United States
%JF • * m Agency
Environmental Protection
Workshop on Monitoring
Oxidation-Reduction
Processes for Ground-water
Restoration
Workshop Summary
Dallas,Texas - April 25-27, 2000
-------
EPA/600/R-02/002
January 2002
on
for
April
Edited by
Richard T. Wilkin
Ralph D. Ludwig
Robert G. Ford
Subsurface Protection and Remediation Division
National Risk Management Research Laboratory
Ada, OK 74820
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH
-------
Notice
The preparation of this report has been funded wholly by the U.S. Environmental
Protection Agency's Office of Research and Development. The views expressed in
this Workshop Summary are those of the individual authors and do not necessarily
reflect the views and policies of the U.S. Environmental Protection Agency (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 recommendation for use.
-------
Foreword
The U.S. Environmental Protection Agency 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
environmental 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
technological and management approaches for preventing and reducing risks from pollution that threatens human
health and the environment. The focus of the Laboratory's research program is on methods and their cost-
effectiveness for prevention and control of pollution to air. land, water, and subsurface resources; protection of
water quality in public water systems; remediation of contaminated sites, sediments and ground water; prevention
and control of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public and
private sector partners to foster technologies that reduce the cost of compliance and to anticipate emerging
problems. NRMRL's research provides solutions to environmental problems by: developing and promoting
technologies that protect and improve the environment; advancing scientific and engineering information to
support regulatory and policy decisions; and providing the technical support and information transfer to ensure
implementation of environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is
published and made available by EPA's Office of Research and Development to assist the user community and to
link researchers with their clients. Monitoring oxidation-reduction processes has direct application to performance
assessments of remedial measures and to risk management approaches that reduce risk through minimizing
exposure. Understanding the complexities associated with oxidation-reduction processes for remediating contami-
nated ground water is a major priority of research and technology transfer for the U.S. Environmental Protection
Agency's Office of Research and Development and the National Risk Management Research Laboratory. This
report summarizes the findings of a workshop held to discuss and summarize the current state-of-the-science with
respect to methods of redox monitoring, data interpretation, and their applications to ground-water remediation.
Stephen G. Schmelling. Acting Director
Subsurface Protection and Remediation Division
National Risk Management Research Laboratory
111
-------
-------
Notice ii
Foreword iii
Introduction vii
Steering Committee ix
Workshop Participants x
Agenda xiii
Acknowledgments xv
Extended Abstracts 1
Role of Redox Processes and Spatial and Temporal Variability in Site Characterization 2
Identifying the Distribution of Terminal Electron-Accepting Processes (TEAPS) in
Ground-water Systems 5
Hydrogen Measurements and Redox Dynamics in Ground-water Systems 19
Geochemical Equilibrium and the Interpretation of EH 21
Abiotic Aquatic Redox Chemistry near the Oxic/Anoxic Boundary 25
Current State of Practice for Evaluation of Oxidation Reduction Processes
Important to the Biological and Chemical Destruction of Chlorinated Organic
Compounds in Ground Water 29
Redox Processes in Petroleum Hydrocarbon Remediation or
Why I Never Want to Measure Eh Again 35
Alteration of Reactive Mineral Surfaces by Ground Water 39
Aqueous Redox Chemistry and the Behavior of Iron in Acid Mine Waters 43
Redox Processes in Inorganic Remediation 49
The Relevance of Redox to Remediation 53
Group Discussion Summaries 57
Redox Processes in Petroleum Hydrocarbon Site Characterization and Remediation 59
Redox Processes for Remediation of Chlorinated Hydrocarbons and
Methyl /tW-Butyi Ether 67
Redox Processes in Inorganic Remediation 79
Field Measurement of Geochemical Redox Parameters 89
Measurement of Dissolved Gases 101
Solid Phase Redox Characterization 105
Appendix A 115
-------
VI
-------
Redox conditions are among the most important factors controlling contaminant transport and fate in ground-water
systems. Oxidation-reduction (redox) reactions, that is. reactions involving electron transfer, mediate the chemical
behavior of both inorganic and organic chemical constituents by affecting solubility, reactivity, and bioavailability.
In recent years, innovative technologies have emerged to meet ground water clean-up goals that take advantage of
the redox behavior of contaminant species. Remedial technologies that strategically manipulate or capitalize upon
subsurface redox conditions to achieve treatment may emphasize reductive processes (e.g., subsurface permeable
reactive barriers, monitored natural attenuation) or oxidative processes (e.g., air sparging). In this context, there is
an obvious need to provide stakeholders with a current survey of the scientific basis for understanding redox
behavior in subsurface systems within the framework of site characterization, selection of remedial technologies.
performance monitoring of remediation efforts, and site closure.
Many elements can exist in nature in more than one valence or oxidation state. Nine of the sixteen inorganic
species for which EPA has set specified maximum contaminant levels (MCLs) in drinking water exhibit multiple
oxidation states. These include antimony, arsenic, chromium, copper, lead, mercury, selenium, and nitrogen
(nitrate and nitrite). Consequently, the reactivity, solubility, and mobility in the environment of these and other
redox-sensitive elements depend, in part, on redox conditions. For example, uranium and chromium are insoluble
(immobile) under reducing (anaerobic) conditions. In contrast, iron and manganese are relatively insoluble under
oxidizing (aerobic) conditions but are quite soluble (mobile) under anaerobic conditions. Environmental mobility
of other potentially hazardous metals, such as cadmium, nickel, and zinc, is indirectly related to redox conditions
because these metals form ionic complexes and solid precipitates with redox-sensitive elements (e.g., sulfur).
Carbon may exist in several oxidation states, from +4 (most oxidized) to -4 (most reduced). Therefore, organic
contaminants in ground water can also be strongly influenced by redox conditions, especially through the metabolic
activity of microorganisms. For example, BTEX compounds are generally more biodegradable in oxidizing
(aerobic) ground water than in reducing (anaerobic) ground water. In contrast, chlorinated solvents such as
trichloroctnylcnc and tctrachlorocthylcnc are more biodegradable under more reducing conditions. Examples are
abundant in each of the major ground-water contaminant classes: inorganics, petroleum-derived hydrocarbons, and
chlorinated solvents, where redox processes have important implications in environmental risk assessment and risk
management.
Despite the importance of redox processes for understanding contaminant transport and fate, these processes
present recurring sources of confusion. Part of the confusion regarding redox processes no doubt relates to the
unfortunate fact that there is no single measurable quantity' that uniquely defines the redox state of natural systems.
Unlike pH measurements, which correlate with the concentration (activity) of hydronium ions in aqueous solution,
there is no equivalent single measurable parameter that correlates well with redox speciation in aquifer systems. At
one point there was promise that Eh measurements (referenced potential of a platinum electrode) could fulfill the
equivalent for redox characterization that pH measurements serve for evaluating acid-base determinations (see, for
example, Garrels and Christ, 1965, Solutions, Minerals, and Equilibria, Harper and Row, New York). Subsequent
research has shown that the Eh-pH framework, while correct within the rigid constraints of thcrmodynamic
equilibrium, is likely to give an inaccurate model of the redox state of ground-water systems. This is primarily
because ground-water systems are rarely, if ever, in thermodynamic equilibrium. The modern view that has
emerged over the past several decades is that proper redox characterization involves a more complete assessment
of the concentrations of potential electron donors and electron acceptors, that is, species that can participate in
redox or electron transfer reactions. Standard electrode (Eh) measurements may still play a useful role in redox
characterization, but such measurements need to be interpreted within the context of other geochemical and
biochemical data.
Vll
-------
Monitoring oxidation-reduction processes in ground-water systems can be viewed as an integrated assessment of
hydrogeochemical processes, microbiological diversity, and aquifer characteristics. A complete assessment of
redox processes will involve developing a system inventory of the most important redox species. This may include
measurement of dissolved oxygen, nitrogen species, iron species, manganese species, sulfur species, carbon
species, and dissolved hydrogen. The fact that redox processes in ground water vary both spatially and temporally
amplifies the complexity of redox assessments. Yet it is recognized that site-specific redox modeling is often a
requirement for predicting future plume migration and for selecting the most successful and cost-effective
remediation approaches.
The main objective of the workshop, therefore, was to provide a forum for the presentation, discussion, and
synthesis of research on monitoring and interpreting ground-water redox conditions in the context of site
evaluation, remedy selection, and performance monitoring. The workshop was attended by researchers, practi-
tioners, and regulators from governmental agencies and non-governmental institutions. Workshop participants
entered into a series of focused group discussions that explored redox processes within the organizational matrix of
contaminant class (petroleum hydrocarbons, chlorinated solvents, and inorganics) and measurement type (geochemical
parameters, dissolved gases, and solid-phase characterization). The principal findings of the workshop are reported
in this document as extended abstracts of the invited platform presentations and summary chapters of the focused
group discussions.
Richard T. Wilkin
Ralph D. Ludwig
Robert G. Ford
U.S. EPA
Office of Research and Development,
National Risk Management Research Laboratory,
Subsurface Protection and Remediation Division,
Ada. OK 74820
Vlll
-------
Michael Barcelona
University of Michigan
Ann Arbor, MI 48109
Frank Chapelle
U.S. Geological Survey
Columbia, SC29210 "
William Deutsch
RR-1 Box 1040
Sedgwick, ME 04676
Robert Ford
U.S. EPA, ORD/NRMRL
Ada, OK 74820
Ralph Ludwig
U.S. EPA, ORD/NRMRL
Ada, OK 74820
Carol Ptacek
Environment Canada
Burlington, ON L7R 4A6
Canada
Robert Puls
U.S. EPA, ORD/NRMRL
Ada, OK 74820
Stephen Schmelling
U.S. EPA, ORD/NRMRL
Ada, OK 74820
John Westall
Oregon State University
Corvallis, OR 973 31
Richard Wilkin
U.S. EPA, ORD/NRMRL
Ada, OK 74820
Richard Willey
U.S. EPA, Region 1
Boston, MA 02114
John Wilson
U.S. EPA, ORD/NRMRL
Ada, OK 74820
IX
-------
Jim Amonette
Pacific Northwest National Lab
Richiand, WA 99353
Ann Azadpour-Keeley
U.S. EPA, ORD/NRMRL
Ada, OK 74820
Jim Barker
University of Waterloo
Waterloo^ Ontario N2L 3G1
Shawn Benner
University of Waterloo
Waterloo! Ontario N2L 3G1
Allison Hacker
SAIC
Cincinnati, OH 45206
Evelyn Hartzell
SAIC
Cincinnati, OH 45206
Janet Herman
University of Virginia
Charlottesville, VA 22903
James Ingle
Oregon State University
Corvallis. OR 973 31
Paul Bradley
U.S. Geological Survey
Columbia, SC 29210
Sonya Jones
U.S. Geological Survey
Austin, TX 78754
Chris Braun
U.S. Geological Survey
Austin, TX 78754
Kyle Cook
SAIC
Cincinnati, OH 45206
Isabelle Cozzarelli
U.S. Geological Survey
Reston, VA20192
Andy Davis
Geomega, Inc.
Boulder, CO 80303
John Fruchter
Pacific Northwest National Lab
Richiand, WA 99352
Peggv Groeber
SAIC
Cincinnati. OH 45206
Donald Kampbell
U.S. EPA. ORD/NRMRL
Ada, OK 74829
Faruque Kahn
U.S. EPA. ORD/NRMRL
Ada, OK 74820
Roger Lee
U.S. Geological Survey
Dallas, TX 75202
Patrick Longmire
Los Alamos National Lab
Los Alamos, NM 87545
Michelle Lorah
U.S. Geological Survey
Baltimore, MD 2123 7"
Daryl Luce
U.S. EPA, Region 1
Boston, MA 02114
Douglas Grosse
U.S. EPA, ORD/NRMRL
Cincinnati. OH 45268
Donald Macalady
Colorado School of Mines
Golden, CO 80401
-------
Charlie Macon
Geomatrix Consultants, Inc.
Austin, TX 78703
Rick McGregor
302-285 Erb St. West
Waterloo, Ontario N21 1W5
Pat McLoughlin
Microseeps, Inc.
Pittsburgh, PA 15238
D. Kirk Nordstrom
U.S. Geological Survey
Boulder, CO 80303
Kirk O'Reilly
Chevron Research & Technology
Richmond, CA 94802
Keith Piontek
Hie Forrester Group
Chesterfield, MO 63017
Robert Pirkle
Microseeps, Inc.
Pittsburgh, PA 15238
Johnette Shockley
U.S. Army Corps of Engineers
Omaha, NE 68114
Tim Sivavec
GE Corp.,
Niskavuna, NY 12309
Gumming Su
ManTech Env.
Ada, OK 74820
Jim Szecsody
Pacific Northwest National Lab
Richland, WA 99352
Caroline Tebes Steven
U.S. EPA, ORD/NERL
Athens, GA 30605
Don Vroblesky
U.S. Geological Survey
Columbia, SC29210 "
Scott Warner
Geomatrix Consultants, Inc.
Oakland, CA 94612
John Washington
U.S. EPA, ORD/NERL
Athens, GA 30605
Eric Weber
U.S. EPA, ORD/NERL
Athens, GA 30605
Todd Wiedemeier
Parsons Engineering Science
Denver, CO 80290
Barbara Wilson
Dynamac Corporation
Ada. OK 74820
XI
-------
Xll
-------
April 25: Plenary Presentations
Morning - Ground Water Redox Chemistry: Abiotic andBiotic Processes
8:15- Introductory Remarks
8:30 - Role of Rcdox Processes and Spatial and Temporal
Variability in Site Characterization
9:05 - Identifying the Distribution of Terminal Electron-Accepting
Processes (TEAPS) in Ground-Water Systems
9:40 - Hydrogen Measurements and Redox Dynamics in
Ground-water Systems
10:15- Break
10:30 - Geochemical Equilibrium and the Interpretation of Eh
11:05 - Abiotic Aquatic Redox Chemistry Near the Oxic/Anoxic
Boundary
11:40 - Discussion/Question & Answer
12:00- Lunch
Afternoon Session - Redox Processes in Contaminant Remediation
1:15- Current State of Practice for Evaluation of Oxidation Reduction
Processes Important to the Biological and Chemical Destruction
of Chlorinated Organic Compounds in Ground Water
1:50 - Redox Processes in Petroleum Hydrocarbon Remediation
2:25 - Alteration of Reactive Mineral Surfaces by Ground Water
3:00- Break
3:15- Aqueous Redox Chemistry and the Behavior of Iron
in Acid Mine Drainage
3:50 - Redox Processes in Inorganic Remediation
4:25 - The Relevance of Redox to Remediation
5:00- Discussion
Richard Wilkin
Michael Barcelona
Frank Chapelle
Don Vrobleskv
John Westall
Don Macalady
John Wilson
Jim Barker
Janet Herman
D.K. Nordstrom
Bill Deutsch
Andy Davis
Xlll
-------
April 26: Breakout Sessions
Morning - Redox and Remediation
Session 1. Redox Processes in Petroleum Hydrocarbon Remediation
Session 2. Redox Processes in Chlorinated Hydrocarbon and MTBE Remediation
Session 3. Redox Processes in Inorganic Remediation
- Redox Measurement
Session 4. Measurement Issues: Geochemical Parameters
Session 5. Measurement Issues: Dissolved Gases
Session 6. Measurement Issues: Solid Phase
April 27: Summary Presentations
Session 1. Michael Barcelona, Ralph Ludwig
Session 2. Frank Chapelle
Session 3. Robert Puls
Session 4. Richard Wilkin
Session 5. Frank Chapelle
Session 6. Robert Ford
XIV
-------
The success of the workshop and the completion of this document are in no
small way attributed to the steadfast determination of many individuals. The
contributions and efforts of each workshop participant and facilitator are greatly
appreciated. A special thanks goes to the authors of all the plenary session
presentations, Robert Pirkle (Microseeps. Inc.) for suggesting the idea of a Redox
Workshop, and Bruce Pivetz, Scott Warner, Kirk Nordstrom, Rick McGregor,
Keith Piontek, and Jim Amonette for providing technical reviews of the Session
Summary chapters. Douglas Grossc (EPA/ORD/NRMRL) provided direction
throughout the planning and development of the conference. Science Applications
International Corporation (SAIC) handled conference logistics and provided
support during and after the workshop. Martha Williams (CSC) provided support
in document preparation. Bill Lyon (MERSC) collected and assembled the
information presented in Appendix A.
XV
-------
Invited speakers who gave oral presentations on the first day of the Workshop prepared the following extended
abstracts. The speakers and general topics of these presentations were selected by the steering committee. The
plenary presentations covered a range of topics relating to redox chemistry in ground water as they relate to contaminant
remediation. These presentations were intended to provide a foundation for discussion at the Workshop.
-------
of and and In Site
Characterization
Michael J. Barcelona
The University of Michigan
Ann Arbor, MI 48109-2099
Site characterization represents the initial phase of
active monitoring undertaken as part of active or passive
organic contaminant remediation efforts. Initial
characterization work sets the stage to evaluate the
progress of transformations of contaminants. There have
been frequent observations of: parent compound
disappearance, active microbial populations with
biotransformation capabilities, and the appearance/
disappearance of organic and inorganic constituents
which provide indirect evidence of bioremediation at
contaminated sites. However, quantitative evidence for
sustained removal of toxic compounds from complex
mixtures solely by biological processes is lacking. This
is due largely to the reliance on monitoring well samples
for evidence of biological activity rather than
identification of the mass of contaminants (and total
reactive organic carbon), redox capacities, and
estimations of the net removal/transformation of reactive
compounds over time.
A dynamic approach to quantitative site
characterization is needed which recognizes intrinsic
bioremediation as an active cleanup approach. Careful
attention must be paid to the identification of the three-
dimensional distribution of contaminant mass. Then, the
correspondence between the contaminant distribution and
favorable physical, geochemical and microbial conditions
in the subsurface over time provide a basis for net
contaminant removal estimates. Mere adaptations of
detective ground-water monitoring networks are
insufficient for the quantitative evaluation of intrinsic
bioremediation technologies.
The practice of site characterization for remediation
of subsurface organic contaminants has evolved slowly
in the past decade. Early guidelines (Scalf et al, 1981;
Barcelona et al., 1985;"u.S. EPA, 1986), for minimal
ground-water contamination detection monitoring (i.e.,
monitoring wells upgradient and downgradient) have
been applied to many sites of potential concern from
detection through remedial action selection phases.
This minimal approach has been used widely
regardless of the physicochcmical characteristics of
contaminant mixtures or the complexity of hydrogeologic
settings. For soluble inorganic constituents, this approach
may be adequate for detection purposes, but assessment
efforts require substantially more comprehensive
approaches. For organic contaminant assessment efforts.
(i.e., determinations of the nature and extent of
contamination), wells alone have been found to be
inadequate monitoring tools. Recognition of the value of
subsurface soil vapor surveys for volatile organic
components of fuel and solvent mixtures has led to
modified, monitoring well-based site characterization
approaches (Eklund, 1985). These approaches to site
characterization and monitoring network design often
suffer from the failure to identify the total mass of
contaminant in the subsurface. There are three main
reasons for this outcome.
First, although volatile organic compounds (VOC's)
are mobile in ground water and frequently early indicators
of plume movement (Plumb, 1987), their detection in
vapor or well samples and apparent aqueous concentration
distribution does not identify the total mass distribution
of organic contaminant (Robbins, 1989). Secondly, efforts
to correlate observed soil vapor or ground water VOC
concentrations with those in subsurface solid cores have
often been unsuccessful. This is because previous bulk
jar collection/refrigeration at 4°C guidelines for core
samples for VOC analyses lead to gross negative errors
(U.S. EPA, 1992). Thirdly, "snapshots" (i.e., one-time
surveys) of background and disturbed ground-water
chemistry conditions have been interpreted as "constant"
ignoring temporal variability in subsurface geochemistry,
particularly redox conditions.
More recent guidelines and recommendations on
network design and operations have led to more
comprehensive, cost-effective site characterization advice
(U.S. EPA, 1992; U.S. EPA, 1994). Also, excellent
reviews of characterization and long term monitoring
-------
needs and approaches in support of in-situ remediation
efforts should guide us in this regard. Site characterization
efforts provide a basis for long term monitoring design
and actually continue throughout the life of a remediation
project.
A dynamic, ongoing site characterization effort
includes objectives to:
1) identify the spatial distribution of contaminants,
particularly their relative fractionation in subsurface
solids, water, and vapor, along potential exposure
pathways recognizing that the mass of contaminants
frequently resides in the solids;
2) determine the corresponding spatial distribution of
total reactive organic matter (e.g., overall reduction
capacity; degradable normal, aliphatic and aromatic
hydrocarbon compounds, etc.). Overall microbial
biomass and activity in the subsurface, (and other
bioremediation indicators) are largely due to the total
mass of reactive organic carbon;
3) estimate the temporal stability of hydrogeologic and
geochemical conditions which favor microbial
transformations in background, source and
downgradicnt zones during the first year of
characterization and monitoring;
4) derive initial estimates of net microbial
transformations of contaminants, under known redox
conditions, which may be refined in an efficient long-
term monitoring network design. Over long time
periods, redox conditions and rates of degradation
can vary substantially near source areas.
The first three objectives establish the environment
of major contamination and geochemical conditions under
which bioremediation may occur. The latter two
objectives are vitally important since the evaluation of
the progress of intrinsic bioremediation processes
depends on distinguishing compound "losses" due to
dilution, sorption and chemical reactions from microbial
transformations. This approach has been suggested
emphatically by Wilson (Wilson, 1.993) and was later
developed into technical protocols by Wicdemeicr. ct al.
(Wiedemeier et al., 1995; Wiedemeier et al., 1996) (See
characterization detail. Table I).
The initial site characterization phase should be
designed to provide spatially dense coverage of critical
data over volumes corresponding to ten to one-hundred
year travel times along ground water flow paths. If the
flowpath intersects a discharge zone in less than 100 years,
then the volume should be scaled accordingly. For
example, if the flowpath discharges after 10 years, the
critical volume would be one year of travel time. The
"volume-averaged" values of the contaminants,
hydrogeologic and geochemical parameters within zones
along the flow path(s) should be derived from large
enough datasets to permit estimation of statistical
properties (e.g., mean, median, correlation distance,
variance, etc.). In general, this means that the datasets
for derived mass loadings of contaminants, aquifer
properties, and geochemical constituents (Table 1)
derived from spatial averages of data points must include
approximately 30 or more data points (Journel, 1986;
Hoeksema and Kitanidis, 1985; Gilbert and Simpson.
1985). Indeed, this minimum dataset size strictly applies
to points in a plane. Two major decisions which must be
made with regard to how spatially averaged masses of
contaminants, electron donors (e.g.. organic carbon. Fe2+,
S~, NH3, etc.) and electron acceptors (e.g., O2, NO3~, NO2~,
Fe, and Mn oxides, SO4=, etc.) are to be estimated. Some
attention should be given to field screening methods for
free phase or high residual NAPL Phase. (Xieetal. 1999).
The dynamic approach to site characterization for
chlorinated hydrocarbons is more demanding in review7
and data analysis than that for hydrocarbons. There are
very few published examples of site characterization
effects for these contaminants which have estimated mass
loadings in specific media.
1. Site Specific Parameters fo be Determined during
Site Characterization (Modifiedfrom Wiedemeier
etal., 1995)
FRACTIONATION AND SPATIAL EXTENT
OF CONTAMINATION
1. Extent and type of soil and ground water
contamination,
2, Location and extent of contaminant source
area(s).
(i.e., areas containing free- or residual-
phase product),
3, The potential for a continuing source due
to leaking tanks or pipelines.
HYDROGEOLOGIC AND GEOCHEMICAL
FRAMEWORK
4, Ground water geochemical parameter
distributions.
5. Regional hydrogeology including:
- Drinking water aquifers, and
-------
- Regional confining units.
6. Local and site-specific hydrogeology,
including:
- Local drinking water aquifers,
- Location of industrial, agricultural, and
domestic water wells,
- Patterns of aquifer use,
- Lithology,
- Site stratigraphy, including identification of
transmissive and nontransmissive units,
- Grain-size distributions, (sand vs. silt vs.
clay),
- Aquifer hydraulic conductivity deter-
mination and estimates from
grain-size distributions,
- Ground water hydraulic information,
- Preferential flow paths,
- Location and type of surface water bodies,
- Areas of local ground water recharge and
discharge,
7. Definition of potential exposure pathways
and receptors,
References
Barcelona, M. J., Gibb, J. P.. Helfrich, J. A., and Garske,
E. E. 1985. Practical Guide for Ground-Water
Sampling, Illinois State Water Survey, SWS Contract
Report 374. Ada. OK, U.S. Environmental Protection
Agency.
Eklund, B. 1985. Detection of Hydrocarbons in Ground
Water by Analysis of Shallow Soil Gas/Vapor, API
Publication No. 4394, Washington, D.C.
Gilbert, R. O. and Simpson, J. C. 1985. Kriging for
estimating spatial patterns of contaminants: Potential
and problems. Environmental Monitoring and
Assessment, v. 5, pp. 113-135.
Hoeksema, R. J. and Kitanidis. P. K. 1985. Analysis and
spatial structure of properties of selected aquifers.
Water Resources Research, v. 21, pp. 563-572.
Journel, A. G. 1986. Geostatistics: Models and tools for
earth sciences. Mathematical Geology^. 18, pp. 119-
140.
Plumb, R. H. 1987. A comparison of ground-water
monitoring data from CERCLA and RCRA sites.
Ground Water Monitoring Review, v. 7, pp. 94-100.
Robbins, G. A. 1989. Influence of using purged and
partially penetrating wells on contaminant detection,
mapping and modeling. Ground Water, v. 27, pp.
155-162.
Scalf, M. R., McNabb, J. R, Dunlop, W. J., Cosby, R. L.,
and Fryberger, J. S. 1981. Manual Of Groundwater
Sampling Procedures. National Water Well
Association, 93 p.
U.S. Environmental Protection Agency. 1986. RCRA
Technical Enforcement Guideline Document,
OSWER-9950.1, U.S. Government Printing Office,
Washington, B.C., 208 p.
U.S. Environmental Protection Agency. 1.992. RCRA
Ground Water Monitoring: Draft Technical Guidance
Document EPA/530-R-93-001, U.S. Government
Printing Office, Washington. D.C.
U.S. Environmental Protection Agency. 1994.
Proceedings of the Ground Water Sampling
Workshop, Dallas, TX, December 8-10, 1993, U.S.
Environmental Protection Agency, U.S. Government
Printing Office, Washington, D.C.
Wiedemeier, T. H., Wilson, J. T, Kampbell, D. H., Miller,
R. N., and Hansen, J. E. 1995. Technical protocol
for implementing intrinsic remediation with long-
term monitoring for natural chemical attenuation for
fuel contamination dissolved in groundwater.
Prepared for the Air Force Center for Environmental
Excellence, Brooks Air Force Base, Texas.
Wiedemeier, T. H., Swanson, M. A., Moutoux, D. E.,
Gordon, E. K., Hansen, J. E., Hass, P., Wilson, J. T,
Kampbell, D. H., and Chapelle, F. H. 1996. Draft
protocol for evaluating natural attenuation of
chlorinated solvents in groundwater. Prepared for
the Air Force Center for Environmental Excellence.
Brooks Air Force Base. Texas.
Wilson, J. T. 1993. Testing bioremediation in the field.
In //; Situ Bioremediation When Does It Work?,
Committee on In-situ Bioremediation, Water Science
and Technology Board National Research Council,
National Academy Press, Washington, D.C., pp. 160-
164.
Xie, G., Barcelona, M. J., and Fang, J. 1999.
Quantification and interpretation of TPH in sediment
samples by a GC-MS method and comparison with
EPA 418.1 and a rapid field method. Analytical
Chemistry, v. 71, pp. 1899-1904.
-------
the of (TEAPS) In
Francis H. Chapelle, James E. Landmeyer, and Paul M. Bradley
U.S. Geological Survey
Columbia, SC 29210-7651
Introduction
The distribution of terminal electron-accepting
processes (TEAPs) in ground-water systems can be
assessed by documenting the availability of particular
electron acceptors (oxygen, nitrate, Fe(III), sulfate), by
showing the distribution of characteristic final products
(Fc(II), sulfide, methane), and measuring concentrations
of intermediate products (hydrogen) of microbial
metabolism. When applied to a gasoline-contaminated
site in South Carolina (Laurel Bay), this methodology-
shows that redox conditions changed continuously since
the initial spill occurred in 1990. By 1994, oxygen had
been depleted in the contaminated zone, and Fe(III) and
sulfate-reduction were the predominant TEAPs.
Mcthanogcncsis was not measurable in 1994. By 1996,
depletion of Fe(III) and sulfate resulted in the initiation
of a discrete methanogenic zone in the source area.
Between 1996 and 2000, this methanogenic zone
progressively grew until most of the plume core was
dominated by methanogenic metabolism. These dynamic
spatial and temporal changes in TEAPs, which have been
documented at several other sites as well, appear to be
characteristic of rcdox processes in contaminated ground-
water systems. The methodology for deducing TEAPs,
which is illustrated in this paper by application to a
gasoline-contaminated aquifer, is generally applicable to
all ground-water systems.
Reduction-oxidation (redox) processes affect the
chemical composition of ground water in all aquifer
systems. In addition, ambient redox processes affect
biodcgradation or biotransformation rates of
contaminants such as petroleum hydrocarbons,
chlorinated solvents, and metals. Thus, reliable methods
for characterizing redox processes are crucial to re-
mediating contaminated ground-water systems. The
purpose of this paper is to describe a non-equilibrium,
kinetic framework for documenting the spatial and
temporal distribution of microbially mediated terminal
electron-accepting processes (TEAPs) in ground-water
systems.
The Equilibrium Approach
The traditional approach for characterizing redox
processes in ground water is based on conventions and
methods developed in classical physical chemistry (Sillen,
1952). In particular, Back and Barnes (1965) used
platinum electrode measurements to characterize the Eh
of ground-water samples. This approach was
systematized by Stumm and Morgan (1981), who
suggested that the theoretical concentration of electrons
in aqueous solution (pe), could be used in direct analogy
to hydrogen ion concentration (pH) as a "master variable"
to describe redox processes. In this treatment, the pe of a
water sample is a linear function of Eh (pe = 16.9 * Eh at
25°C).
The definition of Eh, and thus pe, is given by the
Nernst equation (Equation 1), in which the Eh of a solution
is related to concentrations of aqueous redox couples at
chemical equilibrium and the voltage of a standard
hydrogen electrode (E°). For example, when
concentrations of aqueous Fe3 andFe21 are at equilibrium.
Eh is defined as:
£ft = r + MMIlog
nF
aFe2+
(1)
Equation 1 illustrates an important point. Eh is only
uniquely defined when a system is at thermodynamic
equilibrium Q)rQver, 1982, p. 257). If the activity of Fe3+
and Fc2+ ions in Equation 1 are not at equilibrium,
measured Eh (based on electrode potential) represents
electrode response to all reactions active at the platinum
surface, but will not be an "Eh" by the definition Equation
1.
In the 1960's and 1970's, when ground-w7ater
systems were thought to be largely sterile environments
devoid of microbial life, assuming equilibrium or near-
equilibrium conditions seemed to be a reasonable
approach. However, in the early 1980's, it became clear
that ground-water systems contained active, respiring,
reproducing microorganisms (Wilson et al., 1983;
-------
Chapelle etal., 1987). Furthermore, it gradually became
clear that many of the important redox processes
occurring in ground-water systems were catalyzed by
microorganisms (Baedecker et al. 1988; Chapelle and
Lovley, 1992; Baedecker et al., 1993). This realization
coincided with growing evidence that Eh measurements
of ground waters were problematic. In particular, it was
shown that Eh measurements with platinum electrodes
were usually not consistent with Eh's calculated from
the Nernst equation using different redox couples
(Lindberg and Runnels, 1984). There are several
reasons for these problems. These include:
1. The actual concentrations of free electrons in water
are so low (~ 10"55 M) that they arc essentially zero
(Thorstenson, 1984). Thus, while pe (and Eh) is a
thermodynamically definable quantity, it is not
practical to measure (as is pH).
2. The pH electrode responds to aqueous concentrations
of hydrogen ions. However, the Eh electrode does
not respond to aqueous concentrations of electrons,
but rather to electron transfers between solutes
(Thorstenson. 1984). A platinum Eh electrode,
therefore, readily responds to concentrations of Fc2*
and Fe3" because they react rapidly with platinum.
However, because CO, and CH4 react sluggishly on
a platinum surface, the Eh electrode is relatively
insensitive to the CO7CH4 couple.
3. Microorganisms cannot actively respire and
reproduce unless there is available free energy to
drive their metabolism. That is, microorganisms
require that their immediate environment not he at
thermodynamic equilibrium. Thus, using Eh to
describe redox processes driven by microbial
processes violates the underlying equilibrium
assumption of Eh.
In light of these difficulties, it is not surprising that
Eh measurements in ground-water systems are so often
problematic.
The Kinetic Approach
Equilibrium considerations are not the only way to
describe redox processes. The metabolism of
microorganisms is based on the cycling of electrons from
electron donors (often organic carbon) to electron
acceptors such as molecular oxygen, nitrate, ferric iron,
sulfate, carbon dioxide, or other mineral electron
acceptors. This flow of electrons is capable of doing work.
Microorganisms capture this electrical energy, convert it
to chemical energy, and use it to support their life
functions. If it is assumed that redox processes in ground-
water systems are driven predominantly by microbial
metabolism, it becomes possible to describe these
processes by the cycling of electron donors, electron
acceptors, and intermediate products of microbial
metabolism. Because this is an inherently non-
equilibrium, kinetic description, it is termed the "kinetic
approach"' (Lovley. Chapelle, and Woodward, 1994).
A kinetic description of redox processes in ground-
water systems includes two components. These are
(1) documenting the source of electrons (electron donor)
that supports microbial metabolism, and (2) documenting
the final sink for electrons (electron acceptors) that
supports microbial metabolism. In many ground-water
systems, identifying electron donors is not a difficult
problem since particulate or dissolved organic carbon is
the most common source of electrons for subsurface
microorganisms. Amore difficult problem is determining
the terminal electron accepting processes (TEAPs) that
occur in a system. This problem is made even more
difficult by the inherent heterogeneity of ground-water
systems. This heterogeneity causes both spatial (Chapelle
and Lovley. 1992) and temporal (Vroblesky and Chapelle.
1994) variations in TEAPs.
This paper describes methodology for deducing the
spatial and temporal distribution of TEAPs in ground-
water systems. This methodology is illustrated using
long-term monitoring data of a gasoline spill site in
South Carolina.
for TEAJPs
Microbial Competition
Studies in aquatic sediment microbiology have
clearly demonstrated that microbially mediated redox
processes tend to become segregated into discrete zones.
When this happens, the observed sequence of redox zones
follows a predicable pattern. At the sediment-water
interface, oxic metabolism predominates. This oxic zone
is underlain by zones dominated by nitrate reduction,
manganese reduction, and ferric iron reduction (Froelich
et al., 1979). In more organic rich marine sediments, it is
commonly observed that a sulfate-reducing zone overlies
a zone dominated by methanogenesis (Martens and
Bemer, 1977).
For many years, the mechanisms causing the
observed segregation of redox zones were not clear.
However, studies with pure cultures of methanogens and
sulfate reducers (Lovley and Klug, 1983), followed by
studies with aquatic sediments (Lovley and Klug, 1986;
-------
Lovley and Goodwin. 1987), showed that redox zonation
resulted from the ecology of aquatic sediments. In aquatic
sediments, organic matter oxidation is carried out by food
chains in which fermentative microorganisms initiate
biodegradation with the production of fermentation
products such as acetate and hydrogen (Figure 1). These
fermentation products are then consumed by terminal
electron-accepting microorganisms such as Fe(III)-
reducers or sulfate reducers. Because Fe(III) reduction
produces more energy per mole of acetate or hydrogen
oxidation. Fe(III) reducers are able to lower
environmental concentrations of these fermentation
products below levels required by less efficient sulfate
reducers. Thus, when Fe(III) is available, Fe(III)-reducers
can outcompctc sulfate reducers for available hydrogen
(Figure 1), and sequester the majority (although not all)
of the available electron flow. This, in turn, leads to the
observed development of redox zones dominated by
particular redox processes.
