or Sandy Soil. Pro-
ceedings of ASCE National Conference on Environmen-
tal Engineering. Am. Soc. of Civil Eng., Pittsburgh, PA.
Hoigne, J. and H. Bador. 1983. Rate Constants of Reac-
tion of Ozone with Organic and Inorganic Compounds in
Water. Water Research. 17:173.
Kovalick, W. 1998. Innovative Ground-WaterTechnologies:
Development Status and Trends. Proc. Conf. On Advances
in Innovative Ground-Water Remediation Technologies.
USEPATechnology Innovation Office and Ground-Water
Remediation Technologies Analysis Center. Dec. 15,1998,
Atlanta, GA.
Leung, S.W., Watts, R.J., and Miller, G.C. 1992. Degrada-
tion of Perchloroethylene by Fenton's Reagent: Specia-
tion and Pathway. J. Environ. Qua/. 21,377-8 1.
Li, Z.M, et al. 1997. Fenton Oxidation of 2,4,6-Trinitrotolu-
ene in Contaminated Soil Slurries. Environmental Engi-
neering Science, 14:1:55-66.
Lowe, K.S., F.G. Gardner, R.L. Siegrist, and T.C. Houk.
1999. Field Pilot Test of In Situ Chemical Oxidation through
Recirculation Using Vertical Wells at the Portsmouth Gas-
eous Diffusion Plant. Proc. EPA Conf. on Abiotic In situ
Technologies for Groundwater Remediation. August 31-
September 2,1999, Dallas Texas.
Marvin, B.K., C.H. Nelson, W. Clayton, K.M. Sullivan, and
G. Skladany. 1998. In Situ Chemical Oxidation of Pen-
tachlorophenol and Polycyclic Aromatic Hydrocarbons:
From Laboratory Tests to Field Demonstration. Proc. 1st
International Conf. Remediation of Chlorinated and Re-
calcitrant Compounds. May 1998, Monterey, CA. Battelle.
Murdoch, L., B. Slack, B. Siegrist, S. Vesper, andT. Meiggs.
1997b. Hydraulic Fracturing Advances. Civil Engineering.
May1997.pp.10A-12A.
National Research Council (NRG). 1994.
Alternatives for Groundwater Cleanup. Nat. Academy
Press, Washington, D.C.
National Research Council (NRG). 1997.
Innovations in Groundwater and Soil Cleanup. National
Academy Press, Washington, D.C.
NATO. 1998. NATO/CCMS Pilot Study Special Session on
Treatment Walls and Permeable Reactive Barriers. Feb-
ruary 1998, Vienna. EPA 542-R-98-003. May 1998.
Nelson, C.H. and R.A. Brown. 1994. Adapting Ozonation
for Soil and Groundwater Cleanup. Chemical Engineer-
ing. McGraw-Hill, Inc.
Ravikmur, J.X., and Gurol M. 1994. Chemical Oxidation of
Chlorinated Organics by Hydrogen Peroxide in the Pres-
ence of Sand. Environ. Sci. Technol. 28, 394-400.
Riley, R. G. and J. M. Zachara. 1992. Nature of Chemical
Contaminants on DOE Lands. DOE/ER-0547T. Office of
Energy Research, U.S. Department of Energy, Washing-
ton, D.C.
Schnarr, M., C. Truax, G. Farquhar, E. Hood, T. Gonully,
and B. Stickney. 1998. Laboratory and Controlled Field
Experimentation Using Potassium Permanganate to
Remediate Trichloroethylene and Perchloroethylene
DNAPLs in Porous Media. Journal of Contaminant Hy-
drology. 29:205-224.
Sherman, B.M., H.E. Alien, and C.P. Huang. 1998. Cata-
lyzed Hydrogen Peroxide Treatment of 2,4,6-Trinitrotolu-
ene in Soils. Proc. 13th Mid-Atlantic Industrial and Hazard-
ous Waste Conference. Technomic Publishers, Lancaster,
PA.
Siegrist, R.L. and J. J. van Ee. 1994. Measuring and Inter-
preting VOCs in Soils: State of the Art and Research
Needs. EPA/540/R-94/506. USEPA ORD, Washington, D.C.
20460.
Siegrist, R.L. 1998. "In situ Chemical Oxidation: Technol-
ogy Features and Applications." Invited presentation at
68
-------
the Conference on Advances in Innovative Groundwater
Remediation Technologies. 15 December 1998. USEPA
Technology Innovation Office, Washington, D.C.
Siegrist, R.L., K.S. Lowe, L.C. Murdoch, W.W. Slack, and
T.C. Houk. 1998a. X-231A Demonstration of In Situ
Remediation of DNAPL Compounds in Low Permeability
Media by Soil Fracturing with Thermally Enhanced Mass
Recovery or Reactive Barrier Destruction. Oak Ridge
National Laboratory Report. ORNL/TM-13534. March.
Siegrist, R.L. K.S. Lowe, L.C. Murdoch, T.L. Case, D.A.
Pickering, and T.C. Houk. 1998b. Horizontal Treatment
Barriers of Fracture-Emplaced Iron and Permanganate
Particles. NATO/CCMS Pilot Study Special Session on
Treatment Walls and Permeable Reactive Barriers. EPA
542-R-98-003. May 1998. pp. 77-82.
Siegrist, R.L., K.S. Lowe, L.G. Murdoch, T.L. Case, and
D.A. Pickering. 1999. In Situ Oxidation by Fracture
Emplaced Reactive Solids. J. Environmental Engineering.
Vol.125, No.5, pp.429-440.
Struse, A.M. 1999. Mass Transport of Potassium Perman-
ganate in Low Permeability Media and Matrix Interactions.
M.S. thesis, Colorado School of Mines, Golden, CO.
Tyre, B.W., Watts, R.J., and Miller, G.C. 1991. Treatment
of Four Biorefractory Contaminants in Soils Using Cata-
lyzed Hydrogen Peroxide. J. Environ. Qua/. 20, pp. 832-
38.
USEPA. 1997. Cleaning up the Nation's Waste Sites. Mar-
kets and Technology Trends. EPA 542-R-96-005. Office of
Solid Waste and Emergency Response. Washington, D.C.
USEPA. 1998. In Situ Remediation Technology: In Situ
Chemical Oxidation. EPA 542-R-98-008. Office of Solid
Waste and Emergency Response. Washington, D.C.
Vella, P.A., Deshinsky, G., Boll, J.E., Munder, J., and Joyce,
W.M. 1990. Treatment of Low Level Phenols with Potas-
sium Permanganate. Res. Jour. WPCF. 62 (7): 907-14.
Vella, P.A. and B. Veronda. 1994. Oxidation ofTrichloroet-
hylene: A Comparison of Potassium Permanganate and
Fenton's Reagent. 3rd Intern. Symposium on Chemical
OxHatbruIi: In Situ Chemical Oxidation for the Nineties.
Vol. 3. Technomic Publishing Co., Inc. Lancaster, PA. pp.
62-73.
Venkatadri, R. and R.W. Peters. 1993. Chemical Oxida-
tion Technologies: Ultraviolet Light/Hydrogen Peroxide,
Fenton's Reagent, and Titanium Dioxide Assisted Photo-
catalysis. J. Haz. Waste Haz. Materials. 10(2): 107-149.
Watts, R.J., Rausch, P.A., Leung, S.W., and Udell, M.D.
1990. Treatment of Pentachlorophenol Contaminated Soils
using Fenton's Reagent. Haz. Waste and Haz. Mater. 7,
pp. 335-45.
Watts, R.J., Leung, S.W., and Udell, M.D. 1991 Treatment
of Contaminated Soils Using Catalyzed Hydrogen Perox-
ide. Proceeding to the First International Symposium on
Chemical Oxidation. Technomic, Nashville, TN.
Watts, R.J. and Smith, B.R. 1991. Catalyzed Hydrogen
Peroxide Treatment of Octachlorobidenzo-pdioxin (occd)
in Surface Soils. Chemosphere, 23, 949-55.
Watts, R.J., A.P. Jones, P. Chen, and A. Kenny. 1997. Min-
eral-catalyzed Fenton-like Oxidation of Sorbed Chloroben-
zenes. Water Environ. Res., 69,269-275.
West, O.R., Cline, S.R., Siegrist, R.L, Houk,T.C., Holden,
W.L., Gardner, F.G. and Schlosser, R.M. 1998b. A Field-
Scale Test of In Situ Chemical Oxidation Through Recir-
culation. Proc. Spectrum '98 Intern. Conf. on Nuclear and
Hazardous Waste Management. Denver, Sept. 13-18, pp.
1051-57.
Van, Y. and F.W. Schwartz. 1996. In Situ Oxidative Dechlo-
rination of Trichloroethylene by Potassium Permanganate.
Proc. 3rd Intern. Conf. on Advanced Oxidation Technolo-
gies. October 26-29. Cincinnati, OH.
69
-------
Fenton Oxidation, Carbon Regeneration, and Groundwater Remediation
Scott G. Huling and Patrick K. Jones
USEPA
National Risk Management and Research Laboratory
PO Box 1198
919 Kerr Research Dr.
Ada, Oklahoma, 74820
580-436-8610 (fax) 580-436-8614
Huling.Scott@EPA.GOV
Robert G. Arnold and Raymond A. Sierka
Department of Chemical and Environmental Engineering
University of Arizona at Tucson
Tucson, AZ 85721
Dennis D. Fine
ManTech Environmental Research Services Corp.
PO Box 1198
Ada, OK 74820
Abstract
A ground water treatment process is described in which
contaminants are adsorbed onto granulated activated car-
bon (GAG) containing fixed iron oxide. Hydrogen peroxide
(H2O2) is amended to the GAG suspension and reacts with
the iron, forming hydroxyl radicals (-OH). The radicals react
with and oxidize sorbed and soluble contaminants regen-<
erating the carbon surface. Laboratory results are pre-
sented in which 2-chlorophenol (2CP) was first adsorbed
to GAG and subsequently oxidized via the Fenton-driven
mechanism. Transformation of 2CP was indicated by the
formation of carboxylic acids and Cl- release. 2CP treat-
ment efficiency, defined as the molar ratio of Cl' released
to H2O. consumed, increased with increasing amounts of
iron oxide and 2CP on the GAG. The extent of 2CP oxida-
tion increased with H2O2 concentration. Lower treatment
efficiency was evident at the highest H2O2 concentration
utilized (2.1 M) and was attributed to increased -OH scav-
enging by H2O2. Aggressive oxidation procedures used in
sequential adsorption/oxidation cycles did not alter the
GAG surface to a degree that significantly interfered with
subsequent 2CP adsorption reactions. Although process
feasibility has not yet been established beyond bench-
scale, experimental results illustrate the potential utility of
the adsorption/oxidation process in above-ground systems
or permeable reactive barriers for the treatment of con-
taminated ground water.
This paper has been published in the Journal of Environmental Engineering-The complete citation is:
Huling, S.G., R.G. Arnold, R.A. Sierka, KP.Jones and D. Fine. 2000. "Contaminant Adsorption and Oxidation via Fenton
Reaction." Journal of Environmental Engineering, 126(7), 595-600.
70
-------
In Situ Chemical Oxidation for Groundwater Remediation
Near a Waste Injection Well
Robert C. Starr
Lockheed Martin Idaho Technologies Company
Idaho National Engineering and Environmental Laboratory
PO Box 1625
Idaho Falls, ID 83415-3953
208-526-0174; (fax) 208-526-6852
starr@inel.gov
In situ chemical oxidation (ISCO) is being evaluated for
potential use in remediating groundwater contamination
near a former waste disposal well in a fractured basalt
aquifer. The site is Test Area North (TAN), a facility at the
Idaho National Engineering and Environmental Labora-
tory (INEEL), a U.S. Department of Energy laboratory in
southeastern Idaho.
The TAN facility is underlain by a sequence of plains-type
basalt flows intercalated with eolian and alluvial sedimen-
tary interbeds that were deposited between volcanic erup-
tions. Individual basalt flows average about 20 feet in ver-
tical thickness. Basalt flows have a typical morphology
controlled by its cooling history. The interior of a typical
flow is massive, while the upper and lower portions are
highly fractured, vesicular, and scoriaceous. Hence, flow
interiors tend to have relatively low permeability while the
interflow zones (the area at the top of one flow and the
bottom of the overlying flow) have relatively high perme-
ability. On a local scale the spatial distribution of perme-
able units is highly complex due to the complex morphol-
ogy of coalescing basalt flows. On a larger scale, the highly
permeable zones are sub-horizontal features that can be
correlated between well pairs. Sedimentary interbeds are
typically low permeability units that, where laterally exten-
sive, restrict vertical movement of water and solutes
through the system. The water table at TAN lies approxi-
mately 200 feet below ground surface. A laterally exten-
sive, fine-grained interbed at a depth of approximately 400
feet below ground surface provides a base to the con-
taminated upper portion of the aquifer beneath TAN.
Test Area North was developed during the early 1950s for
conducting research on nuclear power reactors, including
nuclear powered aircraft. Between 1953 and 1972, liquid
wastes were disposed by injecting them into the upper
portion (200 to 300 feet below grade) of the Eastern Snake
River Plain aquifer using injection well TSF-05. The wastes
included sanitary sewage, wastes from manufacturing,
machining, and maintenance activities, and from nuclear
reactor research. The injected wastes included
tetrachloroethene (PCE) and trichloroethene (TCE), and
radionuclides. Wastes apparently received little treatment
other than grinding prior to disposal. The wastes included
a solid fraction. This sludge accumulated in the well and
reduced the permeability of the formation, which contrib-
uted to a decision to discontinue use of the well for waste
injection. The remnants of this sewage sludge is still
present in the formation, and is referred to as 'organic
sludge'.
Groundwater contamination at TAN was discovered dur-
ing routine monitoring of water supply wells at TAN in 1987,
when PCE and TCE were detected.'Subsequent investi-
gations performed during a CERCLA remedial investiga-
tion and afterwards revealed that a TCE plume extends
approximately 10,000 feet downgradient from TSF-05.
There are smaller plumes of PCE and other chloroethenes.
Tritium is pervasive at sub-MCL concentrations through-
out the TCE plume, while 90Sr, 137Cs, and 234U are present
above risk-based concentrations in limited areas near the
former injection well. TCE is the compound of greatest
interest due to the large areal extent of the TCE plume.
The TCE plume appears to have been stable for the last
ten years. The highest contaminant concentrations are
found today - 27 years after waste injection ceased - in
the immediate vicinity of the injection well. Coupled with
several lines of evidence that show that organic sludge is
still present in the formation, this indicates that there is a
secondary contaminant source that continues to leach
dissolved contaminants into groundwater.
The secondary source includes several components. Or-
ganic sludge, which would be a good sorbent for
chloroethenes due to its high organic carbon content, is
known to be present in the formation near the injection
well. Although it has not been directly observed, non-aque-
ous phase liquid may be present, as suggested by per-
71
-------
cent range concentrations of chloroethenes in sludge
samples recovered from the injection well. Finally, dissolved
contaminants may have diffused into matrix porewater.
In 1995, a record of decision (ROD) was signed that speci-
fies the remedial actions for contaminated groundwater
beneath TAN. The ROD specifies that pump-and-treat
(P&T) will be used to prevent migration of contaminants in
groundwater that emanates from the hot spot near the
injection well, and also specifies that a separate P&T sys-
tem will be used for extracting dissolved contaminants from
the plume downgradient of the hot spot. The parties to the
ROD recognized that P&T has a poor track record for re-
turning aquifers to pristine conditions, and hence also stipu-
lated that several alternative technologies be evaluated
for possible use in enhancing or replacing one or both
P&T systems. In situ chemical oxidation was identified in
the ROD as a technology that may be applicable for
remediating the hot spot, and hence a treatability study of
ISCO is being conducted.
The objective of ISCO for long term application at the hot
spot is to prevent migration of chloroethenes at concen-
trations above MCLs in groundwater that emanates from
the hot spot.The approach would be to destroy enough of
the secondary source that mass transfer of contaminants
from the secondary source into groundwater would be
substantially reduced, such that dissolved concentrations
in groundwater would be less than MCLs. The goal of the
ISCO treatability study, which is currently in progress, is
to determine the feasibility of using ISCO for hot spot
remediation. The specific objectives are to evaluate the
effectiveness of ISCO for reducing mass transfer from the
secondary source, and to evaluate possible long term ef-
fects of ISCO other than secondary source destruction.
Based on previous work, potassium permanganate was
selected as the oxidant to be evaluated in the treatability
study. It was selected largely because it is a strong oxi-
dant, it is readily available, and it is easy to use under field
conditions.The reaction of potassium permanganate with
TCE is represented by
C2CL3H +2KMiO4 +2H2O -»
2MnO2(s) +3H+ +2HOO3~ + 2K+ +3CT
A laboratory treatability study was performed for the INEEL
by the Oak Ridge National Laboratory. The objectives of
this study were to measure the amount of oxidant con-
sumed (oxidant demand) by reaction with geologic mate-
rials and with anthropogenic wastes, and to measure the
rate of chloroethene oxidation as a function of oxidant and
chloroethene concentration.The oxidant demand of geo-
logic materials was found to be small (-0.1 g MnO4~ /kg),
while the oxidant demand of organic sludge was much
higher (70 g MnO4Vkg). Oxidation destroyed the organic
carbon fraction of organic sludge, thereby removing most
of the ability of the sludge to sorb chloroethenes.
TCE was rapidly oxidized in the lab experiments, which
were performed in sealed, gently agitated, batch reactors.
Initial concentrations of permanganate ranged from 0.01 %
to 3%. TCE concentrations typically declined from the ini-
tial concentration (0.1 to 1000 mg/L) to the MDL (5 to 10
mg/L) when the first (1/2 hour) or second (2 hours) samples
were collected. The data resolution was usually not suffi-
cient to allow calculation of a first order rate constant, and
thus the lower bound of a zero order rate constant was
calculated. These rates are >0.1 to >4000 mgTCE L'1 rr1.
Even the slowest reaction rates are fast relative to the
time scale at which ISCO would be applied at the field
scale, which would have contact times of days or longer.
TCE added as a NAPL to some batch reactors was rap-
idly destroyed. It is thought that oxidation in the aqueous
phase maintained very low dissolved concentrations, which
enhanced the rate of mass transfer from the NAPL into
the aqueous phase. Thus, although TCE is not oxidized in
the non-aqueous phase, the net effect of oxidation was to
reduce the mass of non-aqueous phase TCE.
The lab study results suggest that the chemical reaction
aspects of ISCO are feasible, that oxidation destroys TCE
in each of the secondary source forms thought to be
present at TAN, and that the loss of oxidant to unproduc-
tive reactions with geologic material are not exorbitant.
Hence, a field evaluation is being planned for the 2000
field season.