The principle mechanism of microbial competition
is easily illustrated experimentally. If methanogenic
aquatic sediments are placed in anoxic vials, with some
vials amended with oxygen-free distilled water
(methanogenic treatment), some amended with a 5 mM
nitrate solution (nitrate-reducing treatment), some
amended with a suspension of amorphous ferric
hydroxide suspension (Fe(III)-reducing treatment), and
some amended with a 5 mM' sulfate solution (sulfate-
reducing treatment), hydrogen concentrations are
observed to change over several days to different
characteristic levels (Figure 2). Mcthanogcnesis, the least
efficient hydrogen-consuming TEAP, is characterized by
the highest hydrogen concentrations (5-15 nM) by the
end of the experiment. Sulfate reduction is characterized
by hydrogen concentrations in the 1-4 nM range, Fe(III)
reduction in the 0.2-0.8 nM range, and nitrate reduction
having hydrogen concentrations < 0.1 nM at the end of
the experiment. This leads to the tendency for discrete
redox zones to develop in sediments. This overall
principle, in turn, suggests a methodology for determining
the distribution of redox zones in ground-water systems.
Interpreting Concentrations of Final Products and
Intermediate Products
Ground-water systems are usually more complex
hydrologic systems than aquatic sediments. Part of this
complexity stems from the inherently heterogeneous
lithologic and geochemical nature of ground-water
systems (Cherry et al, 1995; Cozzarelli et al., 1999). Part
Complex
Organic
Matter
Fermentation
Acetate
and other
Organic Acids
Hydrogen
Predominant
TEAP
Fe(lll) Reduction Sulfate Reduction Methanogenesis
CH,,C02
!l), CO2
HS. CO,
figure 1. Development of predominant terminal electron-accepting processes due to m,icrobial competition for electron donors.
-------
c
o
•*=
0)
o
c
o
O
c
0)
O)
2
T3
><
I
24 -
22 -
20 -
18 -
16 -
14 -
12 -
10 -
8-
6-
4-
2-
0-
Methanogenic
Sulfate-reducing
Fe(lll)-reducing
Nitrate-reducing
0
80
100
20 40 60
Time (hrs)
Figure 2. Characteristic response of hydrogen concentrations to methanogenic, sulfate-reducing, Fe(III)-reducing, and
nitrate-reducing conditions.
of this complexity also stems from the predominance of
advective ground-water flow as the principal solute
transport mechanism, in contrast to molecular diffusion-
dominated aquatic sediments. Nevertheless, because
particular TEAPs require the presence of certain electron
acceptors (nitrate or sulfate, for example), and because
particular TEAPs produce characteristic final products
(methane from methanogenesis, sulfide from sulfate
reduction. Fc(II) from Fc(III) reduction), concentration
trends of potential electron acceptors and final products
through time and/or space can be used to deduce ongoing
TEAPs. Similarly, concentrations of intermediate
products such as hydrogen can provide insight into the
spatial and temporal distribution of TEAPs. A schematic
diagram showing the logic behind this methodology is
shown in Figure 3.
If. for example, sulfate concentrations are observed
to decrease with time or along a flowpath segment of a
ground-water system, sulfide concentrations are observed
to increase, and hydrogen concentrations are observed to
be in the range characteristic of sulfate reduction, the
occurrence of sulfate reduction can be identified in that
portion of the aquifer with a high degree of confidence
(Figure 3).
In practice, however, this methodology often
encounters uncertainties. For example, if methane,
sulfide, and Fe(II) concentrations are observed to increase
along a flowpath segment or with time, it can be
concluded that methanogenesis, sulfate reduction, and
Fe(III) reduction are ongoing (Baedecker et al. 1993).
However, because methane, sulfide. and Fe(II) are
actively transported by flowing ground water, it is often
difficult to determine the spatial location of each redox
zone. If methane is present in ground water at a particular
well location, it cannot be concluded that methanogenesis
is occurring in that well's screened interval. Rather, it
can only be concluded that methanogenesis is occurring
somewhere upgradient of that well. In these cases,
concentrations of hydrogen, which is continuously cycled
by microorganisms and thus not subject to significant
advective transport, may more accurately identify redox
zones. However, it is always possible that wells can be
screened across different redox zones, leading to
intermediate, non-diagnostic concentrations of hydrogen.
Moreover, redox zones shift arcally and temporally in
response to recharge events (Vroblesky and Chapelle,
1994). the availability of organic matter, or other
environmental factors. Because of this observed
variability, the terminology 'predominant" TEAPs. rather
than "exclusive" TEAPs, is used. Under such dynamic
conditions, discrete redox zones may not be well-
developed. Such possible complexities must always be
considered when using the methodology of Figure 3.
Evolution, Laurel
A gasoline-contaminated aquifer in Laurel Bay,
South Carolina can be used to illustrate the methodology
-------
rt!
V
iLiCTRON ACCEPTOR ^^
CONSUMPT1-ON -*j FINAL PRODUCT —
=-" I ACeUMULATJON ; PREDOMMANT
t= I TEAR DIAGNOSIS
YES
-^-- S-H RftdudtCrt
-|
I PRODUCT (Hz)
JC-ONCINTRATIOM
NO
I
YES
frt NO;, Ftacut&on
YES
NO
NO
, Per1" R€SLi£,'.ion
YES
| NO
J NO
l!
V *'rS
NO
YES
J
j I'lT^^'r^nnfiensritel
I ........................... NO
figure 3. Flow chart for deducing predominant terminal electron-accepting processes in ground-water systems.
shown in Figure 3. This site has been previously described
(Landmeyeretal.. 1998). The aquifer consists of highly
permeable and relatively homogeneous marine beach
sands underlain by a clay confining bed. Gasoline leaked
into the aquifer from an underground storage tank in about
1990 (Figure 4). The tank was removed in 1992, and an
investigation of ground-water contamination initiated
soon after. Figure 4 shows a cross section of this site
along the direction of ground-water flow with the
locations of monitoring wells used to delineate redox
processes. Figures 5-9 show changes in redox conditions
along the axis of this cross section during the time period
1994-2000.
In 1994, concentrations of benzene in the
contaminant source area were in the 40 mg/L range, and
decreased downgradient (Figure 5a). Upgradient of the
contaminant source area, ground water contained high
concentrations of dissolved oxygen (Figure 5 a) and sulfate
(Figure 5c), which decreased rapidly as ground water
moved into the contaminated zone. Fe(II) concentrations
were low upgradient, and increased in the source area
(Figure 5b). Concentrations of sulfide and methane were
below detection levels (Figure 5c).
Based on the concentrations of electron acceptors
(oxygen, sulfate) and metabolic products (methane,
sulfide, Fe(II)), it can be deduced that Fe(III) reduction
and sulfate reduction were ongoing in the source area in
1994. but that oxic metabolism predominated upgradient
of the source area and downgradient of well LB-8. There
w7as no evidence of methanogenesis in 1994.
-------
Over the next several years, redox conditions
changed markedly. By 1996. the anoxic zone was
extending further downgradient (Figure 6a), and
methanogenesis and sulfate reduction had become
increasingly more prevalent processes. Concentrations
of methane increased markedly (Figure 6b). and the zone
of sulfate depletion which was confined to the source
area in 1994 (Figure 5c), extended 200 feet downgradient
in 1996 (Figure 6b). The initiation of methanogenesis in
and near the source area is indicated both by the
generation of methane (Figure 6b) and by the presence
of hydrogen concentrations in the range characteristic of
methanogenesis (Figure 6d).
This trend toward increasing importance of meth-
anogenesis continued over time. By 1998, concentrations
of dissolved oxygen were below 1 mg/L for the entire
600 foot plume (Figure la), concentrations of methane
(Figure 7b) and sulfide (Figure 7c) continued to increase.
Hydrogen concentrations were in the methanogenic range
near the source area, decreased into the sulfate reducing
range between 200 and 400 feet along the flowpath. and
then decreased into the range characteristic of Fc(III)
reduction (Figure 7d).
By January. 2000. low dissolved oxygen con-
centrations continued to dominate the core of the plume
(Figure 8a), concentrations of methane remained
relatively high (Figure 8b). sulfate concentrations
remained relatively depleted (Figure 8c), and hydrogen
concentrations in the range characteristic of
methanogenesis extended from the source area to 400
feet downgradient.
The picture that emerges from this analysis at the
Laurel Bay site using the logic of Figure 3. is that of an
initially oxic aquifer (pre-1990), which became anoxic
in the contaminant source area soon after the gasoline
release (Figure 9). Once anoxic conditions were achieved
(1994). the predominant TEAP was Fc(III) reduction with
some evidence of sulfate reduction, but no evidence of
methanogenesis. By 1996. the source area had become
methanogenic, and this methanogenic zone gradually
expanded until it reached 400 feet downgradient in 2000.
Similar evolution of redox zones in petroleum
hydrocarbon-contaminated aquifers has been dem-
onstrated previously (Baedcckcr ct al., 1993; Cozzarclli
et al., 1999), and so this overall behavior is expected.
This analysis does illustrate, however, that a kinetic
MLS-
Gasollne-Contamlncited
Sediments
Explanation
Screen or
Sampling Port
'CLAY/ //// / //////
figure 4. Cross section showing location of contaminated zone, orientation of the contaminant plume, and locations of
monitoring wells at the Laurel Ba) > site.
10
-------
March 1994
CD
en
o
tf>
w
—•— Dissolved Oxygen
•••A--- Benzene
50
40
30
- 20
- 10
- 0
CO
100
200
300
400
en
£
g
'I
-i—*
<§
25
20
15
10
5
—•— Dissolved iron
•••A--- Methane
B
100
200
300
400
100 200
Distance Along Flowpath (feet)
300
400
figure 5. Concentrations ofredox-sensitive species at the Laurel Ba) > site, 1994,
11
-------
June 1996
D)
s
x
o
T3
1 >
O
100
200
300
400
—•— Dissolved Iron
• ••A--- Methane
100
200
300
A A-
1.0
- 0.8
ZT
- 0.6 "3)
- 0.4 «"
- 0.2
- 0.0
400
35
30
- 25
- 20
A
>•£- /^
10 fc
- 5
- 0
100
200
300
400
(U
D)
O
"O
>s
I
100 200
Distance Along Flowpath (feet)
Figure 6. Concentrations ofredox-sensitive species at the Laurel Ba,} > site, 1996,
300
D
400
12
-------
CO
b
1.0
0.8 -
0.6 -
0.4 -
0.2 -
0.0 -
0
50
3 40 H
E
- so H
o
- 20 H
T3
site, 1998,
D
13
-------
January 2000
E
o ^-~
w
b
7
6 -
5 -
4 -
3 -
2 -
1 -
0 -
o
CO
b
(U
o)
o
w
CO
25
20
15
10
5
0
10
8
6
4
2
0
(U
D)
(D
"5
W
CO
10 -
—•— Dissolved Oxygen
• -A-- Benzene
A V-U *-
0 100 200 300 400 500 600 700
—•— Dissolved Iron
• -A-- Methane
- 1
t.
-4-
0 B
ro
I
- 0
0 100 200 300 400 500 600 700
100 200
300
400 500 600 700
100 200 300 400 500
Distance Along Flowpath (feet)
Figure 8. Concentrations ofredox-sensitive species at the Laurel Ba,} > site, 1996,
600 700
D
14
-------
approach that considers sequential utilization of electron
acceptors, production of final products, and
concentrations of intermediate products can lead to an
accurate evaluation of redox processes in ground-water
systems.
Conclusions
The spatial and temporal distribution of redox
conditions in ground-water systems are complex, even
at relatively simple sites. For example, the Laurel Bay
site is a relatively homogeneous aquifer, but the
development of discrete redox zones occurred gradually
over many years and involved considerable spatial
variability. Nevertheless, the data show that the overall
evolution of redox processes follows certain predictable
patterns. Available electron acceptors at this site were
consumed in the order (O2>Fe(III)>SO/l>CO2) with
mcthanogcncsis (CO2 reduction) occurring only after
other available electron acceptors were exhausted. This
is consistent with the model (Figure 2) of sequential
electron-acceptor utilization, and the principle of
microbial competition for available electron donors.
Fe(lll)-Reduction Oxygen Reduction
1994
Methanogenesis
\: Fe(lll]-Reduction Oxygen Reduction
' V^
Sulfate Reduction
1996
Methanogenesis Fefllll-Reduction
1998
2000
Figure 9. Schematic diagram showing the evolution of redox zones at the Laurel Bay site, 1994-2000,
15
-------
The data clearly show, however, that the
development of these redox zones is accompanied by
considerable heterogeneity. For example, in 1996
concentrations of sulfate and sulfide were highly variable
at the interface between low-sulfate ground water adjacent
to the contaminant source area (Figure 6c). and high-
sulfatc ground water downgradicnt of this interface.
Clearly, the shift between Fe(III) reduction to sulfate
reduction involved interfingering, poorly defined redox
zones. The development of these spatially complex zones
probably depends on many factors including variations
in contaminant concentrations, hydraulic conductivity,
Fe(III) content of aquifer sediments, and sulfate
concentrations.
In spite of this complexity, it is possible to document
the distribution of redox processes in ground-water
systems in both time and space using a kinetic approach
based on well-known principles of microbial ecology
(Figures 2 & 3). This kinetic approach provides a much
better resolution of ongoing microbial processes than is
afforded by traditional equilibrium approaches (Eh or pe)
to redox processes. This, in turn, makes it possible to
make more precise predictions concerning the fate and
mobility of redox-sensitive contaminants in ground-water
systems.
References
Back. W. and Barnes, I. 1965. Relation of electrochemical
potentials and iron content to ground-water flow
patterns. U.S. Geological Survey Professional Paper
498-C.
Baedecker, M. J., Siegel, D. L., Bennett. P. C., and
Cozzarelli, I. M. 1988. The fate and effects of crude
oil in a shallow aquifer: 1. The distribution of chemical
species and gcochcmical facics. In U.S. Geological
Survey Ibxic Substances Hydrology Program Review,
eds. G. E. Mallard and S. E. Ragone, U.S. Geological
Survey Water-Resources Investigations Report 88-
4220, pp.13-20.
Baedecker, M. J., Cozzarelli, I. M., Eganhouse, R. P.,
Sicgcl, D. I, and Bennett, P. C. 1993. Crude oil in a
shallow sand and gravel aquifer. III. Biogeochemical
reactions and mass balance modeling in anoxic
ground water. Applied Geochemistry, v. 8, pp. 569-
586.
Chapelle, F. H., Zelibor, J. L., Grimes. D. J., and Knobel,
L. L. 1987. Bacteria in deep coastal plain sediments
of Maryland: A possible source of CO2 to ground
water. Waler Resources Research, v. 23, pp. 1625-
1632.
Chapelle, F. H., and Lovley, D. R. 1992. Competitive
exclusion of sulfate-reduction by Fe(Ill)-reducing
bacteria: A mechanism for producing discrete zones
of high-iron ground water. Ground'Water, v. 30, pp.
29-36.
Chapelle, F. H., McMalion, P. B., Dubrovsky, N. M.,
Fujii, R. F., Oaksford, E. T, and Vroblesky, D. A.
1995. Deducing the distribution of terminal electron-
accepting processes in hydrologically diverse ground-
water systems. Water Resources Research, v. 31, pp.
359-371.
Cherry, J. A., Barker. J. F., Feenstra, S., Gillham, R. W.,
Mackay, D. M., and Smyth, D. J. A. 1995. The Borden
site for groundwater contamination experiments:
1978-1995. In Groundwater and Subsurface
Remediation: Research Strategies for In-situ
Remediation, eds. H. Kolbus, B. Barczewski, and H.
P. Kosschitzky, Springer-Verlag, Berlin, pp. 102-127.
Cozzarelli, I. M., Herman, J. S., Baedecker, M. J.. and
Fischer, J. M. 1999. Geochemical heterogeneity of a
gasoline-contaminated aquifer. Journal of
Contaminant Hydrology, v. 40, pp. 261-284.
Drever, J. I. 1982. The Geochemistry of Natural Waters.
Prentice Hall, Englewood Cliffs, New Jersey. 388 p.
Froelich, P. N., Klinkhammer, G. P., Bender, M'. L.,
Luedtke, N. A., Heath, G. R., Cullen, D., and
Dauphin, P. 1979. Early oxidation of organic matter
in pelagic sediments of the eastern equatorial Atlantic:
suboxic diagenesis. Geochimica et Cosmochimica
Ada, v. 43, pp. 1075-1090.
Landmeyer, J. E., Chapelle, F. H., Bradley, P. M.,
Pankow, J. F, Church, C. D., andTratnyek, P. G. 1998.
Fate of MTBE relative to benzene in a gasoline-
contaminated aquifer (1993-98). Ground Water
Monitoring and Remediation, v. 18, pp. 93-102.
Lindberg, R. D. and Runnells, D. D. 1984. Ground-water
redox reactions: An analysis of equilibrium state
applied to Eh measurements and gcochcmical
modeling. Science, v. 225, pp. 925-927.
Lovley, D. R. and Klug, M.. J. 1983. Sulfate reducers can
outcompete methanogens at fresh water sulfate
concentrations. Applied and Environmental
Microbiology, v. 45, pp. 187-192.
Lovley, D. R., and M. J. Klug. 1986. Model for the
distribution of methane production and sulfate
reduction in freshwater sediments. Geochimica et
Cosmochimica Acta, v. 50, pp. 11-18.
16
-------
Lovley, D. R. and Goodwin, S. 1988. Hydrogen
concentrations as an indicator of the predominant
terminal electron-accepting reactions in aquatic
sediments. GeochimicaetCosmochimicaActa, v. 52,
pp. 2993-3003.
17
-------
18
-------
and In
Don A. Vroblesky and Francis H. Chapelle
U.S. Geological Survey
Columbia, South Carolina 29210-7651
The use of hydrogen, a transient intermediate
product of anaerobic microbial metabolism, as an
indicator of predominant terminal electron accepting
processes (TEAPs) was introduced by Lovley and
Goodwin (1988). The standard method used for
measuring H2 in ground water is a gas-stripping procedure
that has previously been referred to as the "bubble strip'"
method (Vroblesky and Chapelle, 1994; Chapelle and
McMahon, 1991; Chapelle ct al, 1995). In this method,
a standard 250 ml gas sampling bulb (Supelco) is
continuously purged with ground water at a rate of about
500 mL/min. A 20-mL bubble of HL-free N,, gas is
introduced into the sampling bulb through the septum.
As the bubble is vigorously agitated by the stream of
inflowing water, slightly soluble gases such as H2 are
stripped from the water to the gas phase over time, and
asymptotically come into equilibrium with the water
flowing through the bulb. After equilibrium is achieved
(-20-30 min), a gas sample is withdrawn from the bulb
by using glass syringes with stopcocks and analyzed for
H Analysis is done using gas chromatography with
reduction gas detection (Trace Analytical. Menlo Park,
California). Other methods that have been used to collect
H2 from ground water include a down-hole sampler, and
a diffusion sampler (Chapelle et al.. 1997).
Measurement of H., using the gas-stripping method
can be affected by a variety of factors. A field investigation
showed that one factor affecting the H, concentration is
the type of pumping equipment. Four different methods
for pumping ground water were evaluated. These were
(1) a peristaltic pump (Geotech Environmental
Equipment. Inc., Denver. Colorado). (2) a stainless steel
piston pump (Bennet Sample Pumps, Inc., Amarillo,
Texas), (3) a bladder pump (Well Wizard Bladder Pump,
QED Environmental Systems, Ann Arbor, Michigan), and
(4) a stainless steel submersible pump (MP-1, Grundfos
Pumps. Corp., Clovis, California). The peristaltic pump
draws water under negative pressure, the piston and
bladder pumps push water under positive pressure, and
the Grundfos pump pushes water under positive pressure
by impellers driven by a Direct Current (DC) electrical
motor.
The various pumps found that peristaltic, piston,
and bladder pumps all gave similar results when applied
to water produced from the same well (Chapelle et al..
1997). It was observed, however, that peristaltic-pumped
water (which draws water under a negative pressure)
enhanced the gas-stripping process and equilibrated
slightly faster (~20 min) than either piston (~25 min) or
bladder pumps (-30 min), which push water under a
positive pressure (Figure 1A).
The direct current-driven submersible pump gave
much higher H., concentrations than the other pumps.
Because H, can be clcctrolytically produced by direct
current, it is probable the high H,, concentrations observed
using this submersible pump reflect the interaction of
direct current with water. Thus, direct current submersible
pumps appear to not be suitable for measuring Ft, in
ground water (Chapelle et al.. 1997).
When the peristaltic pumping rate was increased
from 500 to 1.000 mL/min, equilibration was achieved
in proportionally less time (Figure IB). These
observations indicate that equilibration times, and thus
the accuracy and reproducibility of the gas-stripping
method, can van' depending on pumping methods and
rates. These results indicate that specific combinations
of pumping systems and rates should be individually-
tested to evaluate the length of time needed for
gas-aqueous phase equilibration.
Measurements from two field sites indicate that iron
or steel well casings produce H, .which masks H2
concentrations in ground water. PVC-cased wells, or
wells cased with other materials that do not produce H2,
are necessary for measuring H, concentrations in ground
water (Chapelle et al., 1997).
Measurements of the dissolved H2 and other
biologically active solutes in ground water from a shallow
19
-------
10 20 30 40
Time (min)
50
0 5 10 15 20 25 30 35
Time (min)
Figure 1, (A) Comparison ofequlibration time of the gas-stripping procedure using the bladder, piston, and peristaltic
pumps, and(B) the influence of flow rate on equilibration times for the peristaltic pump (modifledfrom Chape lie
era/., 1997). '
petroleum hydrocarbon-contaminated aquifer indicate
that the distribution of TEAPs is highly dynamic in time
and space (Vroblesky and Chapeile, 1994). During times
when little or no sulfatc or oxygen is introduced to the
ground water by recharge from rainfall or alternate
sources, the available oxygen. Fe(III). and sulfate can be
depleted by respiring bacteria, leaving methanogenesis
as the predominant TEAR Introduction of oxygen from
rainfall infiltration into the methanogenic parts of the
aquifer can cause precipitation of Fe(III) as grain coatings.
Once the oxygen is depleted, Fe(III)-reducing bacteria
can take advantage of the precipitated Fc(III) to sequester
most of the electron flow7 from degradation of organic
compounds. Introduction of sulfate into a methanogenic
part of the aquifer can cause the predominant TEAP to
shift to sulfate reduction. The shift from methanogenesis
to sulfate reduction was observed to range from <10 days
to about 31/2 months, depending on the sulfate and organic
carbon concentrations.
References
Chapeile, F. H., Vroblesky, D. A., Woodward, J. C., and
Lovely, D. R. 1997. Practical considerations for
measuring hydrogen concentrations in groundwater.
Environmental Science and Technology, v. 31,
pp. 2873-2877.
Chapeile, F. H., McMahon, P. B., Dubrovsky, N. M.,
Fujii. R. F.. Oaksford, E. T., and Vroblesky, D. A.
1995. Deducing the distribution of terminal electron-
accepting processes in hydrologically diverse
ground-water systems. Water Resources Research^
v. 31, pp. 359-371.
Chapeile, F. H. and McMahon, P. B. 1991. Geochemistry
of dissolved inorganic carbon in a coastal plain
aquifer: Sulfate from confining beds as an oxidant
inmicrobial CO., production. Journal of Hydrology,
v. 127, pp. 85-108.
Lovley, D. R. and Goodwin, S. 1988. Hydrogen
concentrations as an indicator of the predominant
terminal electron-accepting reactions in aquatic
sediments. GeochimicaetCosmochimicaActa,v. 52,
pp. 2993-3003.
Vroblesky, D. A. and Chapeile, F. H. 1994. Temporal
and spatial changes of terminal-accepting processes
in a petroleum hydrocarbon-contaminated aquifer and
the significance for contaminant biodegradation.
Water Resources Research, v. 30, pp. 1561-1570.
20
-------
and the of E.
H
John C. Westall
Oregon State University
CorvaliisOR97331
Among the most important factors controlling the
persistence, mobility, and biological effects of many
organic and inorganic pollutants are oxidation-reduction
(redox) transformations. For example, it is known that
dehalogenation of aromatic compounds occurs most
readily under anoxic conditions, while ring cleavage and
conversion to CO2 occurs most readily under oxic
conditions. In design of in-situ remediation, control of
microbial metabolism through the proper balance between
carbon substrate and the terminal electron acceptor is
critical in obtaining the desired transformation of an
organic contaminant. In general, degradation rates and
pathways of organic compounds are known to depend
strongly on ambient redox conditions. Among inorganic
chemicals, speciation, and hence transport and biological
effects, depends strongly on redox reactions. For
example, Cr(VI) is generally mobile and toxic, while
Cr(III) is insoluble, immobile, and relatively nontoxic.
Thus, a clear understanding of the redox state and
speciation of chemicals is required in virtually all aspects
of hazardous materials management: evaluation of
disposal options, risk assessments for contaminated sites,
evaluation of clean-up options, and performance
assessment. Thus, it is useful to recast the issue of
'"Gcochemical Equilibrium and the Interpretation of EH"
in terms of the broader underlying questions, (i) what is
the redox state of an environment?, and (ii) does a
particular contaminant undergo a redox transformation
under these conditions?
The
All too often the problem of characterizing the redox
state of an environment is approached in the following
manner: (i) measure the apparent redox potential with a
Pt electrode; (ii) calculate an equilibrium distribution of
the pollutant from the measured redox potential and pH;
(iii) remark in disgust that the calculated distribution does
not match the observed distribution, that the Pt electrode
docs not really work for this application, and that the
system is probably not at equilibrium anyway.
The pitfalls of redox potential measurements in the
environment have been well documented (Morris and
Stumm, 1967; Whitfield, 1974; Lmdberg and Runnelis,
1984): low redox buffer capacity of the aqueous phase,
slow kinetics among the environmental redox reactions,
and slow kinetics of the electrode itself. Still. Pt
electrodes continue to be used, and in spite of these
problems, appear to give useful data under some
circumstances, as will be discussed.
Alternatives to the Pt-electrode include (i) direct
determination of concentrations of all significant
redox-active species in water samples; (ii) redox titrations
of aquifer materials to determine an effective differential
or integral "redox-capacity;" and (iii) active control of
redox conditions in laboratory "redox-stats."
Stumm (1984) advances the concept of determining
the concentrations of major redox active species as a
means of characterizing the redox status of an
environment. Frevert (1984) used commercially available
metal and O2 electrodes and a feedback system to attempt
to control the redox status of a system. Although he
considered slow kinetics of redox reactions in his study,
he concluded that it is impractical to characterize the
redox status of a sediment-water system in terms of a
single parameter. Related work on redox capacities was
carried out by Whitfield (1972), who investigated the
voltammetric properties of sediment-water systems.
In summary, if one chooses to advance the argument
that it is impossible to characterize the redox status of a
natural system, there is certainly evidence to support the
argument. On the other hand, if one recognizes that redox
transformations are extremely important in transport of
many elements and in degradation of many organic
compounds, that some redox reactions are coupled, and
that additional information on characterizing the redox
status of the environment would be helpful, then there is
evidence that much insight could be gained from further
experimentation.
21
-------
Approach to Characterizing the
Our general conceptual approach to
characterization of redox transformations of contaminants
in the subsurface environment involves interactions along
the three legs of a triangle, at the apices of which are: (i)
the geochemical matrix, (ii) the redox sensor, and (iii)
the rcdox-activc contaminant. Coupling of redox
processes must be understood at two different linkages:
contaminant and matrix, and sensor and matrix. The
contaminant-matrix linkage must be investigated on a
species-by-species basis, without the a priori assumption
of equilibrium.
The Matrix
The "geochemical matrix" drives the redox trans-
formations. One way to approach redox conditions in
aquifers is to consider major oxidant and reductant
species. In some regards, these species can be considered
"redox poising" species.
O2 - In the presence of significant concentrations of O2,
experimental EH values (Pt electrode vs SHE) arc
relatively high (300 - 400 mV).
MnO,(s)/Mn(II) - Mn(III) and Mn(IV) oxide/hydroxide
minerals are present in some aquifers; equilibrium
potentials of this couple are relatively high, about
500-650 mV at pH 7, although appearance of Mn(II)
is observed at lower potentials (200 to 300 mV).
Fe(OH)3(s)/Fe(II) - Fe(III) oxide/hydroxide minerals are
found in many near surface aquifers. Iron is about
ten times more abundant than manganese, and is
therefore important in poising redox potential. Fe(III)
is a considerably weaker oxidant than either O2 or
Mn(X,(s). The equilibrium redox potential is about
100-200mVatpH7.
SO42VS°/S2" - The sulfur system is the most important
inorganic system poising redox potentials in reducing
conditions "(EH -100 to -300 mV).
Organic compounds are the ultimate reductants in
natural systems and have a major role in defining redox
conditions.
The
To answer the underlying question, "will the
contaminant undergo a redox transformation under the
prevailing environmental conditions," the ideal sensor
would obviously be based on the sensor itself. For
example, if the contaminant could be either
(i) immobilized on a suitable electrochemical or
spectrochemical transducer, or (ii) injected, withdrawn,
and chemically detected in a push-pull test, then one
would have directly the information on the extent of
conversion. As an alternative to the contaminant itself,
redox indicators have been immobilized to form such
redox sensors (Lemmon et ah, 1996).
As surrogates for this direct information one can
determine the concentration of chemical species that
reflect the redox status of the environment. In
environments with significant amounts of organic carbon
and microbial activity, the redox status is probably best
defined by the terminal electron acceptor for degradation
of organic carbon (e.g., O2, Mn(IV), NO;, Fe(III), SO/-,
CO2. H2, etc.). Microbial degradation of organic
compounds is generally most closely linked to this
measure of redox status. Over the last several years
Lovley, Chapelle and coworkers have advanced the idea
that H_(g) is the best indicator of the metabolic state of a
subsurface microbial system (Lovley et ah, 1988, 1994;
Chapelle et ah. 1996); hence H2(g) could be the best
indicator of the likelihood that a contaminant will undergo
a redox transformation, particularly under reducing
conditions. In environments in which transport processes
are rapid compared to redox processes, the presence.
absence, or concentration of a particular species may not
alone be sufficient to indicate the primary terminal
electron acceptor and the redox status of the environment.
Therefore, rates of redox processes relative to rates of
transport processes are important ancillary information.
From a practical point of view, the availability of
facilities, personnel, and time for chemical analysis are
important. If a method is impractical, it will not be used,
even though it might give the best answer. If a simple
method (e.g., Pt electrode) gives the same information
as a more complex method (e.g. complete chemical
analysis), is there any need for the complete chemical
analysis? Associated with the use of the Pt electrode are
an array of questions that ultimately determine its utility:
What sorts of species poise the Pt electrode at each of the
major terminal electron acceptor steps in a biologically
active medium (e.g., O2, Mn(IV), NO;, Fe(III), SO42-,
CO2, H2 etc.)? What is the exchange current density, i.e.,
how well is the electrode poised (Bard and Faulkner,
1980)? Does it vary dramatically with redox level? What
role is played by "mediators", i.e., soluble molecules that
shuttle between particle surfaces and the electrode
surface? Docs the performance depend strongly on
whether or not the electrode is in direct contact with the
particles?
Contrary to what one may be led to believe by a
superficial review of the literature, the Pt electrode can
be of value in determining environmental redox
22
-------
conditions. Data from experiments of Patrick and co-
workers (e.g.. Masscheleyn et al., 1990. 1992) show that
redox transformations of inorganic species, and
appearance and disappearance of terminal electron
accepting species track very well Pt electrode potentials.
This discussion is not to advance the argument that the
Pt electrode even approaches an ideal indicator for
environmental redox reactions, but just to say that its
performance warrants serious consideration.
A complicating factor in almost all of the methods
is that the majority of redox buffer capacity in the
subsurface environment is associated with biotic, biogenic
or mineral particles, not with the aqueous solution. Then
one of the major issues in understanding redox
transformations of environmental contaminants is the
distinction between transformations that occur directly
at particles (biotic or abiotic) vs. transformations that
occur in solution, presumably via "mediator molecules"
that conduct between the particle surfaces and the
contaminant being transformed. This issue is important
not only in characterizing transformations of
contaminants, but also in how sensors work. Particle-
free media are much more amenable to chemical analysis
such as spectrophotometry, cyclic voltammetry, etc.
A further complication is the apparent redox
"disequilibria" observed in many subsurface
environments. These conditions arise when transport
kinetics are rapid compared to reaction kinetics.
Associated with this observation is the question, how do
each of the various "redox sensor technologies" described
above track these changes?
Summary
In assessing the redox status of the subsurface
environment, it is important to pose the right question.
Previous work has indicated that the question, "what is
the EH or pe of a particular environment?" is not optimal;
a better approach is to pose the question, "will a certain
redox transformation take place under a certain set of
conditions, and how can this set of conditions best be
recognized?" Judging performance of redox sensors on
this criterion may lead to the best decision.
References
Bard, A. J. and Faulkner, L. R. 1980. Electrochemical
Methods. Wiley, New York, pp. 391 398.
Chapelle, F. H, Hack, S. K., Adnens, P., Henry, M. A.,
and Bradley. P. M. 1996. Comparison of EH and H7
Measurements for delineating redox processes in a
contaminated aquifer. Environmental Science and
Technology, v. 30, pp. 3565-3569.
Frevert. T. 1984. Can the redox conditions in natural
waters be predicted by a single parameter?
Schweizerische Zeitschrift Fur Hydrologie, v. 46,
pp. 269 290.
Lemmon, T. L., Westall, J. C., and Ingle, J. D., Jr. 1996.
Development of redox sensors for environmental
applications based on immobilized redox indicators.
Analytical Chemistry, v. 68, pp. 947-953.
Lindbcrg, R. D. and Runnclls, D. D. 1984. Groundwater
redox reactions: An analysis of equilibrium state
applied to Eh measurements and geochemical
modelling. Science, v. 225, pp. 925-927.
Lovley, D. R. and Goodwin, S. 1988. Hydrogen
concentrations as an indicator of the predominant
terminal electron-accepting reactions in aquatic
sediments. GeochimicaetCosmochimicaActa,\. 52,
pp. 2993-3003.
Lovley, D. R., Chapelle, F. H., and Woodward, J. C. 1994.
Use of dissolved H, concentrations to determine
distribution of microbially catalyzed redox reactions
in anoxic groundwater. Environmental Science and
Technology, v. 28, pp.1205-1210.
Masscheleyn. P. H., DeLaune, R D., and Patrick, Wr. H.. Jr.
1990. Transformations of selenium as affected by
sediment oxidation-reduction potential and pH.
Environmental Science and Technology, v. 24, pp. 91-
96.
Masscheleyn, P. H., Pardue, J. H., DeLaune. R. D.. and
Patrick, W. H., Jr. 1992. Chromium redox chemistry
in a lower Mississippi bottomland hardwood wetland.
Environmental Science and Technology, v. 26,
pp. 1217-1226.
Morris, J. C. andW. Stumm. 1967. Redox equilibria and
measurements of potentials in the aquatic
environment. In Equilibrium Concepts in Natural
Water Systems, ed. W. Stumm, Advances in
Chemistry Series No. 67, American Chemical
Society, Washington, D.C., pp. 270-285.
Stumm, W. 1984. Interpretation and measurement of
redox intensity in natural waters. Schweizerische
Zeilschrift Fur Hydrologie, v. 46, pp. 291-296.
Whitfield. M. 1972. The electrochemical characteristics
of natural redox cells. Limnology and Oceanography,
v. 17, pp. 383-393.
Whitfield, M. 1974. Thermodynamic limitations on the
use of the platinum electrode in Eh measurements.
Limnology and Oceanography, v. 19, pp. 857-865.
23
-------
24
-------
the
Donald L. Macalady
Colorado School of Mines
Golden, CO 80401
Considerations of ground water oxidation/reduction
chemistry are often predicated on chemical measurements
designed to determine the "redox state" of the system
under investigation. There has been a long-standing
tendency to seek one or more universal parameters that
will serve as uniformly applicable indicators of the redox-
state of any ground-water system. The desired outcome
is a set of measured characteristics that enable prediction
of redox processes. There is abundant literature to
document that such a universal mechanism for prediction
of redox processes is not forthcoming (see for example
Lyman et al, 1987 and references therein).
For any measured set of activities of the constituent
chemicals of a redox couple, there is a corresponding
Nernstian electrochemical potential. The disturbing fact
relative to the chemistry of natural waters is that parallel
determinations of electrochemical potentials calculated
from different redox couples in the same water sample
often imply distinctly different redox conditions (White
et al., 1990; Walton-Day et al., 1990). These and other
studies imply that the lack of a definable "redox state" of
an aquatic system, is a consequence of the lack of
thermodynamic equilibrium among the components of
the system.
This condition is particularly prevalent in a class of
systems that is especially important in considerations of
contaminant transformations in ground water. By this is
meant those ground-water systems that are substantially
depleted in dissolved oxygen (DO). The most commonly
used methods of analyses for DO have detection limits
greater than 2 (imolar, a concentration much greater than
the detection limits for most other rcdox-activc species
in natural and contaminated ground waters.