The objectives of the ISCO field evaluation are to evalu-
ate the effectiveness of ISCO for reducing the concentra-
tions of dissolved chloroethenes in and downgradient of
the hot spot, and to evaluate the effects of ISCO other
than secondary source destruction.The potential second-
ary effects include introduction of heavy metals into the
aquifer, and mobilization of previously injected contami-
nants. Industrial grade potassium permanganate contains
heavy metals as minor constituents, and if sufficiently high
permanganate concentrations are used the MCLs forthese
metals could be exceeded in the oxidant solution. Radio-
nuclides (90Sr and 137Cs) that were previously disposed
are now immobilized in the hot spot. The effect of ISCO on
the mobility of these contaminants will be investigated.
The approach to be utilized in the field evaluation is to
make a series of oxidant solution injections into the former
injection well (TSF-05), and then to assess the effect of
oxidation on VOC mass transfer from the secondary source
into groundwater. This cycle of oxidation - performance
assessment will be repeated several times.
Oxidant solution will be injected using the single well push
pull test (SWPPT) approach, in which both permangan-
ate and a conservative tracer (bromide) will be injected,
allowed to react, and then extracted. The recovery of oxi-
dant relative to bromide can be used to deduce reaction
rates and the mass of oxidant consumed by reactions in
the subsurface. A series of SWPPTs will be performed.
72
-------
The trend of oxidant demand (in a single test) as a func-
tion of the cumulative amount of oxidant injected will be
used to infer whether ISCO is having a significant effect
on the mass of oxidizable material (assumed to be pre-
dominantly organic sludge) in the hot spot.
The effect of ISCO on mass transfer from the secondary
source will be periodically assessed by injecting VOC-free
potable water into the aquifer, and monitoring VOC con-
centration rebound afterwards. Previous studies at TAN
have shown that VOC concentrations rise following injec-
tion of VOC-free water, presumably due to mass transfer
from the secondary source. Differences in pre-ISCO to
post-ISCO rebound behavior should reflect differences in
mass transfer from the secondary source, and hence be
a direct measurement of the effect ISCO on the second-
ary source.
In summary, laboratory studies suggest that the chemical
aspects of ISCO are favorable for remediating the sec-
ondary source of contamination in the aquifer near a former
waste injection well at Test Area North. A field evaluation
that will address the effectiveness of ISCO under field
conditions is planned.
73
-------
Groundwater Remediation Using the CleanOX In Situ
Chemical Oxidation Process
Robert A. Briggs
ManTech Environmental Corporation
14290 Sullyfield Circle
Chantilly.VA 20151
703-814-8364; (fax) 703-378-3396
rbriggs® mantech.com
In 1998, ManTech acquired the assets of CleanOX Envi-
ronmental Services, Inc. of Houston, Texas including the
equipment, personnel, and patents for implementation of
the CleanOX Process. CleanOX is a patented technology
that is applicable for in situ remediation of organic con-
tamination in groundwater and saturated soils, including
fuels, solvents, and pesticides. The CleanOX Process re-
duces groundwater contaminants to harmless carbon di-
oxide and water through an injection method that costs
significantly less than conventional environmental
remediation technologies. Since 1993, this innovative tech-
nology has been successfully used on scores of sites in
the United States.
The CleanOX Process oxidizes hydrocarbon and other
organic contamination in groundwater and saturated soil
using Fenton's Reagent. The CleanOX technology is an in
situ process utilizing the injection of proprietary liquid
chemical formulations that generate hydroxyl radicals, via
a Fenton Reaction, in the subsurface. The reagents are
applied to the subsurface through wells that are installed
within the contaminated portion of an aquifer. The reagents
are formulated based on site-specific conditions, but typi-
cally include acids, catalysts, and peroxide. Prior applica-
tions of the process have demonstrated immediate reduc-
tions in the concentrations of the following constituents:
Aromatic Compounds (BTEX)
Total Petroleum Hydrocarbons (TPHs)
Chlorinated Solvents
Polynuclear Aromatic Hydrocarbons (PAHs)
Nitro-aromatic Compounds
Organic Pesticides
Alcohols (phenols)
Mineral Oil Products
Polychlorinated Biphenyls (pcbs)
The CleanOX Process has been developed over the last
7 years for the remediation of organic contaminants in
groundwater, saturated soil and contamination within a
natural attenuation solution is viable, and to rapidly
remediate sites that are part of a real estate transfer or
brownfield project.
Key Advantages for applying the CleanOX Process include:
• Contaminant reduction in weeks to months;
• Mobile in situ treatment system that has limited dis-
ruption to on-site operations;
• CleanOX reagents applied to two-inch diameter moni-
toring wells;
• Applied under buildings and within operational areas;
• Requires no capital equipment purchase; and
• Eliminates long-term operation and maintenance
(O&M) costs.
The CleanOX Process is applied in a phased program:
• Bench test to determine the effectiveness of CleanOX
reagents;
• Field pilot test performed within source area to reduce
contaminant concentrations and obtain data for full-
scale application; and
• Full-scale remediation of groundwater plume to re-
duce or eliminate organic contamination.
The CleanOX technology has been applied at over 80 sites
within the United States including Superfund sites, For-
tune 500 industrial sites, commercial retail properties,
gasoline service stations, land disposal facilities, DoD in-
stallations, and dry cleaners.
Field applications usually require one to two weeks on-
site, with the chemical reactions being complete 48 hours
after application. Groundwater can be sampled from weeks
to months after the CleanOX application.
74
-------
CleanOX Process Chemistry
The basis of the CleanOX Process chemistry is related to
the well-known Fenton's Reaction wherein hydrogen per-
oxide reacts with a ferrous ion to produce an hydroxyl radi-
cal in the acidified aqueous medium containing the
contaminants(s) according to the equation:
H2O2+Fe+2-
>OH-+OH-+Fe+
The hydroxyl free radical OH$ is an extremely powerful
oxidizer which progressively reacts with organic contami-
nants yielding carbon dioxide and water. The basic oxida-
tion reaction (without inhibitor, initiator, enhancer, or cata-
lytic chemistry included) is:
Hydrogen Hydrocarbon Carbon
Peroxide Contaminant Water Dioxide
H,O
2^2
CnHx
H,O + CO,
This chemical form of the CleanOX Process indicates the
environmental compatibility of the technology. Laboratory
test data from post-treated groundwater grab samples has
not indicated the generation of harmful chemical bi-prod-
ucts from the application of the process. Some transitory
pH shift toward acidic occurs during treatment; however,
the acidic groundwater will equalize over time.
The oxidation process generates progressively shorter
chemical chains, cleaving chemical bonds, and eventu-
ally resulting in carbon dioxide and water. A proprietary,
empirically-derived computer modeling program developed
from laboratory and field applications over the last several
years, Geo-Environmental Modeling Software (GEMS), is
used to design each process application. GEMS gener-
ates a custom, site-specific treatment design and dosage
for the application of the process from the relevant quan-
tified engineering parameters and the hydrogeology, wa-
ter chemistry, and contaminants at a given site.
The design and treatment cost will depend upon the site-
specific parameters including contaminant concentration,
groundwater flow rate, porosity and permeability of sub-
surface soil, and the vertical and lateral extent of contami-
nation. The CleanOX Process is primarily directed toward
remediating dissolved-phase and adsorbed-phase organic
contamination, but has been applied to treat free-product
phases and capillary fringe soils. CleanOX can be applied
for augmentation of existing treatment technologies such
as pump and treat, air sparging, and other remediation
methods.
CleanOX Process Field Application
The field application of the CleanOX Process includes the
following elements:
• installation of application wells of similar construction
as monitoring wells;
• performing sampling and analyses of soil and/or
groundwater samples within the treatment area to es-
tablish baseline contaminant concentration conditions;
• mobilization of reagents and field technicians to the
site;
• attaching well head apparatus and chemical feed lines
to each application well;
• application of the site-specific reagent dosage to each
application well;
• field monitoring of groundwater and application well
parameters during reagent injection to ensure safety,
efficiency, and effectiveness of the process;
• performing sampling and analyses of soil and/or
groundwater within one month following application
to determine application effectiveness; and
• completing additional rounds of reagent application,
as needed, to effect further contaminant concentra-
tion reductions.
There are several critical elements that contribute to the
success of a in situ chemical oxidation application. The
most important element is safety. The application of
Fenton's Reaction chemistry to the subsurface environ-
ment results in vigorous, exothermic, pressurized subsur-
face environment. CleanOX applications are designed and
implemented so that the reagents are applied in a con-
trolled, conservative manner to avoid dangerous pressur-
ization and runaway reactions. Another important factor
that contributes both to safety and effectiveness is perox-
ide delivery rate and concentration. Again, our initial de-
signs for a given site are based on observations made
during the laboratory bench-scale testing of site soil and
groundwater and on the output of the GEMS model.These
design formulations of reagent concentration and delivery
rate are never immediately applied at a site. ManTech
personnel begin peroxide application at a site with a slow
flow rate of dilute reagents until monitoring of the subsur-
face confirms that hydroxyl radical formation reaction and
subsequent oxidation reactions have initiated and are at
steady-state. Reagent concentrations and delivery rates
are then slowly increased until design parameters are
achieved.
CleanOX Case Studies
Light, Non-Aqueous-Phase Liquids at a
DOD Fuel Farm in Puerto Rico
The CleanOX in situ chemical oxidation process was ap-
plied at an active military fuel farm for remediation of light,
non-aqueous-phase liquids (LNAPL) detected at the site.
Soil and groundwater impacts from petroleum product re-
leases (jet fuel and diesel) had been studied for several
years at the site under RCRA requirements. ManTech was
contracted by the prime consultant to conduct a pilot test
at the site to determine the applicability, feasibility, effec-
tiveness, and cost to address the LNAPL as a full-scale
remedial action at the site. Results of the pilot test will be
included in the prime consultant's corrective measures
study (CMS).
75
-------
The facility is located in Puerto Rico and has been oper-
ated by the DoD since the early 1950s. A combination of
fuel tank leaks, piping leaks, overfills, spills, and past main-
tenance practices has resulted in contaminated surface
and subsurface soil and groundwater at the site. DoD is
currently preparing an evaluation of remediation alterna-
tives that will meet USEPA and Puerto Rico regulatory
requirements. Cleanup of the LNAPL impacted areas at
the site by mechanical pumping (skimming) has been pro-
jected to take as long as 60 to over 100 years due to the
low hydraulic conductivity of the site soils (1O^4 to 10'5 cm/
sec) which originated from weathered volcanic bedrock.
ManTech conducted a bench test of the CleanOX Pro-
cess on groundwater and free product samples provided
by the prime contractor and prepared a Work Plan for
implementing a pilot test of the CleanOX Process. The
Work Plan provided detailed information regarding our
understanding of the site conditions and the results of the
bench test. The Work Plan was presented to the prime
contractor, DoD, the USEPA, and the Puerto Rico Envi-
ronmental Quality Board (EQB) in order to obtain regula-
tory approval and to negotiate the pre- and post-treatment
monitoring requirements. Further, the Work Plan detailed
the various DoD requirements and plans needed for the
project (e.g. spill control, environmental protection, qual-
ity assurance, etc.).
The pilot test consisted of installing two, two-inch diam-
eter stainless-steel application wells and four, two-inch
diameter PVC monitoring wells; collecting pre-and post-
treatment site data; and applying CleanOX Process re-
agents to degrade LNAPL. The pilot test wells were in-
stalled in an upgradient area of the LNAPL plume where
LNAPL measurements varied from several inches to sev-
eral feet thick in two existing monitoring wells. ManTech
collected baseline soil and groundwater samples to es-
tablish the initial contaminant mass loading adsorbed to
site soils, dissolved in site groundwater, and as a sepa-
rate phase LNAPL. CleanOX Process reagents were ap-
plied over a two week period while monitoring field-mea-
sured parameters such as dissolved oxygen, oxidation-
reduction potential, pH, temperature, specific conductance,
and LNAPL thickness. No measurable LNAPL was de-
tected in either the off-set monitoring wells (located 12 to
15 feet from application wells) or the application wells by
the conclusion of the pilot test.
Following the reagent application phase of the pilot test,
ManTech conducted a one-week, three-week, and 60-day
post-treatment monitoring program. No measurable
LNAPL was detected in any of the application or monitor-
ing wells during the one-week and three-week monitoring
events. Due to the size of the LNAPL plume, re-infiltration
of LNAPL was observed at the 60 day monitoring event
when LNAPL was measured at one to two inches in outly-
ing monitoring wells. Contaminant mass balance calcula-
tions, based on soil and groundwater analyses and on
LNAPL measurements, indicated that as much as 1,000
pound of petroleum product was oxidized during the pilot
test. Pilot test data are being compiled to estimate the
scope and cost of full-scale remediation of the site.
Light Non-Aqueous-Phase Liquids at an
Abandoned Industrial Facility: Boston,
Massachusetts
The CleanOX in situ chemical oxidation process was ap-
plied at an abandoned facility for remediation of light, non-
aqueous-phase liquids (LNAPL) detected at the site. The
former industrial facility was undergoing redevelopment
by a property developer, and LNAPL removal was required
to keep the redevelopment schedule on-track. ManTech
was contracted by the developer to address the LNAPL
as an interim remedial measure. Only four months were
required to complete the project, from the time ManTech
was first contacted through the completion of the field
application of the CleanOX process.
LNAPL was encountered during removal of the under-
ground storage tank (UST) used to store no. 2 fuel oil.The
LIST was located outside the boiler room wall and subse-
quent assessment activities showed that the LNAPL had
migrated under the building foundation.The LNAPL plume
was estimated at approximately 2,000 square feet. LNAPL
thickness was highest at the former tank bed at 1.5 feet
and was measured at 2 to 4 inches over the rest of the
impacted area. The geology underlying the site was char-
acterized as fine-grained sand with lenses of silty, fine sand
and silt, and trace clay. The hydraulic conductivity of sub-
surface soil was estimated using slug tests at 3.5 x 10~s
cm/sec. Regionally, bedrock exists at an estimated depth
of 50 to 100 feet below ground surface (bgs). The depth to
groundwater at the site was measured at approximately
20 feet bgs. Further, the basement of the on-site building
was approximately 15 feet below grade, making the depth
to groundwater under the building five feet.
Based on the geology of the site and the results of bench
testing, seven application wells were used during the
course the project. ManTech mobilized to the site with
materials and equipment to apply one cycle of treatment
to all six wells over a two week period. Monitoring during
the application indicated that oxidation-reduction poten-
tial (ORP) and dissolved oxygen levels increased signifi-
cantly in treatment area monitoring wells.
LNAPL measurements were made over a two-week sta-
bilization period. Most wells were found to have a slight
sheen of LNAPL, and the tank bed well had less than one
foot of LNAPL. A second cycle of CleanOX treatment was
implemented focusing more on the tank bed area where
LNAPL thickness was the greatest. Monitoring that was
performed following the second application cycle indicated
that no measurable product was detected in any of the
site wells. Groundwater sampling and analysis was then
conducted for use in risk analysis. Based on the risk analy-
sis results, the site owner will either apply for site closure
or may perform additional CleanOX process application
cycles as a polishing step to achieve site closure.
76
-------
Chlorinated Hydrocarbons at a Former
Manufacturing Facility in Arkansas
A CleanOX in situ chemical oxidation project was con-
ducted at a former manufacturing facility in Arkansas to
demonstrate that the process could rapidly degrade hy-
drocarbon contamination in groundwater and saturated
soil underlying the plant site. The perched water table
had been contaminated with chlorinated hydrocarbons
from previous material handling practices associated with
chlorinated solvents including trichlorethylene (TCE) and
1,2- dichloroethylene (1,2-DCE).
After a successful bench test, the CleanOX Process was
applied using two on-site groundwater monitoring wells.
These application wells were selected where the highest
concentration of chlorinated solvents had been detected
on-site. During the application of the CleanOX reagents,
elevated photoionization (PID) readings were not observed
during monitoring of the on-site wells. Post-application
head space monitoring determined that the vapor con-
centrations in the on-site monitoring wells had been re-
duced.
Analysis of groundwater samples following application of
the CleanOX Process indicated a 93% to 97% reduction
of the chlorinated solvent constituents in the two demon-
stration wells. Based on this dramatic reduction in con-
taminant concentration and the absence of environmen-
tal or human receptors, the site owner petitioned for clo-
sure pending the results of the on-going groundwater
monitoring program.
VOLATILE ORGANIC
Compounds (ug/L)
Trichloroethene
1 ,2-Dichloroethene
Vinyl Chloride
1,1-Dichloroethene
Toluene
Tetrachloroethene
Acetone
Chloromethane
Bromomethane
2-Butanone
Carbon Disulfide
Total VOCs
PRE-PILOT STUDY
MW-15 MW-108
20,000 1,100
11,000 340
730 180
41 <10
13 <10
<10 <10
<10 " <10
<10 <10
<10 <10
<10 <10
<10 <10
31,784 1,620
POST-PILOT (7 DAYS)
MW-15 MW-108
2,900
1,600
120
<10
<10
<10
150
83
36
14
<10
4,903
9,500
35
20
<10
<10
33
120
92
35
<10
24
454
POST-PILOT (56 DAYS)
MW-15 MW-108
1,700 <10
540 <10
22 <10
<10 <10
<10 <10
<10 55
<10 <10
<10 <10
<10 <10
<10 <10
<10 <10
2,262 55
77
-------
Degradation-Desorption Relationships in the Treatment of Contaminated
Soils Using Modified Fenton's Reagent
Richard J. Watts
Department of Civil & Environmental Engineering
Washington State University
Pullman, WA 99164-2910
509-334-3761; 509-335-7632
rjwatts ©wsu.edu
Despite over two decades of research and implementa-
tion of innovative technologies for the destruction of
biorefractory organic hazardous wastes, some contami-
nants are still not effectively treated, particularly if they
are sorbed to soils or if they are transformed into refrac-
tory products. For example, 2,4,6-trinitrotoluene (TNT)
remains a challenge to both biological and chemical treat-
ment processes, primarily because some of its degrada-
tion products polymerize or are degradative end points
(Bradley etal., 1994). For most hydrophobic contaminants,
desorption often limits the rate of treatment in soils and
groundwater. In pump and treat groundwater remediation,
hundreds of pore volumes of clean water are often re-
quired to achieve acceptable contaminant concentrations.
Although soil washing may be effective for enhancing con-
taminant desorption, the wash water (which usually con-
tains high concentrations of surfactants) must then be
treated, often in a separate reactor. Although a recent
theme in site remediation has been to deemphasize pro-
cess engineering in favor of natural attenuation as a solu-
tion for the clean up of hazardous wastes, the long-term
monitoring required for natural attenuation is not always
cost effective; furthermore, responsible parties remain li-
able until site closure is reached. Therefore, there is an
increased interested in systems for the active treatment
of hazardous wastes, such as advanced oxidation pro-
cesses (AOPs), in which soils can be treated rapidly.
The use of modified Fenton's reactions for the treatment
of contaminated soils and groundwater has recently seen
increased emphasis because of its ability to rapidly oxi-
dize a range of biorefractory contaminants. Such Fenton-
like reactions are based on the catalyzed decomposition
of hydrogen peroxide (H2O2) by iron (II) to form hydroxyl
radicals (OH»):
H,O, + Fe2+ 0 OH • +OH~ + Fe3+
(1)
Hydroxyl radical, a non-specific oxidant that reacts with
most organic compounds at near-diffusion controlled rates
(i.e., > 109 M'1 sec*1) (Haag and Yao, 1992), readily attacks
even highly chlorinated aromatic and olef inic compounds.