Several examples from the literature will confirm
that many ground waters in the regime between truly
anoxic and 2 umolar DO are characterized by pronounced
spatial and/or temporal disequilibria (Peiffer et al., 1999;
Walton-Day et al, 1990: Walton-Day, 1991; Eary and
Schramkc, 1990). In these systems, both electrode
measurements of Eh and determinations of specific redox -
activc species such as ferrous iron are likely to
erroneously imply extremely low oxygen partial
pressures.
A seldom-used colorimetric method is described that
uses a proprietary redox-sensitive dye and allows accurate
detection of DO levels as low as 0.2 jimolar. Use of this
technique illustrates the importance of this low oxygen
tension regime as far as comprehensive understandings
of ground-w7ater redox conditions are concerned.
Detectable DO in waters containing measurable ferrous
iron and sulfide, for example, is not uncommon. Eh
measurements in such systems typically imply oxygen
partial pressures 10-20 orders of magnitude lower than
measured values (White et al., 1990).
In other words, redox disequilibrium is perhaps most
commonplace in the types of ground-water systems most
relevant to considerations of contaminant transformation.
Furthermore, the causes of such disequilibria can
commonly be the presence or absence of precisely those
chemical or biological processes that are most important
in considerations of the rates and mechanisms of
contaminant transformations.
Many of these disequilibria are "intentionally"
maintained by the selective mediation of electron transfer
processes by microbial systems. Thus, sulfate/sulfide
ratios in systems containing active sulfate reducing
bacterial populations may represent Nernstian potentials
vastly different from those calculated from "abiotic"
couples present in the same system. There is little
evidence for any universal electron transfer mediator that
rapidly reacts to assure redox equilibria among the many
redox active species in a natural or contaminated ground-
water system.
Perhaps the closest approach to this notion of a
"universal" electron transfer mediator is the complex
mixture of substances commonly called natural organic
matter (NOM). NOM in both the dissolved and particulate
phases of ground-water systems is capable of mediating
25
-------
a wide range of electron transfer processes. NOM in this
context refers to the suite of extracellular organic
chemicals present in any natural water system. The
components of NOM' can vary from molecules recently
excreted by active microorganisms to facilitate electron
transfer process of utility to the microbial community to
'"abiotic" degradation products from the debris of long
inactive living systems. The ability to mediate electron
transfer processes seems to be ubiquitous among NOM
samples (Macalady and Ranville, 1998; Tratnyek and
Macalady, 2000).
NOM has been shown to be an effective mediator
of electron transfer in a wide variety of redox processes.
Included are the reduction of nitroaromatic compounds,
azo compounds, halogenated organic chemicals, metals
and metal/organic complexes (see for example. Macalady
and Ranville, 1998; Larson and Weber, 1994; Curtis and
Remhard, 1994; Wolfe etal., 1986; Wittbrodt and Palmer,
1996; Skogerbee, 1981; Matthiessen, 1996). NOM can
also serve as an electron transfer mediator in microbial
processes, e.g. the microbial reduction of solid iron (III)
oxyhydroxidcs (Lovley ct al., 1996). The role of NOM
in mediating the transfer of electrons in the opposite
direction has received less attention, but the oxidation of
sulfide and ferrous iron is clearly enhanced in the presence
of oxidized NOM. In fact, one of the few redox processes
that is apparently not mediated by NOM is the reduction
of molecular oxygen (Peiffer, unpublished data).
The mcchanism(s) by which NOM facilitates
electron transfer processes is only partially understood.
Clearly quinone-like functional groups within NOM
structures are an important part of this reactivity.
However, in certain pH ranges, other functional groups
may be important (Dunnivant et al., 1992; Perlinger et
al., 1996; Schwarzenbach et al., 1990; Gantzer and
Wackett 1991; Schindleretal., 1976).
Attempts to determine the role of NOM in specific
ground-water redox processes are related to both the
solution phase and particulate NOM fractions in the
aquifer matrix. The potential roles of NOM include direct
participation as an electron transfer mediator and an
indirect role as a transport inhibitor (particulate NOM)
or facilitator (dissolved and/or colloidal NOM). Specific
examples of such processes serve to illustrate the
importance of NOM in such considerations.
References
Curtis, G. P. and Reinhard, M., 1994. Reductive
dehalogenation of hexachlorethane, carbon
tetrachloride, and bromoform by anthrahydroquinone
disulfonate and humic acid. Environmental Science
and Technology, v. 28, pp. 2393-2401.
Dunnivant, F. M., Schwarzenbach, R. P. and
Macalady, D. L. 1992. Reduction of substituted
nitrobenzenes in aqueous solutions containing natural
organic matter. Environmental Science and
Technology, v. 26, pp. 2133-2141.
Ear\', L. E. and Schramkc, J. A. 1990. Rates of inorganic-
oxidation reactions involving dissolved oxygen. In
Chemical Modeling of Aqueous Systems, II, Chapter
30, eds. D. C. Melchior and R. L. Bassett, American
Chemical Society. Washington. B.C., pp. 379-396.
Gantzer, C.J. and Wackett, L.P., 1991. Reductive
dechlorination catalyzed by bacterial transition-metal
coenzymes. Environmental Science and Technology,
v. 25, pp. 715-722.
Larson, R. A. and Weber, E. J. 1994. Reaction
Mechanisms in Environmental Organic Chemistry.
Lewis Publishers, Chelsea, Michigan, pp. 169-273.
Lovley, D. R., Coates, J. D., Blunt Harris, E. L.,
Phillips, E. J. P. and Woodward, J. C. 1996. Humic
substances as electron acceptors for microbial
respiration. Nature, v. 382, pp. 445-448.
Lymaii, W. J., Bodek, L, Reehl, W. F., Rosenblatt, D. H.
1987. Electron transfer reactions. In Methods for
Estimating Phys'icochenrical Properties of Inorganic
Chemicals of Environmental Concern, Final Report,
Chapter 2, U.S. Army Medical Research and
Development Command, Contract DAMD
17-83-C-3274.
Macalady, D. L. and Ranville, J. F. 1998. The chemistry
and geochemistry of natural organic matter. In
Perspectives in Environmental Chemistry, Chapter
J, ed. D. L. Macalady. Oxford University Press.
New York, pp. 94-137.
Matthiessen. A. 1996. Kinetic aspects of the reduction of
mercury ions by humic substances. Fresenius
Journal of Analytical Chemistry, v. 354, pp. 747-749.
Perlinger, J. A., Angst, W, and Schwarzenbach, R. P. 1996.
Kinetics of the reduction of hexachloroethane by
juglone in solutions containing hydrogen sulfide.
Environmental Science and Technology, v. 30,
pp. 3408-3417.
Peiffer, S., Walton-Day, K., Macalady, D. L., 1999. The
interaction of natural organic matter with iron in a
wetland receiving acid mine drainage. Aquatic
Geochemistry, v. 5, pp. 207-223.
26
-------
Schindler, J. E., Williams, D. J., Zimmerman, A. P. 1976.
Investigation of extracellular electron transport by
humic acids. In Environmental'Biogeochenristry, Vol.
/., ed. J. O. Nriagu. Ann Arbor Science, Ann Arbor,
Michigan, pp. 109-115.
Schwarzenbach, R. P., Stierli, R., Lanz, K. and Zeyer, J.
1990. Quinone and iron porphyrin mediated
reduction of nitroaromatic compounds in
homogeneous aqueous solution. Environmental
Science and Technology, v. 24, pp. 1566-1574.
Skogerboe, R. K. 1981. Reduction of ionic species by
fulvic acid. Analytical Chemistry, v. 53, pp. 228-232.
Tratnyek, P. G. and Macalady, D. L. 2001. Oxidation-
reduction reactions in aquatic systems. In Estimation
of Chemical Properties for the Environmental and
Health Sciences: A Handbook of Methods,
eds. B. Boethling and D. Mackay, Ann Arbor Press.
Ann Arbor, Michigan.
Walton-Day, K., Macalady, D. L., Brooks, M. H., Tatc,
V. T. 1990. Field methods for measurement of ground
water redox chemical parameters. Ground Water
MonitoringReview, v. 10, pp. 81-89.
Walton-Day. K. 1991. Hydrology and Geochemistry of a
Natural Wetland Affected by Acid Mine Drainage,
St. Kevin Gulch, Lake County, Colorado, Ph. D.
Thesis, Colorado School of Mines. Golden, CO,
300 p.
White, A. R, Peterson, M. L,, Solbau, R. D. 1990.
Measurement and interpretation of low levels of
dissolved oxygen in ground water. Ground Water,
v. 28, pp. 584-589.
Wittbrodt, P. R. and Palmer, C. D. 1996. Effect of
temperature, ionic strength, background electrolytes,
and Fe(III) on the reduction of hexavalent chromium
by soil humic substances. Environmental Science and
Technology, v. 30, pp. 2470-2477.
Wolfe, N. L., Kitchens, B. E., Macalady, D. L., and
Grundl, T. J. 1986. Physical and chemical factors
that influence the anaerobic degradation of methyl
parathion in sediment systems. Environmental
Toxicology and Chemistry, v. 5, pp. 1019-1026.
27
-------
28
-------
Current of Practice for of Oxidation Reduction to the
of in
John T. Wilson
U.S. Environmental Protection Agency
Ada, OK 74820
The most important oxidation reduction process
involved in the destruction of chlorinated organic
compounds in ground water is biological sequential
reductive dechlorination (Vogel and McCarty, 1985). In
some cases this process does not provide benefit to the
organisms that carry it out, the process is entirely
accidental, and may be considered a form of co-
metabolism. In other cases the biological process yields
energy to the microorganisms and can support their
growth and proliferation (Maymo-Gatell et al., 1995).
In this circumstance the process functions as a respiration,
and has been termed halorespiration (See discussion in
diaper 6, Wiedemeieretal., 1999). Chlorinated organic
compounds may also be destroyed by chemical reaction
in aquifers, usually involving direct chemical reaction
with sulfide or ferrous iron. Examine Butler and Hayes
(1999), and Devlin and Mullcr (1999) for illustrations of
recent research.
In ''The Technical Protocol for Evaluating Natural
Attenuation of Chlorinated Solvents in Ground Water",
the U.S. Environmental Protection Agency used a scoring
system to identify sites where geochemical conditions
were appropriate or not appropriate for biological
reductive dechlorination of chlorinated solvents. Weights
were given to the concentrations of important electron
acceptors, including oxygen, nitrate, and sulfate, to the
concentrations of important electron donors including
molecular hydrogen, volatile fatty acids, dissolved native
organic carbon (TOC), and petroleum derived
monoaromatic hydrocarbons (BTEX), and the
concentrations of metabolic end products including
methane, ferrous iron, sulfide, chloride and carbon
dioxide represented as increases in carbon dioxide or
alkalinity. Weights were also given to general descriptions
of the oxidation/reduction environment such as electrode
potential, pH, and temperature. Finally, weight was given
to the accumulation of metabolic daughter products
including dichloroethylene, dichloroethane, vinyl
chloride, ethene, or ethane. Weight was only given when
the daughter products were not originally present in the
material that was released to the environment. The
weighted scores were totaled, and compared to a table
that interpreted the scores as providing either inadequate
evidence, limited evidence, adequate evidence, or strong
evidence of anaerobic biodegradation of chlorinated
organic compounds.
The National Research Council (2000) in the report
Natural Attenuation for Groundwater Remediation noted
on pages 210 and 211 that
Unfortunately, this scoring system is being
widely adopted for uses that the authors never
intended. For example, many states are using it
to evaluate natural attenuation for all types of
chlorinated solvents. Tables of natural
attenuation scores are showing up in remedial
investigation reports at Super/und sites. Maps
and cross sections showing natural attenuation
scores are being included in final reports as a
key line of evidence. Some regulators are
accepting this inappropriate use of scoring.
* The method applies only to chlorinated
ethenes.
* The scores emphasize reducing
environments more than dehalogenation
reactions.
* A reduced geochemical environment does
not guarantee that natural attenuation will
occur, because geochemical environments
can be very reduced without reductive
dehalogenation ofchlorocarbons occurring
(for example, if dehalogenafing bacteria are
not present).
* The scoring system included items that are
of current research interest (for example,
hydrogen concentration), but that may have
limited practical impact on making
remediation decisions.
29
-------
/, Analytical Parameters and Weighting for Preliminary Screening for Anaerobic Biodegradation processes used in
the Technical Protocol for Evaluating Natural Allenualion of Chlorinated Solvents in Ground Water, U.S. EPA,
1998
Analysis
Oxygen*
Oxygen*
Nitrate*
Ferrous iron
Sulfate*
Sulfide*
Methane*
Oxidation Reduction
Potential* (ORP)
against Ag/AgCI
electrode
pH*
TOC
Temperature*
Carbon Dioxide
Alkalinity
Chloride*
Hydrogen
Hydrogen
Volatile Fatty Acids
BTEX*
Tetrachloroethene
Trichloroethene*
DCE*
VC*
1,1,1 -Trichloroethane*
DCA
Carbon Tetrachloride
Chloroethane*
Ethene/Ethane
Chloroform
Dichloromethane
Concentration in
Most Contaminated
Zone
<0.5 mg/L
>5 mg/L
<1 mg/L
>1 mg/L
<20 mg/L
>1 mg/L
<0.5 mg/L
>0.5 mg/L
<50 millivolts (mV)
<-100mV
5 pH >9
> 20 mg/L
>20°C
>2x background
>2x background
>2x background
>1 nM
<1 nM
> 0.1 mg/L
> 0.1 mg/L
>0.01mg/L
>0.1 mg/L
Interpretation
Tolerated, suppresses the reductive pathway at higher
concentrations
Not tolerated: however, VC may be oxidized aerobically
At higher concentrations may compete with reductive pathway
Reductive pathway possible; VC may be oxidized under Fe(lll)-
reducing conditions
At higher concentrations may compete with reductive pathway
Reductive pathway possible
VC oxidizes
Ultimate reductive daughter product, VC Accumulates
Reductive pathway possible
Reductive pathway likely
Optimal range for reductive pathway
Outside optimal range for reductive pathway
Carbon and energy source; drives dechlorination; can be natural or
anthropogenic
At T >20°C biochemical process is accelerated
Ultimate oxidative daughter product
Results from interaction between CO2 and aquifer minerals
Daughter product of organic chlorine
Reductive pathway possible, VC may accumulate
VC oxidized
Intermediates resulting from biodegradation of aromatic compounds;
carbon and energy source
Carbon and energy source; drives dechlorination
Material Released
Material released
Daughter product of PCE
Material released
Daughter product of TCE.
If cis is > 80% of total DCE it is likely a daughter product
1,1 -DCE can be chemical reaction product of TCA
Material released
Daughter product of DCE
Material released
Daughter product of TCA under reducing conditions
Material Released
Daughter product of DCA or VC under reducing conditions
Daughter product of VC/ethene
Material Released
Daughter Product of Carbon Tetrachloride
Material Released
Daughter Product of Chloroform
Value
3
-3
2
3
2
3
0
3
1
2
0
-2
2
1
1
1
2
3
0
2
2
0
0
2a/
0
2a/
0
2a/
0
2
0
2
2
3
0
2
0
2
' Required analysis, a/ Points awarded only if it can be shown that the compound is a daughter product (i.e., not a constituent of the source NAPL).
30
-------
* The system identifies interactions between
contaminants only for electron donors,
The National Research Council (2000)
recommended on page 15 of their report that national
consensus guidelines be created for evaluation of natural
attenuation, and that
* The national consensus guidelines and
Juture protocols should eliminate the use of
"scoring systems "for making decisions on
natural attenuation. The evaluation
methods outlined in Chapter 4 f of their
report/, using conceptual and footprints of
natural attenuation, should replace scoring
systems.
Nyer et al. (1998) cautioned against the use of the
scoring system as a primary method to substantiate natural
attenuation, because the system failed to recognize natural
attenuation at a site where subsequent sampling and
evaluation revealed that natural attenuation was believed
to be occurring. In other words, the scoring system
produced false negatives.
To improve the site-screening methodology for
evaluating sites where natural reductive dechlorination
may be selected as a remedial option, Stiber et al. (1999)
constructed a causative model for the reductive
dechlorination process, then solicited expert knowledge
from twenty-two experts on reductive dechlorination in
ground water. He solicited expert opinion on fourteen
types of evidence in "The Technical Protocol for
Evaluating Natural Attenuation of Chlorinated Solvents
in Ground Water" (EPA, 1998). The experts were asked
to provide their estimates of the probability that reductive
dechlorination was occurring at a site based on site
conditions for each type of evidence. The expert opinions
were scaled by comparing the difference in the logarithm
of the odds ratios where
ALOR = log [prior probability (true | evidence)/ posterior
probability (false evidence)] - log [posterior
probability (true | evidence)/ prior probability
(false evidence)]
The experts put the greatest weight and greatest trust
on data showing the accumulation of metabolic
2. Values for the change in the log of the odds ratio (DL OR)for the probability that anareobic degradation is occurring
by reductive dechlorinalion associaledwilh findings from screening data wilh respect to different types of evidence,
The more positive the change with positive findings, the greater the weight the experts put in that type of evidence.
The more negative the change with negative findings, the greater the weight the experts put in that type o/'evidence,
These data are extrapolated from the Average Model in Figure 4 of Stiber et al., 1999.
Type of Evidence
Terminal Electron Accepting Process
Hydrogen
Oxidation Reduction Potential
Dissolved Organic Carbon
BTEX
TPH
Oxygen
Temperature
pH
Dichloroethene
Vinyl Chloride
Ethene and Ethane
Methane
Chloride
Positive Findings
0.07
0.10
0.07
0.04
0.02
0.02
0.03
0.01
0.01
0.63
0.95
0.78
0.43
0.50
Negative Findings
-0.15
-0.15
-0.15
-0.10
-0.10
-0.05
-0.15
-0.01
-0.10
0.80
0.40
0.25
0.38
0.30
31
-------
transformation products, and methane and chloride (Table
2). They put much less weight on data concerning electron
acceptors, hydrogen, oxidation reduction potential,
dissolved oxygen. TPH, and BTEX. They put little weight
on data concerning pH and essentially no weight on data
concerning temperature. Stiber et al. (1999) noted that
'"Simply identifying that prerequisite conditions arc
adequate is not sufficient; experts want to see products
to have confidence that reductive dechlorination is
occurring." They also noted that measuring the
concentration of molecular hydrogen was a particularly
divisive issue. "Some experts believe that, if performed
correctly, hydrogen measurements provide the best means
to assess the reducing conditions. Others feel that there
are too many variables in microbial communities for
hydrogen to be a reliable measurement."
Measurement of hydrogen in ground water is a
particularly powerful technique to understand electron
flow in microbial communities (Lovley et al., 1994).
However, measurement of hydrogen requires careful
sampling as well as state-of-art laboratory analytical
measurements (Chapellc et al., 1997). At this point the
use of hydrogen to understand reductive dechlorination
has had mixed results. It is not clear if this mixed
performance is a result of inadequate field protocols for
collection and analysis of hydrogen, or is a result of real
variations in the concentrations of hydrogen from one
time to another in the same plume, or from one plume to
another.
Recent work has shown that reductive
dechlorination may proceed in ground water at relatively
low concentrations. Smatlak et al. (1996) observed that
the physiological threshold for the concentration of
hydrogen that allowed growth of dechlorinating
microorganisms in their laboratory cultures was less than
2 nM. Yang and MacCarty (1998) compared the presence
or absence of cis-dichlorocthylcnc on the equilibrium
concentration of hydrogen in batch microcosms fed with
benzoate as a fermentable source of hydrogen. In the
presence of cis-dichloroethylene, reductive dechlorination
to ethene dominated electron flow7, and the concentration
of hydrogen was poised at 2.2 +/- 0.9 nM'. In the cultures
that did not have cis-dichloroethylene, the cultures were
methanogenic and the equilibrium concentration of
hydrogen was 10.9+/-3.3 nM. The dechlorinating
microorganisms out-competed methanogenic micro-
organisms, and poised the concentration of hydrogen gas
in the ground water.
Jakobsen et al. (1998) showed from theoretical
calculations and field data from Denmark, that the
equilibrium concentration of hydrogen in cold ground
waters (near 10° C) should be approximately half of the
equilibrium concentration of hydrogen expected in
warmer ground waters and laboratory studies (near 20° C
to 25° C). A concentration hydrogen of at least 1.0 nM
would be required to sustain dechlorinating
microorganisms in colder ground waters.
One of the most problematic issues in evaluation of
natural attenuation is the tendency of reductive
dechlorination to '"stall"' in many plumes at the level of
dichloroethylene, presumably because they have depleted
their sources of reducing power. In these plumes the
dichloroethylene persists without further metabolism to
vinyl chloride, ethene, or ethane. Concentrations of
hydrogen significantly below 1.0 nM may indicate that
further reductive dechlorination should not be expected.
In other plumes dichloroethylene and vinyl chloride
do not accumulate because these compounds are directly-
oxidized by aerobic microorganisms (Davis and
Carpenter, 1990; Bradley and Chapelle, 2000; Hartmans
and dc Bont, 1992; Klier et al., 1998), or iron reducing
bacteria ( Bradley and Chapelle, 1996, 1997; Bradley et
al., 1997) or manganese reducing bacteria (Bradley et
al., 1998), or by bacteria using native humic material as
the electron acceptor (Bradley et al., 1998). A
concentration of hydrogen above 1.0 nM' may indicate
that reductive dechlorination is on-going in these plumes.
and the reduction transformation products are not
accumulating because they arc destroyed by the oxidativc
processes described above. Analysis for hydrogen could
be used to distinguish between active plumes and inactive
plumes that have low concentrations of vinyl chloride,
ethene, and ethane because reductive dechlorination is
limited and these compounds were not produced in the
first place.
for by the Work
Groups
The use of geochemical parameters to determine
the "footprint" of a plume can be useful in determining
whether an apparent attenuation in the field is real, or
whether it is an artifact of an inadequate network of
monitoring wells. The primary' use of information on
the distribution of oxidation/reduction parameters should
shift. Instead of using the oxidation/reduction parameters
to identify plumes or regions of plumes where conditions
32
-------
for natural attenuation are favorable, the parameters
should be used for establishing "footprints" of plumes.
The most direct limitation on microbial activity is
the supply of the requisite substrates for metabolism.
However, the parameters that address the requisites for
reductive dcchlorination (such as hydrogen. BTEX,
volatile fatty acids, and dissolved native organic matter)
have little weight in the evaluation of natural attenuation
by subject matter experts. Data use and data quality-
objectives should be developed for the geochemical
parameters that measure the requisites for metabolism
of contaminants in aquifers. Field protocols should be
developed for sampling, preservation, and analysis of
these parameters.
The most convincing evidence of natural attenuation
is direct evidence of transformation of the contaminants.
The "Technical Protocol for Evaluating Natural
Attenuation of Chlorinated Solvents in Ground Water"
(U.S. EPA, 1998) relied on the accumulation of
transformation products as evidence of transformation.
Biodcgradation of organic compounds may also result in
isotopic fractionation of the contaminants remaining.
Natural attenuation due to biodegradation could be
distinguished from simple dilution or dispersion because
the isotopic ratio of the contaminant remaining in the
plume would become progressively heavier as
biodegradation proceeded. Data on stable isotope ratios
could be used as a second line of evidence to support an
evaluation of natural attenuation (sec Hunkclcr ct al., 1999
for a recent application of this approach to a PCE spill in
Canada). Data use and data quality objectives should be
developed for the use of stable isotope ratios to document
contaminant biodegradation in ground water.
References
Bradley, P. M. and Chapcllc, F. H. 1996. Anaerobic
mineralization of vinyl chloride in Fe(III)-reducing
aquifer sediments. Environmental Science and
Technology, v. 30, pp. 2084-2092.
Bradley, P. M. and Chapelle, F. H. 1997. Kinetics of DCE
and VC mineralization under methanogenic and
Fe(III)-reducing conditions. Environmental Science
and Technology, v. 31, pp. 2692-2696.
Bradley, P. M. and Chapelle, F. H. 2000. Aerobic microbial
mineralization of dichloroethene as sole carbon
substrate. Envimnmental Science and Technology,
v. 34, pp. 221-223.
Bradley, P. M., Chapelle, F. H, and Lovley, D. R. 1998.
Humic acids as electron acceptors for anaerobic
microbial oxidation of vinyl chloride and
dichlorethene. Applied and Environmental
Microbiology, v. 64, pp. 3102-3105.
Bradley, P. M., Chapelle, F. H., and Wilson, J. T. 1997.
Field and laboratory evidence of intrinsic
biodegradation of vinyl chloride contamination in a
Fe(III)-reducing aquifer. Journal of Contaminant
Hydrology, v. 31, pp. 111-127.
Bradley, P. M., Landmcyer, I.E., and Dinicola, R. S.
1998. Anaerobic oxidation of [1,2-
14C]dichloroethene under Mn(IV)-reducing
conditions. Applied and Environmental Microbiology,
v. 64, pp. 1560-1562.
Butler, E.C. and Hayes, K. F. 1999. Kinetics of the
transformation of trichloroethylene and
tetrachloroethylene by iron sulfide. Environmental
Science and Technology, v. 33, pp. 2021-2027.
Butler, E.C. and Hayes, K. F. 2000. Kinetics of the
transformation of halogcnatcd aliphatic compounds
by iron sulfide. Environmental Science and
Technology, v. 33, pp. 422-425.
Chapelle, F. H., Vroblesky, D. A., Woodward, J. C., and
Lovely. D. R. 1997. Practical considerations for
measuring hydrogen concentrations in groundwater.
Environmental Science and Technology, v. 31.
pp.2873-2877.
Davis, J. W. and Carpenter, C. L. 1990. Aerobic
biodegradation of vinyl chloride in ground water
samples. Applied and Envimnmental Microbiology,
v. 56, pp. 3878-3880.
Devlin, J. F., and Muller, D. 1999. Field and laboratory
studies of carbon tetrachloride transformation in a
sandy aquifer under sulfate reducing conditions.
Environmental Science and Technology, v. 3 3,
pp. 1021-1027.
Hartmans, S., and de Bont, J. A. M. 1992. Aerobic vinyl
chloride metabolism in mycobacterium aurum Li.
Applied and Environmental Microbiology, v. 58.
pp. 1220-1226.
Jakobscn, R., Albrcchtscn, H-J., Rasmussen, M., Bay, H.,
Berg, P. L,, and Christensen, T. H. 1998. H,
concentrations in a landfill leachate plume
(Grindsted, Denmark): In situ energetics of terminal
electron acceptor processes. Environmental Science
and Technology, v. 32. pp. 2142-2148.
Klier, N. J., West, R. J., and Donberg, P. A. 1998. Aerobic
biodegradation of dichloroethylenes in surface and
subsurface soils. Chemosphere, v. 38, pp. 1175-1188.
33
-------
Lovley, D. R., Chapelle, F. H., and Woodward, J. C. 1994.
Use of dissolved H2 concentrations to determine
distribution of microbially catalyzed redox reactions
in anoxic groundwater. Environmental Science and
Technology, v. 28, pp.1205-1210.
Maymo-Gatell, X., Tandoi, V., Gossett, J. M., and
Zinder, S. H. 1995. Characterization of an H,-utilizing
enrichment culture that rcductivcly dcchlorinatcs
tetrachloroethylene to vinyl chloride and ethene in
the absence of methanogenesis and acetogenesis.
Applied and Environmental Microbiology, v. 61,
pp. 3928-3933.
National Research Council. 2000. Natural Attenuation
for Groundwater Remediation, 219 p.
Nyer, E., May-field, P., and Hughes, J. 1998. Beyond the
AFCEE protocol for natural attenuation. Ground
Water Monitoring Review, Summer, pp. 70-77'.
Smatlak, C. R., Gossett, J. M., and Zinder, S. H. 1996.
Comparative kinetics of hydrogen utilization for
reductive dechlorination of tetrachloroethene and
methanogenesis in an anaerobic enrichment culture.
Environmental Science and Technology, v. 30,
pp. 2850-2858.
Stiber, N. A., Pantazidou, M., and Small, M. J. 1999.
Expert system methodology for evaluating reductive
dechlorination at TCE sites. Environmental Science
and Technology, v. 33, pp. 3012-3020.
U.S. Environmental Protection Agency. 1998. The
Technical Protocol for Evaluating Natural
Attenuation of Chlorinated Solvents in Ground Water,
EPA/600/R-98/128, U.S. Government Printing
Office, Washington, B.C.
Vogel,T. M., andMcCarty, P. L. 1985. Biotransformation
of tetrachloroethylene to trichloroethylene,
dichloroethylene, vinyl chloride, and carbon dioxide
under methanogenic conditions. Applied and
Environmental Microbiology, v. 49, pp. 1080-1083.
Wiedemeier, T. H., Rafai, H. S., Newell, C. J., and
Wilson, J. T. 1999. Natural Attenuation of Fuels and
Chlorinated Solvents in the Subsurface,
John Wiley & Sons, New York, 617 p.
Yang, Y and McCarty, P. L. 1998. Competition for
hydrogen within a chlorinated solvent dehalogenating
anaerobic mixed culture. Environmental Science and
Technology, v. 32, pp. 3591-3587.
34
-------
in or I to
Eh
J. F. Barker
University of Waterloo
Waterloo, ON, Canada N2L3G1
The for In
Hydrocarbons
The study of transport and fate of petroleum
hydrocarbons in ground water is "mature" and has
evolved into a remedial technology - monitored natural
attenuation (MNA). For example, the US EPA OSWER
Directive 9200.4-17P (April, 1999) states: "Natural
attenuation processes, particularly degradation, arc
currently best documented at petroleum fuel spill sites.
Under appropriate field conditions, the regulated
compounds benzene, toluene, ethylbenzene, and xylene
(BTEX) may naturally degrade through microbial activity
and ultimately produce non-toxic end products (e.g.,
carbon dioxide and water).'" The natural BTEX
degradation is "through microbial activity", so microbial
activity is the key process. Microbial activity must be
considered in the context of the field conditions.
Classification of subsurface microbial activity is
commonly via the dominant terminal electron accepting
process (TEAP). Thus we can identify sulfate reduction
as being the dominant TEAP and can discuss BTEX
biodegradation by a community in which sulfate is the
dominant electron acceptor (EA). This is the link between
petroleum hydrocarbon MNA and rcdox processes.
Geochemically, a TEAP is considered a redox process.
The OSWER. Directive requires monitoring
programs that, among other things, demonstrate that
natural attenuation is occurring and that deleterious
"changes in environmental conditions" are not. Collection
of site-specific data is required "to estimate with an
acceptable level of confidence both the rate of attenuation
processes and the anticipated time required to achieve
remediation objectives." The current distribution of
contaminants must be adequately defined and it must be
sufficiently clear, from field monitoring data, that
potential receptors are protected. A three-tiered approach
is recommended, each providing successively more
detailed information. These "three lines of evidence" are:
1. historical data that demonstrate loss of contaminant
mass and/or concentrations over time,
2. hydrogcologic and gcochcmical data to demonstrate
indirectly the type(s) of NA processes active at the
site and their rate(s),
3. data from field or microcosm studies which directly
demonstrate the occurrence of a particular N A process
at the site.
For selecting MNA for petroleum hydrocarbons,
demonstrating an understanding of how the current
contamination was produced and how it will likely evolve
is useful, if not essential. While contaminant distribution
is the primary' evidence, the need for a credible story
should not be dismissed. It both relies upon and
demonstrates the scientific underpinning for MNA.
Not being microbial ecologists, hydrogeologists and
geochemists have had to simplify the evaluation of
microbial activity. Since microbial ecologists seemed to
be willing to discuss biodegradation in terms of TEAPs,
that compromise has been accepted and subsurface
environments and geochemical processes have been
grouped in terms of TEAPs. Under current guidance (e.g.,
Wiedemeier et al, 1999), we are directed to examine the
apparent utilization of EAs and to define the distribution
of dominant TEAPs in plumes. That has led us to redox
processes.
Thus, rcdox processes and measurements arc not
of primary importance in MNA for petroleum
hydrocarbons. They are only important as they affect the
supporting "story".
for MNA for
Hydrocarbons
Site
The critical need at research and "real" sites is
insight into what has occurred and what is occurring in
the subsurface. A noted contaminant hydrogcologist
characterised MNA, especially for chlorinated solvents
as "1990's microbiology; 1970's hydrogeology".
35
-------
Chapelle (1999) states: "'The most important lesson that
can be gleaned from the history of petroleum hydrocarbon
bioremediation is the critical importance of accurately
assessing the hydrogeology of each site." It is clear to
hydrogeologists that the critical weakness in the
application of MNA lies within their discipline.
't
If MNA is really mature, shouldn't we expect
quantitative balances between EAs consumed and
petroleum hydrocarbons biodegraded, at least at '"research
sites"? A controlled, in place source experiment at CFB
Borden has been recently reported (King and Barker,
1999). About 70 kg of coal tar creosote (CTC) was
emplaced below the water table and the plumes of
dissolved chemicals that developed were monitored with
six "snapshot" samplings over four years, each using 450
to 2800 samples from multilevel monitoring wells. The
analysis of background/plume redox-sensitive species
indicated that the major TEAPs were aerobic respiration
and sulfate reduction. Mass balance calculations
suggested the apparent consumption of EAs would
support biodegradation of about 300 g of CTC organics
in the plumes. However, more than 1200 g of CTC
organics was apparently biodegraded in the plume.
Usually, it is expected that more EAs will appear to
be utilized than can be balanced by organic contaminant
loss. This is because of the presence of other, non-target
organics that exert a demand for EAs. In the CTC case,
the likely problem is the incomplete transformation of
organics. The above mass balance assumed complete
mineralization. Significant concentrations (up to 18mg/L)
of aromatic acids, the probable metabolites, were found
in a very superficial sampling of nine points. In this, the
biodegradation reactions would need to be re-written
based on actual transformations; not a quantitative process
at present.
While depletion of electron acceptors and donors
and the appearance of reaction products can be
qualitatively supported at most sites, acceptable balance
between EAs and organics consumed is unlikely to be
demonstrated, especially when ''real world" monitoring
(5 - 40 points, typically) is the basis. That is not a very
secure scientific underpinning for MNA. Some credible
quantitative mass balances need to be produced.
0fEA
Delineation of redox environments or TEAPs
appears useful/essential in developing the story
supporting MNA of petroleum hydrocarbons.
Albrechtsen et al. (1999) captured the problem very well:
"Delineations of redox environments have been based
mainly on the concentrations of redox sensitive
components in ground water. However, since most of the
reactive components may be transported by ground water.
for example, methane and Fc(II), their presence in a
sample does not necessarily reflect redox conditions at
the sampling location but perhaps conditions at an
upgradient location."
ofH2
Hydrogen appears to this desperate researcher as
potentially the "ideal" indicator of the TEAP active at
the monitored location. While H2 can be transported, it is
sensitive to local microbial activity and so its
concentration is adjusted to reflect local conditions
(Lovley et al., 1994). As such, it overcomes the transport
problem of the other geochemical indicators. I sense a
healthy reluctance to accept this hypothesis and look
forward to further demonstration such as reported by
Albrechtsen ct al. (1999). The key is to compare H2
concentrations directly to microbially-significant aspects
of TEAP and hydrocarbon degradation, not to flawed
(transported, reactive) aqueous geochemical data.
MNA for petroleum hydrocarbons is underpinned
by a demonstrated understanding of the controlling
processes. The key is the transport and fate of
contaminants, but EAs and TEAPs (perhaps defined as
redox processes) are involved.
One aspect is to determine the distribution, in space
and perhaps in time, of TEAPs. The researchers from the
Groundwater Research Centre at the Technical University
of Denmark (TUD) have provided the most
comprehensive characterization of ground-water redox
environments, not in BTEX plumes, but in landfill
leachate plumes. They have used essentially all the
techniques, many of which are referenced by Albrechtsen
etal. (1999).
To demonstrate understanding of interactions of
inorganics/organics/mincrals/microbial communities
insightful geochemical interpretation must be employed.
Since transport is significant, I think we are admitting
that the processes might be more complex than can be
handled by "typical" site monitoring, supported by simple
spreadsheet models. We need at least a few detailed field
studies in which all important parameters are measured
and interpretation is quantitative. If the problem is as
36
-------
complex as I have suggested, then qualitative
demonstration of process understanding will require
sophisticated flow-transport-reaction models.
Using reactive transport modeling, certain key redox
features can be reproduced, but it is clear that matching
all features, especially pH in poorly buffered systems,
strains our conceptual understanding. At actual field sites
major hydrogeological uncertainties, unknown source
functions, and temporal variability will be added. If we
are to demonstrate a comprehensive understanding of the
redox processes involved in biodegradation of petroleum
hydrocarbons, we must do so through matching results
from such models with detailed field observations from
at least a few sites.
Do we need this for MNA of petroleum
hydrocarbons? We needed it five years ago to provide
the underpinning for current guidance. Do we still need
it? MNA is being selected, errors will be made, and the
blame shouldn't reside with bad scientific underpinning.