Therefore, Fenton-like reactions have the potential to oxi-
dize a large number of hazardous organic compounds
when they are present in aqueous solutions.
The majority of environmental contaminants are hydro-
phobic and are therefore sorbed to soils and subsurface
sediments. Because most contaminants are unavailable
for transformation processes in the sorbed state, desorp-
tion often controls the rate at which they are treated by ex
situ and in situ processes. For example, Sedlak and Andren
(1994) found that desorption rates of sorbed PCBs were
not enhanced by Fenton-like reactions and that the PCBs
were not available for oxidative attack by hydroxyl radi-
cals. However, the enhanced degradation of sorbed con-
taminants has been documented using more vigorous
Fenton-like reactions (hydrogen peroxide concentrations
• 2%) (Watts et al., 1993; Watts et al., 1994; Spencer et
al., 1996). For example, Watts et al. (1994) found that at
successively high hydrogen peroxide concentrations,
sorbed hexachlorobenzene was degraded more rapidly
than it was naturally desorbed. Gates and Siegrist (1995)
also found enhanced desorption of trichloroethylene at
hydrogen peroxide concentrations • 2%. Watts and Stanton
(1999) documented that, although hexadecane sorbed to
soils showed no measurable gas-purge (GP) desorption
over 72 hr, under Fenton-like conditions it was oxidized >
90% within 24 hr (with 82% recovery of 14C-labeled
hexadecane as 14C-CO2), confirming that sorptive pro-
cesses can be overcome by vigorous Fenton-like reac-
tions. Although the overall dynamics of the enhanced deg-
radation of sorbed contaminants are evident from recent
studies, the mechanisms occurring in these systems have
not previously been elucidated. The contaminants could
78
-------
be displaced from the soil surface by hydrogen peroxide,
hydroxyl radicals, or reductanis; alternatively, the contami-
nants could be oxidized in the sorbed state by a reactant
that is generated through vigorous Fenton-like reactions.
The standard Fenton's procedure for generating hydroxyl
radicals involves the slow addition of dilute hydrogen per-
oxide to a degassed solution containing the substrate and
excess iron (II). This procedure is well documented and
produces nearly quantitative stoichiometric generation of
hydroxyl radicals. However, the standard Fenton's process
is ineffective in degrading sorbed contaminants (Watts et
a/., 1993).
Hexachloroethane, which has negligible reactivity with
hydroxyl radicals, was transformed more rapidly in modi-
fied Fenton's reactions («2 % hydrogen peroxide) than it
was lost by gas-purge desorption, suggesting the exist-
ence of a non-hydroxyl radical mechanism. Addition of
excess isopropanol to scavenge hydroxyl radicals slowed,
but did not stop, the desorption and degradation of
hexachloroethane. In the presence of the reductant scav-
enger chloroform, hexachloroethane did not desorb and
was not degraded, suggesting that a reductive pathway in
vigorous Fenton-like reactions is responsible for enhanced
contaminant desorption. Fenton-like degradation of
hexachloroethane yielded the reduced product
pentachloroethane, confirming the presence of a reduc-
tive mechanism. In the presence of excess isopropanol,
toluene, which has negligible reactivity with reductants,
was displaced from the soil but not degraded. The results
are consistent with enhanced contaminant desorption by
reductants, followed by oxidation and reduction in the aque-
ous phase.
The modified Fenton's reactions that have been most ef-
fective in degrading sorbed contaminants and compounds
that are not reactive with hydroxyl radicals (e.g., hexachlo-
roethane) use excess hydrogen peroxide, which may pro-
mote the following reactions to generate other reactive
species, such as, perhydroxyl radical (HO •), superoxide
radical (O •-), and hydroperoxide anion (HO •):
Fe2+ + H2O2
OH « +OH~ + Fe3+
OH • +H202 -» HO2 • +H2O
HO2 • <-» O2 •- +H+ pKa = 4.8
HO • +Fe2+ ->• HO- + Fe3+
(1)
(2)
(3)
(4)
Hydroperoxide anion has been shown to be an effective
reductant of a number of oxidized species (Farhataziz and
Ross, 1977). Alternatively, hydroperoxide may be involved
in another redox couple to produce hydrated electrons.
Our initial results on Fenton-like reductions show that a
compound's reactivity with hydrated electrons correlates
well with Fenton-like reductions.
The use of vigorous Fenton-like reactions to treat sorbed
contaminants in soils and groundwater has a number of
advantages over other soil and groundwater treatment
processes, including 1) a combined enhanced desorption-
degradation process that can be conducted in situ or ex
situ, and 2) coexisting oxidative and reductive mechanisms
that may provide the potential to treat a wider range of
contaminants than can be treated by OH» mechanisms
alone. The potential for generating both oxidants and re-
ductants in modified Fenton's reactions has important
implications in waste treatment. A Fenton's system that
generates both species has the potential to desorb con-
taminants, oxidize reduced contaminants (e.g., monocy-
clic aromatic hydrocarbons, alkenes, PAHs), and reduce
oxidized contaminants (e.g., carbon tetrachloride, 1,3,5-
trinftrobenzene). Furthermore, many degradation products
that are relatively unreactive with OH» may be transformed
by reductants in the system, which may enhance the po-
tential for contaminant mineralization by Fenton-like reac-
tions. Therefore, vigorous Fenton-like reactions in which
reductants are generated may provide a universal treat-
ment matrix in which contaminants are rapidly desorbed
from solids and sludges, followed by transformation
through both oxidative and reductive mechanisms.
References
Bradley, P.M., F.H. Chapelle, J.E. Landmeyer, and J.G.
Schumacher. 1994. Microbial transformation of
nitroaromatics in surface soils and aquifer materials. Appl.
Environ. Microbiol. 60:2170-2175.
Farhataziz, M.J., and A.B. Ross. 1977. Selected specific
rates of reactions of transients from water aqueous solu-
tion. NSRDS-NBS 59. U.S. Department of Commerce.
Gates, D.D., and R.L Siegrist. 1995. In situ chemical oxi-
dation of trichloroethylene using hydrogen peroxide. J.
Environ. Engr. 121:639-644.
Spencer, C.J., P.C. Stanton and R.J. Watts. 1996. A cen-
tral composite rotatable design for the catalyzed hydro-
gen peroxide remediation of contaminated soils. Jour. Air
Waste Manage. Assoc. 88:971 -979.
Watts, R.J., RH. Chen, and A. Kenny. 1997. Mineral cata-
lyzed peroxide oxidation of chlorobenzenes. Water Environ.
Res. 69:269- 275.
Watts, R.J., S. Kong, M. Dippre and W.T. Barnes. 1994.
Oxidation of sorbed hexachlorobenzene in soils using
catalyzed hydrogen peroxide. J. Haz. Mater. 39:33-47.
Watts, R.J., and P.C. Stanton. 1999. Process conditions
for the mineralization of a hexadecane in soils using cata-
lyzed hydrogen peroxjde. Water Res. 33:1405-1414.
Watts, R.J., M.D. Udell and R.M. Monsen. 1993a. Use of
iron minerals in optimizing the peroxide treatment of con-
taminated soils. Water Environ. Res. 69:839-845.
79
-------
Containment Technology and Monitoring
Milovan S. Beljin
M.S. Beljin & Associates
9416 Shadyoak Court
Cincinnati, OH 45221
513-729-1602
Mbeljin@aol.com
Randall R. Ross
USEPA
National Risk Management Research Laboratory
Subsurface Protection and Remediation Division
P.O. Box 1198
Ada, OK 74820
580-436-8611; (fax) 580-436-8614
Ross.Randall@epamail.epa.gov
Subsurface vertical barriers have been used to control
ground-water seepage in the construction industry for
many years. Recently, much attention has been focused
on the use of containment technologies as supplemental
and stand-alone remedial options for hazardous waste
sites to prevent or reduce the impact of contaminant
sources on ground-water resources. Containment systems
can be classified as active (e.g., ground-water extraction
to control hydraulic gradient) or passive (e.g., physical
barriers only). Frequently, containment systems employ a
combination of active and passive components, depend-
ing on the remedial objectives and complexity of the
hydrogeologic setting. Such systems commonly incorpo-
rate low permeability vertical barriers (walls) keyed into
an underlying aquitard (floor), a low permeability cover
(cap) to prevent the infiltration of precipitation, extraction
and/or injection wells, trenches, and a network of moni-
toring wells.
Soil-bentonite (slurry) cutoff walls are the most common
types of vertical barriers used at hazardous waste sites.
Potential failure mechanisms of vertical barriers can be
classified as design errors, construction defects, and post-
construction property changes. Proper design will reduce
the potential for errors associated with wall configuration,
materials incompatibility and other factors. Construction
defects may form high hydraulic conductivity "windows" in
an otherwise low hydraulic conductivity barrier. Post-con-
struction property changes may result from wet-dry cycles
due to water table fluctuations, freeze-thaw degradation
or chemical incompatibility between the barrier compo-
nents and NAPLs.
The performance of hazardous waste containment sys-
tems has generally been evaluated on the basis of con-
struction specifications. Specifically, most systems are
required to maintain hydraulic conductivity of a vertical
barrier below a specified value, typically less than 1x10-
9 meters per second (m/s). During construction, the use of
appropriate field quality assurance (QA) and quality con-
trol (QC) testing is essential to ensure that the design
performance specifications are satisfied. Despite rigorous
field QA/QC, the unintentional formation of preferential
pathways within a vertical barrier is still possible. Whereas
the success of a construction dewatering vertical barrier
system may be judged by the ability of the barrier to limit
ground-water leakage to quantities that can reasonably
be extracted, there are no uniform methods to reliably
measure and document the hydrologic performance of
existing and proposed hazardous waste containment sys-
tems.
To determine whether a containment system is protective
of human health and the environment, leakage from the
system into the environment must be evaluated. Several
geophysical techniques have been identified as potentially
applicable for the indirect detection of defects associated
with vertical barriers. These techniques include ground
penetrating radar, electrical resistivity and continuous-wave
microwave technologies. Unfortunately, the resolution nec-
essary to identify small scale, yet potentially significant
breaches in a containment system may be beyond the
ability of much of the instrumentation currently available.
Additionally, the high costs of data acquisition and difficul-
ties associated with data interpretation are some of the
problems that plague many of the geophysical techniques.
80
-------
Hydrogeological techniques may be used to determine
whether leakage is occurring, and if so, estimate the rate
of loss from the system. If it is determined that a signifi-
cant volume of groundwater is exiting the system, then
the location and magnitude of the leak(s) must be estab-
lished to ascertain whether major repair efforts are nec-
essary to maintain a protective remedy. The hydraulic sig-
nature associated with a containment system leak is de-
pendent on the magnitude of the difference in hydraulic
head across the barrier wall, the extent of the leakage,
and the hydraulic conductivities of the vertical barrier, win-
dow and surrounding aquifer materials. Although spatial
variations in water levels have been used to identify gross
construction defects in ground-water containment systems,
no specific protocols/methodologies have been developed
to evaluate whether or not ground-water containment sys-
tems are operating as designed.
By analyzing the relationship of the hydraulic head distri-
bution inside and outside the containment system, it is
possible to assess whether or not the system is operating
as designed and which potential transport mechanisms
are most significant and require further evaluation. Under
ideal conditions, the hydraulic head will be lower inside a
containment system relative to that outside the system.
Relatively small differences in hydraulic heads inside and
outside a containment system indicate the lack of signifi-
cant active hydraulic forces for advective transport of con-
taminants. However, due to the concentration gradient
across the barrier, diffusive flux from the system is still
possible.
While the hydraulic heads inside and outside a contain-
ment system have been used to identify gross system fail-
ures, little attention has been given to monitoring the
changes in hydraulic heads with time. Temporal water level
fluctuations should be evaluated in conjunction with spa-
tial head variations to assess whether or not containment
is effective.
It may be possible to identify the general location of sus-
pected leaks indicated by spatial and temporal water level
fluctuations using existing monitoring systems. However,
the identification of specific leak locations and discharge
rates may require additional three-dimensional hydraulic
characterization and the installation of additional piezom-
eter clusters. Analysis of sufficiently detailed piezometric
head data may allow the identification of subtle changes
in the hydraulic head distribution, thereby indicating the
general locations of potential leaks in a vertical barrier.
Until recently, such an undertaking would be prohibitively
expensive, due to the high cost of installing the large num-
ber of monitoring wells necessary to adequately define
the hydraulic head distribution around a barrier wall. How-
ever, with the development of several relatively inexpen-
sive small diameter piezometer installation technologies,
it may be possible to install a sufficiently large number of
small diameter monitoring points to identify the hydraulic
signatures associated with containment system leaks.
Similarly, information obtained from cone penetrometer
surveys may be useful to identify similar hydraulic signa-
tures.
Additional research is needed to provide a better under-
standing of the complex hydraulics associated with leaky
containment systems. Such insight could be used to en-
hance existing performance monitoring systems and aid
in the design of new monitoring systems and allow the
estimation of monitoring point spacing requirements (ver-
tically and horizontally) necessary to detect containment
system breaches. Given the current interest in contain-
ment systems, as either supplemental or stand-alone re-
medial alternatives, and the lack of adequate performance
monitoring strategies at most existing hazardous waste
sites utilizing containment technologies, there is an im-
mediate need for a general protocol for evaluation of con-
tainment systems.
81
-------
Hydraulic Fracturing Overview and Issues
William W. Slack
FRx, Inc.
PO Box 37945
Cincinnati, OH 45222
513-469-6040; (fax) 513-469-6041
wslack@frx-inc.com
In situ remediation of soil or groundwater depends upon
accessing the contaminant, either to remove it, to destroy
it, or to immobilize it. Hydraulic fracturing provides the
unique opportunity to create zones in situ that effect or
enhance a chosen remedial process. In this sense, it en-
ables remedial processes, but is not a remedial process
by itself.
Hydraulic fracturing involves injecting fluid into an open
well at a pressure such that a crack, or fracture, forms
within the surrounding soil, and continuing injection to di-
late the fracture and fill it with beneficial material. Most
fractures are created with slurries of granular solids that
have desirable chemical or physical properties. The first,
and still most widely used application, couples sand-filled
fractures to soil vapor extraction (SVE) or pump-and-treat
recovery. This technique results in highly permeable lay-
ers that increase the flow rate of a well, typically by one to
two orders of magnitude. However, environmental appli-
cations of hydraulic fracturing have advanced beyond the
initial uses since it was first applied twelve years ago in
EPA research projects.
Uses of Hydraulic Fractures
Increases in discharge improve the performance of soil
vapor extraction, free-product recovery and other fluid-
based technologies in tight formations where they might
otherwise be infeasible. Typically, results increase dis-
charge by factors ranging from 10 to 100, and increase
the distance affected y the well 10 times of more com-
pared to control wells. Multiple fractures can be created to
focus subsurface flow into zones that can not be effected
by conventional wells.
Alternatively, fractures can be used to admit beneficial flu-
ids. Insufficient oxygen too frequently limits in situ
bioremediation, and hydraulic fractures can be used to
improve air injection during bio-sparging or bio-venting.
The flux of injected air around a fracture in low permeabil-
ity soil is particularly widespread and uniform. Addition-
ally, nutrient or oxygen-rich solutions could be injected
through fractures to stimulate biological systems. Likewise,
heated air or steam can be injected to improve the volatil-
ity and mobility of contaminants. Solvents and solvating
solutions, which mobilize or coalesce contaminants, can
also be injected through fractures.
Sand-filled fractures generally are used in these applica-
tions because of the low cost of sand. Other materials
may be considered if a secondary remediation technol-
ogy is to be implemented as a polishing process after pri-
mary recovery.
Hydraulic fracturing applications are not restricted to the
movement of fluid. Fractures can also be used for the
emplacement of granular solids that have the capacity to
destroy or immobilize contaminants. Hydraulic fracturing
methods have been used to create permeable reactive
barriers filled with zero valent iron for the purpose of de-
stroying chlorinated compunds dissolved in groundwater.
Other barriers have been created with granules of potas-
sium permanganate, which create highly oxidizing condi-
tions that destroy organic solvents and other compounds.
Less potent, oxygen-producing peroxides have been
placed with fracturing techniques to stimulate aerobic deg-
radation. Other examples of beneficial materials that can
be injected as fracturing slurries include activated carbon
or other materials for adsorption of contaminants, inocu-
lated porous solids to initiate bioremediation, and electri-
cally conductive materials as electrodes.
Variations in techniques and the spectrum of materials
that can be used permit hydraulic fracturing to target pe-
troleum hydrocarbons such as gasoline, diesel, jet fuel
and motor oils, chlorinated solvents such as trichloroeth-
ylene (TCE) and perchloroethylene (PCE), and inorganic
contaminants such as hexavalent chromium, nitrates, ar-
senic and selenium and even radionuclides such as ura-
nium and strontium. Fractures can be placed to access
82
-------
sub-surface source zones (hot-spots) or intercept contami-
nant plumes in soil for the purpose of destroying or re-
moving the contaminants.
Because hydraulic fracturing is a delivery technique, nei-
ther the list of target contaminants nor the repertoire of
remedial process can be considered exhaustive, and
materials yet to be invented may also be usable. Some of
the features and advantages are listed in Table 1.
Hydraulic fracturing is applicable to a wide variety of site
conditions and contaminants. It can be applied at depths
of 5 to 150 feet or greater in either saturated or unsatur-
ated soils. It has been demonstrated from Maine to Cali-
fornia in a variety of soils including low permeability clays
deposited as glacial drift, lacustrine or overbank sediments,
residuum on limestone, saprolite, in shales and siltstones,
and in sand and gravel aquifers. Fractures can be created
around and under most structures. The technology ap-
pears viable for all but the smallest jobs.
Methods of Creation
The fundamental processes of creating fractures by in-
jecting a fluid is straightforward: the fluid is injected into a
borehole until the pressure exceeds some critical value
and a fracture nucleates. This method can create a frac-
ture in most naturally occurring materials, from rock to
unlithified sediment or soil. Once the fracture nucleates,
fluid continues to be injected, propagating it away from
the borehole.
The different methods of creating hydraulic fractures share
a common feature in preparation and handling of the in-
jected slurry. While fractures can be created with water,
most hydraulic fracturing relies uponviscosified water that
can more effectively transport solid particles. Guar gum
gel is the most typical viscosifier used. Guar gum is a food
additive derived from the guar bean and is composed of
short-chained polysaccharides. Mixed with water, guar
gum forms a fluid of moderate viscosity, similar to mineral
oil. Adding a crosslinker causes the polymer chains to link
and form a thick gel capable of suspending high concen-
trations of heavy solid granules.The non-Newtonian char-
acter of the gel permits it to be pumped and to flow through
fairly small openings. The properties make guar gum gel
ideal for filling fractures with solid material. An enzyme
added to the gel eventually breaks the polymer chains,
allowing recovery or dispersal of the thinned fluid from the
fracture. In some cases the reactivity of the solid material
to be transported into the fracture precludes the use of
guar gum gel. For instance, potassium permanganate
granules are best carried as aqueous slurry viscosified by
mineral additives.