So, I think this research is still essential.
Is it being done? The modeling tools appear to be
sufficiently developed (e.g., BIORXNTRN, Hunter et al,
1998; Mayer. 1.999) and some models are being employed
in this area (e.g., Curtis et al., http://wwwmn.cr.usgs.gov/
bemidji/). Certainly, TUD researchers have the field
studies and experience for landfill leachate plumes. The
USGS is apparently well along at the Bemidji site
(http://wwwmn.cr.usgs.gov/bcmidji) and other
hydrocarbon-contaminated sites (e.g., Wurtsmith) have
major components of this in place. All of these sites
(perhaps less so at Bemidji) suffer from uncertainties
about the source and/or complex hydrogeology. Too bad
the controlled field experiments in simple
hydrogeological settings (e.g., Barker et al., 1987; King
and Barker. 1999) were not more comprehensive. With
simple hydrogeology, simple and known sources, mass
balances, etc. would have been a reasonable expectation
and the concept would have been quantitatively evaluated.
The of Redox In Practical MNA
for
Is proper redox characterization required to
implement MNA for petroleum hydrocarbons?
No. The primary information is the distribution of
contaminants. Redox characterization only contributes
to the credibility of the story of how the plume got to the
current status.
Is proper redox characterization required for remedy
selection? Also, no. I might reject MNA for BTE-X plumes
in naturally anaerobic aquifers, since I don't feel benzene
degrades reliably under such conditions. However. I
would acknowledge others would not share this
pessimism and so my redox-based view would not be
compelling. We would, I am sure, need to rely on the
primary line of evidence - the distribution of BTE.X and
interpretation of apparent attenuation.
Is proper redox characterization required for
performance assessment? Again, no. While, for example,
assessment of deleterious '"changes in environmental
conditions" might be tracked through redox
measurements, direct contaminant measurements will
likely be the assessment criteria. Perhaps redox
contributes to the credibility of the assessment. For
example, if benzene is found beyond the predicted point.
perhaps noting the unexpected extent of reduced
conditions would be useful.
Redox processes are part of the supporting story
for MNA of petroleum hydrocarbons. As such, it is
probably more important that research-level studies
demonstrate understanding of redox-related processes.
Our goal should be to demonstrate this. We should then
assure hydrogcologists that they must rely primarily on
hydrogeological studies, not assessments of redox
processes, in applying MNA to BTEX-contaminated
groundwater.
References
Albrechtsen, H. J., Bjerg, P. L., Ludvigsen, L., Rugge. K.
and Christensen, T. H.. 1999. An anaerobic field
injection experiment in a landfill leachate plume,
Grindsted, Denmark 2. Deduction of anaerobic
(methanogenic, sulfate-, and Fe(III)-reducing) redox
conditions. Water Resources Research, v. 35,
pp.1247-1256.
Barker, J. F., Patrick, G. C., and Major, D. 1987. Natural
attenuation of aromatic hydrocarbons in a shallow
sand aquifer. Ground Water Monitoring Review^ v. 7,
pp. 64-71.
Chapelle, F. H. 1999. Bioremediation of petroleum
hydrocarbon-contaminated ground water: The
perspectives of history and hydrology. Ground Water,
v. 37, pp. 122-132.
Curtis, G. P., Cozzarelli, I. M., Baedecker, M. J., and
Bekins, B. A. Coupled biogeochemical modeling of
ground water contamination at the Bemidji Minnesota
crude oil spill site, from http://wwwmn.cr.usgs.gov/
bemidji/.
37
-------
Hunter, K. S., Wang, Y, and Van Cappeiien, P. 1998.
Kinetic modeling of microbially-driven chemistry of
subsurface environments: coupling transport,
microbial metabolism and geochemistry. Journal of
Hydrology, v, 209, pp. 53-80.
King, M. W. G., Barker, J. E, Devlin, J. E, and B. I Butler.
1999. Migration and natural fate of a coal tar creosote
plume; 2. Mass balance and biodcgradation
indicators. Jotimal of Contaminant Hydrology, v. 39,
pp. 281-307.
Lovley, D. R., Chapelle, E H., and Woodward, J. C. 1994.
Use of dissolved H2 concentrations to determine
distribution of microbially catalyzed redox reactions
in anoxic groundwater. Environmental Science and
Technology, v. 28, pp. 1205-1210.
Mayer, K. U. 1999. A numerical model for
multicomponent reactive transport in variably
saturated porous media, Ph.D. thesis, University of
Waterloo, Waterloo, Ontario, Canada.
Wiedemeier, T. H., Rafai, H. S., Newell, C. I, and Wilson,
J. T. 1999. Natural Attenuation of Fuels and
Chlorinated Solvents in the Subsurface, John Wiley
& Sons, New York, 617 p.
38
-------
of by
Janet S. Herman1
Aaron L. Mills1
Isabelle M. Cozzarelli2
'University of Virginia
Charlottesville, VA 22904
2U. S. Geological Survey
Reston,VA 20192
Two different investigations of shallow sandy
aquifers inform our thinking about the role of reactive
iron minerals in hydrogeological systems. Ground water
in a number of settings has been described as having
elevated concentrations of dissolved iron in anoxic
portions of contaminant plumes (e.g. Baedecker et al.,
1993; Lyngkilde and Christensen, 1992), and the coupling
of microbial reduction of Fe(III) to the degradation of
organic compounds links the study of the biogcochcmistry
of iron to our desire to understand processes influencing
contaminated aquifers. Questions about the transport and
fate of reactive constituents in ground water are linked
to the study of iron through recognition of the importance
of sorption processes on sesquioxide mineral surfaces
common in sedimentary aquifers limiting the migration
of metals, anions, and bacteria. We have been grappling
with questions related to the stability of iron mineral
phases in contaminated aquifers and the role of Fe(III)-
bearing minerals and grain coatings as sorption substrates.
In our ongoing study of the transport of bacteria
through an unconsolidated sandy aquifer in the Coastal
Plain of Virginia, we address a number of questions
related to the potential for bacterial attachment to mineral
solids. In laboratory and field-scale experiments, the
presence of these positively charged metal-oxide coatings
on quartz sand has been shown to greatly affect the
sorption and transport of reactive constituents. The
mineralogy of the metal oxide affects the affinity its
surface has for ion sorption. The presence of more than
one oxide in a system may affect the sorption
characteristics as well. In addition to metal-oxide
abundance, surface area is commonly employed as an
estimator of total sediment reactivity, and the result can
be inconsistent with an assessment of that portion of the
surface that actually participates in reactions.
Our investigation was undertaken on the Eastern
Shore near the village of Oyster where depositionally
similar sediments have been exposed to chemically
distinct ground-water conditions. In a narrow, organic-
rich leachate plume derived from vegetable waste pits.
dissolved oxygen was only 0 to 0.9 mg L"1, whereas it
was 5.0 to 11 mg L"1 in the regionally extensive aerobic,
uncontaminated ground water. In addition, the amount
of dissolved iron in the aerobic ground water was 0.001
to 0.01 mg/L"1, but it was 12 to 42 mg/L"1 in the anaerobic
zone. Samples of sediments from cores were subjected
to metal-oxide extractions and determination of
operationally defined fractions of iron (ferrous and ferric)
and aluminum. The amount of extractable iron was an
order of magnitude higher for the aerobic sediments than
for the anaerobic sediments indicating that reductive
dissolution removed the oxide coatings. The total iron
in the HC1 extracts of the aerobic sediments ranged from
3.2 - 52 umol/g whereas it was only 0.9 - 2.4 umol/g in
the anaerobic sediments (Table 1). There was very little
ferrous iron (Fc(II)) extracted from cither of the two
zones, and the values from each zone only ranged from
0.01 - 0.5 umol/g of Fe(II) extracted. It is likely that
microbial oxidation of organic contaminants is linked to
the reduction of iron from the surfaces of sediments in
the polluted portion of the aquifer. Although the majority
of the iron had been removed from the surfaces of the
anaerobic sediments, there was still an appreciable
amount of surface area measured.
We sought a means to characterize the aquifer solids
for their potential for bacterial attachment, and we
developed a single-point sorption determination for a
reactive anion. In attempting to evaluate the reactivity
of mineral surfaces in an aquifer setting, sorption of
solutes directly probes that portion of the solid surface
that participates in reactions involving other negatively-
charged reactive constituents, including bacteria. The
reactivity of the sediment surfaces, as indicated by the
sorption of 35SO 2~, was an order of magnitude higher in
the aerobic vs. anaerobic sediments. The sulfate sorbed
to the aerobic sediments (2.2 to 50 umol/g) was an order
of magnitude greater than for the anaerobic sediments
(0.007 to 0.25 umol/g). The presence of anaerobic
39
-------
TDlble 1. Concentration of total extractable iron with depth
(below the water table) for samples from aerobic
and anaerobic cores.
Aerobic
Depth (m)
-1.36
-1.44
-1.53
-1.61
-1.69
-1.77
-1.88
-2.32
-2.80
Core
Total Fe
(|imol/g)
2.10
1.54
1.88
2.16
2.12
2.32
2.45
2.40
1.27
Anaerobic Core
Depth Total Fe
(m) ((imol/g)
-0.31
-0.81
-1.31
-1.81
3.35
18.4
5.49
10.5
-4.26
1.97
1.34
-2.31
-2.81
-3.31
-3.81
-4.31
52.4
10.5
4.26
3.16
10.9
conditions did not significantly alter the amount of
extractable aluminum oxides on the surface of the
sediments, and those coatings helped to maintain a high
surface area for the anaerobic sediments. It appears that
the removal of the iron oxides from the surfaces under
anaerobic conditions was solely responsible for the
significant reduction of sediment reactivity observed.
The iron-oxide content was the only one of the
surface properties to be significantly altered by the
anaerobic conditions with a concomitant alteration in the
ability of the sediments to retain the anion sulfate (Knapp
et al., in review). It appeared, therefore, the presence or
absence of iron oxides was the dominant factor controlling
sulfate sorption. These results would indicate an
alteration in the reactive surfaces of minerals when
exposed to anaerobic ground water and the potential for
greater movement of reactive constituents, including
bacteria, through these aquifer sediments.
As redox conditions in a contaminant plume change
with the depletion of oxygen and nitrate as electron
acceptors, ferric iron is consumed as an electron acceptor
coupled to the oxidation of organic matter (e.g. Lovley
et al., 1989). The glacial-outwash, sandy aquifer at the
USGS Toxic Substances Hydrology site near Bemidji,
MN. became contaminated by crude oil in 1979.
Observations since nearly the time of the pipeline burst
show accumulations of Fe(II) in the site ground water
over time, whereas no ferrous iron was found in the
uncontaminated background water (e.g. Baedecker et al.,
1993). Previous field and microcosm results (Baedecker
et al., 1993, Lovley et al., 1989) indicate the degradation
of toluene may proceed according to
CH +36Fe(OH)^+65H+ ->7HCO~+36Fe!++87H 0
7 8 V '3(s> 3 2
In this study, we examined the distribution of Fe in
the sediments of the aquifer in order to assess the
magnitude of impact of Fe(III) reduction in the
contaminant plume on the geochemistry of the sediments.
Sampling and extraction of sediments revealed that the
average HC1 -extractable Fc(III) concentration in the most
contaminated portion of the aquifer was 16.2 umol/g, a
30% reduction from the value in background sediments
of 23.8 umol/g (Tuccillo et al., 1999). Comparison
between HC1 extractions that should capture poorly
crystalline Fe(III) solids and Ti-citrate-EDTA-bicarbonate
extractions mat should reductively dissolve amorphous
and crystalline Fe(III) oxides indicates that the bulk of
the microbially mediated iron reduction is dissolving
amorphous or poorly crystalline oxides.
In contrast to the study in Virginia in which virtually
no Fe(II) was extracted from aquifer solids, at Bemidji
we found 19.2 umol/g Fe(II), as much as 4 times the
background sediments with 4.6 umol/g Fe(II) (Figure 1).
Scanning electron microscopy detected authigenic ferroan
calcitc in the anoxic sediments (Baedecker et al., 1992;
Tuccillo etal., 1999). Likely the contrast can be attributed
to the availability of minerals that buffer pH, giving the
contaminated ground water at Bemidji at pH near 6.9
whereas at the Oyster, VA, site it is closer to 6.3 in the
anaerobic ground water.
One of the most striking features of the distribution
of iron content of the aquifer sediments was the extreme
concentration at the anoxic/oxic transition zone. HC1-
extractable Fe(III) reached values as great as 49.8 umol/g
at this boundary (Figure 1). This 70% increase in total
extractable Fe at the interface indicates the reoxidation
and precipitation of Fe mobilized from aquifer sediments
40
-------
8.
o
a.
50 -
40 -
30 -
20 -
•in -
{
i
« * I*
i i i
HCI-Extractable 1
Fe(lll)
*
1 1 XX I
bO -
40 -
30 -
20 -
.
*
Ti-CEB-
Extractable Fe
* {
1 1 XX 1
O)
•3
^M^
nK
20 -
16 -
10 -
5 -
n -
XX
XX
2 T HCI-Extractable
* f Fe(ll)
I I i i
*
I XX I . .
o
1.6 -
1r\
,£, —
0.8 -
04 -
nn -
f
i
i
*.
±^ 1
XX
Pyrophosphate-
Extractable Fe
I «
| I XX T
800 -
600 -
400 -
200 -
0 -
" 2+
£ Aqueous Fe
I
I
, " , • , xx t
figure 1.
0 40 80 120 160 Bkgd.
Distance from Center of Oil Body (m)
Graphs of various fractions of iron extractedfrorn
the sediments and' oj"aqueous Fe2'*' in pore waters
vs. distance from the center of the oil body. Circles
represent the mean of samples from all depths for
each core. Error bars represent standard errors
(s.d./tf's) which gives an indication of within-core
variability.
located upgradient in the anoxic, contaminated
groundwater plume as that water flows downgradient into
an oxic background ground water. Scanning electron
microscopy confirmed the abundant Fe(III)
oxyhydroxides at the anoxic/oxic boundary (Tuccillo et
all 1999).
The alteration of sediment chemistry has
implications for the availability of electron acceptors and
the biodegradation of contaminant organics. As
availability of Fe(III) decreases, alternate hydrocarbon
degradation processes linked to methanogenesis will be
less efficient and intermediate metabolic products of
biodegradation will persist and be transported in the
ground water. The Fe(II) that is mobilized out of the
sediments in the most contaminated portion of the aquifer
will be reoxidized at some location downgradient and
precipitate as an 'iron curtain" that can alter not only the
reactivity of the solid phase but can also change the
permeability as mineral grains are cemented together and
the interstitial spaces filled with Fe(III) precipitates. The
high Fc(III) content of that anoxic/oxic boundary gives
rise to a sorptive capacity of the transition-zone sediments
that is enhanced above both background and anoxic
sediments. The presence of the Fe(III) oxyhydroxides
has implications for the reactivity of aquifer sediments,
altering the nature and extent of reaction between sorbing
species and mineral surfaces. The contaminants that may-
be retarded in their transport are not just those of the
original crude oil spill but potentially include bacteria,
anions, and metals. Improved understanding of the
biogeochemistry of iron in contaminated sedimentary
aquifers likely will lead to a better ability to anticipate
the rate of transport of a variety of solutes and colloids.
References
Baedecker, M. J., Cozzarelli, I. M., Evans, J. R. and
Hearn, P. P. 1992. Authigenic mineral formation in
aquifers rich in organic material. In Water-Rock
Interaction: Proceedings of the 7th International
Symposium on Water-Rock Interaction, eds. Y. K.,
Kharaka. and A. S. Maest, Balkema, pp. 257-261.
Baedecker, M'. J., Cozzarelli. I. M., Eganhouse, R. P.,
Siegel, D. I., and Bennett, P. C. 1993. Crude oil in a
shallow sand and gravel aquifer - III. Biogeochemical
reactions and mass balance modeling in anoxic
groundwater. Applied, Geochemistry, v. 8,
pp. 569-586.
Knapp, E. P., Herman, J. S., Mills, A. L., and
Hornberger, G. M. 2001. Redox alternation of
reactive mineral surfaces in biologically active
groundwater. Applied Geochemistry, in review7.
Lovley, D. R.. Baedecker, M'. J., Lonergan, D. J..
Cozzarelli, I. M., Phillips, E. J. P., and Siegel, D. I.
1989. Oxi dation of aromatic contain i nants coupled
to microbial iron reduction. Nature, v. 339,
pp. 297-299.
41
-------
Lyngkilde, J. and Christensen, T. H. 1992. Redox zones
of a landfill leachate pollution plume (Vejen,
Denmark). Journal ofContammant Hydrology, v. 10,
pp. 273-289.
Tuccillo, M. E., Cozzarelli, I. M., and Herman, J. S. 1999.
Iron reduction in the sediments of a hydrocarbon-
contaminated aquifer. Applied. Geochemistry.^. 14,
pp. 71-83.
42
-------
the of in
D. Kirk Nordstrom
U.S. Geological Survey
Boulder, CO 80303
Introduction
The earth's surface is in a steady straggle between
oxidation, driven by oxygen in the atmosphere, and
reduction, driven by organisms and organic matter. The
redox cycle established about 600 million years ago by
photosynthesizing and respiring organisms, organic
burial, and weathering processes maintains the
biogeochemical conditions we enjoy today. Without this
constant redox disequilibria in the air and in waters and
in rocks, life would not exist. Plants and petroleum would
spontaneously combust and the atmosphere would
become a sea of nitrogen oxides.
Confusion regarding redox (reduction-oxidation)
potentials and redox chemistry of aquatic systems stems
from misunderstandings of how aqueous redox reactions
occur and from misunderstandings of the assumptions
that enter into any calculations. Specific conceptual
difficulties arise in the following three areas: (1) the
existence and representation of the aqueous electron,
(2) the kinetics of electron transfer reactions, and (3) the
measurement and interpretation of redox in water bodies.
This paper discusses these various issues, shows examples
of equilibria and disequilibria of redox chemistry in
natural waters with particular attention to the ferrous-
ferric redox couple, and explains the challenges of
modeling redox geochemistry.
The
Free aqueous electrons have been synthesized in
solution at micromolar concentrations by intense
radiation, either by flash photolysis or pulse radiolysis
(Hart and Anbar, 1970). These hydrated electrons are
extremely ephemeral, having half-lifes of 230 (is at pH
of 7. They quickly form hydrogen atoms and hydrogen
gas. Their concentrations in natural waters would be
negligible to nonexistent (Hostetler, 1984; Thorstenson,
1984). Hence, discussions of the activity of aqueous
electrons, designated as ae_, do not relate to a physically
meaningful quantity. This fact distinguishes aqueous
electrons from aqueous protons because a + is a
physically meaningful quantity. The mathematical entity
"pH" is physically measurable and has chemical
significance whereas the entity '"pe," or - log a , does
not have physicochemical significance when applied to
aqueous electrons. There is no analogy between pH and
pe in a chemical sense. To draw the analogy in a
mathematical sense can be very misleading.
of
Reactions involving aqueous solutions are divided
into homogeneous (occurring only within the solution
phase) and heterogeneous (occurring between a solid or
gas surface and the solution phase). Homogeneous redox
reactions can go quickly if only one electron is being
transferred at a time and if the activation barrier is not
too large. However, numerous examples show that large
activation barriers do exist - otherwise nitrogen, oxygen,
methane, and carbon dioxide would not all coexist in the
atmosphere as they do. Eary and Schramke (1990) showed
clearly the wide range of oxidation rates for several
inorganic reactions as a function of pH and dissolved
oxygen. Redox reactions involving two electrons (more
than two per reaction step is generally forbidden) or
involving heterogeneous reactions are usually slower
unless catalysts or high temperatures are provided.
Empirical homogeneous redox reactions arc
described by the usual kinetic equations (zero-order,
first-order, second-order, Michaelis-Menten for
enzyme catalysis, etc.) but electrode kinetics are
described by the current-overpotential equation:
-(\-a)nfTtlRT
(1)
where the exchange current. /. is dependent on the
equilibrium exchange current, ig, the concentration
gradient for the oxidized species between the surface,
C (0,1), and the bulk solution, C * . the concentration
0] ' 7, ? 0 ,
gradient for the reduced species. Cr(0,t), and the bulk
solution, C*, the symmetry factor, a, the overpotential,
h, the Faraday constant, F, the ideal gas constant. /?, and
43
-------
the Kelvin temperature, /'(Bard and Faulkner, 1980). If
the surface concentrations do not differ significantly from
the bulk solution concentrations then equation (1) reduces
to the Butler-Volmer equation developed in the 1920's:
t-anFj]/RT
- e
-(l-a)nFt]/RT '
(2)
Equation (2) shows how the current across a metal-
solution interface depends on the difference in potential
between the actual non-equilibrium and equilibrium
potential differences. r\ = E - Eet. When this difference
(the overpotential) goes to zero, the equation reduces to
the well-known Nernst equation of equilibrium
thermodynamics written here for a half-cell with one
reduced and one oxidized species:
(3)
where E° is the standard electrode potential based on the
standard hydrogen electrode, the natural logarithm
contains the ratio of the activity of the oxidized species
to the activity of the reduced species, each taken to the
exponent represented by the stoichiometric coefficient, v.
The equation is also commonly written with a negative
sign before the log term with reduced species activity on
the top and the oxidized underneath. These equations
remind us that the Nenist equation is a special case of the
more general equation of electrode kinetics. The challenge
is knowing when a system is in equilibrium and when it
is not.
There are two systems being considered with respect
to equilibrium conditions. One system is the electrode-
solution interface and the other is the environmental
system composed of gases, waters, rocks, and biological
activity. This section addresses the electrode-solution
system and the next section addresses the environmental
system.
The ease with which electrons are transferred across
the metal-solution interface depends on the nature of the
metal, the nature of the aqueous redox-sensitive species,
and the solvent properties. Metal electrodes need to be
non-reactive and electrically conductive. Thus, gold,
platinum, and graphite have worked well and among these
platinum is most frequently used. The metal is not usually
a limiting factor in reaching equilibrium as long as the
surface is free of electroactive coatings and other adsorbed
impurities. For aqueous redox species, there are some
serious limitations: (1) the net exchange current across
the interface must be effectively zero, i.e. electrochemical
reversibility must be maintained, (2) to maintain
reversibility, redox ions must have individual exchange
currents greater than 10'7 amp cm"2 (Morris and Stumm,
1967), which translates into concentrations of about 10"5m
or greater for two redox constituents in natural waters:
Fe(II/III), S(II-), and possibly U(IV/VI). Several
examples in the literature have shown the equilibrium
maintained between the Pt electrode and the Fe(II/III)
redox couple in laboratory studies (Morris and Stumm,
1967; Macalady et al, 1990; Stipp, 1990). Peschanski
and Valensi (1949) showed that the .Pt electrode responded
quantitatively and reversibly to changes in sulfide ion
activity in the S(II-/0) system in the laboratory.
Experiments designed to examine the electrochemical
reversibility of the As(III/V) and Se(IV/VI) redox couples
showed that they do not reach equilibrium (Kcmpton et
al., 1990; Runnells et al., 1987; Runnells and Skoda,
1990). Hence, only iron and reduced sulfur species
demonstrate electrochemical reversibility for aqueous
conditions likely to be found in the environment.
of In
Waters
Two facets of redox chemistry need to be considered
when examining redox in aquatic systems: the attainment
of equilibrium for a redox couple at a Pt electrode surface
and the attainment of equilibrium between different redox
couples in ground water or surface water. Field
measurements and their interpretation are not simple and
proper precautions should be exercised (Langmuir, 1971;
Nordstrom and Wilde, 1998). Laboratory and theoretical
considerations would lead to the view that only rarely
would redox potential measurements in waters reflect
equilibrium and even then it should only be observed for
iron and reduced sulfur. Field work confirms this view.
For the iron system, an update of the original diagram by
Nordstrom et al. (1979) that compares measured vs.
calculated Eh for more than 60 acid mine waters is shown
below in Figure 1. The calculated Eh values were based
on ferrous and ferric iron determinations (To et al., 1999).
complete water analyses, and speciation based on output
from the WATEQ4F program (Ball and Nordstrom, 1990).
The agreement is generally excellent with most Eh
values agreeing to within 25 mV except at low Eh where
a strong positive deviation occurs. Figure 2 plots the
difference between calculated and measured Eh as a
function of total dissolved iron concentrations. Marked
deviations occur below 10"5 m as predicted from lab
studies. These results reflect the mixed potential effect
as the iron concentrations become too low to be
electroactive and oxygen begins to be sensed by the
44
-------
1.0
• Leviathan samples
o Iron Mountain samples
n Summitville samples
0 0.2 0.4 0.6
Eh calculated from Fe (ll/lll), volts
0.8
1.0
figure 1. Comparison of calculated with measured Eh.
0.6
0.4
0.2
-0.2
-6 -4
log total dissolved iron, molality
2, Difference in Eh as a Junction ofFe molarity,
electrode. However, the O,/H O redox equilibrium
potential is never reached because oxygen is not
sufficiently electroactive. Hence, field studies have been
able to demonstrate the same equilibrium response of a
platinum electrode to the Fe(II/III) redox couple as in
the lab. This study also shows the reliability of the
spcciation code. Similar success has been shown with
over 50 measurements of a .,, compared to Eh electrode
sulfsde *^
measurements by Berner (1963) for anoxic marine
sediments. Boulegue and Michard (1979) made similar
measurements in sulfide-rich springs using acid-base
titrations. The difficulty with these field measurements
is that the slightest trace of oxygen can produce
polysulfides, thiosulfate, and elemental sulfur which will
cause deviations in the Eh measurement. Likewise, for
anoxic ground waters high in iron, the introduction of a
well, pumps, and sampling lines causes oxygen to enter
the sampling system and instantaneously oxidizes any-
reduced ferrous iron if the pH is circumneutral. The result
is the precipitation of hydrated ferric oxides which will
tend to cause a Fe(II)/Fe(OH) redox couple to occur
regardless of whether or not it actually occurs in the
unaltered ground water. The only other example that
might be approaching equilibrium is the U(IV/VI) redox
equilibrium for 6 Finnish ground-water samples from the
work of Ahonen et al. (1994). No other redox couples in
natural aquatic systems are known to produce an
equilibrium potential at an electrode surface.
Furthermore, when two or more redox couples have been
measured on the same water sample, they do not give the
same Eh value (Nordstrom et al., 1979; Lindberg and
Runnells, 1984), and they should not be expected to give
the same value because they tend to react at different
rates with different mechanisms. If redox species all came
to equilibrium it would destroy life. Life depends on redox
gradients and redox disequilibria. Geochemical codes that
force redox equilibrium for all elements are grossly
oversimplified and cannot be expected to reflect realistic
redox conditions (unless temperatures > 100°C). There
are only three approaches that will permit successful
application of geochemical models to redox reactions:
(1) determine individual redox species analytically,
(2) apply independent speciation to reduced and oxidized
forms of the species, and (3) apply mass balances
(Parkhurst and Plummer, 1993). Changes in pH that
reflect pyritc oxidation and aqueous iron oxidation can
be simulated effectively by speciation and mass balances
if a judicious choice of equilibrium constants is made
(Nordstrom, 2000). The only way to determine the redox
conditions of a water is to analyze the sample for those
redox species of specific concern. There are several
examples of the same water containing two or more of
the following constituents O2, Fe(II), Fe(III), N2 NO,,
NH3, H2S, SO4, CH4, CO,, and H2. There can be no such
thing as redox equilibrium or a single redox potential in
these waters. The way to understand redox chemistry is
to learn the rates and mechanisms of reactions, the main
catalysts in the environment, and the main sources and
sinks of redox active species.
Conclusions
Redox studies on aquatic systems in the lab and in
natural waters lead to the following conclusions:
1. The absence of aqueous electrons in natural waters
means they cannot be measured nor defined by
analogy with aqueous protons.
2. Ground waters and surface waters do not have a
'"redox potential," pe, or Eh but they do have a pH.
To speak of a redox potential of an aqueous solution
or natural water does not have any meaning.
3. Redox species in waters do not reach an equilibrium
state readily, if at all. Redox disequilibria is the rale.
45
-------
4. Redox potential measurements with platinum (or
similar) electrodes are generally useless except to
estimate ferrous-ferric activity ratios or sulfide
activities when the concentrations are greater than
about 10~-s m. Otherwise the potential may drift
according to electrokinetic phenomena, mixed
potentials, or impurities at the metal electrode surface.
It is more reliable to measure such concentrations
directly, where possible, and calculate the activities
with a speciation model.
5. To determine the redox chemistry of a water it is
necessary to determine all the relevant redox species
directly. These species can be expected to react in
different ways and at different rates. Homogeneous
redox reactions can be rather slow and heterogeneous
redox reactions can be even slower.
6. Redox species can be very sensitive to atmospheric
contamination. Contamination and interferences
begin with drilling a well, continue with the manner
and rate of pumping, and end with a set of data that
is often inconsistent or discrepant.
7. A judicious choice of redox parameters depends on
site characteristics, budget, and objectives.
8. A monitoring schedule depends on site-specific
factors such as flow rates, types of redox species
present, rates of redox reactions, redox capacity of
the system, and rates and mechanisms of processes
in the vadose zone.
9. Research knowledge is in the process of finding out
how redox works and how fast. The knowledge is
not at a point where we can say definitively what the
protocol for redox is nor can we model any specific
site to predict just what will happen over time except
in a general, but not necessarily helpful, sense.
References
Ahoncn, L., Ervanne, H., Jaakola, T, and Blomqvist, R.
1994. Redox chemistry in uranium-rich groundwater
of Palmottu uranium deposit, Finland. Radiochimica
Acta, v. 20, pp. 1-7.
Ball, J. W. and Nordstrom, D. K. 1990. User's manual
forWATEQ4F, with revised thermodynamic database
and test cases for calculating speciation of major,
trace, and redox elements in natural waters
U.S. Geological Survey Open-File Report,
91-183, 189 p.
Bard, A. J. and Faulkner, L. R. 1980. Electrochemical
Methods. John Wiley and Sons, New York, 718 p.
Berner, R.A. 1963. Electrode studies of hydrogen sulfide
in marine sediments. Geochimica et Cosmochimica
Acta, v. 27, pp. 563-575.
Boulegue, J. and Michard. G. 1979. Sulfur speciation and
redox processes in reducing environments. In
Chemical Modeling in Aqueous Systems, Chapter 2,
ed. E. A. Jenne, American Chemical Society
Symposium Series 93, American Chemical Society,
Washington, D.C., pp. 25-50.
Eary, L. E. and Schramke, J. A. 1990. Rates of inorganic
oxidation reactions involving dissolved oxygen. In
Chemical Modeling of Aqueous Systems, II, Chapter
30, eds. D. C. Melchior and R. L. Bassett, American
Chemical Society, Washington, D.C.. pp. 379-396.
Hart. E. J. and Anbar, M'. 1970. TheHydrated'Electron.
Wiley Interscience, New York, 267 p.
Hosteller. J. D. 1984. Electrode electrons, aqueous
electrons, and redox potentials in natural waters.
American Journal of Science, v. 284, pp. 734-759.
Kempton, J. H., Lindbcrg, R. D., and Runnclls, D. D.
1990. Numerical modeling of platinum Eh
measurements by using heterogeneous electron-
transfer kinetics. In Chemical Modeling in Aqueous
Systems II, Chapter 27, eds. D. C. Melchior and R.
L. Bassett. American Chemical Society Symposium
Series 416, American Chemical Society.
Washington, D.C., pp. 339-349.
Langmuir, D. 1971. Eh-pHdetermination. In.Procedures
in Sedimentary Petrology, Chapter 26, ed.
R. E. Career, Wiley-lnterscience, New7York, pp 597-
635.
Lindberg, R. D. and Runnells, D. D. 1984. Ground water
redox reactions: An analysis of equilibrium state
applied to Eh measurements and geochemical
modeling. Science, v. 225. pp. 925- 927.
Macalady, D. L., Langmuir. D.. Grundl. T., and
Elzerman, A. 1990. Use of model-generated Fe3"ion
activities to compute Eh and ferric oxyhydroxide
solubilities in anaerobic systems. In Chemical
Modeling in Aqueous Systems. Chapter 28,
ed. E. A. Jenne, American Chemical Society
Symposium Series 93, American Chemical Society,
Washington, B.C., pp. 350-367.
Morris, J. C. and Stumm, W. 1967. Redox equilibria
and measurements of potentials in the aquatic
environment. In Equilibrium Concepts in Natural
Miter Systems, Chapter 13, ed. W. Stumm, American
46
-------
Chemical Society Advances in Chemistry Series 67,
American Chemical Society. Washington. D.C.,
pp. 270-285.
Nordstrom, D. K. 2000. Advances in the
hydrogeochemistry and microbiology of acid mine
waters. International Geology Review, v. 42,
pp. 499-515.
Nordstrom, D. K. and Wilde, F. D. 1998. Reduction-
oxidation potential (electrode method). In National
Field Manual for the Collection of Water-Quality
Data, eds. F. D. Wilde, D. B. Radtke, J. Gibs, and
R. T. Iwatsubo. U.S. Geological Survey Techniques
in Water-Resources Investigations Book 9,
Handbooks for Water-Resources Investigations. A6.5.
Nordstrom, D. K., Jenne, E. A., and Ball, J. W. 1979.
Redox equilibria of iron in acid mine waters. In
Chemical Mode ling in Aqueous Systems, Chapter 3,
ed. E. A. Jcnnc, American Chemical Society
Symposium Series 93, American Chemical Society,
Washington, D.C., pp. 51-79.
Parkhurst, D. L. andPlummer, L. N. 1993. Geochemical
models. \a. Regional Ground-lffitter Quality, Chapter
9, ed. W. M. Alley, Van Nostrand Reinhold, New York,
pp. 199-226.
Peschanski, D. and Valensi, G. 1949. Contribution a
1'electrochemie des solutions aqueuses de
polysulfures. Journal de Chimie Physique, v. 46,
pp.'602-619.
Runnclls, D. D. and Skoda, R. E. 1990. Rcdox modeling
of arsenic in the presence of iron: Applications to
equilibrium computer modeling. In Proceedings:
Environmental Research Conference on Ground/water
Quality and Waste Disposal, EPRI Report EN-6749,
Palo Alto, California.
Runnells, D. D.. Lindberg, R. D., and Kempton, J. H.
1987. Irreversibility of Se(VI)/Se(IV) redox couple
in synthetic basaltic ground water at 25°C and 75°C.
Materials Research Society Symposium Proceedings,
v. 84, pp. 723-733.
Stipp, S. L. 1990. Speciation in the Fe(D)-Fe(in)-SO4-
H,O system at 25°C and low pH: Sensitivity of an
equilibrium model to uncertainties. Environmental
Science and Technology, v. 24, pp. 699-706.
Thorstenson, D. C. 1984. The concept of electron activity
and its relation to redox potentials in aqueous
geochemical systems. U.S. Geological Survey Open-
File Report 84-072, 45 p.
To, T. B., Nordstrom, D. K., Cunningham, K. M., Ball, J.
W., and McCleskey, R. B. 1999. New method for
the direct determination of dissolved Fe(III)
concentration in acid mine waters. Environmental
Science and Technology, v. 33, pp. 807-813.
47
-------
48
-------
In
Bill Deutsch
Sedgwick, ME 04676
Rcdox processes can play a role in subsurface
remediation by either removing inorganic contaminants
from ground water and immobilizing them in the solid
phase or enhancing mobility in the aquifer so that the
contaminants can be removed from the system. The redox
process may directly or indirectly lead to remediation.
An example of direct remediation is the reduction of
Cr(Vl) to Cr(III) and precipitation of a less soluble Cr(III)
hydroxide solid. Indirect rcdox remediation would occur
when dissolved ferrous iron is added to ground w7ater
followed by iron oxidation, precipitation of hydrous ferric
oxide (HFO) and adsorption and removal of arsenic or
other inorganic contaminants from ground water.
Conversely, arsenic might be mobilized and removed
from an aquifer by injecting a reductant to dissolve
arsenic-adsorbing solids and release arsenic to the ground
water, which would then be pumped from the aquifer.
The purpose of this paper is to:
4 Identify redox-sensitive contaminants
4 Discuss natural redox conditions that attenuate
movement
4 Describe applied redox methods that immobilize
contaminants
4 Describe applied redox methods that mobilize
contaminants
Additional activities at the Redox Workshop related
to inorganic contaminants will generate site
characterization requirements and methods of measuring
important parameters. Data interpretation and use will
also be discussed separately.
Elements that occur in more than one valence state
in natural environments and may be present as
contaminants are listed in Table 1. These elements may-
be amenable to direct redox manipulation methods to
produce remediation or at least contaminant concentration
reduction. These elements may also be affected by
indirect redox remediation. Elements that arc not redox-
sensitive (such as Pb, Zn and Al) may also be affected by
indirect redox methods.