Hydraulic fracturing with slurries composed of viscosified
water and solid granules requires several specialized
pieces of equipment. Tanks, hoppers, vats, etc are needed
to handle the materials. A mixer is required to blend the
water, viscosifiers, and other reagents with the solids. The
method also requires a pump capable of handling slurry
that contains high concentrations of solid particles, which
may be abrasive. In most cases, positive displacement
pumps are used.
Variants among hydraulic fracturing methods have been
adopted to accomplish various objectives. For instance,
fracturing methods used to create vertical permeable re-
active barriers in an aquifer differ from those that provide
flow enhancement to wells. Generally, the different meth-
ods seek to exploit local geophysical conditions to influ-
ence the form (orientation, extent, aperture, and direction)
of the resultant fracture.
Table 1. Features and Advantages of Hydraulic Fracturing
Product Features
Advantages
It is a delivery system
Enhances in situ remediation & control
Requires fewer surface access points or wells than
conventional processes
Creates in sity remediation zones away from the
creation well
Can be used to enhance a variety of remediation processes
Can place both solids & fluids at desired locations
Can be adapted to new remediation processes; and thus will remain viable in many
situations
Remediation can be accomplished w/o massive & expensive excavation/removal of
soil
Limits exposure to workers/others
Fewer wells=less cost for installation/O&M
Can make a particular remediation economically feasible
Less disruption of surface activities
Remediation can be established underneath buildings/etc., where surface access
is limited
Works in conjunction w/ other remediation processes Operation/monitoring/verification can follow conventional practice
83
-------
Wellbore configuration during nucleation influences the
initial orientation upon nucleation, so methods intended
to create horizontal fractures utilize a short section of open
borehole and a mechanical cutter or jet to create a hori-
zontal notch to focus the nucleating stresses in the hori-
zontal plane. Similarly, a long open borehole and vertical
notch favor nucleation of a vertical fracture.
Hybrid techniques that combine hydraulic fracturing with
high pressure jets are promising methods for creating ver-
tical barriers filled with reactive materials in the saturated
zone, particularly at depths where excavation is infeasible.
A recent demonstration showed that such hybrid tech-
niques were capable of creating a vertical reactive barrier
filled with zero-valent iron at depths below 50 feet.
Some methods attempt to exploit or influence the state of
stress in the formation and thereby control the orientation
of the fracture. Fracture orientation, after the fracture is
nucleated, js strongly controlled by the state of stress in
the formation. A fracture tends to form in a plane perpen-
dicular to the direction of the least stress, i.e. the fracture
faces find the easiest direction to push the soil apart. If
fracturing pressure is exerted simultaneously in parallel
features in a formation, such as two adjacent boreholes,
the stress field between the features can cause the two
fractures to be parallel and to coalesce into a monolithic
unit upon intersection. This feature can be useful for cre-
ation of continuous barriers.
Methods to control of fracture aperture depend upon the
soil. In low permeability media, aperture can be increased
by using slurry of greater viscosity or injecting at a greater
rate, although substantial increases can not be obtained
without increasing the extent of the fracture. In permeable
sand and gravels, penetration of the slurry liquid into the
surrounding soil can stall the propagation of the fracture,
causing subsequent injection to further dilate the fracture.
The process, known as leak-off, requires careful control
of slurry rheology.
Design Considerations
Several factors affect the choice of hydraulic fracturing as
a remediation tool at the site. Generally the target forma-
tions must be below a depth that can be excavated or
surface structures and activities preclude excavation. For
the installation of permeable barriers, trenching equipment
can reach depths of 50 feet or more. If the contamination
at the site is confined to the upper few feet, removal will
probably be more effective than any in situ process, and
hydraulic fracturing would not be required.
The presence of surface structures can complicate the
placement of fractures. Although fracturing methods can
be employed in close quarters, such as inside buildings or
crawl spaces, consideration must be given to the interac-
tion between the structures and the fractures. Building foun-
dations or weight may restrict the propagation of the frac-
ture. Light buildings, such as wooden frame constructed
on a slab, will effect only the shallowest fractures whereas
multistory masonry buildings can deflect fractures that are
intended to penetrate underneath the foundation. Sensi-
tive structures, such as rotating process equipment, pre-
cision aligned machines, or storage tanks for dangerous
chemicals, may not tolerate movement of the ground sur-
face, which is an inevitable result of creating a space in
the subsurface.
In order for fracturing to enhance remedial processes that
involve fluid delivery or recovery, the fracture must be sig-
nificantly more permeable than the enveloping formation.
Therefore, the relative improvement resulting from hydrau-
lic fractures increases as the permeability of the forma-
tion decreases. In most cases, rock or formations of clay
and silt are best suited to hydraulic fracturing because
they have the lowest permeability.
84
-------
In Situ Chemical Treatment Using Hydraulic Fracturing To
Emplace Fe° Metal and KMnO4 Reactive Solids
Robert L Siegrist and Amanda M. Struse
Colorado School of Mines
Environmental Science and Engineering Division
Golden, CO 80401-1887
303-273-3490; (fax) 303-273-3413
rsiegris@mines.edu
Kathryn S. Lowe
Oak Ridge National Laboratory
Life Sciences Division
Grand Junction, CO
Lawrence C. Murdoch
Clemson University
Clemson, SC
FRx, Inc.
Cincinnati, OH
Introduction
Petrochemicals (e.g., benzene) and organic solvents (e.g.,
trichloroethene (TCE)) are common and problematic con-
taminants of concern (COCs) at federal facilities and in-
dustrial sites across the U.S. and abroad (Riley et al. 1992;
Siegrist and van Ee 1994: API 1995; DOE 1996; USEPA
1997). They are often present in source areas near stor-
age tanks and land treatment sites and in associated soil
and groundwater plumes, where they can exist as vapor,
dissolved orsorbed phase constituents as well as light- or
dense- nonaqueous phase liquids (LNAPLs or DNAPLs).
When these COCs are present in low permeability media
(LPM) such as silt and clay deposits, there are major chal-
lenges with assessment of their behavior and implemen-
tation of effective in situ remediation technologies (API
1995; DOE 1996). Despite a low bulk permeability (K <
10-s cm/s), these COCs contaminate LPM deposits by
moving into and through natural pore and fracture net-
works where they partition into multiple phases (Figure
1). Long-term exposures and unacceptable risk can re-
sult from drinking contaminated water, ingesting surface
soils, or inhaling vapors that enter basements or that are
emitted during showering. In recognition of the need for
effective in situ remediation methods, DNAPL compounds
in LPM deposits was recently ranked as a top environ-
mental restoration need across the DOE Complex (DOE
1996). Similarly, nearly 40% of the underground storage
tanks in the world are located on clay soils and in situ
remediation of LNAPL contamination in these settings has
been a major challenge for the petroleum industry (API
1995).
In situ remediation by conventional methods such as soil
vapor extraction or biodegradation are often ineffective at
LPM sites due to poor accessibility to the contaminants
and severe mass transfer limitations (AP11995; DOE 1996;
Freeze and McKay 1997). Alternative techniques to en-
hance in situ remediation have been developed utilizing
subsurface disruption by soil mixing (e.g., Siegrist et al.
1995, Cline et al. 1997) or the alternative driving forces of
electrokinetics (e.g., Probstein and Hicks 1993; Shapiro
and Probstein 1993; Ho et al. 1995; Murdoch and Chen
1997). In seeking less intensive methods that could be
used over larger areas, multipoint injection and perme-
ation methods have been tested and advanced soil frac-
turing techniques have been developed (Murdoch et al.
1997a,b; Siegrist et al. 1998a,b, 1999).
This paper presents a synopsis of laboratory and model-
ing studies and field demonstration and testing activities
carried out to evaluate in situ chemical treatment using
hydraulic fracturing with zero valent iron metal (Fe°) or
potassium permanganate (KMnO4) solids to achieve in
situ destruction of TCE. For a general discussion of ad-
vanced fracturing employing reactive media, the reader is
referred to Murdoch et al. (1997a,b). A more detailed de-
scription of the field demonstration may be found in Siegrist
et al. (1999) while details regarding related work with ther-
mally enhanced mass recovery may be found in Siegrist
et al. (1998a). Information on supporting laboratory ex-
perimentation may be found in Case (1997) and Struse
(1999), and in forthcoming papers.
85
-------
Fracture emplacement
casing )
Natural fractures
ion of A
anctfor (
t' *
,
COCai)d/or. 1
treatment agent t
t*
f-
Fractures C. Contaminated
filled with LPM deposit
reactive with TCE or
media other organics
Figure 1. Chemical treatment zones installed using hydraulic fracturing and reactive solids.
Hydraulic Fracturing To Emplace Reactive
Solids
Background on Hydraulic Fracturing and
Reactive Media
With soil fracturing, it is feasible to create fractures in LPM
deposits by injecting a viscous liquid containing a high
content of suspended granular solids (USEPA 1993;
Murdoch et al. 1994). The granular solids remain in the
fracture after injection and the properties of the solids can
be exploited for in situ remediation. Hydraulic fracturing
normally employs sand as a proppant to fill and support
the fracture opening and prevent fracture closure during
natural healing processes in unconsolidated deposits.
Fractures filled with sand can enhance the rate of recov-
ery of fluids from an LPM deposit or aid the delivery of
reactive fluids into it. Alternatively, fractures filled with
chemically reactive media such as iron metal or potas-
sium permanganate solids can be used to immobilize or
degrade organic contaminants in situ (Murdoch et al.
1997a,b; Siegristetal.1999).
Zero valent iron metal (Fe°) as a treatment media has been
extensively studied during the past five years (Gillham and
O'Hannesin 1994; Matheson and Tratnyek 1994; Agfawal
andTratnyek 1996; Muftikian et al. 1996; Liang et al. 1997).
Full-scale applications of permeable iron walls for in situ
treatment of contaminated groundwater are rapidly in-
creasing (Gavaskar et al. 1997; O'Hannesin and Gillham
1998). The reductive dechlorination of TCE to
dichloroethene by Fe° is shown in equation 1. The com-
plete dechlorination reaction forTCE to ethene and ethane
involves single and multiple electron transfers following
pseudo-first order kinetics. Half-lives are reported to be
on the order of 30 to 60 min. as normalized for the solution
to solid surface area typical of a packed bed of iron par-
ticles (Liang et al. 1997; Gavaskar et al. 1997; O'Hannesin
and Gillham 1998). The pH of the reacting system tends
to rise, but appears to stabilize in the pH 9 to 10 range
due to iron hydroxide precipitation. Key reaction products
can include daughter chlorocarbons such as
dichloroethenes and vinyl chloride (leading to ethenes and
ethanes), hydrogen gas, chlorides, and iron oxide and
hydroxide precipitates. Accelerated reaction rates and re-
duced production of vinyl chloride have been reported with
palladized iron metal (Muftikian et al. 1 996)
2Fe° +C2HC13 +3H2O -> C2H2C12 +H2 +
(1)
The use of potassium permanganate for oxidative degra-
dation of TCE and other organics at contaminated sites
has more recently evolved through initial laboratory ex-
perimentation with volatile and semivolatile organics in soil
and water (Vella and Veranda 1 992; Gates et al. 1 995; Van
and Schwartz 1996; Case 1997; Struse 1999). Recently
there have been several field applications including sub-
surface delivery by deep soil mixing or by well injection
(e.g., Cline et al. 1997; Schnarr et al. 1998; West et al.
1998). The destruction of TCE by KMnO4 (eqn. 2) can oc-
cur by direct electron transfer or free radical advanced
oxidation. The rate of TCE degradation is 2nd-order with
respect to TCE and permanganate (MnO4~) but can be
approximated by pseudo-first order kinetics. The reaction
is rapid with half-lives on the order of 1 to 2 min or less.
The pH of the reacting system can decline to strongly acidic
conditions (e.g., pH 2 to 3) depending on the buffering
capacity of the reaction system. Key reaction products can
include intermediate organic acids along with production
of manganese oxide solids (MnO2)and chlorides. Daugh-
ter products such as chlorinated alkanes or chlorinated
acids may also conceivably be produced under some con-
ditions but research has yet to fully resolve the occurrence
and significance of this.
2KMnO4+C2HCl3
2KC1 + HC1
2CO2 +2MnO2
(2)
86
-------
Highlights of Research and Development
During the past three years, the authors have been in-
volved in research and development efforts focused on
coupling hydraulic fracturing with chemical treatment meth-
ods. The methods and results of this work are described
in existing and forthcoming publications (Murdoch et al.
1997a,b; Case 1997; Siegrist et al. 1998a; Struse 1999;
Siegrist et al. 1999) and only highlights are provided herein.
Laboratory and Modeling Studies
Laboratory and modeling studies have been focused on
permanganate oxidation chemistry and delivery/transport
processes since a body of research was already avail-
able and/or in progress regarding Fe° metal reduction.
Laboratory studies have included (1) batch tests to define
permanganate oxidation kinetics, (2) development and
testing of a permanganate oxidative particle mixture, and
(3) intact core studies to quantify permanganate diffusive
transport and matrix interactions. Results of this work have
revealed that the rate of oxidative destruction of a
chlorocarbon such asTCE is 2nd-order with respect to TCE
and MnO4\ In groundwater at 20C without appreciable
natural organic matter, the reaction rate constant is in the
range of 0.6 L moMs"1. Lower temperatures will reduce
the rate of organic chemical destruction as will the pres-
ence of other oxidant-demanding substances in the sys-
tem (e.g., natural organic matter). As a proppant for hy-
draulic fracturing, a permanganate oxidative particle mix-
ture (OPM) was developed that was a pumpable solid
comprised of -50 wt.% potassium permanganate in a
mineral gel. The OPM had a chlorocarbon degradation
rate that was equal to or greater than that of permangan-
ate alone. This OPM was used in the field trial noted in the
following section. Transport studies have also been com-
pleted with intact cores of silty clay soil obtained from the
field site described below. Diffusive transport of potassium
permanganate from a 5 g-KMnO4/L source zone through
the silty clay soil was studied to determine oxidation-in-
duced changes in tortuosity due to matrix and contami-
nant interactions as well as to evaluate degradation effi-
ciency for residual TCE within the media. Studies were
made over periods up to 2 months or more with uncon-
taminated cores, cores contaminated with pure phase
TCE, and with glass beads. In uncontaminated cores, dif-
fusing permanganate did oxidize some but not all of the
natural organic matter in the soil (-10 to 30% destruction)
and did yield MnO2 deposits but they did not alter system
tortuosity. In contaminated cores (-100 to 1000 mg/kg),
permanganate transport was retarded due to reaction with
the TCE. Oxidation of residual TCE in the silty clay soil
was almost complete but tortuosity appeared to be in-
creased, possibly due to more focused and intense MnO4"
oxidation of residual TCE. Modeling work has included
development of a spreadsheet-based screening level
model that predicts permanganate diffusive transport away
from its location of initial introduction into the subsurface.
For example, with a concentration of 40 g/L in a fracture
upon emplacement, the enveloping reactive zone was pre-
dicted to have extended to 60 cm in total width after 6
months of emplacement (Struse 1999). These modeling
results compared well with the field observations made
during the field trial noted below.
Field Demonstration and Testing
A field trial was recently completed to evaluate in situ
remediation using hydraulic fracturing to emplace zero
valent iron metal and potassium permanganate solids in
the subsurface to chemically treat TCE (Siegrist et al. 1999)
(Figure 1). At an old land treatment site, two test cells were
installed in silty clay soils with hydraulic fractures filled with
either iron metal or permanganate solids at 1.8, 2.4, and
3.6-m depths.The saturated zone was encountered at 3.6-
m depth below ground surface (bgs). Fracture emplace-
ment was monitored and soil and groundwater conditions
were characterized. After 3,10 and 15 months of emplace-
ment, continuous cores were collected and morphologic
and geochemical data were taken across the fracture
zones. Controlled degradation tests were completed us-
ing site groundwater with TCE concentrations near 53,
144, and 480 mg/L, equivalent to 0.5,1.2, and 4.1 g TCE
per kg media, respectively. The iron-filled fractures formed
a discrete reactive seam less than 1 cm thick, wherein the
Eh decreased and reductive dechiorination could occur,
but effects in the adjacent silty clay soils were negligible
(Figure 2). While the emplaced iron exhibited some sur-
face corrosion after extended emplacement in the sub-
surface, its reactivity was unaffected. Iron from the frac-
tures degraded TCE at efficiencies of as much as 36%
after 24 to 48 hr of contact, which is consistent with Fe°
packed bed degradation half-lives of 1 to 2 hr. The per-
manganate-filled fractures yielded a diffusereactive zone
that expanded over time reaching 40 cm in thickness after
10 months (Figure 3).Throughout this oxidizing zone, the
degradation efficiency was >99% after 2 hr of contact for
dissolved TCE at 0.5 and 1.2 mg TCE per g of media.
When exposed to higher TCE loadings (i.e., 4.1 mg per
g), degradation efficiencies after 10 months dropped to
70% as the TCE load exceeded the oxidant capacity re-
maining. These efficiencies and rates are consistent with
oxidation stoichiometry and previously determined half-
lives of <2 min. for permanganate oxidation of TCE. In both
test cells there were no marked effects on the chemistry
or contamination levels in the groundwater beneath the
cells.
Discussion and Conclusions
Discussion
The feasibility of in situ remediation of TCE and other
DNAPL compounds at LPM sites using reactive solids
emplaced by hydraulic fracturing requires consideration
of the horizontal continuity, degradation capacity, and lon-
gevity of the treatment agents. The results obtained from
the work to date enable an interesting contrast between
reactive fractures created with surface reactive media (i.e.,
Fe° particles in a guar gum gel) and those created with
reactive media that dissolves and permeates into the sur-
rounding soil to produce a wide reactive zone (i.e., KMnO4
87
-------
30
20
10
0-
3J
-10
-20
-30
5J
11.
30
20
10
0
-10
-20
-30
0.05
0,1
0.
Redox (mV)
pH
TOC(%drywt)
Reduction (%)
Reduction (%)
Figure 2. Geochemical properties and TCE degradation potential of Fe° metal fracture zones 10 mon. after emplacement in a silly clay deposit
(Slegrist et al. 1999). Notes: TCE degradation measured using 5 g of media in 40 mL of GW1 (initial TCE = 477.0 mg/L) or GW2 (initial TCE = 53.7
mg/L)
particles in a mineral gel). During the field trial described,
both types of reactive media were handled and emplaced
by conventional hydraulic fracturing equipment and meth-
ods. Handling of the permanganate was more problem-
atic in some respects but modifications to fracturing equip-
ment or development of encapsulation methods should
resolve this issue. In general, the geometry of the reactive
media 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 fractur-
ing fluids (i.e., iron particles in guar gum gel; permangan-
ate particles in mineral-based gel; sand in guar gum gel).
Hydraulic fractures may bifurcate to form offset segments,
which could produce local areas that are avoided by in-
jected material (Murdoch 1995). This challenges the frac-
ture emplacement to be continuous and uniform horizon-
tally with limited breaches through it, a requirement that
may require overlapping fractures created at several
depths. The Fe°-filled fractures are discrete layers and
appear to have limited effect on the soil deposit beyond
the fracture boundaries. Thus any in situ degradation of
TCE or related compounds must rely on contaminants
being mobilized to a fracture and then reacting with the
Fe°within the fracture. The degradation rates observed in
this study were consistent with previous studies suggest-
ing a half-life on the order of 1 to 2 hr for TCE degradation
in Fe°-filled fractures. While slow, this is still rapid enough
for high treatment efficiencies to be achieved during a day
or less of contact that is achievable for most LPM depos-
its. For example, if groundwater percolation through the
fracture is controlled by the surrounding LPM which has a
Ksat of 10'6 cm/s and a hydraulic gradient of unity, then the
retention time in an Fe°-filled fracture of 5 mm thickness
would be on the order of 1 to 5 days depending on the
effective porosity. As a treatment zone that relies domi-
88
-------
30
20
10
8 -10
& -20
-30
KMnO 4 fracture...
2d» 6
-------
about 0.4 g KMnO4 per cm2 of fracture horizontal area.
Based on complete oxidation and a stoichiometric TCE
demand of 2.5 wt./wt., each cm2 of fracture can treat about
0.16 g of TCE.This oxidant loading is sufficient to degrade
an initial TCE concentration of 1000 mg/kg within a zone
of LPM that is 90 cm thick. Alternatively it is sufficient to
treat 16 L of percolate with a concentration of 10 mg/L of
TCE which is equivalent to a 50-yr life at a deep percola-
tion flux of 1 crn/d. Realistically though, it is anticipated
that the oxidant demand of natural organic matter or the
advective loss of oxidant out of the treatment region could
markedly diminish this life. Based on direct observation in
this study, the oxidation capacity within and around the
permanganate fractures was striking even 15 months af-
ter emplacement.
The cost of remediation using hydraulic fracturing to cre-
ate chemical treatment zones was roughly compared to
that of soil vapor extraction (SVE) enhanced by sand-filled
fractures. In general, the reactive chemicals are more ex-
pensive than sand and material handling costs are some-
what higher, so the reactive media filled fractures are more
expensive to create. The costs for media per fracture
amounts to roughly $1000 for iron and $1500 for perman-
ganate as compared to $100 for sand. However, during
passive operation, the treatment zones require less re-
source consumption (e.g., power), less sampling and
analysis (e.g., no off-gas), and reduced manpower com-
pared to many SVE systems. As a result, the total esti-
mated costs of the horizontal treatment zone systems are
similar to the costs for SVE systems. For a 2.2-hectare
site contaminated to 5-m depth, the costs for implement-
ing horizontal treatment zones were estimated to be in
the range of $25 to $35/m3 for iron and permanganate
zones, respectively.This cost was based on 6-m diameter
fractured zones with 1-m overlap and fractures installed
at depths of 1.7 and 3.5 m. Assuming no bulk media dis-
counts this yields an estimated materials cost of $2,700K
for the iron metal and $3,800Kforthe permanganate. The
installation time was estimated at 140 days for a 3-person
crew and equipment, yielding a labor cost including travel
and per diem of $450K, plus a mobilization/demobiliza-
tion cost of $50K.
Conclusions
The work to date has explored two types of chemical treat-
mentscnss far/n situ remediation at LPM sites contami-
nated by TCE. Laboratory and modeling studies have been
completed and a field trial has been conducted. In the field
trial, hydraulic fracturing equipment and methods were
used to create reactive zones of Fe° metal or KMnO4 OPM
in horizontally oriented layers within silty clay soils at depths
up to 5 m.The Fe°-filled fractures produced a reactive seam
with limited effect on the surrounding LPM while the
KMnGyfilled fractures yielded a broad zone of reactivity
within the LPM. With both types of fracture zones, degra-
dation potential for high levels of TCE was sustained even
after 10 months of emplacement in the subsurface. Both
types of horizontal treatment zones may reduce risks as-
sociated with exposure to TCE from a contaminated site.
Although the system using iron-filled fractures may leave
immobile contaminants in the ground untreated, data from
this study suggest that it is capable of degrading mobile
TCE and thus may reduce risk by effectively eliminating
TCE release from a low permeability unit to the atmosphere
or an underlying aquifer. The system using permangan-
ate-filled fractures where MnO4~ ions are diffusively dis-
tributed through a broad region, offers the possibility to
both curtail TCE release to the atmosphere or an underly-
ing aquifer as well as destroy TCE throughout a low per-
meability formation. However, diffusive transport is slow
and the rate and extent are highly dependent on the physi-
cal and chemical properties of the formation. This approach
to remediating low permeability formations using fracture
emplaced reactive solids is extremely encouraging. While
TCE has been the target contaminant in for studies to date,
other organic compounds and also redox sensitive met-
als might be amenable to such an in situ remediation strat-
egy. While the results to date are promising, this in situ
remediation approach is still early in its development and
further work is necessary and appropriate to provide
needed design, implementation, and performance data for
a range of site and contamination conditions.
Acknowledgments
Sponsorship of the work upon which this paper is based
was provided through the Subsurface Contaminants Fo-
cus Area of the DOE Office of Science & Technology and
the Office of Environmental Restoration at the DOE Ports-
mouth Gaseous Diffusion Plant. Mike Urynowicz and Ri-
chard Harnish of the Colorado School of Mines (CSM)
provided assistance with the laboratory studies. Dr. Helen
Dawson of CSM assisted with the development and appli-
cation of the screening level transport model. Mr.Tom Houk
of Bechtel Jacobs Company (formerly of Lockheed Mar-
tin Energy Systems), Mark Mumby of ORNL, and Dr. Bill
Slack of FRx, Inc. are acknowledged for their assistance
in completion of the field demonstration and testing. Dr.
Baohua Gu of ORNL completed micromorphologic analy-
ses of the field-emplaced iron.
References
Agrawal, A. and P.G. Tratnyek. 1996. Reduction of Nitro
Aromatic Compounds by Zero-Valent Iron Metal. Environ.
Sci.Technol. 30(1 ):153-160.
American Petroleum Institute. 1995. Petroleum Contami-
nated Low Permeability Soil: Hydrocarbon Distribution
Processes, Exposure Pathways and In Situ Remediation
Technologies. Publ. No. 4631. September 1995.
Case, T. L. 1997. Reactive Permanganate Grouts for Hori-
zontal Permeable Barriers and In S/'fc/Treatment of Ground-
water. M.S. Thesis, Colorado School of Mines, Golden,
CO.
90
-------
Cline, S.R., O.R. West, N.E. Korte, F.G. Gardner, R.L.
Siegrist, and J.L. Baker. 1997. KMnO4 Chemical Oxida-
tion and Deep Soil Mixing for Soil Treatment. Geotechnical
News. December, pp. 25-28.
DOE. 1996. In Situ Remediation of DNAPL Compounds
in Low Permeability Media: Transport/Fate, In Situ Con-
trol Technologies, and Risk Reduction. Oak Ridge National
Laboratory Report, ORNL/TM-13305. August, 1996.
Freeze, R.A. and D.B. McKay. 1997. A Framework for As-
sessing Risk Reduction Due to DNAPL Mass Removal
from Low Permeability Soils. Groundwater. 35(1 ):111 -123.
Gates, D.D., R.L Siegrist, S.R. Cline. 1995. Chemical Oxi-
dation of Volatile and Semi-volatile Organic Compounds
in Soil. Proc. Air and Waste Management Association Conf.
June.
Gavaskar, A., N. Gupta, B. Sass, T. Fox, R. Janosy, K.
Cantrell, and R. Olfenbuttel. 1997. Design Guidance for
Application of Permeable Barriers to Remediate Dissolved
Chlorinated Solvents. U.S. Air Force Armstrong Labora-
tory. AL/EQ-TR-1997-0014.
Gillham, R.W. and S.F. O'Hannesin. 1994. Enhanced Deg-
radation of Halogenated Aliphatics by Zero-Valent Iron.
Groundwater. 32(6):958-967.
Ho, S. V., Sheridan, W., Athmer, C. J., Heitkamp, M. A.,
Brackin, J. M., Webber, D., and Brodsky, P. H. (1995). Inte-
grated In situ Soil Remediation Technology—The Lasa-
gna Process. Environ. Sci. Technol., v. 29, No. 10, 2528-
2534.
Liang, L., O.R. West, N.E. Korte, J.D. Goodlaxson, D.A.
Pickering, J.L. Zutman, F.J. Anderson, C.A. Welch, M.J.
Pelfry, and M.J. Dickey. 1997. The X-625 Groundwater
Treatment Facility: A Field-Scale Test of Trichloroethylene
Dechlorination Using Iron Filings for the X-120/X-749
Groundwater Plume. Oak Ridge National Laboratory Re-
port, ORNL/TM-13217.
Matheson, L.J. and P.G. Tratnyek. 1994. Reductive
Dehalogenation of Chlorinated Methanes by Iron Metal.
Environ. Sci. Technol. 28:2045-1053.
Muftikian, R., Q. Fernando, and N. Korte. 1996. A Method
for the Rapid Dechlorination of Low Molecular Weight
Chlorinated Hydrocarbons in Water. Water Res. 29:2434.
Murdoch, L.C., D.Wilson, K. Savage, W. Slack, and J. Uber.
1994. Alternative Methods for Fluid Delivery and Recov-
ery. USEPA/625/R-94/003.
Murdoch, L.C. 1995. Forms of Hydraulic Fractures Cre-
ated During a Field Test in Fine-Grained Glacial Drift. Quar-
terly Journal of Engineering Geology. 28:23-35.
Murdoch, L. C. and Chen, J-L. 1997. Effects of Conduc-
tive Fractures During In situ Electroosmosis. Journal of
Hazardous Materials.
Murdoch, L., W. Slack, R. Siegrist, S. Vesper, and T. Meiggs.
1997a. Advanced Hydraulic Fracturing Methods to Cre-
ate In Situ Reactive Barriers. Proc. International Contain-
ment Technology Conference and Exhibition. February 9-
12,1997, St. Petersburg, FL.
Murdoch, L., B. Slack, B. Siegrist, S. Vesper, andT. Meiggs.
1997b. Hydraulic Fracturing Advances. Civil Engineering.
May1997.pp.10A-12A.
NATO. 1998. NATO/CCMS Pilot Study Special Session on
Treatment Walls and Permeable Reactive Barriers. Feb-
ruary 1998, Vienna. EPA 542-R-98-003. May 1998.
O'Hannesin, S.F. and R.W. Gillham. 1998. Long-term Per-
formance of an In Situ Iron Wall for Remediation of VOCs.
Groundwater. 36(1 ):164-170.
Probstein, R. F. and Hicks, R. E. 1993. Removal of Con-
taminants from Soils by Electric Fields. Science. 260:498-
503.
Riley, R. G. and J. M. Zachara. 1992. Nature of Chemical
Contaminants on DOE Lands. DOE/ER-0547T. Office of
Energy Research, U.S. Department of Energy, Washing-
ton, D.C.
Schnarr, M., C. Truax, G. Farquhar, E. Hood, T. Gonully,
and B. Stickney. 1998. Laboratory and Controlled Field
Experimentation Using Potassium Permanganate to
Remediate Trichloroethylene and Perchloroethylene
DNAPLs in Porous Media. Journal of Contaminant Hy-
drology. 29:205-224.
Shapiro, A.P. and R.F. Probstein. 1993. Removal of Con-
taminants from Saturated Clay by Electro-osmosis.
Environ. Sci. Technol. 27:283-291.
Siegrist, R.L. and J.J. van Ee. 1994. Measuring and Inter-
preting VOCs in Soils: State of the Art and Research
Needs. EPA/540/R-94/506. USEPA ORD, Washington, D.C.
20460.
Siegrist, R.L., O.R. West, et al. 1995. In Situ Mixed Region
Vapor Stripping of Low Permeability Media. 2. Full Scale
Field Experiments. Environ. Science & Technology.
29(9):2198-2207.
Siegrist, R.L, K.S. Lowe, LC. Murdoch, W.W. Slack, and
T.C. Houk. 1998a. X-231A Demonstration of In Situ
Remediation of DNAPL Compounds in Low Permeability
Media by Soil Fracturing with Thermally Enhanced Mass
Recovery or Reactive Barrier Destruction. Oak Ridge
National Laboratory Report. ORNL/TM-13534. March.
91
-------
Siegrist, R.L. K.S. Lowe, L.C. Murdoch, T.L. Case, D.A.
Pickering, and T.C. Houk. 1998b. Horizontal Treatment
Barriers of Fracture-Emplaced Iron and Permanganate
Particles. NATO/CCMS Pilot Study Special Session on
Treatment Walls and Permeable Reactive Barriers. EPA
542-R-98-003. May 1998. pp. 77-82.
Siegrist, R.L., K.S. Lowe, L.C. Murdoch, T.L. Case, and
D.A. Pickering. 1999. In Situ Oxidation by Fracture
Emplaced Reactive Solids. J. Environmental Engineering.
Vol.125, No.5, pp.429-440.
Struse, A.M. 1999. Mass Transport of Potassium Perman-
ganate in Low Permeability Media and Matrix Interactions.
M.S. thesis, Colorado School of Mines, Golden, CO.
USEPA. 1993. Hydraulic Fracturing Technology - Technol-
ogy Evaluation Report. EPA/540/R-93/505. Office of Re-
search and Development, Cincinnati, OH.
USEPA. 1997. Cleaning up the Nation's Waste Sites. Mar-
kets and Technology Trends. EPA 542-R-96-005. Office of
Solid Waste and Emergency Response. Washington, D.C.
Vella, P.A. and B. Veranda. 1992. Oxidation of Trichloroet-
hylene: A Comparison of Potassium Permanganate and
Fenton's Reagent. 3rd Intern. Symposium on Chemical
Oxidation. Vanderbilt University, Nashville, TN.
West, O.R., Cline, S.R., Siegrist, R.L, Houk,T.C., Holden,
W.L., Gardner, F.G. and Schlosser, R.M. 1998b. A Field-
Scale Test of In Situ Chemical Oxidation Through Recir-
culation. Proc. Spectrum '98 Intern. Conf. on Nuclear and
Hazardous Waste Management. Denver, Sept. 13-18, pp.
1051-57.
Van, Y. and F.W. Schwartz. 1996. In S/YuOxidative^Dechlo-
rination of Trichloroethylene by Potassium Permanganate.
Proc. 3rd Intern. Conf. on Advanced Oxidation Technolo-
gies. October 26-29. Cincinnati, OH
92
-------
In Situ Treatment of Chromium Source Area Using Redox Manipulation
Faruque A. Khan and Robert W. Puls
USEPA-ORD-NRMRL-SPRD
Ada, OK 74820
580-436-8704; (fax) 580-436-8703
Khan.Faruque @ epa.gov
Introduction
A chrome plating facility was operated at the U.S. Coast
Guard (USCG) Support Center, near Elizabeth City, North
Carolina, for thirty years. Activities in this shop resulted in
the release of chromic acid into the soils below the shop.
A detailed characterization of the underlying soils and
ground water of the chrome plating shop was performed
to provide information on the extent of contamination at
the site and the potential for off-site migration and envi-
ronmental impact.
The National Risk Management Research Laboratory
(NRMRL) scientists, in cooperation with USCG personnel
have completed a project entitled "Use of Permeable Re-
active In-Situ Barrier Wall to Remediate Chromium Con-
taminated Ground Water" at the USCG Support Center,
Elizabeth City, North Carolina. This in-situ permeable re-
active barrier wall technique is a very effective method for
reducing and immobilizing contaminants like Cr(VI) and
TCE present in the ground water, but it has limited effec-
tiveness to reduce contaminants present as sorbed phases
in the vadose zone. Most of the Cr(VI) at the USCG Sup-
port Center site is present in the vadose zone of the aqui-
fer under an oxidizing environment. While some Cr(VI) was
reduced to Cr(lll) in the vadose zone sediments, the re-
ducing capacity of these sediments was eventually over-
whelmed, and resulted in continuous ground water con-
tamination.
Abiotic reduction by manipulating the oxidation-reduction
(redox) status of a vadose zone is a possible approach for
in-situ remediation of a redox-sensitive contaminant like
Cr(VI).The reduction of the mobile and toxic Cr(VI) to less
toxic and insoluble Cr(l 11) precipitates has been demon-
strated to occur in the aqueous system (Palmer and Puls,
1994 and Amonette et at., 1994). Researchers have iden-
tified few inorganic reductants and anthropogenic sources
of electron donors that can transform the toxic Cr(VI) form
to the less toxic Cr(III). Under alkaline to slightly acidic
conditions, this Cr(lll) precipitates as a fairly insoluble hy-
droxide, thereby immobilizing it within the soil. Such "natural
attenuation" of Cr(VI) is of great interest because it sug-
gests that strict water quality standards do not have to be
attained everywhere within a site. Where "natural attenua-
tion" is insufficient to prevent environmental degradation,
it may be possible to rejuvenate the "natural attenuation"
capacity of the vadose zone by incorporating reductant
into the vadose zone.Therefore, the objectives of this study
were to evaluate potential reductants and their effective-
ness in reducing Cr(VI) present in the soils, the effects of
reductants on the geochemical environment, and the in-
teraction between reductants and selected RCRA metals
present in the vadose materials.
Methods and Materials
The field site and the hydrogeology of the site have been
described in detail elsewhere (Puls et al., 1994). Soil cores
were collected from the old plating shop of the USCG
Support Center, Elizabeth City, North Carolina. Samples
were air dried and passed through a 2mm screen. The
<2mm sediment fraction was used in the batch experi-
ments. The amount of Cr(VI) in the soil samples was de-
termined using the phosphate extraction procedure de-
veloped by Bartlet and James (1988). Three soil samples
from three different cores (6-4, 7-4, and 15-4) and three
reductants (sodium dithionite, l-ascorbic acid, and hydroxy-
lamine) were used in this study. A bicarbonate buffer solu-
tion (0.1 M KHCO3) was used with sodium dithionite and I-
ascorbic acid in order to maintain near neutral pH for the
reductant solutions. The procedure used was:
• Weigh replicate 5.0 grams air-dry soil into 50 ml poly-
carbonate centrifuge tubes.
• Add 25 ml 0.05M reductant to the 50 ml centrifuge
tube and shake vigorously.
• Immediately measure Eh and pH.PIace tubes on a
Thermolyne Maxi-Mix III rotating shaker @ 100 rpm
for 24 hrs.
• Measure Eh and pH.
• Centrifuge @ 3600X g for 60 minutes.
93
-------
• Filter supernatant through 0.2 urn Nuclepore mem-
brane filter.