Tm6le 1. Redox-Sensitive Elements Potentially Present as
Contaminants in the Environment
Antimony (Sb)
Mercury (Hg)
Arsenic (As)
Nitrogen (N)
Carbon (C)
Molybdenum (Mo)
Chromium (Cr)
Selenium (Se)
Copper (Cu)
(S)
Iron (Fe)
(V)
(Mn)
Uranium (U)
III, V
0,1,11
III, V
-III, 0, III, V
-IV to IV
IIIJVVVI
III, VI
-II, 0, IF VI
1,11
-II to VI
II,III
III, IV V
II, III IF
IV VI
Many applied remediation methods for inorganic
compounds are based on natural processes that attenuate
or immobilize these compounds in the subsurface. This
discussion of remediation methods will start by looking
at these natural environments that concentrate inorganic-
compounds by redox processes.
to
The concept of a geochemical barrier in the natural
system was first suggested by Perel'man in 1961
(PereFman, 1986. Geochemical barriers: theory and
practical applications. Applied Geochemistry, 1:669-680).
A geochemical barrier is a zone in the subsurface
characterized by a sharp change in physical or chemical
environment along the flowpath and is often associated
49
-------
with the precipitation of elements from solution. Redox
barriers are classified as a type of physico-chemical
barrier, which also includes alkaline, acidic, adsorption,
evaporation and thermodynamic barriers. Three types
of redox barriers have been described: oxidizing, reducing
with sulfide, and reducing without sulfide. Because
barriers represent a change in environmental condition,
each type of barrier must be considered in the context of
the three redox types of water that might enter the barrier.
For example, weakly oxidizing water may encounter a
strongly oxidizing barrier or a reducing barrier with or
without sulfide. Finally, the pH of the aquifer wall affect
the type of element immobilized by the geochemical
barrier. Table 2 provides the potential encounters
developed by Perel'man and the elements that might be
concentrated at the three redox barriers.
Examples of the geochemical barriers listed in Table
2 include oxidation and precipitation of iron from mildly
oxidizing or reducing waters without sulfide
downgradient of a landfill; presence of U, As, and Mo in
roll front deposits formed by reduction of oxidizing water;
and precipitation of metal sulfides when oxidizing water
becomes reduced by organic matter in the presence of
sulfide. Table 2 shows the types of direct redox
remediation that might occur in geochemical barriers.
The barrier will lower concentrations of inorganic
contaminants, however the table does not provide
information on whether the resulting concentration will
be below a remediation cleanup level. The equilibrium
solution concentration will be the solubility of the mineral
in the barrier environment. This can be estimated by
proper characterization of the environment and
calculating the solubility of minerals in that environment.
An indirect natural redox attenuation process is the
precipitation of iron and manganese hydroxides followed
by adsorption of other contaminants. These solids have
a predominantly positive surface charge at neutral to
acidic pH, and in this pH range will preferentially adsorb
elements that are anionic in groundwater such as arsenic
(AsO43\ AsO/'1, chromium (CrO42'), selenium (SeO42~,
SeO32~). and molybdenum (MoO42~). At pH higher than 7
to 8. the mineral surfaces are predominantly negatively
charged and they will be better adsorbents for cations
such as Cu2+. C-d2% Zn2* and Ba2+.
Nitrate is an inorganic contaminant that does not
form minerals and thus would not be removed from
solution and concentrated in a geochemical barrier. The
fate of nitrate is usually tied to biological redox processes
that reduce nitrogen to N2 or NH^.
of
Contaminants
Applied techniques for immobilizing inorganic
contaminants arc generally based on the well-documented
natural processes that impact element mobility,
particularly at geochemical barriers. If manganese
concentrations in mildly oxidizing water downgradient
from a landfill must be lowered, Table 2 suggests either
2. Elements Concentrated at Geochemical Redox Barriers (after Perel 'man 1986}
Barrier Type
Oxidizing
Reducing w/H2S
Reducing w/o H2S
Oxidizing
Reducing w/H2S
Reducing w/o H2S
Oxidizing
Reducing w/H2S
Reducing w/o H2S
PH<3
Oxidizing
Fe
Tl, Cu, Hg, Pb.
Cd, Sn, As, Sb,
Mo,U
Cu, U, Mo
Reducing
S, Se, (Fe)
Reducing
Fe
Tl, Pb, Cd, Sn
Cu, U, Mo
pH 3 - & J
Water Entering
Fe, Mn, Co
Tl, Mn, Co, Ni, Cu,
Zn, Pb, Cd, Hg, Sn,
Cr, Mo, U
Cu, U, Mo
H2S Water Into
S, Se
Water w/o H,S
Fe, Mn, Co
Tl, Fe, Co, Ni, Pb,
Cu, Zn, Cd, Hg, U
Cu, U, Mo
pH 6.5 -8,5
Barrier
Mn
Tl, Cr, Mo, U, Se,
V
Cu, Cr, U, Mo,
Se, V
Barrier
S, Se
Into Barrier
(Fe), Mn, Co
Tl, Fe, Co, Ni,
Cu, Zn, Cd, Hg,
(Mo, U)
Mo, U
pH>8,5
Cu, Ag, Zn, Cr,
Mo, U, V, As
Cu, Ag, Cr, Mo, U,
Se, V, As
S, Se
(Mn)
Tl, Cu, Zn, Cd,
Hg, Mn, (Fe, Co,
Ni,U)
Mo, U
Elements in parentheses are less commonly found in the association
50
-------
J, Potential Oxidants and Reductants for Inorganic Remediation
OxlAants
Dissolved Oxygen
Ozone
Oxygen Release Compound (ORC®)
" [[[
Fenton 's Reagent
Reductants
Fe"
Na2 S
H2S
Hydrogen Release Compound (HRC™)
Organic Material
Elements Tested
Fe
u
Elements Tested
Cr, U, Tc, NO-,', Mo
"'
Zn)
Cr
Cr
Cr
increasing the redox potential to precipitate a less soluble
Mn(IV) oxide or reducing the redox potential and adding
sulfide to form a less soluble sulfide mineral. Some of
the issues to consider in the applied redox approach arc
the type and amount of reactant, means of emplacement,
reaction kinetics, unwanted byproducts, solubility of
contaminant-containing mineral, and geochemical
stability of the imposed barrier environment.
The types of oxidizing and reducing compounds that
have been tested or suggested to lower inorganic
contaminant concentrations in an aquifer are listed in
Table 3. Most of these compounds can be dissolved in
water and injected into the aquifer. Air and ozone arc
added to the subsurface in a vapor phase. Elemental iron
(Fe°) is a solid that is emplaced in a trench that intercepts
the contaminated water. It has been shown to be effective
not only at reducing the concentrations of inorganic
contaminants (particularly chromate and uranium) but
also chlorinated volatile organic compounds such as PCE
and TCE. As shown in Table 3, the current list of reagents
is fairly limited and the elements tested arc not extensive.
This type of remediation is expected to expand as its
effectiveness becomes more widely known and the
injection of reagents to achieve remediation becomes
more commonly accepted by regulatory agencies.
In general, the amount of oxidant or reductant that
must be added to achieve remediation will not exceed
100 mg/L. For example to reduce 50 mg/L Cr(VI) in
CrO,,2" to Cr(III) would require only 12 mg/L sulfide,
assuming all the sulfide is oxidized to sulfate. An
important consideration for some reactions is the rate of
oxidation or reduction, which may be slow in relation to
a remediation timcframc. For example, the half-life for
chromium reduction may be several years at pHs greater
than 7 unless a strong reductant is applied. The solubility
of the reaction product containing the contaminant can
be estimated by ion speciation calculation using the
projected environmental conditions during reagent
injection. The long-term solubility of the mineral must
also be estimated using the aquifer geochemical
conditions that arc expected to prevail after active
remediation has ceased.
of
It is preferable to leave a contaminant in place if its
mobile concentration can be reduced to below a cleanup
level, however in some instances immobilization may
not be possible and removal is necessary. Removal of an
inorganic contaminant requires not just extraction of
contaminated ground water but also dissolution of
contaminant-containing minerals and perhaps desorption
of the contaminant from the surface of solids. Table 2
can be used to develop redox methods of removing
contaminants from the solid phase. For example, if
Table 2 shows that an element is concentrated under
reducing conditions (e.g., Cr, U, Mo, Se), then creating
oxidizing conditions in the aquifer should mobilize the
contaminant. This is exactly the process used to mine
uranium by in-situ leach methods. Oxygen is added to
the subsurface where uranium has been concentrated by
natural processes in the reduced U(IV) mineral uraninite
(UO2). Uraninite is soluble under the imposed oxidizing
conditions leading to its dissolution and the removal of
-------
4. Redox Manipulation to Mobilize Inorganic Contaminants
Mobilizing
Environments
Reducing w/o
Sulfide
Oxidizing
Oxidizing
pH<3
pH 3-6.S
pH 6.5 -8.5
pff>8.5
Initial Oxidizing Environment
Fe, Se, S
Initial
Tl, Cu, Hg,
Pb. Cd, Sn,
As, Sb, Mo,
U
Initial
Cu, U, Mo
Fe, Mn, Co, Se, S
Reducing w/ S2"
Tl, Mn, Co, Ni,
Cu, Zn, Pb, Cd,
Hg, Sn, Cr, Mo, U
Reducing w/o S2"
Cu, U, Mo
Fe, Mn, Co,
Se, S
Environment
Tl, Cr, Mo, U,
Se, V
Environment
Cu, Cr, U, Mo,
Se,V
Mn, Se, S
Cu, Ag, Zn,
Cr, Mo, U,
V, As
Cu, Ag, Cr,
Mo, U, Se,
V, As
Table 4 shows imposed redox conditions that are
expected to mobilize contaminants concentrated in three
types of subsurface environments. The methods of
inducing redox conditions are the same as those given in
Table 3. Mobilization of contaminants can be complicated
by the strong adsorption and high capacity of iron and
manganese oxyhydroxides for other metals. If oxidizing
conditions are generated in a reducing environment to
mobilize metals, iron and manganese solids may
precipitate and remove from solution some of the metals
that are the target of remediation. This occurrence may
require a stepwise mobilization process to first oxidize
the elements in the minerals and then reduce and dissolve
iron and manganese oxyhydroxides to release adsorbed
contaminants.
Natural and applied redox processes can be used to
achieve remediation goals at sites contaminated with
inorganic compounds. This has been documented for
many elements by the existence of natural geochcmical
barriers, and recent applied techniques have shown that
the natural redox processes can be enhanced by chemical
addition. A goal of the Redox Workshop has been to
expand the use of redox processes for inorganic
remediation by providing examples of natural attenuation
analogues and the application of these natural processes
to enhance applied restoration techniques.
52
-------
to
Andy Davis
Geomega
Boulder, CO 80303
Introduction
Redox conditions represent an important control on
the speciation of metals including As (III/V), Cr (III/VI),
Sb (ITT/V), and Se (IV/VI). Although the measurement
and environmental geochemistry of these constituents is
well understood, application of this knowledge in setting
regulatory standards for soils, ground water, surface water
and sediments is rare. .Hence, bulk chemistry is usually
considered in setting standards rather than evaluating the
ameliorating effect of metal valence on toxicity and
mobility. Given the escalating costs of environmental
restitution, it is critical that reasonable options, including
the influence of valence, be considered in setting realistic,
risk-based cleanup goals.
This paper provides examples of the value of
understanding redox conditions and species in the
environment. Arsenic is less bioavailable in the As(III)
form in mine wastes due to sparing dissolution of these
less soluble forms in the short GI tract residence times,
while selenium form in sediment is important because
Se(IV) is less toxic than Se(VI). AVS/SEM analysis on
both metalloids has demonstrated that they are
incorporated in framboidal pyrite. rendering them non-
bioavailable.
In addition, novel approaches to constrain metal
solubility by modifying the valence state are realistic
goals. For example, the mitigating influence of
precipitation reactions in sediments is demonstrated
through sulfate/sulfide reduction/precipitation and
coincident encapsulation of metals in a biologically active
system in Richmond, California. For example, the
solubility of Cr(III) in acid ground waters can be reduced
by careful pH control while remediation of acid pit lakes
can be effected by modification of the pH with
concomitant precipitation of amorphous ferric hydroxide
as the Fe(II)/Fe(III) ratios change, and sequestration of
previously soluble ions.
The practical application of redox conditions to
remedial actions is of interest, first from the perspective
of '"does it matter," and secondly in defining how to
implement closure in the event that redox conditions are
relevant. From the closure/remediation perspective, the
importance of redox is closely related to the implied risk
associated with different valence states. For example,
As(III) is more toxic in groundwater whereas As(V) may-
be more toxic in an estuarine setting, depending on the
exposure pathway; Cr(VI) is generally more toxic than
Cr(III), while Sc(VI) is more toxic than Se(IV).
From a pragmatic perspective, although the relative
risk associated with different valence states may be
known, translating theory into practical closure that meets
regulatory standards is rare. For example, contaminated
environmental systems are complex and heterogeneous.
frequently consisting of multiple compartments and
contaminants, while translating theory into practical
reality is fraught with uncertainty. In addition, the
significance of the issue and a reasonable benchmark
representing success is generally not understood by the
decision maker, while with few exceptions, the remedial
actions do not finesse the benefit to be accrued by altering
valence states in the environment.
Despite such recidivism, where viable, manipulation
of redox conditions may portend significant financial
benefit and should be included as a weapon in the
remedial arsenal. The application of redox can be divided
into the following areas:
1) Active manipulation (e.g., redox control/response)
2) Constraints on risk (e.g., definition of valence in land
application, metal bioavailability), and
3) Passive long-term (e.g., intrinsic remediation by
dcchlorination)
This paper reviews the genesis of the redox-based
standards, describes a few examples where redox states
have been incorporated into closure/remedial measures,
forecasts what may work and whether success or failure
can be predicted, and discusses issues that are likely to
become important in the future.
53
-------
The Evolution of
Promulgation of the Aquatic Water Quality Control
(AWQC) Standards may represent the initial foray that
led to a description of the relative toxicities of metal
valences in the "Gold Book" (EPA, 1986). The later
document identified a range of conditions that affect
different trophic levels. For example, the alga,
seienastmm capricornutum was reported 45 times more
sensitive to As(V) thaiiAs(III), although other data present
conflicting information. In general, the resiliency to
arsenic is species and salinity specific.
Chromium toxicity appears more straightforward,
in that there is general concurrence that Cr(III) is less
toxic than Cr(VI). For example the AWQC standard for
Cr(VI) is 100 |jg/l, while for Cr(III) it is 3,433 (ig/1. In
freshwater, the acute toxicity for Se(IV) is reported as
260 ju.g/1, while for Se(VI) it is 760 jog/l. However, there
has been no directives that clearly identify the relationship
between valence species and the total dissolved
concentration in aqueous or sediment systems, and the
ramifications to risk assessments and, ultimately, to
selection of cleanup standards.
of Chromium Water
At a former chemical facility in Wilmington, a
historical release of chromium sulfate has generated an
acidic plume (pH~3) of Dense Aqueous Phase Liquid
(DAPL) with complex chemical characteristics. A Release
Abatement Measure (RAM) is being instituted to assess
the effectiveness of in-siru precipitation of ground water
solutes using various chemical agents applied in solution
or slurn? through existing wells with the objective of
raising the pH, while avoiding oxidation of the Cr(III) to
soluble Cr(VI). The effectiveness will be determined by
monitoring the arcal extent of DAPL precipitation. The
goal of this treatment is to mitigate the flux of solutes
from the DAPL to the overlying ground water that
discharges into the on-site ditch system, hence reducing
precipitates forming in the ditch as ferrous iron in solution
discharges and precipitates in surface water.
In the target area, the local ground water/diffuse
laycr/DAPL profile consists of approximately 6 feet of
DAPL overlain by a 3 foot thick diffuse layer. Comparison
of ground water surface elevations from co-located wells
in the region of the ditches demonstrates that ground water
discharges to the ditch system resulting in the formation
of floe. It is anticipated that the DAPL RAM will inter
DAPL, reducing the solute flux to the South Ditch, and
eventually floe formation.
Bench-scale studies observed decreases in metal
concentrations attributed both to precipitation reactions
and to changes in solubility due to the more neutral pH
achieved through treatment. For example, chromium
decreased from 2,000 mg/1 to < O.lmg/1, while similar
decreases were observed for Al and Fe. Additionally,
sulfate concentrations decreased by up to 30%. Hence,
neutralization of acidic DAPL to circumneutral (~6)
conditions should result in a decrease in DAPL solubility
by a precipitation reaction of the form:
9 Fe+2 + Cr2[SO4]3 + 24 OH'- => 8 Fe075Cr025(OH)3 +
3FeSO4° + 6 e-
with the relevant redox reaction controlled by oxidation
of ferrous to ferric iron and coprecipitation of Cr(III) in
the amorphous ferric hydroxide matrix. The optimal
liquor concentration determined from the bench scale
testing will be applied to the target area. The neutralizing
liquor and the sodium bromide tracer (at 100 mg/L
effective bromide concentration) will be initially
delivered into the injection well by passive head
differential. As delivery slows, the liquor may be pumped
underpressure if feasible.
by inert
Castro Creek, an industrial drainage ditch that
historically received disposed process fluids, contains
residual arsenic in the sediments. This project initially
identified solubility-constraining reactions and then
sought to enhance the natural conditions to further
influence pore water arsenic.
The molar SEM/AVS ratios for arsenic are well
below 1, ranging from 0.0003 at B2 to 0.62 at F. Although
SEM/AVS data for arsenic has not been directly coupled
with arsenic toxicity, other studies (Davis et al., 1996)
have demonstrated that when the SEM/AVS ratio is <1,
bioaccumulation of arsenic docs not occur in higher
trophic level receptors (fish). Therefore, based on these
studies (Davis et al., 1996), the ambient geochemical
conditions of the creek sediments facilitates sequestration
of arsenic in insoluble mineral phases such as Fe(As)S,
Fe(As)SO4, and AsS2 that are likely to reduce As
bioavailability in a manner akin to the metal-sulfide
complexes described by DiToro (1990 and 1992).
In
Recognizing the benefits to be realized from a
reducing sediment environment resulted in research
(currently ongoing) to facilitate additional removal of
-------
arsenic from pore water at the site. Currently a remedial
design scheme intended to remove arsenic (As) and
control remnant As mobility in marsh sediment using a
Peat-Iron-Gypsum (PIG) blanket to minimize As
solubility by reducing As(V) to As(III) and facilitating
co-precipitation of As in an inert FeS0 (e.g., FexAs1_xS2
and FcxAs, XSO4) phase is being investigated.
Iron oxide and sulfates containing up to 3.000 mg/kg
As have been found precipitating in creek sediments
suggesting that natural conditions in these areas are
conducive to immobilization of incorporated As.
Theoretically, addition of appropriate amendments would
attenuate most of the remaining soluble pore-water As
over the ambient range of redox conditions, also reducing
the flux of As into the creek.
A bench scale study was ran for 24 hours (thought
to be an adequate period for pyrite precipitation; Rickard
1995) under zero-head space conditions in a nitrogen
atmosphere and each bottle sacrificed after 7 days. The
aqueous and solid phases were separated by centrifugation
and samples of each collected and analyzed. Measured
geochemical parameters include SO42", S2~, Fe(II), Fe(III),
pH, Eh, and specific conductivity. A sample was also
analyzed for inorganic arsenic species [As (III)/(V)]. The
results were surprising in that the reaction kinetics were
much slower than has been reported in the literature using
pure materials.
as
of intrinsic
An investigation was undertaken to ascertain the
potential and actual degradation of chlorinated pesticides
(a-, P-, and y-BHC, DDT, and ODD), and the purgeablc
aromatic compounds (ethylbenzene and xylene) at the
Marzone Superfund Site, Tifton, Georgia. The objective
was to test the hypothesis that Monitored Natural
Attenuation (MNA) would provide an effective ground-
water polishing strategy to augment the remedial collector
trench recently installed downgradient of the site. The
trench effectively intercepted over 90% of the mass of
all Compounds of Concern (COCs) while concentrations
downgradient of the trench are at, or close to, the relevant
numerical standards.
The geochemical conditions were characterized by
analyzing redox chemistry (Eh, iron and sulfur valence
states), gas concentrations (hydrogen, methane, ethane,
ethene), and other indicators (chloride, alkalinity, etc).
These data demonstrated that upgradicnt ground water
that was unaffected bv the historical Marzone Site
operations is moderately aerobic and oxidizing, becoming
moderately reducing proximal to the operating facility.
In the core of the plume, conditions become strongly
reducing based on the increase in the relative ratio of
hydrogen, the presence of dissolved iron exclusively in
the ferrous form, and on the concomitant sag in the
oxygen ratio. Downgradient of the plume core, redox
conditions are moderately reducing and methanogenic,
and again become more oxidizing at the furthest
downgradient location, where site ground-water oxygen
concentrations are highest.
The historical COC data demonstrated that in most
wells, COC concentrations have decreased over the 8
years since monitoring began, while a suite of degradation
products provided conclusive evidence for dechlorination.
The results provide evidence that the geochemical
environment is conducive for mineralization, while the
analytical chemical data demonstrate that MNA
represents a viable alternative to reach the final remedial
goals.
Closure of #1
Placer Dome is in the process of obtaining approval
from the Nevada Division of Environmental Protection
(NDEP) to close heap leach Pad # 1 at the Bald Mountain
Mine (BMM), located in White Pine County, Nevada.
Specifically, the closure includes:
4> Applying heap effluent to site soils and possibly-
increasing the application rate by enhancing
evaporation,
4 Re-grading the pad to stabilize the pad slopes for
erosion control and to control runoff, and
4 Covering the heap with site soils and revegetating to
minimize runoff and infiltration of meteoric waters,
while meeting risk objectives for plants, livestock.
and wildlife.
Groundwatcr in the vicinity of the site is
approximately 500 feet below the surface (Hydro-Search,
1981), and is therefore unlikely to be impacted at
reasonable application rates.
Water quality in the barren and pregnant ponds
associated with Pad #1 are shown for constituents that
have on occasion exceeded standards since 1998.
Concentrations of a few constituents have decreased since
1998 to below or near NDEP standards (CN, Hg, Ni),
while others have consistently remained above the
standards (Sb, As, NO3-NO2, Se, SO4, pH).
-------
A batch attenuation test indicated Langmuirean
solute behavior, from which n and Kd were obtained. For
some compounds (e.g.. SO4) sorption is minimal, while
for others (e.g., As) sorption will effectively attenuate
solutes in the soil horizon.
Results from the spcciation analyses for As. Se, and
Sb show that the predominant species are As(V), Se(VI),
and Sb(V). For As and Sb, the predominant valence state
is the less toxic of the forms, while Se (VI) is of similar
toxicity to Se (IV). The predominant species detected for
As, Se, and Sb are consistent with the generally oxidizing
and alkaline geochemical conditions measured for the
heap effluent. For example. As(V) constituted >80% of
the total As, Sc(VI) was >90% of the total Se, and Sb(V)
was >90% of the total Sb analyzed. The total solute
concentration based on the sum of the two valence states
is also in good agreement with the total solute
concentration reported by SVL Analytical. The effect of
attenuation on solute concentration where solute lost from
solution by sorption to soils is illustrated by the difference
between the sorption data and the theoretical 1:1
equivalence curve.
Conclusions
Redox can play a significant role in modifying metal
solubility. The applications are far-reaching, ranging from
incorporation in human health and ecological risk
assessments, the development of novel methods to
remediate sites, to demonstration of natural attenuation.
However, while understanding and applying redox
strategies is important, the evolving technology must be
applied within a regulator}? framework that will allow
their incorporation into the panoply of environmental data
interpretation and restitution. At the same time, it is
incumbent on industry to demonstrate that redox
strategies really do provide meaningful and realistic
solutions to environmental problems.
References
Davis. A. and Olsen, R. L. 1995. The geochemistry of
chromium migration and remediation in the
subsurface. Ground Water, v. 33. pp. 759-768.
DiToro, D., Malioney, I, Hansen, D., Scott, K., Hicks, M,
Mayr, S., and Redmond, M. 1990. Toxicity of
cadmium in sediments: The role of acid volatile
sulfide. Environmental Toxicology and Chemistry,
v. 9, pp. 1487-1502.
DiToro, D., Malioney. J.. Hansen, D., Scott, K.,
Carlson, A., and Ankley, G. 1992. Acid volatile
sulfide predicts acute toxicity of cadmium and nickel
in sediments. Environmental Science and
Technology, v. 26. pp. 96-101.
PTI. 1997. Evaluation of Reactive Barrier Effectiveness
Batch and Column Tests at the Wilmington,
Massachusetts Site. December, 1997.
Rickard, D. T. 1995. Kinetics of FcS precipitation: Part
1. Competing reaction mechanisms. Geochimica et
CosmochimicaActa, v. 59, pp. 4367-4379.
56
-------
Six breakout sessions were organized to summarize and evaluate key considerations during assessment and
implementation of remedial strategies for redox sensitive contaminants. Three concurrent sessions took place
in the morning and afternoon of the second day of the Workshop. The focus of the three morning sessions
centered on identification of approaches and parameters necessary to evaluate remediation of petroleum
hydrocarbons, chlorinated hydrocarbons, MTBE, and inorganics. The focus of the three afternoon sessions
centered on aqueous, gas phase, and solid phase measurement issues. Items emphasized during the afternoon
sessions, in part, depended on the outcome of the morning sessions.
Before arriving at the workshop participants were delivered a set of questions that along with the Plenary
Presentations, served as a launch point for the small group discussions. These questions are provided below.
/.
(R. Ludwig, M. Barcelona, A. Azadpour-Keeley, I Barker, P. Bradley, I. Cozzarelli, R. Ford, S. Warner,
E. Hartzell, D. Macalady, K. O'Reilly, K. Piontek, D. Vroblesky, D. Kampbell)
What remediation processes specific to petroleum hydrocarbons require the need for rcdox measurements
during site characterization and/or remediation (e.g., MNA. electron acceptor addition, thermal treatment,
surfactant flushing, etc)? Do we need to distinguish between categories of petroleum hydrocarbons (e.g..
aromatics, PAHs, alkenes. alkanes, etc.)? What specific redox measurements are required (e.g., Eh, pH, DO.
Fe2+, etc) and how arc the measurements used (e.g., determining feasibility vs. non-feasibility of specific
remedial processes, predicting rates of clean-up, determination of enhancements that could potentially be used
to accelerate clean-up, etc.)?
2. in
(J. Wilson, F. Chapelle, J. Amonette, C. Braun, J. Ingle, S. Jones. R. Lee. M. Lorah, C. Macon. P. Groeber.
R. Pirkle, T. Sivavcc, J. Szecsody, C. Tcbcs Steven, J. Washington, E. Weber, T. Wicdcmcicr, B. Wilson)
What remediation processes specific to chlorinated hydrocarbons and MTBE require the need for redox
measurements during site characterization and/or remediation (e.g.. MNA, enhanced bioremediation, zero
valent metal treatment, in-siru oxidation, in-situ redox manipulation, etc.)? Do we need to distinguish between
categories of chlorinated hydrocarbons (e.g., chlorinated methanes, chlorinated ethenes, chlorinated ethanes,
etc.)? What specific redox measurements are required (e.g., Eh, pH, DO, Fe2% etc.) and how are the
measurements used (e.g., determining feasibility vs. non-feasibility of specific remedial processes, predicting
rates of clean-up, determination of enhancements that could potentially be used to accelerate clean-up, etc.)?
3. in
(R. Puls, B. Deutsch, J. Westall, S. Benner, A. Davis, J. Fruchter, J. Herman, P. Longmire, D. Luce,
R. McGregor, O.K. Nordstrom, C. Ptacek, S. Schmelling, J. Shockley, C. Su, R. Wilkin, D. Willey, K. Cook)
What remediation processes specific to inorganic contaminants require the need for redox measurements
during site characterization and/or remediation (e.g., MNA, zero valent metal treatment, in-situ redox
manipulation, etc.)? What parameters must be measured during site assessment and remediation? Are the
required parameter measurements universal or contaminant-specific? Is there a universal set of parameters that
is sufficient for all remedial strategies? What specific redox measurements are required (e.g., Eh, pH, DO, Fe2",
etc) and how are the measurements used? Is there a need to evaluate redox characteristics of the solid matrix
in addition to the mobile phase?
57
-------
4. 'Parameters
(R. Wilkin, C. Ptacek, A. Azadpour-Keeley, B. Deutsch, J. Frachter, J. Ingle, P. Longmire, M. Lorah, D. Luce,
K. Cook, D. Macalady, D.K. Nordstrom, T. Sivavec, C. Su, J. Westall, B. Wilson, J. Wilson, T. Wiedemeier,
K. Piontek)
Participants in this breakout group will identify and discuss the techniques used to determine the concentrations
of key redox indicator species, such as Dissolved Oxygen. Ferrous Iron, and Hydrogen Sulfide, as well as the
Oxidation-Reduction Potential (ORP/Eh). Group discussion will focus on identifying the best practices for
sample collection, sample handling, documentation, and measurement techniques for these key parameters.
We will attempt to identify rigorous QA/QC procedures that can be adopted in the field, and determine how
these parameters may be practically applied to understand the migration and evolution of contaminant plumes.
5.
(D. Kampbell, F. Chapelle, J. Barker, P. Bradley, C. Braun, S. Jones, R. Lee, R. Ludwig, P. McLoughlin,
K. O'Reilly, P. Groeber, R. Pirkle, J. Shockley, D.' Vroblesky, S. Warner)
Participants in this breakout group will identify' the acceptable methods for determining the concentrations of
dissolved gases in ground water. Discussion will be primarily focused on hydrogen, ethene/ethane, and
methane. Group discussion will center on the best practices for sample collection, sample handling, documen-
tation, and measurement techniques for these gas species. A specific question to be discussed is how is gas
phase characterization used to recognize the footprint and dynamics of a contaminant plume?
6. Measurement Issues: SolidI%ase
(R. Ford, J. Herman, J. Amonette, M. Barcelona, S. Benner, I. Cozzarelli, A. Davis, C. Macon, R. McGregor,
R. Puls, S. Schmelling, J. Szecsody, C. Tebes Steven. J. Washington. E. Weber, E. Hartzell)
Participants in this breakout group will identify' the key solid phase parameters relevant to redox-controlled
processes active during in-silu remediation of ground water. In addition, we will identify' acceptable methods
for sample collection and characterization. Emphasis will be placed on solid-phase iron. An example question
to be addressed is what is the most appropriate method(s) for quantifying the fraction of Fe-bearing minerals
that function as electron acceptors during microbially mediated redox transformation of a contaminant?
Discussion will also address analytical characterization and the role of solid phase sulfur and carbon during
remedial assessment and implementation.
58
-------
1
in
Remediation
Ralph Ludwig, Michael Barcelona, and Keith Piontek
1.1
The objective of this session was to evaluate the value and potential role of redox parameter measurements in
petroleum hydrocarbon site characterization and remediation. (Redox parameter measurements are herein
defined as any parameter measurement that provides insight into the electron transfer status of an aquifer
system or the propensity of an aquifer system to donate or accept electrons.) General conclusions of the session
included:
1. The need to measure redox parameters at petroleum hydrocarbon impacted sites will depend on
site-specific questions that need to be answered and the remedial strategies that are under
consideration. Knowledge of the electron transfer (redox) status within an aquifer will not always
be required.
2. Collection of at least some redox parameters such as dissolved oxygen, dissolved (ferrous) iron,
and dissolved manganese can be very beneficial in the initial stages of site characterization to help
develop a conceptual model for the site and to aid in screening potential remedial strategies. The
success or failure of some remedial strategies can be impacted by the electron transfer status of the
aquifer.
3. Additional redox parameter measurements including dissolved nitrate/nitrite, dissolved sulfate,
methane, dissolved carbon dioxide (as measured by acidimetric titration), and oxidation-reduction
potential (e.g., electrode-based ORP measurements) may be useful particularly in cases where a
natural or enhanced bioremediation strategy is under consideration.
4. In cases where an electrode-based redox potential (ORP) measurement is made, the measurements
should be used for qualitative purposes only to support other redox parameter measurements in the
development of the site conceptual model. No direct effort should be made to link electrode-based
redox potential (ORP) measurements to specific electron acceptor pathways or to the speciation of
redox sensitive species in the aquifer.
5. Redox parameter measurements should not be used as surrogate measurements to reduce the
number of hydrocarbon monitoring locations or reduce the frequency of hydrocarbon sample
collection. Redox parameter measurements should be used strictly in support of direct hydrocarbon
measurements to better understand (conceptualize) subsurface processes and plume behavior at the
site.
1.2 of
Extensive research addressing petroleum hydrocarbon behavior and fate in ground water has been conducted
over the past twenty years. Various in situ and ex situ technologies for remediating petroleum hydrocarbon
impacted sites have been developed and utilized with varying degrees of success. Evaluation of the feasibility
of using one or more remedial technologies at a particular site may in some cases benefit significantly from
knowledge of the redox conditions in the subsurface. Knowledge of specific redox parameters and conditions
at a site can aid in establishing baseline conditions from which the existing behavior or anticipated future
behavior of a hydrocarbon plume can be evaluated. In some cases, redox parameters may also be used as a
cost-effective means of evaluating the performance of specific remedial technologies.
59
-------
Three key questions will generally need to be answered in assessing the role of redox parameter measurements
at petroleum hydrocarbon impacted sites.
1. Will knowledge of the redox status in the aquifer be of any value in evaluating the feasibility of
potential remedial strategics at the site?
2. If so, which redox parameter measurements might be useful?
3. Will the redox parameter measurements be used for qualitative or quantitative purposes?
Knowledge of the redox status of a hydrocarbon impacted aquifer system could certainly be important should
in-situ bioremediation technologies be under consideration. Specifically, the presence/absence and distribution
of electron acceptors can have important implications with regard to biodegradation pathways and plume
behavior. Petroleum hydrocarbons (aromatics, PAHs) as well as coal-derived hydrocarbons (e.g., cresols.
phenols) generally undergo much more rapid and efficient biodegradation under aerobic (oxidizing) condi-
tions. Thus, knowledge of the presence/absence of oxygen (and other potential electron acceptors) provides
insight into existing electron and biological activity and the means by which bioactivity can potentially be
enhanced. Knowledge of redox conditions can also be important in evaluating the biodegradation potential of
gasoline additives such as methyl-tert butyl ether (see Session 2 Summary).
The potential use of redox parameter measurements in the evaluation of technologies other than bioremediation
for hydrocarbon impacted aquifer systems will not be so obvious and will generally need to be considered on
a case by case basis. For example, in the case where a thermal based remediation technology might be the
preferred remedy, redox parameter measurements are not likely to be of any value. Alternatively, if a remedy
involving delivery of oxygen to the subsurface is to be considered, then knowledge of specific redox
parameters (e.g., dissolved iron and manganese concentrations) in the ground water and their potential impact
on system performance can be important.
The use of redox data for qualitative or quantitative purposes will depend on how the redox data are to be
utilized. If the objective is simply to determine which redox processes might be dominant at a particular
location, then establishment of the relative presence/absence of selected electron acceptors within and outside
a hydrocarbon plume will generally suffice. However, if an effort is being made to answer questions regarding
electron acceptor capacity and longevity (e.g., for a monitored natural attenuation remedy), men a quantitative
approach may be needed. This would require a more rigorous and defensible (and generally more sophisti-
cated) data collection and analyses process than would be required for a qualitative analysis. The significant
spatial and temporal variability of many redox parameters can make quantitative analyses at sites extremely
challenging.
1.3 Characterization Applications
The need to collect redox parameter measurements at petroleum hydrocarbon impacted sites will depend on the
site-specific questions that need to be answered. The most important information regarding petroleum
hydrocarbon plume migration and behavior will generally be gained from direct knowledge of both the
temporal and spatial distribution of specific contaminants of concern (e.g., BTEX compounds). Redox
parameter measurements are best used as secondary lines of evidence for locating residual source areas and for
better understanding, predicting, and/or confirming dissolved hydrocarbon plume migration and behavior. It is
known that a strong correlation generally exists between petroleum hydrocarbon concentrations and the
relative distribution of various electron acceptors and reduction products including dissolved oxygen, ferrous
iron, manganese, nitrate, sulfate, and methane. This correlation results from the utilization of electron
acceptors in a favored sequential order by microorganisms during oxidation (metabolism) of hydrocarbon
compounds. The use of oxygen as an electron acceptor allows for the greatest energy gain (and thus the most
efficient and rapid degradation) during oxidation of hydrocarbons by microorganisms. Thus, as long as oxygen
is available, oxygen will be the predominant and preferred electron acceptor utilized by microorganisms.
However, once oxygen is depleted, microorganisms must turn to less favorable electron acceptors, which
results in the oxidation process becoming increasingly less efficient. The relative relationships between
petroleum hydrocarbon concentrations, electron acceptors, and redox potential are illustrated in Figure 1.1. It
should be noted that the relative relationships are equilibrium tendencies and may or may not be correlated
directly.
Collection of selected redox parameter measurements in the initial stages of site characterization may aid in
establishing the feasibility of one or more technologies at a given site as well as evaluating the stability of the
60
-------
TPH-
BTEX
Low
High
Oxygen
Nitrate Manganese Iron
Sulfate
CO2
2^ CH4
Order of electron acceptor use during biodegradation
^
High
O,
Relative aqueous-phase concentrations
High Low Low High
Mn2+ Fe2+ SO42'
Low
CH4
Low
Low
High
High
Low
High
Redox
Potential
High
Low
Figure 1.1 Relative relationships between TPH-BTEX concentrations and electron acceptor/reduction product
concentrations and redox potential in ground water within a petroleum hydrocarbon plume.
area of affected ground water. For example, the absence of oxygen, presence of high levels of ferrous iron, and
low ORP readings in the core of a BTEX plume at a site (relative to background locations) can suggest strong
bioactivity and therefore indicate conditions that might be conducive to an aerobic bioremediation-based
remedial strategy. Low Eh ground water with a high dissolved iron and manganese content (which can lead to
clogging of pumping wells and above ground treatment systems) could signify greater maintenance attention
should an ex-situ remedial strategy such as pump and treat be considered. Low Eh ground water with a high
dissolved iron and manganese content could also present problems for in situ systems such as air sparging. The
collection of redox information can thus be important in contributing to the initial screening and evaluation of
candidate remedial technologies.
Redox parameters that should be measured both within and outside the plume in the initial stages of site
characterization should at a minimum include dissolved oxygen, ferrous iron, and manganese. ORP measure-
ments may also be considered, but should be used solely as a secondary line of evidence in support of these
more reliable redox parameter measurements. If bioremediation is to be considered as a candidate remedy,
then the redox parameter measurements should be expanded to include nitrate/nitrite, sulfate, and in some
cases, alkalinity.
Measurement of dissolved oxygen, dissolved (ferrous) iron, and manganese within a hydrocarbon plume
relative to background concentrations at the site will provide an indication of whether biodegradation processes
are in effect and consequently whether the aquifer system is amenable to natural or enhanced bioattenuation.
Measurement of these parameters will also provide an indication of design features that may need to be
incorporated into candidate in situ and ex situ treatment systems to avoid potentially adverse reactions (e.g..
precipitation). Measurement of nitrate/nitrite and sulfate/sulfidc will provide an indication of additional
electron acceptor availability/capacity within the aquifer while alkalinity measurements can be used as a
61
-------
measure of microbial CO2 production during biodegradation in the absence of other potential sources of
alkalinity. In addition to these parameters, field measurements should generally include temperature, pH, and
conductivity to further determine the extent to which the ground water may or may not be favorable to selected
treatments.
Specific examples where redox parameter measurements may play a useful role in site characterization
activities at petroleum hydrocarbon contaminated sites include the following.
The magnitude of biodegradation activity in ground water may be an important factor in the selection of
remedial approaches such as enhanced bioremediation or monitored natural attenuation. Biodegradation by-
products (e.g., degraded hydrocarbons, enzymes, residual inorganic constituents) of petroleum hydrocarbons
are not easily quantified in ground water. Thus, there is a need to rely on alternative evidence such as the
presence or absence of electron acceptors and reduction products to support the presence/absence of biodegra-
dation activity at sites. For example, the absence of oxygen and presence of dissolved (ferrous) iron in a zone
of ground-w7ater contamination provides evidence that biodegradation is occurring. Increases in alkalinity may
be indicative of carbon dioxide production associated with microbial respiration.
Source Area
Decreasing dissolved oxygen, nitrate, and sulfate concentrations and increasing manganese and ferrous iron
concentrations generally coincide with increasing proximity to hydrocarbon source areas. These measure-
ments, as well as methane concentrations and ORP measurements, can be used as secondary lines of evidence
to complement TPH and BTEX concentration measurements in verifying the locations of petroleum hydrocar-
bon source areas.
Dissolved Plume
Higher dissolved oxygen and nitrate concentrations and lower dissolved (ferrous) iron and manganese
concentrations will generally be more characteristic of ground water outside the limits of a dissolved petroleum
hydrocarbon plume. Within the limits of a dissolved plume, dissolved oxygen and nitrate concentrations will
generally decrease and dissolved iron and manganese will increase with increasing proximity to the dissolved
contaminant center of mass. These measurements can potentially be used to complement TPH and BTEX
measurements in verifying the limits of dissolved plumes and the locations of dissolved contaminant center of
masses.
Shrinking dissolved contaminant plumes would be expected to show gradually increasing oxygen and nitrate
concentrations and decreasing dissolved (ferrous) iron and manganese concentrations around their peripheries
with time. In contrast, expanding dissolved contaminant plumes would be expected to show gradually
decreasing oxygen and nitrate concentrations and increasing dissolved (ferrous) iron and manganese concentra-
tions at their peripheries with time. These measurements can potentially be used to complement TPH and
BTEX measurements in determining whether source areas and dissolved plumes are shrinking, expanding, or at
steady state.
1.4
Measurement of selected redox parameters in ground water may have application in determining the feasibility
of utilizing specific in-situ and ex-situ remedial treatments and in evaluating the performance of certain
remedial systems once implemented. Technologies for remediating petroleum hydrocarbon impacted ground
water currently include ex-situ treatments such as pump-and-treat and in-situ treatments focused on enhanced
biodegradation or monitored natural attenuation. Technologies often considered for remediating residual phase
and non-aqueous phase petroleum hydrocarbon contamination in the subsurface currently include in-situ
surfactant/co-solvcnt flushing, in-situ thermal treatment, and in-situ oxidation. Examples of the potential
applications of redox parameter measurements in determining the feasibility and evaluating the performance of
remedial systems are presented below.
Dissolved oxygen, iron, nitrate, and sulfate concentrations can be used to indicate the presence/effect of natural
bioattenuation processes on the ambient hydrochemical system. High dissolved (ferrous) iron concentrations
62
-------
(e.g., > 1 ppm) coupled with low dissolved oxygen levels (e.g., < 1 ppm) near a source area or within a
dissolved plume would generally indicate active bioattenuation. In contrast, sustained elevated dissolved
oxygen levels and low dissolved (ferrous) iron levels near a source area or within a dissolved hydrocarbon
plume could indicate the absence of natural bioattenuation processes. Estimation of the long-term capacity of
an aquifer system to naturally attenuate petroleum hydrocarbons could be attempted by determining average
up-gradient electron acceptor fluxes into the impacted zone and quantifying the aquifer solid phase "bioavailable"
iron present within the impacted zone. The instantaneous BTEX biodegradation capacity of an aquifer system
can be approximated based on the differences in up-gradient dissolved electron acceptor concentrations relative
to dissolved electron acceptors and by-products near a petroleum hydrocarbon source area using the following
relationship (U.S. EPA, 1996).
Biodegradation capacity (mg/L) =
[(O2concentration up-gradient - O2 concentration in source area)/3.14)] +
[(NO3" concentration up-gradient - NO,~ concentration in source arca)/4.9)] +
L(SO42~ concentration up-gradient - SO,,2" concentration in source area)/4.7)J +
[Fe2" concentration in source area/21.8] +
[CH4 concentration in source area/0.78]
This relationship for BTEX biodegradation is the basis of EPA's BIOSCREEN model used as a screening tool
to assess the potential role of natural bioattenuation at petroleum hydrocarbon contaminated sites.
It should be noted that although solid-phase "bioavailable" iron may serve as a major long-term electron
acceptor source, it may become depleted with time. If the natural attenuation process is highly dependent on
the presence of solid-phase ferric iron as an electron acceptor source, the eventual depletion of the iron source
could result in further expansion of a hydrocarbon plume that might have initially been determined to be stable
or receding. Quantification of "'bioavailable" solid phase iron is addressed in the Session 6 Summary.
Redox parameters such as dissolved oxygen measurements can be used to determine zones of influence for in
situ bioenhancement treatments involving delivery of oxygen into an aquifer. Dissolved oxygen measurements
coupled with pre-treatment and post-treatment redox potential and dissolved (ferrous) iron measurements, for
example, can provide evidence of whether oxidizing conditions favorable for petroleum hydrocarbon biodeg-
radation are being established and maintained. In cases where a monitored natural attenuation process has been
implemented, redox parameter measurements (e.g., dissolved oxygen, dissolved (ferrous) iron) within and
around the periphery of the plume can be used as supporting evidence to determine whether the plume is stable
or is shrinking. This should be supplemented by solid phase ferric iron analysis to ensure adequate long-term
electron acceptor capacity/longevity within the impacted aquifer.
on
Some ex-situ ground-water treatment systems may be adversely impacted by the presence of iron, manganese.
and/or carbonates. This could be the case for any treatment system involving a significant change in redox
potential, pH, or temperature during treatment of the ground water. Vulnerable components of treatment
systems could include filters, air strippers, bioreactors, and re-injection systems. For example, an increase in
redox potential of influent water during air stripping as a result of contact with oxygen can induce precipitation
of dissolved (ferrous) iron and manganese, which may ultimately clog the stripper. Measurement of dissolved
(ferrous) iron, manganese, alkalinity, and dissolved oxygen in ground water supplemented by redox potential
measurements can provide essential information needed in the selection and proper design of ex-situ treatment
systems.
OH
Some in-situ treatment systems may be adversely impacted by various precipitation reactions or dissolution
reactions that occur in response to changes in redox potential and pH. For example, the hydraulic conductivity
of in-situ aquifer treatment zones that involve a change in redox potential or pH may decrease significantly over
time as a result of specific precipitation or mineralization reactions that occur. Air sparging systems, for
example, may induce precipitation of iron and manganese as a result of an increase in redox potential brought
about by the introduction of oxygen into the aquifer. Measurement of parameters (see Session 4 Summary)
63
-------
such as dissolved oxygen, dissolved (ferrous) iron, and manganese, and alkalinity in ground water can be used
in conjunction with geochemical modeling to predict the nature and extent of precipitation/dissolution
reactions that can be expected to occur during treatment.
1.5
The need for redox parameter measurements in the characterization and remediation of petroleum hydrocarbon
impacted aquifer systems will vary based on the site-specific questions that need to be answered. A summary'
of potential site characterization and performance monitoring redox parameter needs for a number of different
remediation technologies is presented in Table 1.1.
Table 1.1 Ground-water Redox Monitoring Guidance for Petroleum Hydrocarbon Site Characterization and
Remediation
Remediation Technology1
Ground-water extraction with
treatment and/or re-injection3
Monitored natural attenuation
(MNA)
Air sparging (mass removal in
source zone)9
In-situ aquifer treatment zones
(e.g., permeable reactive
barriers)12
Enhanced bioremediation -
oxygen addition
Enhanced bioremediation -
addition of other electron
acceptors
In-situ chemical oxidation
Surfactant flooding16
Solvent flooding16
Thermal methods (e.g., steam
injection)
Phytoremediation
Ground-water Redox Monitoring Requirements
Site Characterization2
Iron speciation4, manganese.
Additional parameters5: TSS,
hardness
Protocol parameters7 (e.g., ASTM
E-1943-98)
Variable10
Ferrous iron and manganese
Protocol parameters13
Additional parameters: solid
phase oxygen demand
Protocol parameters13
Protocol parameters
Additional parameters: TPH,
solid phase COD, and solid
phase iron mineralogy15
None
Other parameters: TDS, major
ions17
Variable19
None
Research needed20
Performance Monitoring
Variable6
Variable8
DO and ferrous iron11
Variable depending on specific
technology
DO, ferrous iron, TIC
TIC, others14
DO, ferrous iron, TIC
Additional parameters: vadose
zone CO2
Variable18
Variable
None
Research needed
Only saturated zone remedial technologies are considered. Surface soil and vadose zone remedial technologies (e.g., bioventing) are not
addressed.
Includes data needed for remediation technology assessment and/or design.
Characterization of ground-water geochemistry is important to the design of virtually any remedial system involving extraction of ground
water. Fouling of ground-water treatment and handling systems with iron precipitates, scale, and other solids is a significant design
consideration. Potentially \itlnerabie systems include filters, strippers, bioreactors, re-injection systems (e.g., wells, french drains, etc.) and
ancillary components (e.g., in-line monitors such as flow meters are particularly vulnerable).
Total iron loading is important, but it is also important to quantify- dissolved (ferrous) iron to assess how much iron precipitation can be
avoided by maintaining reduced conditions in the treatment system.
"Additional parameters " refers to important ground-water geochemical parameters not typically considered redox parameters.
Performance monitoring requirements will depend on system design, but will typically include analyses for one or more of the parameters
specified under site characterization. However, performance monitoring is typically performed on extracted ground water rather than the in
situ grouncl-waler svsiem.
64
-------
7 A variety of protocols describe the range of redox parameters that should be included in characterization of natural attenuation at petroleum
hydrocarbon sites. The first round of ground- water monitoring should typically involve monitoring for a relatively complete spectrum of redox
parameters. The suite of analytical parameters used in subsequent monitoring can be optimized based on the initial findings.
8 This should he assessed on a site-specific basis.
9 Air sparging can also be used to create a "treatment wall" (with stripping and-'or aerobic biodegradation providing the treatments) to
minimize volatile hydrocarbon plume migration. For this air sparging application, also refer to the monitoring requirements associated with
"in-situ aquifer treatment zones ".
10 Ij the intent oj the application is stripping oj volatile constituents of concern, no redox characterization is req^iired. If /he intent is to promote
aerobic biodegradation (for non-volatile constituents that will not be "stripped" from the ground water), it is prudent to confirm the premise
of oxygen-limited conditions through analyses for MNA protocol parameters.
11 Performance moniforing will (ypically include assessmeni. of fhe zone of sparging influence. This can be assessed through determination of
spatial and temporal changes in dissolved oxygen (DO) and ferrous iron (a decrease in ferrous iron can be used as a supplemental indicator
of aerobic conditions).
12 Includes any treatment technology relying on ground-water flow through a subsurface zone in which ground-water composition/conditions are
altered (e.g., continuous wall systems, funnel and gate systems, sparging barriers, ground-water circulation wells, etc.)
13 When careful "mapping " of naturally occurring biological activity is critical to system design, use of hydrogen analyses in site characteriza-
tion should be considered.
14 Performance monitoring requirements will be a function of the specific electron acceptor used. For example, performance monitoring for
nitrate addition will typically include nitrogen speciation. See Session 2 Summary for definition of total inorganic carbon (TIC).
IS TPff measurement is used in calculating the upper hfmndary on aqueous phase oxygen demand. COD - chemical oxygen demand.
16 Focuses on the data needs associated with the in situ aspects of the technology. 'This technology typically includes ground-water extraction,
treatment, and/or re-injection; also refer to the data needs specified for this category of remediation technologies.
17 Surfactant system design/effectiveness (e.g., inlerfdcial tension reduction, potential surfactant precipitation) is a function of ground-water
geochemistrv. TDS ~ total dissolved solids.
18 Driven by the chemistry of the specific surfactant system being used.
19 There may be no need for redox data to assess the feasibility or design of a solvent flushing system. However; MNA will typically be used as a
"polishing step " to deal with residual constituents oj concern and the solvent. Also, use oj biodegradable solvents may drastically lower the
ground- water redox potential.
20 Given the focus of this document on ground-water geochemistry, the phytoremediation application of interest is the use of trees or other
vegetation to sen'e as solar driven ground-water extraction systems. From a practical perspective, since this phytoremediation application is
most often implemented as an "incremental boost" to natural attenuation processes, the ground-water redox will already have been character-
ized as part of the natural attenuation assessment. However, research is needed to allow interpretation of the effects of ground-water redox
conditions on potential phytoremediation effectiveness.
References
American Society for Testing and Materials (ASTM). 1995. Standard guide for risk-based corrective action
applied at petroleum release sites. ASTM E 1739-95. West Conshohocken. Pennsylvania.
American Society for Testing and Materials (ASTM). 1998. Standard guide for remediation of ground water by
Natural Attenuation at petroleum release sites. ASTM E 1943-98. West Conshohocken. Pennsylvania.
Chapelle, E H. 1999. Bioremediation of petroleum hydrocarbon-contaminated ground water: The perspectives
of history and hydrology. Ground Water, v. 37. pp. 122-132.
U. S. Environmental Protection Agency. 1988. Cleanup of Releases from Petroleum USTs: Selected Technologies,
EPA/5 3 O/UST-8 8/001.
U. S. Environmental Protection Agency. 1996. BIOSCREEN - Natural Attenuation Decision Support System,
User's Manual Version 1.3, EPA/600/R-96/087.
65
-------
-------
2
for of
Frank Chapelle, John Wilson, and Ralph Ludwig
2.1
The purpose of this session was to develop recommendations concerning redox measurement parameters for
remediation of ground water contaminated with chlorinated solvents and MTBE. Information on redox
processes is necessary for site managers to plan and implement site characterization, select remedies, and
optimize and monitor the performance of the selected remedies. The participants in the session made
recommendations concerning the redox parameters that should be evaluated for various treatment technologies,
and on the appropriate methods to collect the data.
The technologies include various techniques for destruction of the contaminants, including physical removal.
chemical destruction, enhanced biological degradation, and monitored natural attenuation. The information
needs for each technology are discussed within these categories.
2.2 for the and of
for
The information needed to characterize a site and select a remedy is often different from information needed to
optimize a technology or monitor its performance. The tables in this section present a brief description of the
major issues associated with redox parameters, grouped by major category of technology.
The "Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in Ground Water"
(Wiedemeier et al. 1998) presents a comprehensive list of redox parameters that are useful for evaluating
monitored natural attenuation in ground water. The parameters are listed in Table 2.1 of the Technical
Protocol, and include alkalinity, arsenic, chloride, conductivity, ferrous iron, hydrogen, manganese, methane.
nitrate, oxidation reduction potential (ORP), oxygen, pH, sulfate, sulfide, temperature, and total organic carbon
in water.
Some of the parameters are useful for assessment of the system redox status, including dissolved oxygen.
hydrogen, methane, sulfide, oxidation reduction potential, pH, and temperature. Some of the parameters are
useful for assessment of the potential for microbial metabolism of contaminants. Oxygen can support
microbial metabolism of certain chlorinated hydrocarbons including cis-dichloroethylene and vinyl chloride.
In contrast, the presence of sulfate, nitrate, or oxygen can inhibit anaerobic biological processes such as
reductive dcchlorination of the chlorinated hydrocarbons. During reductive dechlorination, the contaminant
serves as the electron acceptor in microbial metabolism. If sulfate, nitrate or oxygen are present they can
compete with the contaminant for supplies of electron donors. Total organic carbon in water is a bulk
parameter that can be used to estimate the supply of electron donors for reductive dechlorination.
Other redox parameters in the EPA Technical Protocol are measures of the degradation process. These include
chloride and alkalinity. Chloride is an ultimate biodegradation product of chlorinated hydrocarbons. The
analysis for alkalinity in the Technical Protocol was originally intended as an estimate of the amount of carbon
dioxide that has been produced in the ground water by biological processes. In many aquifers, carbon dioxide
produced by biological processes cannot be measured directly due to reactions with carbonate minerals in the
aquifer matrix. Alkalinity measurements may also provide an overestimate of microbial CO2 production in
contaminant plumes with high concentrations of titrable organic acids. To address this later issue, a better
estimate of the amount of carbon dioxide that has been produced is the change in total inorganic carbon (TIC)
in ground water. The TIC can be measured directly, or calculated from measurements of alkalinity, pH, and
dissolved carbon dioxide (see section 2.8).
67
-------
Certain of the redox parameters can be used to establish the "footprint" of the contaminated water (National
Research Council, 2000). If the geochemical "footprints" are similar between contaminated ground water in
source areas and water produced from monitoring w7ells farther downgradient, they can be used to establish a
hydrological connection between the source and the monitoring well. If the geochemical "footprints" in down
gradient monitoring wells are distinctly different from the geochemistry of ground water at the source, they are
not in the flow path from the source area of the plume. Geochemical "footprints" can also reveal the major
biological and chemical oxidation/reduction processes that are responsible for removing chlorinated hydrocar-
bons from ground water. Chloride, methane, specific conductance, and TIC are particularly useful as
"footprints."
Some of the redox parameters are required to evaluate risk from materials that were not the original
contaminants of concern. Arsenic and manganese are included because the geochemical conditions that allow
natural biodcgradation of chlorinated hydrocarbons also allow biological reduction of natural aquifer materials
that contain arsenic and manganese. Under conditions that promote natural anaerobic biodegradation of
chlorinated hydrocarbons, the concentration of total arsenic and manganese produced from natural sources may
exceed drinking water standards.
The appropriate regulatory authorities will determine on a site specific basis when it is necessary to analyze for
arsenic and manganese. The circumstances where analysis for hydrogen is useful are discussed in Section 2.7.
The following redox parameters should be routinely applicable to most sites, and will be referred to as the
Routine EPA Protocol Parameters: dissolved oxygen, nitrate, sulfate, chloride, ferrous iron, sulfide, hydrogen,
methane, oxidation/reduction potential, pH, temperature, specific conductance, total inorganic carbon, and
total organic carbon.
Chlorinated hydrocarbons can be removed from aquifers by air sparging, bioventing or phytoremediation (see
Table 2.1). They can be chemically destroyed by dithionite reduction, oxidation by the Fenton reaction.
oxidation by permanganate, or by reduction by zero valence iron in a permeable reactive barrier (see Table 2.2).
They can be biologically degraded through substrate additions, phytoremediation in the rhizosphcre, and
through monitored natural attenuation (see Table 2.3).
Air sparging is a process of stripping volatile compounds from ground water through the injection of air.
Redox chemistry is not involved in this purely physical process. In theory precipitation of Fe (III) by oxidation
of Fe (II) in ground water could plug the pore space in the aquifer. In practice this is not an important problem.
The success of air sparging is controlled by the physics of flow of the injected air as controlled by the geology
of the site.
Bioventing is technology applied to the unsaturated zone that is intended to treat ground water. It treats the
residual hydrocarbons present as a NAPL mat act as a long term source of contamination to ground water. It
is successful at sites where the water table moves down during the time period of treatment to expose the oily
phase at residual saturation to air in the unsaturated zone. Treatment is a combination of volatilization and in
situ biodegradation. The recommended redox parameters are important to understand the biodegradation
processes.
Phytoremediation has particular application to chlorinated hydrocarbons in shallow7 ground water in fine
textured geological materials. Water is removed by evapotranspiration. Hie chlorinated hydrocarbons can be
transpired with the water, or they can volatilize into air spaces created when the water is extracted. The relevant
geochemical parameters determine whether plant roots can grow in the contaminated ground water (Pivetz.
2001).
In conventional practice for site characterization, the major emphasis is on sampling and analysis of ground
water. Much less attention has been paid to characterization of aquifer solids. An analysis of aquifer solids is
critical to the successful use of non-biological transformation technologies such as dithionite reduction.
permanganate oxidation or the Fenton reaction to treat organic contaminants. The natural aquifer solids play an
important role in the chemistry of the treatment process. The availability and reactivity of iron minerals in the
aquifer are particularly important.
This is less true for passive reactive barriers using zero-valent iron in the treatment process. Only ground water
is treated, and it is treated in the reactive wall as opposed to the aquifer. Passive reactive barriers are designed
to allow adequate residence time of contaminated ground water to reduce the contaminant concentrations to the
design goals. The design is constrained by the kinetics of reaction of the contaminants of concern with the
reactive medium in the barrier, and with the seepage velocity of ground water through the barrier. In theory,
68
-------
the geochem istry of the ground water should directly affect the long term capacity of a barrier to treat water. As
ail example, reaction with oxygen in ground water converts zero valent iron to higher oxidation states. The
reaction of zero valence iron with oxygen and contaminants raises the pH. A variety of reaction products can
form precipitates that may plug the pore space. These concerns are discussed in more detail in Powell et al.
(1998) on pages 25 and 26. At present there is a research program coordinated between private industry and
several federal agencies to evaluate the life cycle of passive reactive barriers. However, the present state of
practice is to design them to work well, but not necessarily to design them to last for a predetermined time. As
a consequence, geochemical considerations are less important to their design and implementation.
In technologies that add substrates (such as vegetable oil. lactate. molasses, or hydrogen) to support reductive
dechlorination of chlorinated hydrocarbons, the success of the technology may be related to the availability of
mineral nutrients for growth of bacteria, in particular phosphate and reduced fixed nitrogen in the form of
ammonium ion. The primary source of reducing power is hydrogen, which is provided directly as dihydrogcn,
or as organic substrates that can be fermented to form dihydrogen. Organisms that use nitrate and nitrite as
electron acceptors can effectively compete with organisms using chlorinated hydrocarbons for the available
pool of dihydrogen. The presence of measurable nitrate or nitrite will preclude or strongly inhibit reductive
dechlorination.
Chlorinated hydrocarbons can be biochemically degraded in soil or aquifer material in direct proximity to plant
roots (the rhizosphere). The oxidation parameters are used to predict microbiological activity in the
rhizosphere. In this application, total organic carbon is an indirect measure of plant root exudates that stimulate
microbial activity against the chlorinated hydrocarbons.
Monitored Natural Attenuation is an important remedial technique for chlorinated hydrocarbons. The EPA
protocol parameters have proved themselves useful as a qualitative tool to recognize geochemical environ-
ments where reductive dechlorination is possible (Wiedemeier et al., 1998). However, the role of alkalinity in
the original EPA protocol should be replaced with an analysis or calculation of TIC (see section 2.8).
Table 2.1 Rcdox Parameters for Evaluation of Treatment by Physical Removal
Treatment Technology
Air Sparging
Bioventing to remove
oily phase residuals of
chlorinated
hydrocarbons mixed
with petroleum
hydrocarbons.
Phytoremediation by
evapo-transpiration of
water.
Site Characterization
None
Concentrations of oxygen and carbon
dioxide in soil gas. Concentrations of
methane in soil gas to define background
conditions. An in situ respiration test to
measure the instantaneous rate of oxygen
consumption.
Will plant roots grow in the contaminated
ground water?
Conductivity, pH, Dissolved Oxygen,
Sulfide.
Performance Monitoring/
Remediation Optimization
None
Same
None
-------
Table 2.2 Redox Parameters for Evaluation of Chemical Destruction
Treatment Technology
Site Characterization
Performance Monitoring/
Remediation Optimization
Dithionite
Inherent stability of Dithionite in the
aquifer matrix. Dissolved Fe(III)
extractable with 1 N HC1. Mineralogy of
Iron.
ORP, reducing capacity of
sediments measured as the
demand for dissolved oxygen.
Other Oxidation
Processes
Fenton's reaction
Permanganate
Routine EPA Protocol Parameters,
Estimate volume of aquifer material in
source area contaminated with organic
materials. Permanganate demand of
contaminated material.
Routine EPA Protocol
Parameters, carbon dioxide in
soil gas, Residual permanganate
using colorimetric test.
Zero Valent Iron
Not Critical
Dissolved Fe (II), dissolved
oxygen, sulfide, sulfate, ORP,
conductivity, pH
(measured either quarterly or
semi-annually)
Table 2.3 Redox Parameters for Evaluation of Biological Degradation
Treatment Technology
Site Characterization
Performance Monitoring/
Remediation Optimization
Substrate Additions
Vegetable Oil
Lactate
Molasses
Hydrogen
Routine EPA Protocol Parameters
hydrogen, nitrite, ammonium ion,
phosphate.
(Have the redox processes
changed as a result of the
substrate addition?)
Routine EPA Protocol
Parameters, acetate, formate,
propionate, hydrogen, nitrite,
ammonium ion, phosphate.
Phytoremediation
through
biodegradation in the
Rhizosphere
Routine EPA Protocol Parameters with
emphasis on Total Organic Carbon.
Routine EPA Protocol
Parameters with Emphasis on
change in oxidation/reduction
status.
Monitored Natural
Attenuation of
Chlorinated Ethenes
Chlorinated Ethanes
Chlorinated Benzenes
Routine EPA Protocol Parameters plus
Potential Daughter Products. Hydrogen if
necessary to interpret biodegradation
processes.
Routine EPA Protocol
Parameters plus Potential
Daughter Products. Hydrogen
if necessary to confirm
continued biological activity.
70
-------
2.3 for
Mcthyl-/e'/'/-butyl-cthcr (MTBE), an oxygenate added to petroleum hydrocarbon fuels to produce cleaner
engine emissions, has been increasingly detected in ground water throughout the United States since the early
1990s. The increased presence of MTBE in ground water has been of particular concern due to the high
mobility, relative recalcitrance, and suspected carcinogenic properties of this compound.
Although initially believed to undergo negligible or no biodegradation in the subsurface, there is an increasing
body of evidence indicating mat aerobic biodegradation can be used to remove MTBE from ground water.
Salinitro et al. (2000) used bioaugmentation and aerobic in situ bioremediation to reduce MTBE in an aquifer
to concentrations below 10 jig/liter. Redox parameters that are important for design of in-situ aerobic MTBE
bioremediation systems include dissolved oxygen, sulfide and ferrous iron in ground water. Oxygen
enhancement will stimulate biodegradation of MTBE if the original ambient oxygen concentration is less than
1 mg/liter. The presence of high concentrations of ferrous iron can cause problems with precipitates and
plugging of pipes and distribution systems. The presence of ferrous iron or sulfide in ground water is often
indicative of a high non-biological demand for oxygen associated with precipitates of iron and sulfide in the
aquifer solids. Redox parameters important for monitoring in-situ aerobic bioremediation include dissolved
oxygen, ferrous iron, and TIC. Dissolved oxygen concentrations should exceed a minimum concentration of
2 mg/liter to support degradation of MTBE, and should exceed 5 mg/liter for optimum performance. The
increase in TIC over time can be used to monitor oxidation of the organic contaminants in the aquifer. In
particular, redox parameters will be useful in monitoring the performance (e.g., zone of influence) of any
in-situ enhanced bioremediation treatment system.
Bradley et al. (2001) have reported significant aerobic mineralization of MTBE in microcosm studies using
MTBE and non-MTBE contaminated stream and lake bed sediments from eleven sites throughout the
United States. The study indicated that even microbial communities indigenous to newly contaminated
sediments exhibited some innate ability to mineralize MTBE under aerobic conditions. The magnitude of
MTBE mineralization in the study was observed to vary inversely with the sediment grain size distribution.
Landmeyer et al. (2001) showed that aerobic metabolism of MTBE in the bed sediments of a stream prevented
transfer of MTBE from an anaerobic plume in ground water to aerobic surface water. Natural aerobic
biodegradation of MTBE in surface waters and sediments probably make a substantial contribution to surface
water quality.
Natural aerobic biodegradation of MTBE is probably not important for plumes of MTBE contamination in
ground water. Kolhatkar et al. (2000) examined seventy-four MTBE plumes from spills of gasoline in the
eastern United States. The ground water that contained MTBE contamination was strongly reducing. The most
contaminated wells at forty-three of the seventy-four sites had no detectable sulfate and an accumulation of at
least 0.5 mg/liter of methane. Only eighteen of the seventy-four sites had detectable concentrations of sulfate
and methane concentrations less than 0.5 mg/liter. Nitrate was depleted in contaminated wells at every site. In
the survey reported by Kolhatkar et al. (2000), water samples were collected according to the standard practice
for sampling monitoring wells at underground storage tanks as approved by the local regulatory authority. In
almost every case, the water samples were collected with bailers (see warnings in section 2.6). The water as
sampled from the contaminated wells contained oxygen, even though the same water contained substantial
concentrations of ferrous iron, sulfide, or methane. When water samples are collected with bailers, it is almost
impossible to prevent contamination of the sample with oxygen. The widespread use of bailers for sampling
wells at UST spills has created a mistaken impression at some MTBE spills mat the ground water is aerobic,
which can lead to a mistaken conclusion that aerobic natural degradation of MTBE in that particular aquifer is
possible.
Other studies have indicated some biodegradation of MTBE under oxygen limiting conditions including
methanogenic conditions (Wilson et al., 2000), iron reducing conditions, or nitrate reducing conditions
(Bradley et al., 2001). However, it is unlikely that monitored natural attenuation will be an acceptable solution
for MTBE in ground water at many sites. Kolhatkar et al. (2000) applied an objective test for natural
attenuation of MTBE to the seventy-four stations they sampled in the eastern United States. Based on one
round of sampling from existing monitoring wells, natural attenuation of MTBE was shown to be statistically
significant at the 90% confidence level at only four of the seventy-four stations.
In many cases, an ex-situ approach may be preferred for addressing MTBE in ground water that is produced
and distributed for water supply (Keller et al., 2000). As with any potential ex-situ ground-water treatment
71
-------
Table 2.4 Appropriate Field and Laboratory Methods for Critical Redox Parameters
Parameter
Dissolved
Oxygen
Ferrous Iron
Fe (II)
Manganese
Total
Inorganic
Carbon
(TIC)
Sulfide
Methane
Oxidation/
Reduction
Potential
Specific
Conductance
pH
Temperature
Hydrogen
Nitrate
Sulfate
Chloride
Field Method
Colorimetric methods are
recommended. Electrodes can be used
successfully in a flow-through or
overflow cell. See discussion in
section 2.4 below and Session 4
Summary.
Colorimetric analysis, use of field
spectrophotometer is preferred. (See
Session 4 Summary)
Colorimetric analysis, use of field
spectrophotometer is preferred.
Determine alkalinity (endpoint
titration to pH 4.5) and pH. For water
with pH < 6.5, also use a Colorimetric
field kit for free carbon dioxide.
Calculate TIC from the sum of
carbonate and bicarbonate alkalinity
plus free carbon dioxide. (See section
2.8)
Colorimetric analysis, use of field
spectrophotometer is preferred.
Preserve in the field, ship to the
laboratory for analysis. (See Session 5
Summary)
Platinum electrode in a flow- through
cell or an overflow cell. Readings
should be corrected to the standard
hydrogen electrode. (See Session 4
Summary)
Electrode in a flow-through cell or an
overflow cell. The electrode may not
automatically correct for temperature;
measure and report temperature.
Electrode in a flow-through cell or an
overflow cell. The electrode should
be calibrated in the field prior to use.
Thermometer in a flow-through cell or
an overflow cell.
Collect using bubble Stripping
Technique. Analysis in a field
laboratory is preferred, using gas
chromatography with a reducing gas
detector. (See Session 5 Summary)
Not recommended. Adequate field
methods exist for these parameters;
however, they are not cost-effective
compared to analysis in a laboratory.
Laboratory Method
Analysis on samples shipped to the laboratory is
not recommended.
Analysis on samples shipped to the laboratory is
not recommended.
Analysis on samples shipped to the laboratory is
not recommended.
The samples should be shipped without
preservative, and sealed into the container
without headspace. Laboratory analysis by
purging TIC from an acidified sample and
measuring CO2 by infrared absorption.
Analysis on samples shipped to the laboratory is
not recommended. If analysis on samples
shipped to the laboratory is necessary, the
samples should be shipped with an appropriate
preservative, such as zinc acetate.
Headspace analysis using gas chromatography
with a flame ionization detector (GC/FID)
Analysis on samples shipped to the laboratory is
not recommended.
Electrode in a water sample that has not been
preserved. The samples should be sealed into the
container without headspace to prevent the
oxidation and precipitation of iron.
Electrode in a water sample that has not been
preserved. The samples should be sealed into the
container without headspace.
Not Possible.
Analysis using gas chromatography with a
reducing gas detector. A flame ionization
detector is not sufficiently sensitive to hydrogen.
Ion Chromatography. The following preservation
and holding times are recommended: nitrate, 4 °C
with 48 hour holding time; sulfate, 4 °C with 28
day holding time; chloride, ambient temperature
with 28 day holding time (see U.S. EPA, 1999).
72
-------
process, the geochemistry of the ground water (e.g., iron and manganese content) can be an important factor in
determining which ex situ treatment process is likely to be most cost-effective. Low redox ground waters
containing high levels of dissolved iron and manganese may pose problems (e.g., precipitate clogging or
coating) for treatment systems in which increases in redox potential occur during treatment. In such cases, a
pre-treatment or modified treatment process may be required to achieve effective overall treatment. Thus, in
cases where an ex-situ treatment process is under consideration, the measurement of redox parameters such as
dissolved iron/manganese may be beneficial in evaluating the feasibility of using selected above ground
treatment technologies.
2.4 of Processes to Phytoremedlation
In the root zone (rhizospherc) of plants, especially trees, it has been shown that cxudates from roots supply
organic carbon, especially organic acids, that stimulate bacterial growth (Lee et al., 2000; Schnoor et al., 1995).