• Analyze Cr(VI) using the Hach method 8023 and Hach
® DR/2010 spectrophotometer, and trace metals with
the Inductively Coupled Plasma Spectrometry(ICP).
Phosphate extractions were conducted on the residual
sediments following the modified method of Bartlet and
James (1988) in order to determine if any Cr(VI) remained
on the sediments.These samples were analyzed for Cr(VI)
in the same procedure described above.
Results and Discussion
Each reductant used in this study was effective in reduc-
ing Cr(VI) in the soils (Figure 1). In this figure, pre-treat-
ment Cr(VI) represents the Cr(VI) concentration extracted
with phosphate solution before the reductant was applied.
Post-treatment Cr(VI) represents Cr(VI) concentration
extracted by phosphate solution determined after the treat-
ment with reductants. Each of the reductants was able to
reduce 100% of very high Cr(VI) contaminated samples.
Total Cr represented the amount of all chemical forms of
Cr present in the extraction after reductant treatment and
determined by ICP.The total Cr content in these extracts
was negligible except for the l-ascorbic acid. Figure 1b
shows that there was 286 to 111 mg/Kg of some form of
soluble Cr present in the l-ascorbic acid solution. After 24-
hours extraction period, Cr(VI) was not reduced to solid
phase of Cr(lll) by l-ascorbic acid. O'Brien and Woodbridge
(1997) indicated that Cr(VI) may go through step-wise re-
duction to Cr(V) and Cr(IV) then Cr(lll) with the progres-
sion of time.
The pH of the reductant solutions without a bicarbonate
buffer was less than 4.0 for sodium dithionite and l-ascor-
bic acid. To maintain neutral pH, the reductants (sodium
dithionite and l-ascorbic acid) were made up in 0.1 M
KHCO3 solution. The KHCO3 buffer was not used with hy-
droxylamine to maintain neutral pH because hydroxylamine
is a weak base. Post treatment pH of all the extractants
was very stable after 24 hours shaking with soil samples
and remained in the range of near neutral. It was impor-
tant to maintain neutral pH because adsorption of Cr(VI)
on clay particles decreases under neutral or higher pH.
Earlier it was mentioned that 100% Cr(VI) has been re-
duced by these reductants.Therefore, sorbed Cr(VI) must
have been released during extraction periods and even-
tually reduced. After treatment with reductants, the residual
sediment samples were used to extract total and insoluble
forms of Cr(VI) with hot 0.28M Na2CO3-0.5MNaOH solu-
tion using a method developed by James et al. (1995).
This extraction method has the capability to extract any
soluble mineral forms of Cr(VI) in the sediment sample.
No extractable Cr(VI) was determined by this method which
confirmed that all Cr(VI) in the sediment samples had been
reduced by the selected reductants.
The Eh of the solution was monitored after adding reduc-
tants at time zero hour (initial Eh) and 24 hours (final Eh).
Figure 2 shows that Eh values obtained in the extraction
with sodium dithionite were different from the two other
reductants. In the sodium dithionite extraction, the initial
Eh was more than -400 mV as compared with +80 mV or
less for l-ascorbic acid and hydroxylamine.The final Eh in
the sodium dithionite extractions was more than -300 mV.
The final Eh values did not change under l-ascorbic acid
and there were slightly higher Eh values obtained after 24
hours in the hydroxylamine extractions.Therefore, the re-
sult shows that sodium dithionite has the highest reduc-
tion capacity among these reductants. These reductants
not only reduced Cr(VI) but also redox sensitive metals
like Fe and Mn. Figure 3 shows that sodium dithionite ex-
tractions contained higher amounts of Fe and less Mn as
compared with other extractants (l-ascorbic acid and hy-
droxylamine).
The concentration of RCRA metals (As, Ba, Cu, Cd, Ni,
Pb, and Se) in the sodium dithionite extractants were ei-
ther below the detection limit of the ICP or the NPDWR's
MCLG standards for ground water. The result suggests
that these reductants did not mobilize these trace elements
during the batch experiments.
Conclusions
All reductants reduced Cr(VI) present in the studied va-
dose materials, and each provided a unique geochemical
environment. High concentrations of soluble species of
Cr in the l-ascorbic acid extraction suggest that Cr(VI) may
have been reduced to lesser oxidation states and remained
in the solution. Hydroxylamine reduced a large portion of
Mn but negligible amounts of Fe. In contrast, a large
amount of Fe has been reduced by sodium dithionite which
is effective in reducing Cr(VI). It will be beneficial to have a
large reservoir of reduced Fe which will sustain a reactive
zone for a period of time and continue to enable the re-
duction of Cr(VI) present in the vadose zone. Analyses of
RCRA metals in the dithionite extractants suggest that
there will be no deleterious effect on ground water.There-
fore, sodium dithionite has an advantage over other re-
ductants used in this study and it would be the choice
reductant for the field project. This research has contrib-
uted to the development of a field study, planned for the
summer of 1999, to attenuate Cr(VI) present in the va-
dose zone.
References
Amonette, J.E., J.E Szecsody, H.T. Schaef, J.C.Templeton,
Y.A Gorby, and J.S. Fruchter. 1994. Abiotic reduction of
aquifer materials by dithionite: A promising in-situ
remediation technology. Battelle Press, Columbus, Ohio.
Bartlett, R. J. and B. James. 1988. Mobility and
bioavailability of Cr in soils. In J. Nriagu and Nieboer (eds.)
Chromium in the Natural and Human Environment. P. 267-
304. John Wiley & Sons, New York
94
-------
James, R.B, J.C. Petura, R.J., Vitale, and G.R. Mussoline.
1995. Hexavalent chromine extraction from soils: A com-
parison of five methods. Environ. Sci. Technol. 29:2377-
2381.
O'Brien, P. and N. Woodbridge. 1997. A study of the kinet-
ics of the reduction of chromate by ascorbate under aero-
bic and anaerobic conditions. Poluhederon: 16:2081 -2086.
Palmer, C.D. and R.W. Puls. 1994. Natural attenuation of
hexavalent chromium in ground water and soils. EPA/540/
S-94/505.
Puls, R.W., C.J. Paul, D.A. Clark, and J. Vardy. 1994. Trans-
port and transformation of hexavalent chrome through soil
and into ground water. J. Soil Contam. 3:203-224.
Disclaimer
This is an extended abstract of the proposed presenta-
tion and does not necessarily reflect EPA policy.
.05M Sodium Dithionite
b.
c.
O
I
O
[Q PnJnutmentCrfVO
Post-tn*Sm<>ntCr(V(
J3 TotilCrinraductanl
700-
BOO-
500
400
300-
200
100-
0-
700-,
600-
500
400-
300-
200-
100-
0
16.4 7-4 9-4
.05M L-ascorbic add
15-4 7-4 6-4
.05M Hydroxylamine
1S-4 7-4 6-4
Figure 1. Concentrations of initial Cr(VI) compared with post-treatment Cr(VI) and total Cr using three reductants.
95
-------
In Situ Redox Manipulation for Treatment of Chromate and
Trichloroethylene in Groundwater
John Fruchter
Pacific Northwest National Laboratory
Battelle Northwest
PO Box 999
Richland.WA 99352
509-376-3937; (fax) 509-372-1704
john.fruchter@pnl.gov
As a legacy from weapons production at the U. S. Depart-
ment of Energy (DOE) Hanford Site in south central Wash-
ington State, several chromate groundwater plumes are
currently impinging upon the Columbia River. Pump and
treat systems are operating at two of the plumes. Because
permeable reactive barriers appear to offer several ad-
vantages over pump and treat, DOE has decided to test
this concept on a third plume. Because these contami-
nant plumes average about 80 feet below the ground sur-
face, an alternative to "trench and fill" permeable reactive
barriers was developed. This alternative, called In Situ
Redox Manipulation (ISRM), can be installed through con-
ventional groundwater wells.
The goal of the ISRM method is to create a permeable
treatment zone in the subsurface to remediate redox-sen-
sitive contaminants. Redox sensitive contaminants in the
plume are immobilized or destroyed as they migrate
through the manipulated zone. A permeable treatment
zone is created by reducing the ferric iron in the aquifer
sediments to ferrous iron.The 50-ft treatment zone is cre-
ated by injecting appropriate reagents and buffers (e.g.,
sodium dithionite and potassium carbonate) to chemically
reduce the structural iron in the sediments. The ISRM
approach extends the permeable treatment zone concept
to sites where the groundwater contaminant plumes are
deep (tens of meters below the ground surface) to be
treated by excavation or by trench-emplaced permeable
barriers.
For the ISRM process, conventional groundwater wells
are placed at the site to be treated. A chemical solution of
sodium dithionite is injected into the wells for about 10 to
20 hours. Once the solution reaches the groundwater, it
reacts with iron in the soil to form a large barrier. When the
groundwater flows through the barrier, the targeted con-
taminants are destroyed or immobilized. Based on field
measurements and calculations, it is anticipated that bar-
riers formed in this manner will remain effective for up to
30 years and require minimal maintenance.
Initial bench-scale batch and column experiments con-
ducted on Hanford sediments evaluated of potential re-
ducing reagents, including dissolved sulfur dioxide, hy-
droxylamine hydrochloride and sodium dithionite. Once
sodium dithionite was selected as a preferred reagent, a
variety of batch and column experiments with sediment
and dithionite were performed by Amonette et al 1994.
These bench-scale studies were used to develop an un-
derstanding of the important reactions, final reaction prod-
ucts (i.e., residuals), and nature and fate of any ions re-
leased from the sediments and sediment surface coat-
ings under reducing conditions (e.g., mobilization of trace
metals).
Dithionite is a sulfur-containing oxyanion which breaks
down quickly in aqueous solution to form two sulfoxyl radi-
cals. These radicals react rapidly to reduce ferric iron in
minerals and oxides which occur naturally in most aquifer
sediments. Amonette et al. (1994) have shown that, within
the aquifer, the injected dithionite reacts with structural
iron in oxyhydroxide and iron-bearing layer silicate min-
eral phases, reducing Fe(lll) to Fe(ll) according to the over-
all reaction described by equation 1.
S204
2-
(aq) + 2Fe(III)(s) + 2H 2O
2S03z-(aq) + 2Fe(II)(s)
(1)
The reduced sediments in the treatment zone can remove
redox-sensitive contaminants from groundwater flowing
through the zone. For example, the chromate ion (HCrO4-
) is a common groundwater contaminant at many sites,
but it is only significantly water-soluble under oxidizing
conditions. Within the zone of dithionite-reduced sedi-
ments, aqueous chromate reacts with Fe(ll) produced by
the dithionite reaction (equation 1) and is precipitated as
a solid hydroxide (e.g. Cr(OH)3) according to the example
reaction described in equation 2:
HCrO4"(aq) + 3Fe(III)(s) + 4H+ ->
Cr(OH)3(s) + 3Fe(III) + 2H2O (2)
96
-------
Similar precipitation reactions will occur for other oxidized
redox-sensitive metal species.
mine the amount of ferric iron reduced.The results to date
are summarized in Williams et al, 1999.
A proof-of-principle field experiment was conducted in
September 1995 at a chromate (hexavalent chromium)
contaminated groundwater site at the Hanford Site in
Washington State. The proof-of-principle field experiment
was a single-well injection test. Its main objective was to
determine the feasibility of creating a reduced zone in an
aquifer using the ISRM method and to determine the lon-
gevity of the reduced zone in a natural environment. The
test created a 50-ft (~15-m) diameter cylindrical treatment
zone. The three phases of the test consisted of 1) injec-
tion of 20,500 gallons (77,000 L) of buffered sodium
dithionite solution in 17.5 hours, 2) an 18.5-hour reaction
phase, and 3) an 83-hour, 99,600 gallons (375,000 L),
withdrawal phase that recovered 87 to 90% of the reac-
tion products. Analysis of post experimental sediment
cores indicated that 60 to 100% of the available reactive
iron in the treated zone was reduced. Three years later,
the treatment zone remains anoxic, and total and
hexavalent chromium levels have been reduced from 70
ppb to below the detection limit (<7 ppb). Additionally, no
significant permeability changes have been detected dur-
ing any phase of the experiment. The results are summa-
rized in Fruchter et al, 1996.
During fiscal year 1998, a larger five-well treatability test
was installed at the Hanford 10OD Area. The test created
a permeable reactive barrier 150 feet long and 50 feet in
width. The barrier was placed about 500 feet from the
Columbia River in a chromate plume with concentrations
in the 1,000 to 2,000 ppb range (as chromium).The depth
to groundwater was about 85 feet. An average of about
27,000 gallons of buffered sodium dithionite solution was
injected in each of the five wells. The sodium dithonite
concentration averaged approximately 0.08M. The aver-
age reaction phase was 35 hours. Recent monitoring data
have shown that chromate concentration in the reduced
zone have decreased to below detection limits. In addi-
tion, chromate concentrations have also begun decreas-
ing in several down gradient monitoring wells. Monitoring
is continuing. In addition to monitoring wells near the site,
a series of sampling tubes have been placed along the
bank of the Columbia River. Sediment cores will be taken
in the reduced zone during the summer of 1999 to deter-
Based on the success of the treatability test, DOE de-
cided to deploy a full-scale barrier at the 100D site. Cur-
rent plans call for the expanded barrier to approximately
1,000 feet in length. It will be constructed at the same site
as the treatability test barrier. Construction will begin in
late fiscal year 1999, and be completed in fiscal year 2000.
Bench-scale tests have shown that dithionite treated soils
should also be effective for treatment of dissolved trichlo-
roethylene (TCE) contamination. Based on the bench-
scale results, a single well field test was conducted for
TCE at Fort Lewis, Washington. Preliminary results indi-
cate that although TCE is being removed, the half-life for
the TCE destruction is longer than anticipated from the
laboratory data. Additional laboratory tests are being de-
termined to determine the cause, and an additional field
test is scheduled for late summer, 1999.
References
Amonette, J. E., J. E. Szecsody, H. T. Schaef, J. C.
Templeton, Y. A. Gorby and J. S. Fruchter (1994) "Abiotic
Reduction of Aquifer Materials by Dithionite: a Promising
In situ Remediation Technology In Proceedings of the 33rd
Hanford Symposium on Health and the Environment - In
Situ Remediation: Scientific Basis for Current and Future
Technologies,, GW Gee and NR Wing, Ed., Battelle Press,
Richland. pp. 851-882.
Fruchter, U.S., J.E. Amonette, C.R. Cole, Y.A. Gorby, M.D.
Humphrey, J.D. Istok, F.A. Spane, J.E. Szecsody, S.S.Teel,
V.R. Vermeul, M.D. Williams and S.B. Yabusaki (1996) In
situ Redox Manipulation Field Injection Test Report -
Hanford 10OH Area, Subsurface Contaminant Focus Area,
Office of Technology Development, U.S. Department of
Energy, Washington, D.C. PNNL-11372.
Williams, MD, VR Vermeul, JE Szecsody, JS Fruchter and
CR Cole (1998) 100-D Area In situ Redox Treatability
Test for Chromate - Contaminated Groundwater FY1998
Year-End Report, Subsurface Contaminant Focus Area,
Office of Technology Development, U.S. Department of
Energy, Washington, D.C. PNNL-12153.
97
-------
In Situ Remediation of Chromium Contamination of Soil and Groundwater
Jim V. Rouse
Montgomery Watson
370 Interlocken Blvd, Suite 300
Broomfield, CO 80021
303-410-4000; (fax) 303-410-4100
jim.rouse@mw.com
Chromium is a widely distributed element natural to the
environment. In its natural trivalent form, chromium has
limited solubility. Chromium is also widely used in indus-
trial applications, usually in the soluble hexavalent form.
Industrial uses include metal plating, wood treating, and
previously in cooling towers. Chromium is one of the most
wide-spread metallic contaminants, often dating from in-
dustrial applications as far back in time as World War II.
The paper will discuss the geochemical behavior of chro-
mium, as an important basis for the understanding of in
situ remedial measures to address chromium contamina-
tion of soil and groundwater. It then will discuss various
remedial alternatives, with a focus on the strengths and
weaknesses of each alternative. The alternatives which
will be discussed include soil excavation, natural attenua-
tion, ground-water pump and treat systems, zero-valent
iron wall reactive barriers, and in situ geochemical fixa-
tion. Geochemical fixation discussions will include bore-
hole formation of reactive zones, direct-push reductant
injection, and a modification of the pump and treat ap-
proach. Case histories of the various approaches to in
situ geochemical fixation are presented, with operational
data.
98
-------
In Situ Reduction of Hexavalent Chromium in Groundwater and
Surface Soil Using Acidified Ferrous Suifate
William J. Walker and Lara E. Pucik-Ericksen
Geomega
530 21st St.
Sacramento, CA 95814
916-444-1934
billdigna@earthlink.net
Background
The reduction of Cr(VI) by ferrous iron can be described
by the overall reaction:
Cr(VI) + 3Fe(II) -> Cr(ffl) + 3Fe(HI)
This reaction appears to be appropriate for pH less than
10 and for PO4 concentrations less than 0.1 mM. Above
pH 10, the rate of oxidation of the ferrous iron by dissolved
oxygen is greater than the rate of oxidation of ferrous iron
by CrO4.Cr(VI) is a strong oxidant and is reduced in the
presence of electron donors such as Fe(ll). The Cr (VI)
state is generally considered to pose the greatest human
health risk because it is more toxic, more mobile, and more
soluble than Cr(lll). Therefore, the reduction of Cr(VI) to
Cr(lll) is an important process with regard to aquifer
remediation.
The study site consisted of a facility that discharged waste
water from spent chromium (Cr) plating solutions to a se-
ries of collection ponds. Overtime, Cr, principally in the
hexavalent form, leached from the ponds into the shallow
groundwater aquifer below (typically 20 - 35 feet below
ground surface). As a result, soils in the area have total Cr
concentrations ranging from a few ppm to over 30,000
ppm in localized areas of the former wasteponds. Ground-
water impacts are highest downgradient from the former
ponds and reach levels approaching approximately 4 mg/
L (Cr VI). Because the plume is fairly well defined and
because the aquifer is composed largely of sand with in-
termittent lenses of clay and silt, this site represented a
potential candidate for an in situ approach to reducing and
precipitating hexavalent Cr. Based on these considerations,
an approach was developed for performing a field-based
in situ Cr reduction study that considered the following
elements:
1. Effective chemical reduction and fixation of Cr (VI) can
be accomplished using ferrous sulfate, FeSO4. The
reaction by which this occurs is:
Cr(VI)(aq)+3Fe(H)(aq) = Cr(III)(aq) + 3Fe(m)(aq)
If the pH of the solution is near neutral than the following
precipitates can form rapidly:
Cr(HI) + 3OH = Cr(OH)3 and if excess Fe
(1 - x)Fe(EI) + xCr(IH) + 3OH = CrxFe(!_x) (OH)3 (solid)
2. The reduction and precipitation of Cr (VI) can theo-
retically be carried out to treat both groundwater and
soil at the site. This is due in large part to the ease
with fluids can be introduced and dispersed within the
aquifer. Soil treatment should be possible since most
hexavalent Cr still remains in the top 2 feet of soil at
the site.