In addition, the decay of root biomass can also add organic carbon to the vadose zone and shallow ground
water. Where near-reducing conditions have been established in ground water near phreatophytes such as
cottonwood trees, reductive dcchlorination of TCE has been observed in the rhizosphere (Lee et al., 2000).
Alternatively, some plants can provide some oxygen to roots, or create avenues for atmospheric oxygen to enter
ground water or the deep vadose zone through root holes, perhaps enhancing the aerobic environment in
shallow ground water.
Thus, biodegradative processes can occur in the rhizosphere of plants, especially phreatophytes in addition to
the well known removal of contaminated ground water by transpiration and degradation of the contaminants
within the plant biomass (Schnoor et al.. 1995).
2.5 for
Many redox parameters are unstable; they start to change as soon as the ground water is exposed to the
atmosphere (see Session 4 Summary and Appendix A). As a consequence, many of these redox analyses must
be done at the well head. The following recommendations are based on the combined experience of the
participants at the session. They represent a group consensus on methods and techniques that can be expected
to provide data of adequate quality at hazardous waste sites.
2.6 Considerations for Ground-Water
The method used to collect a sample of ground water is often more important for data quality than the method
used to analyze the sample. Recommendations for improved practice are discussed in detail in Puls and
Barcelona (1996). If ground water is exposed to the atmosphere while it is being collected into the sampling
well, or when it is pumped from the well, the sample can be seriously compromised. The ground water may
gain oxygen from the atmosphere. The oxygen introduced into the sample may chemically react with ferrous
iron or sulfide and transform these materials. Oxygen will also disturb measurements of oxidation reduction
potential.
The standing water in a well is exposed to the atmosphere; it is usually oxygenated with respect to the native
ground water in the aquifer. In most applications, the standing water within the well is purged and replaced
with ground water that has not been altered. To avoid aeration, water should be pumped from the well using a
gentle device such as a peristaltic pump or bladder pump. Bailers should only be used when there is no other
option for geochemical sampling. If a bailer must be used, the bailer should be slowly immersed in the standing
column of water in the w7ell to minimize aeration. After sample collection, the water should be drained from the
bottom of the bailer through tubing into the sampling container.
Ground water can also be aerated if the well is pumped down too far. The ground water will enter the well, then
cascade down the inside of the screen as the water level in the well drops. To avoid aeration, wells that are
screened below the water table should not be pumped at a rate which lowers the standing water in the well to
a level below the top of the screen. If practical, wells screened across the water table should be pumped at a rate
that lowers the total height of the water column in the well no more than five to ten percent.
An overflow cell or a flow-through cell should be used for the measurement of well-head parameters such as
pH, temperature, specific conductance, dissolved oxygen, and oxidation-reduction potential. Experts in the
area of field sampling of ground water differ in their preference for flow-through cells or overflow cells. Flow-
through cells effectively prevent contact of a water sample with oxygen, particularly when they don't trap a
bubble in the cell. Depending on the position of the outlet tube and flow restrictions in the outlet tube, water in
-------
the flow-through cell can be under positive pressure. This will preclude the proper use of certain electrodes that
depend on gravity flow of electrolyte to a salt bridge in the electrode. The user s manual for an electrode should
be consulted to determine if this may be a problem. Overflow cells can function as w7ell as flow-through cells
when the flow from the pump is presented directly to the sensor surface of the electrodes, and then flows up past
the body of the electrode before it overflows. The position of an electrode can be adjusted in the overflow cell
to make the elevation of the reservoir of electrolyte higher than the overflow of water being monitored.
The "Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in Ground Water"
(Wicdemeir et al, 1998, EPA/600/R-98/128) depicts a flow-through cell on page A4-17. The authors of the
Technical Protocol used an Erlenmeyer flask as an overflow cell because the wide base of the flask kept the
leads to the electrodes from tipping the flask over. However, this configuration provides a reservoir for mixing
of the ground water in the flask and does not provide directed flow of ground water across the sensor surface of
the electrodes. The overflow cell should be a cylinder that is only slightly larger than the space needed to
contain the electrodes. A graduated cylinder works well. It can be held in place with clamps or a lead
doughnut.
2.7 Wfcen to In Ground
Measurement of hydrogen in ground water should be considered when the redox processes cannot be
interpreted from the more conventional Protocol Parameters. As an example, a trichlorethylene plume may
show a marked reduction in the concentration of trichloroethylene dow7ngradient of the source, with accumula-
tion of minor concentrations of the first reduction transformation product, cis-dichloroethylene, and no
detectable vinyl chloride. This distribution can be easily interpreted in two ways. Reductive dechlorination
may be weak, resulting in transformation of a minor amount of trichloroethylene to cis-dichloroetheylene. The
reduction in concentrations of trichloroethylene are caused by physical processes such as dilution and
dispersion. In the second interpretation, reductive dechlorination is strong, and is the cause of the loss of
trichloroethylene. The transformation product cis-dichloroethylene does not accumulate because it is further
transformed to carbon dioxide, without going through vinyl chloride as an intermediate. If concentrations of
hydrogen are low, less than 1 nanomolar, the second interpretation is excluded and the first interpretation is
supported. If hydrogen concentrations arc high, greater than 1 nanomolar, the second interpretation is
supported, although not necessarily proven. Hydrogen concentrations should be measured when they are
necessary to understand the redox processes. When the redox processes can be interpreted from the more
conventional parameters, measurement of hydrogen is not necessary. A recent publication compares two
techniques for sampling hydrogen (Mclnnes and Kampbcll, 2000).
2.8 or Calculation of Total Inorganic Carbon (TIC)
Microorganisms degrade chlorinated hydrocarbons through a variety of mechanisms including aerobic
respiration, iron and manganese reduction, or use as the electron acceptor in reductive dechlorination. MTBE
has been shown to be degraded under aerobic, nitrate reducing, iron reducing, sulfate reducing and methanogenic
conditions. The most universal end product of biodegradation of chlorinated hydrocarbons or MTBE is carbon
dioxide that is produced from the contaminant itself, or from the compound used as an electron donor in
reductive dechlorination. This biogenic carbon dioxide can be seen in plumes of contaminated ground water as
an increase in the TIC above the background TIC in uncontaminated water in the aquifer. It is often one of the
most revealing "footprints" of ground water that has been remediated by natural biological processes.
This geochemical parameter is termed TIC because it is measured on water samples that have not been filtered.
In most cases the measured TIC is equivalent to the dissolved inorganic carbon (DIC) in the ground water. It
is best not to filter ground water samples before analysis for inorganic carbon. Often ground water is not in
equilibrium with the atmosphere, and it is difficult to avoid exchange of carbon dioxide between the water
samples and the atmosphere during filtering.
The concentration of TIC can be determined directly by adding acid to the water sample, partitioning carbon
dioxide to a head space and measuring the carbon dioxide directly with an infrared detector (see methods for
total inorganic carbon in U.S. EPA, 1999). This is the preferred method. The concentration of TIC can also be
calculated from measurements of alkalinity, free carbon dioxide, and pH as described later in this section.
Methods for alkalinity are described in U.S. EPA (1999), and for free carbon dioxide in APHA (1999) and
ASTM (1996).
These calculations are generally accurate; however, there are important categories of plumes where these
calculations can produce major errors. The calculations assume that there is no hydroxide alkalinity associated
74
-------
with strong bases. Occasionally an unnaturally high pH is caused by disposal of caustic waste materials in the
subsurface. A high pH can also be caused by improper grouting or cementing of monitoring wells. The
calculations also assume that there is no alkalinity associated with other weak acids. Ground waters often
contain organic acids that contribute to alkalinity (Hemond. 1990). Plumes of leachate from solid municipal
waste landfills often have large amounts of alkalinity associated with organic acids such as acetate and
butyrate.
The correspondence between carbon dioxide produced during metabolism and the increase in TIC is not one-
for-one. In acid ground water, some of the carbon dioxide may be lost across the water table. This effect is
probably not important in most plumes, but it should be considered and evaluated. In aquifers that contain
carbonate minerals in the aquifer matrix, the biogenic carbon dioxide can ultimately react with calcium or
magnesium carbonate minerals to dissolve them as calcium or magnesium hydrogen carbonate. In this reaction
there are two moles of TIC produced in ground water for each mole of biogenic carbon dioxide consumed.
Once formed. TIC can precipitate with ferrous iron as siderite, or sorb to the anion exchange complex. In many
ground waters these reactions do not appear to be in equilibrium. As a consequence, it is usually impossible
to directly convert a change in TIC to an amount of carbon dioxide produced by biological activity. However.
it is often possible to calculate a range of carbon dioxide concentrations that could have produced the measured
TIC.
There are two calculations that can be used to estimate the total inorganic carbon (TIC) from pH. alkalinity,
and free carbon dioxide.
TIC from CO?
Dissolved inorganic carbon is the sum of dissolved carbon dioxide, carbonic acid, bicarbonate, and carbonate.
For convenience we will express the sum of dissolved carbon dioxide and carbonic acid as carbonic acid
[H2C03*].
LTICJ = [H2C03*J + LHC03-'J + LCO/J
Alkalinity is conventionally determined as the amount of acid required to titrate water to a colorimetric end
point near pH 4.8. It is conventionally expressed as the amount of CaCO3 that would be consumed by that
amount of acid. The meq/liter of acid required is the alkalinity in nig/liter of CaCO3 divided by 50.
Alkalinity as [H+] consumed = [HCO,;1] + 2[CO3'2]
Each 100 mg/liter of CaCO3 alkalinity accepts 2 meq/liter of FT.
CaCO3 alkalinity (mg/liter)/50 = 0.5 [HCO;1] + [CO3-2]
Free carbon dioxide is calculated from the amount of base needed to convert all the H2CO3* and HCO^1 to
CO3"2. It is usually expressed in mg/liter of CO2.
Free carbon dioxide as [FT] provided = 0.5 [HCO;1] + [H,CO3*]
Each 44 mg/liter of Carbon Dioxide provides 2 meq/liter of FT
Free Carbon Dioxide (mg/liter)/ 22 = 0.5 [HCO,-1] + [H2CO31
CaCO3 alkalinity/50 + free CO2122 = LH,CO3*J + [HCO,-1] + [CO3'2J = LTICJ
TIC (mg Carbon/liter) = 12 *(CaCO3 alkalinity/50 + Free CO2/22)
This method should only be applied to waters with pH > 6.5 and easily measurable CaCO, alkalinity.
[TIC] = [H2CO3*] + [HCO;1] + [CO3-2]
15
-------
The calculation proceeds in three steps:
(1) Calculate [HCO;1] from measured alkalinity and pH.
(2) Then calculate [H,CO3*] and [CO;2] from the pH, the [HCO;1] as calculated previously, and the
governing pKa.
(3) Tlien add the calculated [H2CO,"] + [HCO;1] + [CO;2] and multiply by 12,000 to convert to TIC
(mg C/liter).
Alkalinity is conventionally determined as the amount of acid required to titrate water to a colorimetric end
point near pH 4.8. It is conventionally expressed as the amount of CaCO3 mat would be consumed by that
amount of acid. The meq/liter of acid required is the alkalinity in mg/liter of CaCO3 divided by 50.
CaCO3 alkalinity as [FT] consumed = [HCO./1] + 2[CO;2]
CaCO3 alkalinity (mg/liter)/50 = [HCO;1]" + 2[CO;2]"
The ratio of [HCO;1] and [CO;2] can be predicted from the pKa of [HCO;1] and the pH.
The pKa of [HCO;1] to [CO;2] and [ff] is 10.49 at 10 °C.
In other words: ([H+][CO;2] )/[HCO;]] = 10 -10-49 and [CO;2] = [HCO;1]/ lO'10-49-^
Substituting [HCO;1]/ .W10-49*1 for [CO;2] in equation (1) yields
CaCO3 alkalinity (mg/liter)/50 = [HCO;1] +2([HCO;']/Wia49-PH-1).
Solving for [HCO;1] yields
[HCO;1] = (mg/liter CaCO3/50)* W10-49^ / (2+ lO'10-49^).
Again the pKa of [HCO;1] to [CO;2] and [H+] is 10.49 at 10 °C.
In other words: ([H+][CO;2] XfHCO;1] = 10 -10-49 and [CO;2] = [HCO,;1] * 10 -I"-«+PH
Hie pKa of [H2CO3*] to [HCO;1] and [H^ is 6.42 at 10 °C.
In other words: ([H+][HCO;'] )/[H2CO3*] = 10 ~6-42 and [H2CO,"] = [HCO;1]/ 10 -&42 + PH
Reference!
American Public Health Association, American Water Works Association,Water Environment Federation (APHA).
1999. Standard Methods for the Examination of Miter and Wastewater. Titrimetric method for free carbon
dioxide 4500- CO,, C, American Public Health Association, Washington, B.C.
American Society for Testing and Materials (ASTM). 1.996. Standard Test Methods for Total and Dissolved
Carbon Dioxide in Water, ASTM D 513-92. Philadelphia, Pennsylvania.
Bradley, P. M., Landmcycr, J. E., and Chapcllc, F. H. 2001. Widespread potential for microbial MTBE degradation
in surface-water sediments. Environmental Science and Technology, v. 35, pp. 658-662.
Hemond. H. F. 1990. Acid neutralizing capacity, alkalinity, and acid-base status of natural waters containing
organic acids. Envimnmental Science and Technology, v. 24. p. 1486-1489.
Keller, A. A., Sandall, O. C., Rinker, R. G., Mitani, M. M., Bierwagen, B., and Snodgrass, M. J. 2000. Evaluation
of physicochemical treatment technologies for water contaminated with MTBE. Ground Water Monitoring
and Remediation, v. 20, pp. 114-126.
Kolhatkar, R., Wilson, J., and Dunlap, L. 2000. Evaluating natural attenuation of MTBE at multiple UST Sites.
16
-------
In Proceedings ofNGWA/API Conference and Exposition on Petroleum and Organic Chemicals in Ground
Water: Prevention, Detection and Remediation. Anaheim, California, November 15-17, 2000. pp. 32-49.
Landmeyer, J. E., Chapeile. F. FL, Herlong, H. FL, and Bradley, P. M. 2001. Methyl /£?/r-butyl ether biodegradation
by indigenous aquifer microorganisms under natural and artificial oxic conditions. Environmental Science
amdTechnology, v. 35, pp. 1118-1126.
Lee, R. W., Jones, S. A., Kuniansky, E. J., Harvey, G., Sherwood-Lollar, B, and Slater, G. F. 2000. Phreatophyte
influence on reductive dechlorination in a shallow aquifer contaminated with trichloroethene (TCE).
InternationalJournalofPhytoremediation, v. 2, pp. 193-211.
Mclnnes, D. M. and Kampbell, D. 2000. The bubble stripping method for determining dissolved hydrogen (H2)
in well water. Field and Analytical Chemistry and Technology, v. 4, pp. 283-296.
National Research Council. 2000. Natural Attenuation for Groundwater Remediation, National Academy Press.
p. 11.
Pivetz, B. 2001. Phytoremediation of Contaminated Soil and Ground Wafer at Hazardous Waste Sites, EPA/540/
S-01/500. U. S. Environmental Protection Agency, Cincinnati, OH. [http://www.epa.gov/ada/download/
issue/epa_540_s01_500.pdf|.
Powell, R. M., Blowes, D. W., Gilham, R. W., Schult, T, Puls, R. W., Vogan, J. L., Powell, P. D., and Landis, R.
1998. Permeable Reactive Barrier Technologies for Contaminant Remediation, EPA/600/R-98/125.
U. S. Environmental Protection Agency, Cincinnati. OH. [http://www.epa.gov/ada/download/reports/
reactbar.pdf].
Puls, R .W. and Barcelona, M. J. 1996. Low-Flow (Minimal Drawdown) Ground-Water Sampling Procedures,
EPA/540/S-95/504. U. S. Environmental Protection Agency, Cincinnati, OH. [http://www.epa.gov/ada/
download/issuc/lwflw2a.pdf].
Salanitro. J. P., Johnson, P.C., Spinnlcr, G. E., Mancr, P. M., Wisnicwski, H. L., and Bruce, C. 2000. Field-scale
demonstration of enhanced MTBE bioremediation through aquifer bioaugmentation and oxygenation.
Environmental Science and Technology, v. 34, pp. 4152-4162.
Schnoor, J. L., Licht, L. A., McCutcheon, S. C., Wolfe, N. L., and Carreira. L. H. 1995. Phytoremediation of
organic and nutrient contaminants. Environmental Science and Technology, v. 29, pp. 318-323A.
U. S. Environmental Protection Agency. 1.999. EPA Methods and Guidance for Analysis of Water, June 1999.
Office of Water, Washington, D.C, Version 2.0, EPA 821-C-99-004.
Wiedemeier, T. H., Swanson, M. A., Moutoux, D. E., Gordon, E. K., Wilson, J. T., Wilson, B. H., Kampbell, D. H.,
Haas, P. E., Miller. R. N., Hansen, J. E.. and Chapeile, F. H. 1998. Technical Protocol for Evaluating Natural
Attenuation of Chlorinated Solvents in Ground Water. EPA/600/R-98/128. U. S. Environmental Protection
Agency, Cincinnati. OH. [http://www.cpa.gov/ada/download/reports/protocol.pdf].
Wilson, J. T., Cho, J. S., Wilson, B. H., and Vardy, J. A. 2000. Natural Attenuation of MTBE in the Subsurface
under Methanogenic Conditions, EPA/600/R-00/006. U.S. Environmental Protection Agency, Cincinnati,
OH. [www.epa.gov/ada/reports].
77
-------
78
-------
3
In
Robert W. Puls and William J. Deutsch
3.1
The purpose of this session was to develop recommendations concerning reduction/oxidation (redox) measure-
ment parameters for inorganic hazardous waste site remediation. It is hoped that the information provided will
be useful to site managers to guide site characterization, data collection efforts, and. in particular, performance
evaluation of in-situ remedial approaches. The following issues and/or questions were addressed in the session:
1) What remediation processes specific to inorganic contaminants require the need for redox
measurements during site characterization and/or remediation (e.g.. MNA. zero valent metal
treatment, in-situ redox manipulation, etc.)?
2) What parameters must be measured during site assessment and remediation? Are the required
parameter measurements universal or contaminant-specific? Is there a universal set of parameters
that is sufficient for all remedial strategies?
3) What specific redox measurements are required (e.g., Eh, DO, Fe, etc.), how are redox measurements
best made, and how are the measurements used? Is there a need to evaluate redox characteristics of
the solid matrix in addition to the mobile phase?
The overall findings and conclusions from this session were:
« Geochemical models are recommended for use at all sites. Site data are needed to support the geochemical
model.
« A set of core aqueous-phase redox measurement parameters are recommended for initial characterization
at all sites to understand the redox status of the aquifer. Supplemental measurement parameters may be
needed on a site-specific basis.
* There are commonalities for site characterization across treatment types and contaminant types.
« Eh measurement is included in the recommended core parameters, even though this measurement is
controversial and may only provide qualitative information.
* Site diversity and differences in treatment applications preclude a "hard" list of redox measurement
parameters applicable to even' site. Measurement parameters will change from site-to-site.
* Solid-phase redox measurements are generally not performed. The solid-phase should be tested for metals
to determine the form of the metal contaminant before and after treatment and to look at metal stability.
* Solid-phase sample preparation and analytical methods are lagging behind aqueous-phase methods and
need more development.
* We need to be able to predict the performance of a selected site remedy (treatment) and its influence on the
risk assessment. A decision tree for guidance on site evaluation would be useful.
* WTe need to define parameters/criteria for site closure (post-treatment monitoring).
3.2
Redox processes are chemical reactions that include a transfer of electrons and consequently a change in
valence state of elements that are either oxidized to a higher valence state or reduced to a lower valence state.
This change in valence state of the elements impacts remediation through three mechanisms:
* Change in speciation to a lower/higher toxicity
« Solubility decrease to immobilize contaminants or increase in solubility to mobilize contaminants
« Adsorption/dcsorption impacts on contaminant mobility
79
-------
By definition, rcdox processes involve a change in valence state of the elements whereby one or more elements
are oxidized and one or more are reduced. If the inorganic contaminant participates in these electron transfer
reactions, then its valence state will change. The toxicity of an element is a function of its valence state because
different valence states of the elements react to a greater or lesser degree with biological functions. For
example, the acute toxicity of As(III) is substantially greater than that of As(V) (Frankc and Moxon, 1936).
Furthermore, As(III) is the initial substrate in the methylation of arsenic (Thomas, 1994); thus, changes in the
arsenic valence state will have an effect on the production of the methylated arsenic species.
The relative toxicities of the different valence states of other redox-sensitive elements appear to be related to
both availability and toxic effects. For example, the trivalent form of chromium [Cr(III)] is considered much
less toxic than Cr(VI), however, this may be due as much to the lower solubility of Cr(III) solids in aquifer
environments compared to Cr(VI) solids as to the inherent toxicity of the two valence states of chromium
(Losi et al., 1994). In the case of selenium, availability and toxicity appears to be related to the presence of the
more oxidized, soluble selenate [Se(VI)] and selenite [Se(IV)] redox states compared to the reduced, less
soluble forms of elemental selenium (Se°) and selenide [Se(-II)J (Neal. 1990).
OH
Under equilibrium conditions, the solubility of a mineral limits the dissolved concentration of one of its
constituent elements. Hie solubility of minerals containing the redox-sensitive elements can van' by orders of
magnitude depending on the valence state of the element in the mineral, the redox status of the aquifer and in
particular, the pH of the ground water. For example, the solubility of the Cr(III) mineral Cr(OH)3 under typical
ground-w7ater conditions is in the part per billion range whereas the solubilities of Cr(VI) minerals under these
same conditions are in the part per million range (Richard and Bourg, 1991). As a consequence, reducing
Cr(VI) to Cr(III) either by adding a reductant to the aquifer or by allowing the natural system sufficient time to
reduce the chromium valence state and reach equilibrium has the potential for achieving a ground-water
cleanup level by precipitation of the relatively insoluble Cr(OH)3 mineral.
The use of precipitation of low solubility minerals as a remediation measure is directly related to the concept of
geochemical barriers in natural systems (Perelman. 1986). A geochemical barrier is a zone in the subsurface
characterized by a sharp change in physical or chemical environment along the flowpath and is often associated
with the precipitation of elements from solution. Redox barriers are classified as a type of physico-chemical
barrier, which also includes alkaline, acidic, and adsorption barriers. Because barriers represent a change in
environmental condition, each type of barrier must be considered in the context of the types of water with
different redox characteristics that might enter the barrier. In addition to being a major control on solubility, the
pH of the aquifer will affect the type of element immobilized by the geochemical barrier.
Applied techniques for immobilizing inorganic contaminants are generally based on the well-documented
natural processes that impact element mobility, particularly at geochemical barriers. If manganese concentra-
tions in mildly oxidizing water downgradient from a landfill must be lowered, men either increasing the redox
potential to precipitate a less soluble Mn(IV) oxide or reducing the redox potential and adding sulfide to form
a less soluble sulfide mineral might be considered. Some of the issues to consider in the applied redox
approach are the type and amount of reactant, means of emplacement, reaction kinetics, unwanted byproducts,
solubility of contaminant-containing mineral, and geochemical stability of the imposed barrier environment.
Some of the complexity of the situation has been shown by Seaman et al. (1999) who found that in situ
immobilization of Cr(VI) by using Fe(II) solutions was complicated by the hydrolysis of Fe(III) which lowered
the pH and mobilized Cr(VI) in column experiments.
The types of oxidizing and reducing compounds that have been tested or suggested to lower inorganic
contaminant concentrations in an aquifer are listed in Table 3.1. Most of these compounds can be dissolved in
water and injected into the aquifer. Air and ozone are added to the subsurface in a vapor phase. Elemental iron
(Fe°) is a solid that can be emplaced in a trench or as a slurry that intercepts the contaminated water. It has been
shown to be effective not only at reducing the concentrations of inorganic contaminants (particularly chromatc
and uranium), but also chlorinated volatile organic compounds such as PCE and TCE. This type of remediation
is expected to expand as its effectiveness becomes more widely known and the injection of reagents to achieve
remediation becomes more commonly accepted by regulatory agencies.
80
-------
Table 3.1 Potential Oxidants and Reductants for Inorganic Remediation
Oxidants/elements treated* Reductants/elements treated*
Dissolved oxygen / Fe
Air / U, As
Oxygen Release Compound / As
Potassium permanganate
Fenton's reagent
Fe°(PRB) / Cr, As, U, Tc, Se, Mo
Colloidal Fe° / Cr, As, U, Tc
Dithionite / Cr, U, Tc
Sodium sulfide / Cr, U
Hydrogen sulfide / Cr
Fe2+ / Cr
Ferrous sulfate / Cr
Hydrogen Release Compound / Cr
Organic materials / sulfate, nitrate, ARD
* known elements treated or proposed following successful lab testing
The adsorption capacity of an aquifer can be increased to immobilize contaminants by adding dissolved ferrous
fFe(II)] iron to an aquifer and then oxidizing the ferrous iron to precipitate ferric (oxy)hydroxide. The large
surface area of this solid and its strong affinity for both dissolved cations and anions has the potential for
substantially lowering dissolved contaminant concentrations.
It is preferable to leave a contaminant in place if its mobile concentration can be reduced to below a cleanup
level and rendered non-toxic; however, in some instances immobilization may not be possible and removal is
necessary. Removal of an inorganic contaminant requires not just extraction of contaminated ground water, but
also dissolution of contaminant-containing minerals and perhaps desorption of the contaminant from the
surface of solids. If an element is concentrated under reducing conditions (e.g.. Cr, U. Mo, Se), then creating
oxidizing conditions in the aquifer should mobilize the contaminant. This is exactly the process used to mine
uranium by in-situ leach methods. Oxygen is added to the subsurface where uranium has been concentrated by
natural processes in the reduced U(IV) mineral uraninite (UO ). Uraninite is soluble under the imposed
oxidizing conditions leading to its dissolution and the removal of dissolved uranium in production wells.
Mobilization of contaminants can be complicated by the strong adsorption and high capacity of iron and
manganese oxyhydroxides for other metals. If oxidizing conditions are generated in a reducing environment to
mobilize metals, iron and manganese solids may precipitate and remove from solution some of the metals that
are the target of remediation. This occurrence may require a stepwise mobilization process to first oxidize the
elements in the minerals and then reduce and dissolve iron and manganese oxyhydroxides to release adsorbed
contaminants. The kinetics of these reactions will also control the oxidation/reduction of the elements and can
greatly impact any remedial scheme.
on
Adsorption is the attachment of dissolved ground-water constituents to the surfaces of the aquifer solids and
desorption is the release of the constituents back into the water. Redox reactions that affect the affinity or the
capacity of the solid surfaces for dissolved constituents can impact contaminant mobility. Adsorption is very
pH dependent because cations (or anions) of the same charge will adsorb at very different pH values depending
on the element (Jain et al., 1999; Langmuir, 1997). Any remedial scheme which involves a rcdox manipulation
that affects the pH of the system or alters the nature of sorbent surfaces can have profound impacts on element
adsorption/desorption reactions and thus contaminant mobilization/immobilization.
81
-------
3.3
It is necessary to collect data during site characterization on the aqueous and solid phase geochemistry of the
system, model the data, and construct a site conceptual model. This is an iterative, evolutionary process.
Numerous redox measurement parameters required to characterize site geochemistry and to evaluate treatment
were discussed in the session. Table 3.2 shows which parameters were considered to be core measurements that
are recommended for all sites, as well as additional site-specific parameters that may be useful.
There was considerable debate about the usefulness and necessity of the Eh measurement. It is considered by
some to be more of a qualitative measurement because it can give inconclusive results. The electrode will often
not stabilize in oxygenated waters, or in reduced waters without sulfur being present. Many members of the
panel felt that Eh determinations using electrodes were of limited usefulness in mixed waters (different ground-
water sources). Discrepancies between Pt electrode measurements and Eh calculated from redox couples have
been well documented (Lindberg and Runnells, 1984). The time required for stabilization of the electrode can
also make the measurement impractical in the field. It was noted that because of this, Eh probably should not
be a stabilization parameter for low-flow ground-water sampling. Others felt that in some settings the
measurement has been shown to be of value (acid rock drainage) and research involving long-term perfor-
mance of permeable reactive barriers has focused on Eh as a potential indicator of system performance. Eh can
be helpful in context with other data, such as iron speciation. Some participants felt that it is a good idea to
collect Eh data even though it may be a questionable measurement.
Finally, several members felt that the most valid approach to evaluate redox was collection of multiple
parameters, including Pt electrode measurements, such as iron speciation and other site-specific redox couples
(e.g., arsenic, sulfur). Table 3.3 lists some important characteristics of many rcdox-sensitivc elements.
Table 3.2 Core and Site-Specific Redox Measurement Parameters
Core Parameters
Site-Specific Parameters
Ground Water
Ground Water
Solid-Phase
Temperature
Eh, pH, alkalinity
Turbidity
Total and Ferrous Iron
Dissolved Oxygen
Specific Conductance
Dissolved Organic Carbon
Major ions (Ca2+, Mg2+, Na+,
K+, Cl , S2, SO42, NOs ,
Ammonium, Phosphate )
Aluminum, Silica, Manganese
Dissolved Inorganic Carbon
Total Dissolved Solids
Fluoride
Hydrogen
Methane
Sulfide
Priority Metal Contaminants
(As, Cd, Cr, Cu, Hg, etc.)
Oxidizing Capacity
Reducing Capacity
Neutralization Capacity
Adsorption Capacity
Cation Exchange Capacity
Acid Generation Capacity
Total Metal(s) Concentration
Iron/Aluminum Extraction
Mineralogy (XRD)
Total Organic Carbon
Sulfur Extraction
82
-------
Table 3.3 Sampling Parameters for Inorganic Characterization
Parameter
Comments
Field Measurements -
Temperature
pH
Eh
Dissolved Oxygen
Turbidity
Specific Conductance
Alkalinity
Redox Indicators
Laboratory Analyses — Water
Major Cations
Major Anions
Redox Indicators
Minor/Trace Constituents
Metals of Potential Concern
Must be done in the field; downhole (monitoring
wells) preferred
Must be done in the field, robust, easily measured,
flow cell recommended
Must be done in the field, variable, more standard
checks recommended, flow cell recommended
Must be done in the field, variable, accuracy of
probes limited at less than 1 mg/L
Must be done in the field
Recommended for field
Recommended for field
S2, H,
Ca, Mg, Na and K
HCO3/CO3, Cl, SO4, NO3/NO2
Mn, Fe, S2, As(III)/As(V), CH4, DOC,
Al, Si, PO4, Br, F
As, Cd, Co, Cr, Cu, Hg, Mo, Ni, Pb, Tl, Zn
Collection of turbidity data was seen as important for a variety of reasons including its usefulness as a sampling
equilibration parameter using low-flow techniques where metals are contaminants of concern, and to explain
data anomalies (e.g., high total iron concentrations under oxidizing conditions) and evaluate the influence of
particulates on metals concentrations.
The collection of major anion and cation data is important in order to properly evaluate the aqueous
geochemistry of the system and perform accurate geochemical modeling in support of a site conceptual model.
Additional information on this topic is covered in the next section. Fluoride was seen as important in some
hydrogeologic settings because of its high complexing capacity and thus, its importance in adequate determina-
tion of ion balances for geochemical modeling purposes.
There was a general consensus among the group that more solid phase characterization was required for site
characterization and performance assessment of in-situ technologies for remediation of inorganic-contami-
nated sites. Some of these arc listed in Table 3.2. The solid phase should be tested to confirm the form of the
metal associated with the solid phase (mineralogy) to determine its stability or mobilization potential; this is
particularly important for monitored natural attenuation (MNA) assessments. If immobilization is invoked for
MNA, then it is imperative to know what the immobilization mechanism is and under what conditions it may
be reversed (i.e., mobilized). Assessment of the oxidiation and reduction capacities of sediments can be
important for several in-situ remedial technologies including monitored natural attenuation and the creation of
'redox barriers" using injections of liquid reductants (Palmer and Puls, 1994; Fruchter et al., 2000). Sulfur
extractions are useful to assess the nature of the redox environment and also the stability of precipitated solid
phases. Similarly, extractions for amorphous and crystalline iron oxides are important in terms of metal
stability and iron bioavailablity.
There was consensus among the group that certain parameters must be measured in the field due to stability
issues and other measurements are recommended for the field but could be done in the laboratory. Table 4
83
-------
attempts to address these concerns and provides at least a partial listing of some of the more commonly
measured parameters.
The development of a site conceptual model is imperative for any site characterization effort, particularly as a
project moves closer toward remedial selection. It involves the incorporation of a comprehensive assessment of
physical, chemical, and biological processes operative within a defined geologic framework at a site. The group
felt strongly that an iterative evolutionary process be employed to develop and refine a site conceptual model
regarding ground-water flow, hydrogeochemical setting or definition, and contaminant fate and transport. The
developing model must continually be tested against new data and be consistent with data collected throughout
the site characterization process.
3.4 Geocliemical Modeling
There are two primary reasons for using geochemical computer models at a site. The first is to better
understand the processes active at the site and be able to explain observed compositions of the various phases
and trends in composition across the site. The second reason is to aid in the remedial design for cleanup of a
contaminated site. Accurate geochemical modeling of a contaminated site can be used to estimate the outcome
of various remedial actions. The model can be used to select the best remediation technologies and to optimize
application of the technology. This applies to the evaluation of both natural and applied redox processes as well
as to other geochemical-based remediation measures such as pH neutralization.
The development of a geochemical model of a system starts with the compilation of available information on
water composition, presence of solid phases, hydrogeology, etc. The existing site conditions are evaluated in
terms of the known types of geochemical and physical processes that occur in the environment. These
processes include mineral dissolution/precipitation, oxidation/reduction, adsorption/desorption, ground-water
recharge, and mixing of ground waters from two or more aquifers. The combination of site specific conditions
and known, common geochemical/physical processes leads to the development of a conceptual model of how
the site operates. The conceptual model can be combined with a computer code to calculate the impacts of the
various processes and provide quantitative estimates of these impacts on the composition of the various phases.
The result is a geochemical computer model consisting of the conceptual model and the computer code that
together can simulate existing conditions and predict the impact on the system of changes in environmental
conditions.
3.4
The presence of dissolved constituents in water is a result of chemical interactions between water and the vapor
and solid phases in contact with the water. A competent geochemical model of a site must be able to simulate
the processes that occur between these phases. They include gas dissolution and exsolution from ground water,
aqueous speciation, mineral dissolution/precipitation, oxidation/reduction, and adsorption/desorption. Two of
the most commonly used equilibrium mass transfer codes are MINTEQA2 (Allison et al., 1991) and
PHREEQC (Parkhurst, 1995). MINTEQA2 is available on the EPA website at ftp://ftp.cpa.gOv/cpaj:cam/
wwwhtml/softwdos .htm and PHREEQC at the USGS website http:Mi2o.usgs.gov/software/geochemical .html.
The ability to simulate changing environmental conditions opens up the entire realm of remediation design to
geochemical modeling. Once an accurate model of an existing contaminated site has been developed, it is
relatively simple to vary site parameters and evaluate the impact of the new condition on contaminant
concentrations in the solution and solid phases. Hie parameters that might be varied include pH, redox
potential, ionic strength, and concentrations of complexing compounds. For remediation purposes, it is
possible to evaluate both in-situ fixation and immobilization of contaminants as well as enhanced mobility and
removal of contaminants from a system. For example, arsenic is strongly adsorbed onto iron oxyhydroxide
minerals in shallow, oxidizing aquifers. If the dissolved arsenic concentration must be reduced in order to
attain a cleanup goal, then adding additional iron minerals to the aquifer might enhance adsorption of arsenic
and its removal from ground water. Alternatively, arsenic might be removed from the aquifer by circulating a
reducing solution through the aquifer that dissolves the iron mineral adsorbents and releases arsenic to ground
water so that it can be flushed more easily from the aquifer. Both of these remediation alternatives can be
simulated with a geochemical model.
84
-------
Table 3.4 Characteristics of Rcdox-Scnsitivc Inorganic Contaminants
Element Redox States Water Mobility
Antimony
Arsenic
Chromium
Copper
Iron
Manganese
Mercury
Nitrogen
Selenium
Sulfur
Vanadium
Uranium
III, V
III, V
III, VI
I, II
II, III
II, III, IV
0, I, II
-III, 0, III, V
-II, 0, IV, VI
-II to VI
III, IV, V
IV, VI
Sb solids of both valence states are very
soluble.
As (III) sulfide minerals insoluble at low
Eh, As(V) minerals soluble.
Relatively immobile due to low solubility
of Cr (III) hydroxides. Cr(VI) minerals
highly soluble.