Methods
Soil Test Pit Construction and Character-
ization
Five test pits were constructed at the site. All locations
were based on a transect sampling done earlier to deter-
mine areas with different amounts of total and hexavalent
Cr.
The test pits were used in several ways to determine the
effectiveness of in situ Cr reduction for the site. First, two
of the test pits were set up as long-term control cells. These
cells also had the highest total Cr concentrations of the
five, yet were virtually devoid of hexavalent Cr. Thus they
presented a unique opportunity to observe changes in
speciation patterns over time. These cells have been
leached with water (approximately 40 gallons) each month
and leachate collected for total and hexavalent Cr deter-
mination.
The other test pits served as a "paired test" for observing
the effects of ferrous sulfate addition on changes in
hexavalent Cr in the soils. One test pit was treated initially
99
-------
with 8 grams of reducing agent, dissolved in 10 gallons of
water and adjusted to pH 4.5. This solution was sprayed
on with a hand sprayers. After a 24 hour incubation pe-
riod, the soil was sampled, extracted with phosphate buffer
and Cr (VI) measured. Two days later, the cell was given
another treatment of 400 grams of reducing agent dis-
solved in 10 gallons of water. Again, soil was collected
and analyzed after a 24 hour incubation period. Another
test pit remained untreated and allowed a comparison of
leachate concentrations of Cr (VI) in the treated and un-
treated cell.
Groundwater Well Installation and Experi-
mental Design:
For treatment of Cr (VI) impacted groundwater, a set of
injection and extraction wells were installed. Observation
wells were placed between the injection wells and each
extraction well at a distance of 2 feet. The observation
wells were screened at 25 to 30 feet. Samples were col-
lected at 30 minute intervals after the start of each test
and were analyzed for pH, electrical conductivity, dissolved
oxygen, hexavalent Cr and chloride. In the first test, 300
gallons of ferrous sulfate solution (200 mg/L buffered at
pH 4.5) was delivered to one injection well. At the same
time, 300 gallons of tap water with a 50 mg/L spike of
chloride (as NaCI) was injected in the other injection well
This was also followed with a tap water chaser after all the
initial solution had been injected. In this way it was pos-
sible to compare the effect of the ferrous sulfate reducing
solution to a solution of clean water. The difference (if any)
in Cr (VI) concentration after injection and chasing of the
two solutions would reflect reduction of hexavalent Cr to
Cr (III) by the reducing solution.
A second test was performed about 12 hours after the
completion of the first test. In this test, a much higher con-
centration of reducing solution was injected (almost 2000
mg/L). The purpose of this test was to add excess reduc-
ing solution to determine how low hexavalent Cr concen-
trations could be reduced.
Cr Solubility After Reduction
One of the key factors in assessing whether an in situ Cr
reduction is a viable remediation alternative will be to es-
tablish whetherthe reduction of Cr(VI) to Cr(lll) reduction
and subsequent precipitation is stable. Aquifer sand from
the area near MW 33 (field moist-saturated) was placed
in a plastic acid-washed column fitted with a stop-cock via
a small diameter hole to allow water to flow by gravity.
Ferrous sulfate reducing solution was added to the col-
umn and allowed to mix with the contents over an extended
period of time after which the precipitate in the column
was analyzed in a pH solubility test. By comparing the pH
vs solubility of Cr in the samples and comparing to known
Cr and Fe solids we could determine the solid phase pre-
cipitated and then predict the dissolved Cr concentration
expected at any pH. By observing solution concentration
over time, it would also be possible to determine if any
increase in solubility could occur. Solubility results were
supported by computer controlled scanning electron mi-
croscopy.
Results
Soil Chromium (VI) Reduction
Hexavalent Cr was reduced from approximately 600 mg/
kg to about 350 ug/kg in a 24 hour period. When the fer-
rous sulfate addition was increased to 400 grams per plot,
however, almost all the hexavalent Cr in the test pit soil
was reduced. Leachate samples from the drain tubes col-
lected about one month after the test pit construction shoed
that the treated pits had very low Cr (VI) concentrations
on the order of 10 mg/L or less. Leachate from the un-
treated pit, on the other hand, had hexavalent Cr concen-
trations on the order of 250 to 300 ug/L.
Groundwater Treatment
The results of the two different groundwater treatment tests
shows that where where a 200 ppm ferrous sulfate solu-
tion was introduced Cr (VI) decreased from about 3300
ug/L toabout 800 ug/L. Thus ferrous sulfate at 200 ppm
will reduce about 60 to 70% of the hexavalent Cr in Unit 1.
In the second test, a higher concentration of ferrous sul-
fate effectively reduced Cr (VI) to below detectable levels.
Thus treatment of Cr (VI) in groundwater with a ferrous
sulfate reducing solution appears to be a reasonable
means of reducing Cr (VI) to levels below detection (<20
ppb).
Results of Stability Experiments
The apparent solubility of Cr in the precipitates is much
lower than Cr(OH)3. It appears that a solid solution of Fe1.
xCrx(OH), controls the solubility as predicted from the re-
view of the scientific literature for the conditions of this
experiment. The solubility of the precipitated Cr is very
low, especially above pH 4.5 where 45 day solubility ob-
servations indicated that Cr(lll) was only 35 ug/L. At pH 5
the solubility was observed to be only 3 ug/L. Since the
aquifer sand and groundwater have a pH near 6 (about
5.7 to 6), it can be assumed that Cr(lll) concentrations will
remain less than detectable (<1 ppb) after treatment with
ferrous sulfate. This assumes that the aquifer pH will not
decrease substantially below pH 4.5.The Cr precipitate
appears to be stable under ambient pH conditions.
100
-------
In Situ Remediation of Hexavalent Chromium in Groundwater:
Practical Implementation
Jesse Tremaine and Nora L. Keel
Harding Lawson Associates, Inc.
1400 Center Point Boulevard, Suite 158
Knoxville, Tennessee 37932
423)-531 -1922; (fax) 423-531 -8226
jtremain @ harding.com
nkeel@harding.com
Overview
Hexavalent chromium [Cr (VI)] presents a problem in our
environment due to its high toxicity and carcinogenic ef-
fects. In addition, the generally high solubility and mobility
of Cr (VI) species tend to increase risk of Cr (VI) transport
in groundwater. At a Superfund site (Site) in the South-
eastern United States, a conventional pump-and-treat
groundwater remediation system is currently serving as
an interim action for the hydraulic control and treatment of
Cr (Vl)-contaminated groundwater. However, an innova-
tive in situ remedy for the full-scale remediation of ground-
water has been successfully field-tested at the Site and is
now being implemented on a full-scale basis. The in situ
remedy, which utilizes an acidic solution of ferrous sul-
fate, is a subsurface chemical reduction and precipitation
technology, whereby sorbed Cr (VI) is displaced from aqui-
fer solids and soluble Cr (VI) is reduced to the trivalent
state [Cr (III)]. The residual Cr (III) in groundwater is then
immobilized via the formation of an insoluble iron-chro-
mium hydroxide solid solution. Thus, both Cr (VI) and total
chromium are effectively treated to below applicable regu-
latory levels. The acidic ferrous sulfate reagent is injected
into the aquifer using chemical metering pumps and verti-
cal injection wells. Data indicate that groundwater con-
centrations between 0.1 and 4.4 milligrams per liter (mg/l)
of total chromium have been effectively treated to below
the Safe Drinking Water Act (SDWA) maximum contami-
nant level (MCL) of 0.1 mg/l.
Introduction
The Site is a former saw chain manufacturing facility that
operated between 1966 and 1982. Waste rinse waters
generated during metal-plating operations and other pro-
cesses were disposed of by direct discharge to the ground
surface in low-lying areas at the Site, referred to as "waste-
water ponds". These discharges caused contamination of
site groundwater. The main contaminant of concern at the
Site is Cr (VI).The plume of Cr (Vl)-contaminated ground-
water currently has a maximum width of approximately
550 feet and extends approximately 1,750 downgradient
from the past source area, and ranges in depth from 20 to
40 feet. Approximately one-half of the groundwater plume
is located beyond the property boundary of the Site. The
maximum concentration of Cr (VI) within the plume is be-
tween 4 and 5 mg/l.
In accordance with the Record of Decision (ROD) for the
Site, the groundwater remedy for the Site consists of in
situ chemical treatment of the Cr (VI) plume to reduce the
mobility and toxicity of the chromium in groundwater.The
implementation of the in situ remedy will reduce the over-
all treatment time needed to achieve remediation of the
plume as compared to conventional pump-and-treat tech-
nologies. Prior to selecting in situ chemical treatment as
the preferred groundwater remedy, site-specific engineer-
ing studies, including a demonstration study (pilot-scale
test), were conducted at the Site. The site-specific engi-
neering studies concluded that in situ chemical reduction
of Cr (VI) to the trivalent state, Cr (III), using ferrous sul-
fate would be the optimum remedial technology.
By using ferrous sulfate versus other possible reagents,
such as ferrous chloride, the ferrous iron provides for re-
duction of Cr (VI), while the sulfate works to displace
sorbed Cr (VI) from aquifer solids. Thus, not only is the
groundwater portion of the aquifer effectively remediated,
but the solid matrix of the aquifer is also treated. Further-
more, once the Cr (VI) is chemically reduced the residual
Cr (111) is then immobilized via the formation of an insoluble
iron-chromium hydroxide solid solution.The following equa-
tion shows the stoichiometry of the in situ chemical reac-
tion:
reduction precipitation
Cr(VI) + 3Fe(H) -» Cr(III) + 3Fe(III) -»
101
-------
RD/RA Program
The Remedial Design/Remedial Action (RD/RA) program
for the Site has been executed using the observation ap-
proach.The observational approach applied to RD recog-
nizes that a complete set of data, which fully character-
izes an entire Site, is not feasible, practical, or necessary
to obtain. Because of the variability of the aquifer at the
Site (e.g., grain size distribution, hydraulic conductivity,
stratigraphy, hydraulic gradient, etc.), and the multiple pa-
rameters controlling the in situ chemical reactions (e.g.,
pH, oxidation/reduction potential, ionic strength, buffering
capacity, etc.), it is not possible to develop a meaningful
predictive model except at the "macro" scale. Therefore,
the RD for the Site is based on a data set that is only
sufficient, and contains multiple contingencies for modify-
ing the design as new data become available.
Underthe observational approach, two separate treatment
tasks and two full-scale injections related to in situ ground-
water remediation have been conducted at the Site thus
far. The first treatment task consisted of a small-scale in-
jection of ferrous sulfate reagent into a single vertical in-
jection well in the area where the original source of ground-
water contamination was located. Because the first treat-
ment task was situated in an area having relatively low Cr
(VI) concentrations (0.2 to 0.3 mg/l), a second treatment
task, also consisting of a small-scale injection of ferrous
sulfate reagent into a single vertical injection well, was
conducted in an area of highest groundwater contamina-
tion [2 to 4 mg/l of Cr (VI)].The data obtained from the.first
two groundwater treatment tasks were used to design the
first full-scale injection, which was situated along the
upgradient edge of the plume. Data obtained from the first
full-scale injection was then used to plan the second full-
scale injection, which was situated approximately 200 feet
downgradient of the first injection.
Results of Small-Scale Treatment Tasks
The first treatment task consisted of one vertical injection
well and a network of seven monitoring wells. An acidic
ferrous sulfate (FS) solution was introduced into the sub-
surface via the vertical injection well using the existing
sprayfield infrastructure. The migration of the reagent
plume, along with its effect on the aquifer, was monitored
using four cross-gradient monitoring wells (to a distance
of 20 feet) and three downgradient monitoring wells (to a
distance of 30 feet).The required mass of FS injected was
based on an estimate of the mass of Cr (VI) in groundwa-
ter that required treatment in the selected treatment area
and a pre-determined stoichiometric excess (to account
for Cr (VI) that would be desorbed from the aquifer). A
chloride tracer was also used during the test to monitor
the progression of the reagent plume. The final reagent
mixture (as injected) contained ferrous iron at 240 mg/l
and chloride at 115 mg/l with a pH of 2.6.The reagent was
injected over a period of 8 hours.
Overall, the first treatment task was successful in reduc-
ing total and hexavalent chromium concentrations in
groundwater to below the remediation goal (RG) of 0.1
mg/l. The injected reagent dispersed cross-gradient at a
distance of at least 5 feet, but less than 10 feet from the
injection point. The reagent plume was observed 12 feet
downgradient between 14 and 22 hours after injection and
continued moving downgradient, reaching the last moni-
toring point, 30 feet downgradient of the injection point, by
Day 22. The presence of the reagent at each downgradient
well was evidenced by a decrease in pH, increase in con-
ductivity, and increase in total and ferrous iron [Fe (II)]
and sulfate concentrations. In addition, at each
downgradient monitoring location, the reduction of Cr (VI)
was instantaneous. A temporary increase (4-to 8-fold) in
total chromium was also observed indicating desorption
. of chromium from aquifer solids. As expected, the reduc-
tion of soluble Cr (VI) and desorption of Cr (VI) from aqui-
fer solids happens quickly. In addition, Fe (II) concentra-
tions decrease as desorbed Cr (VI) is reduced. As Cr (III)
and Fe (III) start to form an insoluble complex, concentra-
tions of total chromium and total iron decrease.
The second treatment task was situated in an area of high-
est Cr (VI) contamination and consisted of one vertical
injection well and a network of six monitoring wells. The
acidic FS solution was again introduced into the subsur-
face via the vertical injection well using the existing
sprayfield infrastructure. The migration of the reagent
plume, along with its effect on the aquifer, was monitored
using four downgradient monitoring wells (to a distance of
70 feet). The last monitoring point was an extraction well
that is part of the current pump-and-treat system at the
Site. Deep monitoring wells were installed at 20 and 40
feet downgradient of the injection point to monitor vertical
dispersion of the reagent plume. The required mass of FS
injected was again based on an estimate of the mass of
Cr (VI) in groundwater that required treatment in the se-
lected treatment area and a pre-determined stoichiomet-
ric excess (to account for Cr (VI) that would be desorbed
from the aquifer). The reagent mixture was again injected
over an 8-hour period using the same chemical param-
eters used during the first injection.
Overall, only limited data were obtained during the sec-
ond treatment task, as compared to the first task, due to
an unexpected flow path observed in the vicinity of the
extraction well. The reagent plume did not follow the ex-
pected linear path to the extraction well, but rather fol-
lowed an arched path around the downgradient monitor-
ing wells.The unexpected flow path has been attributed to
small-scale spatial heterogeneity within the aquifer. As a
result, the data collected during the second treatment task
(from temporary direct-push monitoring points installed
after injection) represented only post-treatment conditions.
However, the data that was collected provided pertinent
information, including confirmation that Cr (VI) was again
successfully treated to below the RG. Results also indi-
cated that approximately 10 feet of vertical dispersion was
achieved.
102
-------
Practical Implementation of the Full-Scale
Remedy
Based on data generated during the two small-scale treat-
ment tasks and an engineering evaluation that identified
the most effective and feasible method of injection, the
initial phase of full-scale treatment of groundwater was
designed and constructed. Full-scale implementation of
the groundwater remedy will consist of several lines of
vertical injection wells situated approximately 200 feet
apart, proceeding in the downgradient direction, with the
first injection line located at the upgradient edge of the
contaminant plume.
Data from the small-scale tasks indicated that, at a mini-
mum, a 20-foot spacing for injection wells would effec-
tively provide for lateral dispersion of the reagent. How-
ever, to provide the most feasible alternative for injection,
a 40-foot spacing of vertical injection wells was preferred.
Therefore, the first line of injection wells consisted of 11
injection wells, spaced 20 feet apart, which provided a
total treatment width of 240 feet. However, to evaluate the
effectiveness of lateral dispersion using a 40-foot spac-
ing, FS solution was initially only injected into every other
injection well.
The migration of the reagent plume, along with its effect
on the aquifer, was monitored using 13 monitoring wells,
located 25,43,50,75,100,150, and 200 feet downgradient
of the injection line. Eight of the 13 wells were situated so
that lateral dispersion of the reagent could be evaluated.
Three of the 13 wells were installed below and outside the
contaminant plume to determine whether the injection
causes a displacement of the plume into "clean" areas.
The required mass of FS injected was again based on an
estimate of the mass of Cr (VI) in groundwater that re-
quired treatment in the selected treatment area and a pre-
determined stoichiometric excess (to account for Cr (VI)
that would be desorbed from the aquifer). The reagent
mixture was injected over a 96-hour period using the same
chemical parameters used during the first injection.
After the first six weeks of post-injection monitoring, data
suggested that injection at 40-foot centers was not effec-
tive in dispersing the reagent. This was evidenced by the
appearance of reagent at a well located 43 feet directly
downgradient of an injection point, but the absence of re-
agent at dispersion wells (that straddle two injection points),
located 25 and 43 feet downgradient of the injection line.
Therefore, injection into the remaining wells was initiated
to provide data to complete an evaluation of full-scale treat-
ment effectiveness. However, in the same week that the
second injection was initiated, a trend of increasing sul-
fate concentrations in the first four dispersion wells was
confirmed. The increasing sulfate concentrations in dis-
persion wells indicated that the reagent plume, although
exhausted (i.e., lacking ferrous iron), had successfully
reached the dispersion wells.Thus, the 40-foot spacing of
injection wells appeared to be effective in laterally dispers-
ing the reagent into the aquifer.
Although the injection along Line 1 was successful in lat-
erally dispersing the reagent, effective treatment over the
entire width of the treatment cell was not accomplished
due to the exhaustion of the reagent. The exhaustion ap-
pears to be due to the buffering capacity of the aquifer in
the vicinity of the injection wells. Data for the monitoring
well located 43 feet directly downgradient of an injection
well support this conclusion. Although treatment was ob-
served at this well (as evidenced by a decrease in Cr (VI)
concentration), the pH of groundwater did not decrease,
ferrous iron was not detected, and total chromium con-
centrations did not increase before they decreased (which
is an indicator of Cr (VI) desorption from aquifer solids). It
appears that the pH of the reagent was buffered, thus caus-
ing the pH to increase from 3.0 (pH of the injected solu-
tion) to ambient levels, prior to the reagent reaching the
well. The increase in pH subsequently caused the oxida-
tion of the ferrous iron to ferric iron. The oxidation of the
ferrous iron to ferric iron was complete by the time the
reagent arrived at the well. This observation showed that
the reducing power of the reagent was exhausted.
A second injection into Line 1 injection wells was then
conducted to attempt to overcome the buffering capacity
of the aquifer. During the second injection, a larger vol-
ume of acidified reagent was injected into the aquifer (over
a 30-day period). The pH of the injection solution was also
lowered from approximately 3.0 to 2.5 to help overcome
the buffering capacity. In addition, enhancements to the
chemical deliver system for the reagent solution were
made. Unfortunately, during the second Line 1 injection,
the aquifer near and around Line 1 continued to buffer the
reagent, causing the iron to precipitate. Furthermore, due
to the mass of iron precipitate accumulating in close prox-
imity to the injection wells, backpressure within each well
increased significantly. The buffering condition near Line
1 is thought to be caused by the proximity of Line 1 to the
existing on-site sprayfield, which is used to discharge
groundwater that has been chemically treated to remove
chromium and other site contaminants. The water that is
discharged via the sprayfield is slightly basic and highly
buffered.
Since the buffering condition that was observed at Line 1
is not considered to be a site-wide problem, a decision
was made to install the second injection line, Line 2, ap-
proximately 200 feet downgradient of Line 1, to obtain the
data necessary to finalize the RD for the Site. The location
of Line 2 is outside the influence of the existing sprayfield
and is near the area where the first small-scale ground-
water treatment task was conducted, which provided posi-
tive evidence of effective groundwater treatment.
Injection along Line 2 is currently underway at the Site.
Line 2 consists of 10 injection wells located on 40-foot
centers. Preliminary data were not available at the time of
abstract submittal.
103
-------
Design of Cosolvent Flooding Solutions for
NAPL Remediation
Ronald W. Falta
Bracket! Hall Room 340C
Clemson University
Clemson, SC 29634-1908
864-656-0125; (fax) 864-656-1041
faltar@clemson.ed
Cosolvent flooding is a method for removing NAPLs from
aquifer systems using water miscible solvents. Typical
cosoivents are low molecular weight miscible alcohols such
as ethanol, methanol, isopropanol, n-propanol, and tert-
butanol. These alcohols tend to increase the NAPL solu-
bility and decrease the NAPL-water interfacial tension, thus
promoting both enhanced dissolution as well as separate
phase mobilization. The relative contribution of the two
removal mechanisms is largely controlled by the ternary
(alcohol/water/contaminant) phase behavior, and either
mechanism can be emphasized. With an enhanced dis-
solution approach, an alcohol that partitions mainly into
the aqueous phase is used, while with a separate phase
mobilization approach, an alcohol that partitions mainly
into the NAPL phase is used.These approaches have been
successfully field tested in EPA-SERDP sponsored ex-
periments at the Hill Air Force Base in Utah and at the
Dover Air Force Base in Delaware. An additional field test
of enhanced dissolution was performed at a dry cleaner
site in Florida.
A special variation of the mobilization approach is cur-
rently under consideration for chlorinated solvent (DNAPL)
remediation. This new approach uses a low density alco-
hol which partitions strongly enough into the chlorinated
solvent phase that it swells it and turns it into an LNAPL.
This technique will be field tested this coming year at the
Dover Air Force Base. The different cosolvent flooding
approaches may benefit from the addition of another
chemical component. This addition could be done for one
of four reasons. First, a high molecular weight alcohol,
such as pentanol or hexanol typically exhibits a stronger
cosolvency effect than the lower molecular weight alcohols.
Although the high molecular weight alcohols are not mis-
cible with water, they can be blended with the lower mo-
lecular weight alcohols to form a miscible flooding solu-:
tion. This approach results in a more effective enhanced
dissolution flooding solution.
Secondly, a high molecular weight alcohol, being hydro-
phobic, tends to preferentially partition into the NAPL
phase, swelling it. Thus a mixture of low and high molecu-
lar weight alcohols can form a more effective separate
phase mobilization flooding solution.
A third reason for adding another component to the flood-
ing solution is to alter its density. It may be difficult to de-
liver and recover a light alcohol flooding solution in some
situations. The addition of a dense salt or sugar can in-
crease the flooding solution density until it is comparable
to the resident pore water, and this approach could be
used for either enhanced dissolution or for NAPL mobili-
zation.
Finally, the partitioning of some low molecular weight, mis-
cible alcohols can be altered by the addition of a salt. The
general tendency is for the addition of the salt to increase
the degree of alcohol partitoning into the NAPL phase,
and this approach could be used to promote NAPL mobi-
lization with NAPL swelling.
It is expected that different sites will require different ap-
proaches, and it is not yet known which formulations will
prove to be the most efficient in general situations. The
issues of contaminant separation, and cosolvent recycling
and reinjection will likely be central to the ultimate choice
of cosolvent flooding solution composition for a given site.
104
-------
In Situ Flushing for Enhanced NAPL Site Remediation:
Metrics for Performance Assessment
P. Suresh C. Rao
School of Civil Engineering
Purdue University
West Lafayette, IN 47907-1284
765-496-6554; (fax) 765-496-1107
pscr@ecn.purdue.edu
M.D. Annable and J.W. Jawitz
Environmental Engineering Sciences Department
University of Florida
Gainesville, FL
Remediation technology performance testing requires a
demonstration that acceptable risk reduction has been
achieved in a cost-effective manner and gathering suffi-
cient evidence to establish that the observed risk reduc-
tion resulted from the application of the cleanup technol-
ogy. Here, by risk reduction we mean a decrease in the
concentrations or the total mass or mobility or toxicity of
the NAPL constituents of concern at the site. We will fo-
cus on technical measures of performance, and do not
consider other factors (e.g., stakeholder acceptance). The
level of site-specific testing required to establish credibil-
ity in the innovative remediation technology is a function
of the ease of treatment of the contaminant matrix and
the hydro-geologic complexity of the site. Greater amount
of testing is usually required at sites with highly complex
hydro-geology (e.g., heterogeneous aquifers) and recal-
citrant contaminants that are difficult to treat
Field-scale tests of in situ flushing techniques involving
alcohols or surfactants can be evaluated based on sev-
eral metrics that measure the technical performance. The
strongest proof of technology performance is based on
consistency among multiple lines of evidence, all pointing
to similar levels of risk reduction. Commonly, the perfor-
mance metrics are based on a decrease in NAPL con-
stituent concentrations in soils orgroundwater, either within
the test zone or at some compliance point/plane down-
gradient from the test zone.
For this purpose, NAPL constituent concentrations in soil
or groundwater samples taken before, during and after
the in situ flushing test can be compared to estimate the
total NAPL mass removed and the spatial distribution of
the NAPL mass remaining within in the target zone. Local
information from such point measurements is quite vari-
able, and therefore introduces significant uncertainty in
performance assessments unless a sufficiently large num-
ber of samples are collected. Alternatively, technology
performance can be based on spatially integrated infor-
mation, as might be the case with the use of inter-well
partitioning tracers tests and the total NAPL mass recov-
ered at extraction wells during the alcohol or surfactant
flood. Information from such integrated metrics is based
on much larger support volumes, and as such is usually
much less variable. However, averaging over much larger
spatial domains can lead to a decrease in the limits of
detection. Data we gathered during several field-scale pi-
lot tests of in situ flushing technology will be used to illus-
trate the performance metrics and their use in technology
assessment.
105
-------
Application of Pervaporation for the Removal of VOCs and Recovery of IPA
from Surfactant-based Soil Remediation Fluids
Franklin R. Alvarez, Leland M. Vane, Lynnann Hitchens and Johnny Springer
USEPA
National Risk Management Research Laboratory
26 W Martin Luther King Dr.
Cincinnati, OH 45268
513-569-7631
Alvarez. Franklin @ epamail.epa.gov
Overview
Aquifers and soils contaminated with non-aqueous phase
liquids (NAPLs), such as chlorinated solvents and gaso-
line/fuel components, pose a significant human health and
environmental risk as well as a significant technical chal-
lenge. Recently, the use of surfactant solutions to enhance
NALP solubility has received significant attention because
of the potential to dramatically reduce remediation times
compared to other technologies. Economic analyses of
the surfactant enhanced aquifer remediation (SEAR) pro-
cess have concluded that reuse of the surfactant is nec-
essary to make the process cost effective.
The EPA was involved in a Department of Defense Envi-
ronmental Security Technology Certification Program
project on surfactant-based soil remediation. The objec-
tive was to integrate the subsurface SEAR process with
above ground surfactant recovery and reuse processes.
The surfactants were used to solubilized subsurface con-
taminants. Once the solution has been pumped out of the
extraction wells, the contaminants will be removed by
pervaporation.To be able to reconcentrate and reuse the
surfactant, micellar enhanced ultrafiltration (MEUF) was
used in the field. Here, we will discuss and present the
design criterias, operating procedures and results for the
pervaporation system of the field demonstration.
The field demonstration took place in spring 1999 at Camp
Lejeune, Jacksonville, North Carolina. This Marine Corps
Military Base was contaminated mainly with tetrachloro-
ethylene due to a dry cleaning operation. The field dem-
onstration was divided into three separate tasks.The first
task was the injection/extraction of the surfactant. The
second task was the recovery of the VOC's by using
pervaporation. The third task was the reconcentration of
surfactant by micellar enhanced ultrafiltration.
The US EPA National Risk Management Research
Laboratory(NRMRL) selected pervaporation as the tech-
nology to recover the VOC's. NRMRL performed a series
of bench and pilot-scale studies using pervaporation to
remove VOC's from surfactant solutions. Based on the
performance and results obtained, it was concluded that
pervaporation could remove volatile NAPLs from surfac-
tant solutions. Another reason for selecting pervaporation,
it was found from previous studies foaming does not ad-
versely affect the performance of this technologies in com-
parison to other technologies. Technologies such as air
stripping, steam stripping, and vacuum extraction experi-
ence foaming problems when applied to surfactant solu-
tions.
Technology
Pervaporation is a separation process used for the purifi-
cation or separation of liquid mixtures. This technology
can be characterized as a membrane separation process
in which a membrane is selected to produce a desired
separation. One side of the membrane is exposed to a
liquid feed stream and a vacuum is pulled on the other
side. The component or components targeted for removal
permeate the membrane, and evaporate into the perme-
ate stream. The reduced pressure of the permeate pro-
vides the driving force for the separation. The slowly per-
meating components remain in the liquid residual, and can
be considered purified.
Pervaporation can be used in a variety of industry and
environmental applications. The selectivity of the mem-
branes dictates the separation, based on the permeability
of the component to be removed from the stream. One
application where pervaporation can be used is the dehy-
dration of organic streams, through the selection and use
of a hydrophilic membrane. Pervaporation technology is
attractive for environmental applications. This field dem-
onstration was a perfect case where pervaporation was
the technology to be used. Specifically one of our main
goals was the separation of organic compounds from
surfactant solutions used in soil flushing technologies. The
106
-------
targeted component to be removed from the surfactant
stream containing dry cleaning solvents was primarily tet-
rachloroethylene (PCE).
Field Site
The field site is an active dry cleaning shop on the grounds
of Marine Corps Base Camp Lejeune in Jacksonville, North
Carolina. Once the surfactant was injected, the extracted
fluid contained mainly PCE, Surfactant (Isalchem 145 4
PO ether sulfate (IC145), isopropyl alcohol(IPA), calcium
chloride, and varsol.
Process
The US EPA NRMRL was responsible for the design of
the pervaporation part of the field demonstration. The
pervaporation system was composed of a novel vibrating
membrane device as well as hollow fiber membrane mod-
ules. The system was designed for flow rates up to 10
gpm.The surfactant solution was pumped from the wells
to a three way valve at the front end of the pervaporation
system. This three way-valve allowed the surfactant solu-
tion to flow into an oil-water separator. The over flow from
the oil-water separator went to a 1,000 gallon fiber-glass
tank. The contents in this tank were fed to the pervaporation
system. This stream was sent to a vibrating pervaporation
system (V-SEP). After this, it went to a series of hollow
fiber membrane modules. The treated surfactant solution
was sent to the MEUF storage tank, where the surfactant
was recovered by MEUF.
The pervaporation system was designed for unattended
operation.The system was set up with automatic controls
to regulate flow, feed pressure, vacuum pressure and
whenever necessary empty the permeate reservoirs. Dur-
ing the entire demonstration the system was able to moni-
tor a variety of process variables and responded with warn-
ings when conditions exceeded acceptable ranges. The
field demonstration lasted around 90 days.
A second pervaporation unit was brought to the site to
study the performance of pervaporation in the recovery of
isopropyl alcohol(IPA). It was determined due to the high
concentration of IPA in the extracted stream that it would
be beneficial to study the application of pervaporation to
recover IPA in a field site demonstration. The permeate
from the MEUF system was the feed to this second
pervaporation unit. This system was operated in a batch
mode for a period of 10 days.
Operating Conditions
The main pervaporation unit was operated at approxi-
mately 1 gpm and 40 °C. The system was in operation for
80 days.The following conditions were monitored tempera-
tures, feed flowrate, residual pressures, and permeate
pressures. Sampling and analysis were performed three
times a week. Some of the analysis was performed on
site to determine the required dilutions of the samples to
be sent and analyzed in our labs in Cincinnati, Ohio. Undi-
luted samples were taken and sent to the labs in case of
any mistakes and/or human error.
The second pervaporation unit was in operation for 10
days. Several operating conditions were tested. Two dif-
ferent flow rates (3&6gpm) and three different tempera-
tures (40, 50 & 60 °C ) were tested for a total of 9 runs.
Feed and residual samples were taken every hour. The
analyses for these samples were performed on site.
In this paper, the results from this demonstration will be
presented. An economic analysis for this technology will
be shown to prove the cost efficiency of the process. Pre-
liminary results for the demonstration showed the main
objective of removing 95% of PCE was met.
Disclaimer
This is an extended abstract of a proposed presentation
and does not reflect USEPA policy.
107
-------
Coupling In Situ Flushing with Other Remediation Technologies
Mike Annable
University of Florida
PO Box 116450
Gainesville, FL 32611
352-392-1951
manna@ufl.edu
In situ flushing technologies that employ cosolvents and
surfactants have been recently tested at a number of sites.
The remediation goal has been to remove non-aqueous
phase liquid (NAPL) and reduce the long-term threat to
groundwater resources. In situ flushing techniques might
be coupled with other technologies in a "treatment train"
approach to extend the remediation goals to meet end-
points such as regulatory maximum concentration levels
(MCLs). Technologies such as free phase recovery prior
to in situ flushing and enhanced bioremediation following
a flood will be discussed. A recent DNAPL in situ flushing
project at a dry cleaner site will be used as an example of
technology coupling.
108
-------
Speaker List
Franklin Alvarez
USEPA
26 W. Martin Luther King
Cincinnati, Ohio 45266
513-569-7631
alvarez.franklin @ epa.gov
Mike Annable
University of Florida
PO Box 116450
Gainesville, FL 32611 -02
352-392-1951
manna@ufl.edu
Milovan Beljin
University of Cincinnati
9416 Shady Oak Court
Cincinnati, OH 45231
513-729-1602
mbeljin@aol.com
David Blowes
University of Waterloo
BFG2217
Waterloo, ON N2L 3G1
Canada
519-888-4878
blowes@sciborg.uwaterloo.ca
Robert Briggs
ManTech Corp.
14290 Sullyfield Circle
Chantilly,VA20151
703-814-8364
rbriggs® mantech.com
Dan Bryant
Geo-Cleanse
4 Mark Road Ste. C
Keniworth, NJ 07030
jdsmbryant@aol.com
Eva Davis
USEPA
PO Box 1198
Ada, OK 748121-11
580-436-8548
davis.eva@epa.gov
Paul Devane
AFRL-MLQE
139 Barnes Drive Ste. 2
Tyndall AFB, FL 32403-5323
850-283-6288
Paul.Devane@mlq.afrl.af.mil
Dominic Digiulio
USEPA
PO Box 1198
Ada, OK 74821-11
580-436-8605
digiulio.dominic@epamail.epa.gov
Tom Early
Oak Ridge National Laboratory
PO Box 2008
Oak Ridge, TN 37831 -6038
423-576-2103
eot@ornl.gov
Carl Enfield
NRMRL/USEPA
26 W. Martin Luther King
Cincinnati, Ohio 45228
513-569-7489
enf ield.carl @ epa.gov
Ron Falta
Clemson University
340C Bracket! Hall
Clemson, SC 29632
864-656-0125
faltar@clemson.edu
Edward Feltcorn
ORIA/USEPA
401 M Street, SW
Washington, DC 20460
202-564-9422
feltcorn.ed@epamail.epa.gov
John Fruchter
Battelle Northwest
PO Box 999
Richland, WA 99352
509-376-3937
john.fruchter@pnl.gov
109
-------
Robert Gillham
University of Waterloo
BFG2116
Waterloo, ON N2L3G1
Canada
519-888-4658
rwgillha@sciborg.uwaterloo.com
Douglas Grosse
ORD/NRMRL/USEPA
26 W. Martin Luther King
Cincinnati, OH 45268
513-569-7844
Grosse.douglas @ epamail.epa.gov
Scott Huling
USEPA
PO Box 1198
Ada, OK 74821-11
580-436-8610
huling.scott@epa.gov
Jaret Johnson
New Mexico Engineering Research Institute
901 University Blvd SE
Albuquerque, NM 87106
505-272-7226
jaretjohnson@nmeri.unm.edu
Roger Kennett
TNRCC
NMED-DOE Oversite Bureau
POBox5400-MS1396
Albuquerque, NM 87185-5400
roger_kennett@ nmenv.state.nm.us
Faruque Khan
USEPA
PO Box 1198
Ada, OK 74821-11
580-436-8704
khan.faruque © epa.gov
Nic Korte
Oak Ridge National Laboratory
2597 B 3/4 Road
Grand Junction, CO 81503
970-248-6210
nek@ornl.gov
Walter Kovalick
TIO/USEPA
401 M Street, SW
Washington, DC 20460
703-603-9910
kovallck.walter@ epa.gov
Frank Lenzo
ARCADIS Geraghty & Miller
3000 Cabot Blvd W. ste 300
Langhorne, PA 19047
f lenzo @ gmgw.com
Katherine Lowe
Oak Ridge National Laboratory
4830 W. 28th Avenue
Denver, CO 80212
303-966-3430
kslowe@ornl.gov
Stephen Luftig
OERR/USEPA
401 M Street, SW
Washington, DC 20460
luftig.stephen @ epa.gov
Tom McKeon
Project Performance Group
16935 SE 39th St
Bellevue, WA 98008
425-643-4634
tmcvkeon @ ppc.com
Alan Moore
ENSR
35 Nagog Park
Acton, MA 01720
978-635-9200 ext 3070
amoore@ensr.com
Norine Noonan
ORD/USEPA
401 M Street,SW
Washington, DC 20460
202-564-6620
noonan.norine @ epa.gov
Stephanie O'Hannesin
Enviro Metal Technologies
745 Bridge Street West, Ste 7
Waterloo, ON N2V2G6
Canada
519-746-2204
sohannesin @ eti.ca
Gary Pope
University of Texas
CMC: C0300U of Texas
Austin, TX 78712
512-471-3235
gary_pope @ pe.utexas.edu
Steve Price
KAI Technologies
199 Constitution Ave Bldg 1
Portsmouth, NH 03801
603-431-2266
Price @xdd-llc.com
110
-------
-------
United States
Environmental Protection Agency
National Risk Management
Research Laboratory, G-72
Cincinnati, OH 45268
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand comer.
If you do not wish to receive these reports CHECK HERED ;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use
$300
EPA/625/R-99/012
------- |