Cu(OH)2 formation limits ground-water
concentration; Cu solubility limited by
hydroxysulfate, hydroxycarbonate.
Fe oxyhydroxides (HFO) form at pH >
2.5 under oxidizing conditions and Fe (II)
is limited by sulfides and carbonate under
reducing conditions depending on pH and
solution composition.
Mn (III, IV) oxides and oxyhydroxides
relatively insoluble when pH > 5 and
oxidizing; Mn (II) carbonate more
soluble but stable in reducing conditions.
Soluble solids when pH > 4 and
oxidizing; relatively insoluble sulfide
minerals under reducing conditions.
No concentration limiting solid phases;
dentrification may limit nitrate.
Soluble selenite/selenate minerals;
selenide less soluble than selenite which
is less soluble than selenate.
Soluble sulfate minerals; insoluble sulfide
minerals.
May substitute for Fe in ferric oxide
lattices; mobile in reducing conditions.
Soluble VI minerals; insoluble IV oxides.
One of the additional benefits of modeling remediation is that secondary reactions that might not be
immediately obvious to the investigator may become clear in reviewing the modeling results. For example.
adding iron to an aquifer to precipitate as iron oxyhydroxide and provide additional adsorption sites may
change the pH of the solution because iron precipitation is a hydrolysis reaction that produces hydrogen ions.
If the system is not well buffered, the pH will decrease, which, in turn, may mobilize other metals. The aqueous
speciation computer codes used to develop the geochemical models can take into account the formation of
hydrogen ions during iron precipitation and will allow for buffering by carbonate/bicarbonate ions in solution.
If carbonate minerals are also present in the system, the modeler may want to allow them to interact with the
solution as iron carbonate precipitates.
Limitations
The utility of a geochemical model for a particular site is limited by errors introduced during development of
the model and our ability to model certain processes. The first opportunity for introducing error into the
process is data collection. The model developed is highly dependent on accurate physical and chemical data
from the site. We need to know flow direction to determine whether concentrations are increasing or
decreasing along the flow path and we need accurate measurements of temperature, pH, Eh. and concentrations
85
-------
of dissolved constituents. For example, an incorrect Eh will affect the speciation of all redox-sensitive species
that are allowed by the modeler to equilibrate with the measured value. If nonequilibrium conditions exist for
some of the allowed redox pairs, then the model calculations will be incorrect. This will, in turn, affect
calculations of mineral equilibrium and solubility. Therefore it is important to collect speciation data on iron
and other redox-sensitive elements which may be present in the aquifer to increase confidence in the
approximate redox poising of the system.
Most of the major inorganic geochemical processes impacting natural systems are well known and incorpo-
rated into the modeling codes. However, certain systems may be relatively unique and some of the processes
occurring may not be standardized in the available computer codes. For example, some of the elements that
may be of interest for evaluating radionuclide migration (such as plutonium and americium) are not in the
standard databases of MINTEQA2 or PHREEQC. Also, a rare mineral or amorphous solid may limit solution
concentrations of a contaminant that docs not normally reach high concentrations in the natural environment. If
this solid phase is not in the code database, its formation cannot be simulated. Most of the available computer
codes allow the user to input new thermodynamic data for elements, solution complexes and solid phases. This
allows the user to develop an appropriate model if concentration and thermodynamic data are available.
The computer code used to make the calculations required by the model has its own limitations. Many of the
equilibrium thermodynamic codes do not have a method of calculating reaction rates. If a mineral forms or
dissolves slowly in a system, the model developed from these codes will not account for these kinetic effects.
This is not a major limitation for most aquifer systems where residence times are measured in years; however,
kinetic effects can become more important in modeling surface water systems or reactions anticipated to occur
during applied remediation methods such as the injection of reactants into an aquifer. PHREEQC is one of the
few models which does have kinetic data as part of the code.
These limitations should be considered in developing a geochemical model, however they will rarely be
sufficient reason for not developing a model for a site. Considering the number of useful applications for a
geochemical model in site characterization, contaminant transport and remediation design, it is often beneficial
to collect sufficient data for modeling and develop a geochemical model for each site.
3.5 to
A number of in-situ remedial methods were discussed where redox processes arc key to implementation and
success. These included the following: physical emplacement (permeable reactive barriers), reagent injection
(potassium permanganate, Fenton's reagent, colloidal iron, sodium dithionite, sodium sulfide), and natural
attenuation. The group chose to focus on permeable reactive barriers due to a preponderance of experience
with this in-situ technology by members of the group. Additional discussions involved the issue of acid rock
drainage (ARD) because of the magnitude of the problem and recognition of ARD as an emerging environmen-
tal isssue.
In terms of initial site assessment for permeable reactive barrier application, the core list referenced above in
Table 3.2 was deemed adequate by the group. The following additional measurement parameters specific to
PRBs were also recommended by the participants:
* verify site ground water/barrier interaction - rate of loading of contaminant (e.g., Cr) onto reactive media
(e.g., iron)
• microbiological activity (biomass estimates and populations [e.g., iron oxidizers, sulfate reducers, etc.])
upgradient, downgradient and within the PRB
« longevity (examination of precipitation and clogging within reactive zone over time)
• metal speciation
* solid-phase analysis (sec Table 3.2)
• hydrology
• conservative tracer to track movement into, and through barrier (bromide, chloride, iodide, etc.)
It is imperative to know the extent of source area contamination to design an appropriate remedial system.
Will/can the source area be removed/treated? If left in place, how long will the source area contribute to
86
-------
contamination of the ground water? Identification of the plume boundaries in three-dimensional space as well
as temporal variation is important in siting and designing the PRB. Verification monitoring to demonstrate
treatment effectiveness is also needed. For long-term usage, one needs to determine how quickly mineral
precipitates are accumulating on the iron surface, which will decrease porosity overtime and will also passivate
the reactive surfaces. Coatings may accumulate faster in some geochemical environments compared to others
(e.g.. high sulfate or high alkalinity waters). Regulators want to know how long the PRB is going to work.
In the specific case of chromium remediation using PRBs, the following list of parameters has been measured:
• Cr speciation, aqueous
« Cr speciation, solid phase (sequential extractions, surface analyses)
• Oxidation capacity, adsorption capacity of aquifer solids
« Permeability assessment of PRB over time (water levels, flow meters, tracer tests)
• Microbiological assessments (phospholipid fatty acid analyses)
Cost-effective methods for long-term performance monitoring of these systems are currently being evaluated
by a number of groups. It is hoped that some direct and inexpensive methods may be employed to accurately
and safely predict performance and longevity.
3.6 Acid
This group discussed numerous technologies used to remediate Acid Rock Drainage (ARD) and many more
that are proposed but not used. Many of these techniques, however, apply to cleanup of surface waters, not
ground waters. Interactions between surface waters and ground waters are important, but the nature of these
interactions and their environmental impacts are not well understood. Redox chemistry is what ARD
production, remediation, and performance assessment is all about. All remediation technologies need to
address neutralization and many utilize reductive processes. ARD can be categorized into six types:
« Underground workings
* Tailings
* Waste Rock
* Contaminated Aquifers
• Heap leach pads
• Pit lakes/Open pits (not necessarily acid, but redox is definitely important)
• Slag
• Backfill
Necessary redox parameters are the same as those outlined for other inorganic contaminant situations. Iron
speciation [Fe(IT)/Fe(TTI)] is necessary instead of optional and prioritization of other parameters will be site-
specific. For example, some sites require measurement of As(III)/As(V), but not H2S and vice versa.
Three groups of mineral assemblages are needed to undergo solid phase analyses: oxidizable (sulfide) minerals.
ganguc minerals, and country rock minerals. Solid phase analyses are needed to evaluate 1) the amounts/types
of acid-generating sulfide minerals, 2) the amounts/types of acid-neutralizing minerals (carbonates, alumino-
silicates), and 3) the reactivity of these minerals (function of crystal form. size, and composition).
An important aspect of remedial technologies for ARD is that there are very few, clear-cut success stories with
cleanup, long-term maintenance, and with monitoring for decades demonstrating compliance. Hence, it is
difficult to look to case histories for guidance.
3.7
The members of this group were acknowledged experts in the area of environmental assessment and
remediation for inorganics in the environment. Members were employed in industry, universities and the
government sector. Many had extensive field experience in monitoring and remediation. Strong consensus
was reached on a number of issues including the following:
• Use of geochemical models is strongly recommended for all sites and sufficient data for model usage needs
to be collected.
87
-------
• Certain core geochemical parameters need to be measured at all sites.
« Supplemental parameters are also needed for all sites.
* The diversity of contaminant behavior and potential remedial technologies precludes a one-size-fits-all list
of parameters.
* Metals are persistent in the environment. If immobilization technologies arc considered, the solid phase
must be analyzed to determine/verify the form of the contaminant and evaluate whether conditions will
persist to contain the contaminant.
* Development and application of analytical methods (extractions) for solid phase analyses lags those for the
aqueous phase and more research is needed in this area.
References
Allison, J. K., Brown, D. S., and Novo-Gradac, K. J. \^\.MINTEQA2/PR.ODEPA2, A Geochemical Assessment
Model for Environmental Systems, U.S. Environmental Protection Agency.
Franke, K. W. and Moxon, A. L. 1936. A comparison of the minimum fatal dose of selenium, tellurium, arsenic
and vanadium. Journal of Pharmacology and Experimental Therapeutics, v. 58, pp. 454-459.
Fruchter, J. S., Cole, C. R., Williams, M. D., Venneul, V. R., Amonette, J. E., Szecsody, J. E., Istok, J. D., and
M. D. Humphrey. 2000. Creation of a subsurface permeable treatment zone for aqueous chromate
contamination using in situ redox manipulation. Ground Water Monitoring & Remediation, v. 20, pp. 66-77.
Jain, A., K. P. Raven and R. H. Loeppert. 1999. Arsenite and arsenate adsorption on ferrihydrite: Surface charge
reduction and net OH" release stoichiomctry. Environmental Science and Technology, v. 33, pp.1179-1184.
Langmuir, D. Aqueous Environmental Geochemistry. Prentice Hall, 600p.
Lindberg, R. D. and Runnclls, D. D. 1984. Ground water rcdox reactions: an analysis of equilibrium state
applied to Eh measurements and geochemical modeling. Science, v. 225, pp. 925-927.
Losi, M. E., Amrhein, C., and Frankenberger, W. T. Jr. 1994. Factors affecting chemical and biological reduction
of hexavalent chromium in soil. Environmental Toxicology and Chemistry, v. 13, pp. 1727-1735.
Neal, R. H. 1990. Selenium. In Heavy Metals in Soils, ed. B. J. Alloway. Blackie, Glasgow, pp. 237-270.
Palmer, C. D. and Puls, R. W. 1994. Natural Attenuation of Hexavalent Chromium in Groundwater and Soils,
EPA/540/S-94/505. U. S. Environmental Protection Agency, Cincinnati, OH.
Parkhurst, D. L. 1995. User's guide to PHREEQC, a computer model for speciation, reaction path, adveclive
transport and inverse geochemical calculations, U. S. Geol ogical Survey.
Perelman, A.I. 1986. Geochemical barriers: theory and practical applications. Applied Geochemistry, v. 1,
pp. 669-680.
Richard, F. C. and Bourg. A. C. M. 1991. Aqueous geochemistry of chromium: a review. Water Resources, v. 25,
pp. 807-816.
Seaman J.C., P.M. Bertsch and L. Schwallie. 1999. In situ Cr(VI) reduction within coarse-textured, oxide-
coated soil and aquifer systems using Fe(II) solutions. Environmental Science and Technology, v. 33,
pp. 938-944.
Thomas, D. J. 1994. Arsenic toxicity in humans: research problems and prospects. Environmental Geochemistry
and Health, \. 16, pp. 107-11L
88
-------
4
of
Richard T. Wilkin and Carol J. Ptacek
4.1
Conceptual models of contaminant transport and fate in ground water critically depend on geochemical measure-
ments and their interpretation. Transport and fate processes are often dependent on redox processes because in
many cases oxidation-reduction reactions cause changes in spcciation that have a marked effect on contaminant
mobility and toxicity. The importance of reliable field characterization is especially significant when evaluating
redox-sensitive speciation because of the potential for sample alteration (oxidation) during sample collection and
handling.
This group considered methods used by environmental scientists to characterize the redox chemistry of ground
water in support of site characterization and remedial performance monitoring. The group recognized that a wide
variety of analytical techniques are available to quantitate redox-sensitive elements (e.g., Fe, S, N). Selection of
the appropriate technique for a given parameter will depend on specific project objectives. It was recognized that
project objectives and principal data uses collected in monitoring programs change with time as more site-specific
information is obtained and conceptual models evolve. Although a wide variety of techniques are available for
redox characterization (see Appendix A), members of this group agreed that standards for improved field practices
are needed (see, for example, U.S. Geological Survey, 1997). Methods of analysis identified during the group
discussion are briefly reviewed in this chapter, and include Eh, dissolved oxygen, iron speciation, sulfur speciation,
nitrogen speciation, and alkalinity.
4.2
Sample collection programs are designed around goals associated with specific project objectives. Data Quality
Objectives (DQOs) define the types, quality, and quantity of data that are required by the various aspects of a
project. With DQOs in place, appropriate sampling methodologies, analytical protocols, and specific methods may
be considered and selected. For example, DQOs may be different if project goals put emphasis on detection and
monitoring of contaminant concentrations or whether geochemical spcciation and the development of site
conceptual models are targeted goals.
The principal objective of geochemical characterization in site assessment and remedial performance monitoring is
to obtain water quality information, with no alteration of water chemistry, from a chosen sampling point. Such
uncompromised data are critical in developing accurate conceptual and quantitative models of contaminant
transport and fate. There was strong agreement among the group that sample collection practices and measurement
techniques are among the greatest challenges to developing reliable site conceptual models. Several categories of
ground-water sampling methods were recognized: accumulation or diffusion samplers, purge techniques with flow-
through cells, and bailing. The generally recommended and most commonly practiced method is low-flow,
minimal draw-down purging and sampling. Low-flow methods minimize chemical and hydrological disturbances
in and around the well. In some situations, multiple-tube bundle piezometers or diffusion samplers may provide the
best method for collecting ground-water samples without chemical alteration. Bailing techniques were not
recommended for purposes of redox characterization because of the likelihood of changing water chemistry by
unavoidable reaction with air.
Measurements of geochemical parameters can be divided into three categories: laboratory, in situ, and purging or
field measurements. Laboratory analyses are carried out on preserved samples for characterization of total or
selected metals, anions, and organic constituents. In some cases redox integrity of dissolved components can be
89
-------
preserved by acidification (e.g., for NO3VNH4') because oxidation rates generally decrease substantially with
decreasing pH. In situ measurements such as down hole probes or contained sampling devices are desirable for the
measurement of unstable parameters that are ideally made at well conditions. In-situ measurements are potentially
advantageous for the determination of dissolved oxygen, temperature, and oxidation-reduction potential. Project
objectives may or may not require the extra cost and effort required to collect data using in-situ techniques.
Measurements carried out in the field include electrode measurements of DO. pH, specific conductance, Eh, and
other ions using ion-selective electrodes. Field-deploy able techniques for ions also include UV-Vis spectrometry
and ion chromatography. Field analyses arc made to gather information in the field to guide activities and/or
because the measured parameters are considered to be too unstable to transport samples without compromising
their chemical integrity. There was general agreement that unstable parameters, best measured in the field at the
time of sample collection, include pH. turbidity. DO, ferrous iron, alkalinity, sulfide. and oxidation-reduction
potential. Measurements of these parameters are ideally made during well purging. Members of the group reported
a variety of demonstrated techniques for preserving the redox integrity of samples pumped to the surface, such as:
minimizing tubing length, use of high-quality tubing with low oxygen diffusion coefficients, use of syringes with
luer-lok fittings, preventing direct exposure of tubing and all electronic equipment to sunlight, and ensuring air-
tight seals. Specific methods arc discussed below for the measurement of Eh, DO, and spcciation of iron, sulfur,
and nitrogen (Table 4.1).
Table 4.1 Redox Parameters and their Common Methods of Measurement
Parameter
Oxidation-Reduction
Potential (ORP)
Dissolved Oxygen (DO)
Iron Speciation
Sulfur Speciation
Nitrogen Speciation
Alkalinity
Common Field Methods
Combination Platinum
electrode with Ag/AgCl
reference electrode; KC1 filling
solution
Membrane-covered electrode
Colorimetric (high range,
indigo carmine)
Colorimetric (low range,
rhodazine D)
Modified Winkler titration
Ferrous iron Colorimetric
indicators (1,10
phenan thro line; ferrozine)
Sulfate (turbidimetric);
Sulfide (methylene blue
Colorimetric method)
Nitrate+Nitrite (cadmium
reduction)
Ammonia (Nessler method;
Salicylate method)
Acid titration
Standards for
Performance
Assessments
ZobelPs solution;
Light's solution
Air-saturated
water
Prepared ferrous
solutions
Prepared sulfate
and sulfide
solutions
Prepared nitrate,
nitrite, or
ammonia
solutions
Prepared
bicarbonate or
carbonate
solutions
References
Langmuir (1971)
Nordstrom (1977)
Standard Methods for
the Examination of
Water and Wastewater
(1999)
Hitchman(1978)
Gilbert et al. (1982)
White etal. (1990)
Standard Methods for
the Examination of
Water and Wastewater
(1999)
Tamuraetal. (1974)
Stookey (1970)
Standard Methods for
the Examination of
Water and Wastewater
(1999)
Cline(1969)
Standard Methods for
the Examination of
Water and Wastewater
(1999)
Standard Methods for
the Examination of
Water and Wastewater
(1999)
90
-------
4.3
Eh
The Eh measurement is the most commonly used technique to characterize the oxidation-reduction state of ground
water. Eh is the measured potential of a platinum electrode corrected to the Standard Hydrogen Electrode (SHE).
Eh is not the same as the oxidation-reduction potential (ORP), which is the direct, uncorrected potential reading of
the reference electrode. It was recognized during the group discussion that although measured Eh values usually do
not correspond to Eh values calculated using the Nernst Equation with measured concentrations of aqueous redox
pairs, e.g., N(V)/N(-III) or S(-II)/S(VI), there is benefit to measuring this parameter in the field. Eh measurements
are a relatively easy approach for determining overall redox conditions. Eh measurements should be thought of as
qualitative indicators of the state of oxidation or reduction of a natural system (Langmuir, 1971). In rare situations,
such as acid-mine waters, high activities of both Fe(III) and Fe(II) provide a dominant and reversible couple that
poises Eh measurements (Nordstrom et ah, 1979). Many ground waters are not well poised and speciation
calculations for ground waters have shown that redox couples typically exist in a state of disequilibrium (e.g..
Stumm, 1966; Lindberg and Runnells. 1984). Such systems in disequilibrium yield mixed potentials that may or
may not represent the redox distribution of any specific redox couple. The reasons for this include disequilibria in
low-temperature ground water (Lindberg and Runnells, 1984; Thorstenson, 1984), poisoning or coating of Pt
electrodes (Jackson and Patterson, 1982; Whitfield, 1974), potential measurements reflect mixed potentials with
little thermodynamic significance (Morris and Stumm, 1967).
The recommended procedure for measurement of Eh involves the use of a small-volume flow-through cell into
which a temperature and Eh electrode is inserted. The type of electrode recommended for case of use is a high-
quality combination Pt electrode typically with a Ag/AgCl reference electrode. After purging the flow-through cell
with several volumes of ground-water. Eh readings are recorded until they stabilize or a minimum value is
obtained. Often 30 minutes or longer is required for electrode stabilization, but the group generally agreed that
waiting >30 minutes for electrode stabilization was not a time-effective practice. Proper preparation and mainte-
nance of the Pt electrode is required, including frequent polishing of the electrode surface, changing of the internal
filling solution, and frequent checks with standard solutions mat are maintained at ground-water temperature
(American Public Health Association et ah, 1999). The most common standard solution for reference checks is
Zobcll's solution (Nordstrom, 1977). Additional check solutions include Light's solution (Light, 1972). and others
recommended by electrode manufacturers (i.e.. Orion ORP standard). The importance of correcting field measure-
ments of ORP to reference the values to the Standard Hydrogen Electrode was emphasized. To properly reference
measured values, correction factors must be available as a function of temperature for the specific type of reference
electrode and reference solution used in the measurement assembly. To determine Eh of a sample relative to the
Standard Hydrogen Electrode, measure the ORP of both sample and standard solution (at the same temperature).
Eh of the sample may then be calculated:
Eh = ORP (mV) - ORP , .. (mV) + Eh .. ,. (mV) [1]
% ' reference solution % ' reference solution ^ ' *- -*
The group discussed the limitations on the use of corrected Eh readings. It was recommended that the readings be
used in three ways: 1) as a general guide to interpret other redox data. 2) for speciation of Fe(II) and Fe(III) in acidic
mine drainage waters, and 3) for gcochcmical speciation calculations to determine mineral stability for selected Fe
and Mn phases that are stable under intermediate Eh conditions. Views from the session participants varied, but it
was recognized that in the absence of a field Eh reading, there was no widely available method of calculating
saturation indices for phases such as Fe(oxy)hydroxides, Mn oxides, and Mn and Fe carbonates. It was further
recommended that Eh values should not be used to determine saturation indices for phases such as FeS, which
incorporate species (HS~) that are not generally responsive to measurement by Pt electrodes (Whitfield, 1974; Walton-
Day et ah, 1990).
Dissolved oxygen (DO) is often the principal oxidizing chemical component in ground water; consequently, DO
can be a key parameter governing the mobility of redox-sensitive contaminants, including organic compounds,
transition metals, and transuranic metals. DO is also important in governing the nature and level of microbial
activity. It is therefore directly tied to the fate of dissolved nutrients and organic contaminants consumed through
metabolic processes (e.g., Baedecker and Back, 1979). The concentration of DO in ground water is controlled by-
local inputs of oxygen-rich meteoric water, microbial respiration, biodegradation of organic matter, and reaction
with reduced mineral phases in the aquifer (Champ et ah, 1979; White et ah, 1990).
91
-------
The group recognized four major methods for determining DO concentrations: 1) the Winkler titrimetric method.
2) the membrane-covered electrode method. 3) the indigo carmine colorimetric method, and 4) the rhodazine D
colorimetric method. Of these methods, the Winkler method (and its modifications) was regarded as the most
accurate and precise technique for determining DO. This method has served as a benchmark for laboratory
comparisons and calibration of the electrochemical and colorimetric techniques (e.g., Reynolds, 1969; Hitchman,
1978; Gilbert et al, 1982). From the field perspective, the Winkler method is cumbersome and it uses reagents that
can be hazardous to manage in the field. Seven different Winkler modifications exist and selection of the
appropriate modification requires prior knowledge of a sample that is often not available during field investiga-
tions.
Because of the difficulties associated with carrying out Winkler titrations in the field, alternative methods such as
membrane-covered polarographic electrodes are widely used (Hitchman, 1978). Membrane-covered electrodes are
simple to use and calibration checks arc usually limited to determining oxygen concentrations in water-saturated air
(American Society for Testing Materials. 1992). Electrodes are ideally suited for in-situ measurements of DO
concentrations and for continuous monitoring of oxygen levels. Dissolved oxygen electrodes do not function at
temperatures greater than about 50 °C, which is not a limitation for most ground waters. Membrane fouling can be
a common cause of difficulty, and electrode measurements can be inaccurate without any indication that poor
results are being obtained. Hydrogen sulfide, thio-organic, and other organic compounds are the most problematic
to continued reliable electrode performance. These species react irreversibly at the electrode surface and
detrimentally affect electrode response and accuracy.
Several colorimetric methods have been developed for the measurement of DO. The indigo carmine (blue) and
rhodazine D (reddish violet) methods are the most prominently used techniques (Gilbert et al., 1982; White et al.,
1990; American Society for Testing Materials, 1995a). These colorimetric procedures provide quick and
convenient methods for accurate field measurements of dissolved oxygen. Colorimetric reagents utilize oxidation-
reduction indicators that upon reaction with dissolved oxygen in water transform from reduced, colorless forms to
oxidized, colored forms. The extent of color formation is proportional to the concentration of dissolved oxygen and
can be measured by visual comparison to sets of liquid color standards or with a spectrophotometer. Color
development by reaction with dissolved oxygen is independent of salinity so that a wide range of sample types from
seaw7ater to fresh water can be analyzed with these methods without correction (Gilbert et al., 1982).
Because the colorimetric reagents involve oxidation-reduction reactions to indicate the concentration of DO, other
redox species in ground water can influence the result of colorimetric determinations (Gilbert et al., 1982).
Concentrations of the easily reduced species Fe(III), Cr(VI), and Cu(II) can lead to erroneously high DO values
when rhodazine D is used (White et al., 1990). A potential cause of false-positive readings using the rhodazine-D
reagent in ground-water studies may be the presence of very fine-grained ferric hydroxide colloids. Both Fc(II) and
Fe(III) as well as nitrite were shown to lead to inaccurate determinations of DO using the indigo carmine reagent
(Gilbert et al., 1982). Hydrogen sulfide does not interfere with either colorimetric technique. The effects of
dissolved organic carbon (DOC) on the use of the rhodazine D and indigo carmine reagents are incompletely
understood.
Winkler, electrode, and colorimetric methods for determining DO have different strengths and weaknesses. For
projects requiring the highest available quality (in terms of accuracy and precision over a wide range of DO
concentrations), the extra effort associated with the Winkler methods is warranted. Where data quality objectives
are not as strict, both electrode and colorimetric methods are acceptable. Electrodes are generally reliable at DO
levels above 1 mg/L, but are less accurate at levels below 1 mg/L. Multiple methods are preferred if high quality-
data are required to meet project objectives (Wilkin et al., 2001). Figure 4.1 shows a decision tree discussed by the
group that may be useful for selecting an appropriate method or methods for DO determination.
The most common and convenient quality check for field measurements of DO is air-saturated water. Air-
saturation is achieved using an aerator or through manual agitation. Members of the group pointed out that using
such methods can cause water to become supersaturated with respect to air. Air-saturated DO values are mainly
dependent on temperature and less so on salinity (Figure 4.2). Thus, when quality check values are reported the
temperature must be specified for the check to be meaningful.
Iron is present in ground water as dissolved ferrous iron (Fe(II)), colloidal particles containing either ferric or
ferrous iron, and as dissolved ferric iron (Fe(III)). The relative and absolute abundances of these iron forms vary
widely depending mainly on pH and redox state of ground water. Dissolved forms may be unassociatcd or
92
-------
Use either
electrodes or
coiorimetric methods
Winkler titration not
recommended
Wlnkler (azide) and
HR coiorimetric
(indigo carmine)
DO Electrode
IWinkler
(permanganate)
LR cobrimetric
(rhodazine D)
Figure 4.1 Decision tree for selecting method for determining concentration of DO.
14
12-
10-
I 8^
o
o
6
4
2 -I
0
Air-saturated, 0,001 m Nad
(TDS=67 mg/L)
Air-saturated, 0,1 m NaCI
(IDS=5818 mg/L)
Most ground waters
0 10
Figure 4.2 Air-saturated DO concentration in water.
20
30
T, °C
93
-------
complexed with organic or inorganic ligands. Dissolved iron is generally considered to be that fraction passing
through a 0.45 urn filter paper, however, water filtered in this way may still contain colloidal iron that will dissolve
upon addition of an acid.
Ground waters containing dissolved oxygen at concentrations above 1 mg/L will seldom contain concentrations of
dissolved iron above 1 mg/L. Exceptions to this general rale are more acidic ground waters (pH<4) in which the
solubility of ferric oxyhydroxides is sufficiently high to be detected using standard methods of analysis. The
solubility of ferrous iron increases under reducing conditions, where oxygen is depleted and Eh values are low
(<0 mV). In zones of intense iron reduction, concentrations of ferrous iron may reach levels as high as several
hundred mg/L.
Dissolved Fe(II) concentrations in ground waters are commonly measured using colorimetric techniques with
complexing reagents, such as 1,10 phenanthroline (Tamura et al., 1974), ferrozine (Stookey, 1970 as modified by
Gibbs, 1976; To et al., 1999), and bipyridine (e.g., Baedecker and Cozzarelli, 1992). There was general agreement
among the group that ferrous iron measurements using colorimetric reagents should be completed soon after
sample collection. Minimizing the amount of time between sample collection and Fe(TI) measurement is necessary
because of the relatively fast oxidation rate of ferrous iron at near-neutral pH. Ferrous iron oxidizes following the
overall reaction
Fc~ + 5/2H2O + 1/4O, = Fe(OH)3 + 2W
Hie abiotic rate of ferrous iron oxidation is principally controlled by the partial pressure of O2, pH. and alkalinity.
At pH<3.5. the oxidation rate of ferrous iron is slow and independent of pH. At pH below about 5, however, iron-
oxidizing bacteria greatly increase the rate of iron oxidation. For purposes of sample preservation biotic ferrous
iron oxidation is avoided by filtration and acidification. At pH>3.5, a pH-dependent rate equation is given by Eary
and Schramke (1990):
-d[Fe2+l/dt = ^Fe2+]P O2[OH-12
Figure 4.3 shows the rate of ferrous iron consumption via oxidation as a function of pH (6-8) and PQ2 (0.02 and
0.2 bars) estimated using this equation. Note that at pH 7, ferrous iron concentrations will decrease by a factor of
two in < 15 minutes at air-saturated conditions. A sample recovered from a well and brought to the surface does not
necessarily experience the full atmospheric P O2 of 0.2 bars unless it is continuously agitated. A more realistic
value of P O2 = 0.02 bars still imparts a fast loss of Fe(II), especially at pH > 7 (Figure 4.3). These calculations
illustrate the importance of carrying out Fe(II) measurements immediately after sample collection.
A common practice for estimating Fe(III) is to measure total iron (EFe) on filtered and acidified samples and
subtract from that quantity' the concentration of ferrous iron from measurements made in the field. The group
agreed that no assumptions should be made about the speciation of the iron fraction represented by SFe-Fe(II). At
near-neutral pH and in the absence of significant concentrations of DOC there should be reasonable agreement
between total iron and Fe(II). However, XFe on filtered samples should not be assumed to be equal to Fe(II) due to
the potential formation of soluble Fc(III) - DOC complexes. Measurement of total iron can in many instances be
a useful quality check for Fe(II) measurements made in the field.
Acid mine drainage represents a special case with respect to iron measurements because at low pHthe solubility-" of
Fe(III) increases significantly. An accurate determination of the abundances of Fe(II) and Fe(III) is especially
important because such waters are typically rich in iron and charge balance calculations strongly depend on the
Fe(II)/Fe(III) ratio (see To et al., 1999). When dealing with low pH waters it is desirable to directly measure Fe(III)
concentrations (To et al., 1999).
Field quality checks can be carried out using purchased or prepared iron solutions. Mohr's salt (ferrous ammonium
sulfate) can be used as a standard ferrous solution. Standard solutions should be prepared daily and ideally should
be made from dcoxygcnated water, prepared by purging with nitrogen gas.
Dissolved sulfur may be present in ground water as sulfate (most oxidized form), sulfide (most reduced form), or
as species with oxidation states intermediate between sulfate and sulfide including polysulfides, sulfitc, and
thiosulfate. Concentrations of the intermediate sulfur species are typically low in ground water and their
occurrence is believed to be restricted to redox interfaces that separate reducing hydrogen sulfide-bearing waters
from oxygenated waters. Such interfaces can occur within aquifers or in mixing zones between ground and surface
waters. Sampling techniques and methods for determining the concentration of intermediate sulfoxyanions can be
94
-------
120
0 5 10 15 20 25 30
Figure 4.3 Ferrous iron oxidation kinetics at pH 6-8 and PO2 at 0.02 and 0.2 bars.
246
Ferrous iron, ppm
10
Figure 4.4 Iron monosulfide solubility at pH 6-8 (Ionic strength = 0.0).
95
-------
found in Moses et al. (1984). Sulfate concentrations in ground water vary widely from below 1 to several thousand
milligrams per liter. The highest concentrations of sulfate are typically found associated with acidic surface
drainage produced from the oxidative weathering of metal sulfides. Sulfate concentrations in ground water are
measured using gravimetric methods (precipitation as barium sulfate). the barium dichloride turbidimetric method.
or using laboratory techniques such as ion chromatography or capillary electrophoresis. Preservation is generally
not a problem for sulfate unless samples are especially rich in DOC that can be respired by sulfate reducing
bacteria. Because samples collected for anions are typically not acidified, ferrous iron-rich samples may be
problematic if sulfate becomes incorporated into iron hydroxide precipitates that form prior to sample analysis.
Hydrogen sulfide is produced in anoxic systems through the bacterial reduction of sulfate. Measurements of total
sulfide in ground water include H2S(aq), HS~, and acid-soluble metal sulfides present in colloidal suspended
materials. These species can react rapidly in the presence of oxygen so samples must be collected with a minimum
of aeration. Methods for determining hydrogen sulfide concentrations include electrode, gravimetric, iodometric,
and colorimetric techniques.
The handling and storage methods are critical for water samples containing concentrations of dissolved sulfide.
Ideally, measurements are carried out immediately after sample collection. Preservation of samples will otherwise
be necessary to prevent losses through volatilization or oxidation. At pH less than ~6, volatilization of H2S can be
a problem, i.e., H2S(aq)^H2S(gas). Sulfide anti-oxidant buffers (SAOB) are alkaline solutions that contain
complexing agents such as EDTA and a reductant such as ascorbate. Hydrogen sulfide is then released from the
alkaline solutions by acidification and purging with an inert gas such as nitrogen or argon.
The most commonly used methods for determining dissolved sulfide concentrations are modifications of the
mcthylcnc blue technique described by Cline (1969). In this method, reduced sulfur species (H,S, HS", S2~) react
with diamine (N,N-dimethylphenyl-l,4-diamine) in the presence of ferric chloride to form a blue complex that is
measured spectrophotometrically at a wavelength of 670 nm. Visual colorimetric comparison kits are also
available. Detection limits down to 0.01 ppm sulfide are obtainable. Standard solutions can be prepared by
dissolving into dcoxygcnatcd water quantities of sodium sulfide (Na2S x 9H,O). The concentration of sulfide is
then measured by adding excess iodine and back-titrating with thiosulfate. To minimize oxidation of sulfide
species, reagents should be added to samples immediately after sample collection followed by spectrophotometric
measurement.
Sulfide electrodes that use silver sulfide as the sensing element are generally not used in field studies. Specific ion
electrodes are better suited for laboratory studies conducted under controlled conditions. Long response times and
poisoning of the reference electrode by sulfide have limited the use of sulfide-selective electrodes (Lawrence et al.,
2000). Gravimetric determinations of sulfide concentrations are appropriate in situations where concentrations of
dissolved sulfide are > 5 rng/L. Solutions can be directly collected into bottles containing quantities of Zn acetate
or Cd acetate solutions. A 1 L sample of solution containing 5 mg/L dissolved sulfide will produce a mass of
15.2 mg or 22.5 mg of ZnS or CdS precipitate, respectively.
In ground-water systems containing dissolved sulfide, dissolved oxygen concentrations will be generally 0 or
< 1 mg/L. If hydrogen sulfide is present in ground water above 1 mg/L, concentrations of ferrous iron will generally
be negligible. The concentrations of ferrous iron and hydrogen sulfide in surface and subsurface environments are
generally governed by the solubility of iron monosulfide minerals, such as mackinawite. i.e.,
O j O ./ - - °
FeS + H1 = Fe21 + HS'
The equilibrium constant (K^) for this reaction is 10"3-1 where FeS is taken to be crystalline mackinawite (Davison,
1991). In Figure 4.4 the solubility curves of mackinawite are plotted at three pH values from 6 to 8. Note that
solubility decreases as pH rises, and that when concentrations of either iron or sulfide are low the other is
necessarily high.
The common nitrogen species present in ground water include nitrate (NO3~), nitrite (NO2~), and ammonia (NH3).
At near-neutral pH typical of most ground water, ammonia is present predominantly as the ammonium ion (NH +).
Although colorimetric methods are available to quantitate nitrate, nitrite, and ammonia, field determination using
these methods is not generally needed as long as laboratory measurements can be made within specified holding
times.
96
-------
The cadmium-reduction method can be used to determine the amount of nitrate+nitrite in ground water. In this
method, cadmium metal is used to reduce nitrate to nitrite. The nitrite ions then react with sulfanilic acid and
gentistic acid to form an amber-colored compound (American Society for Testing Materials, 1995b). Strong
oxidizing and reducing substances are possible interfering components; ferric iron, in particular, will cause high
results.
Ammonia is produced during the microbiological decay of organic matter. Two colorimetric methods for ammonia
are the Nessler method and the Salicylate method. The Nessler method is probably the most commonly used
method and develops a yellow color proportional to the concentration of ammonia. High concentrations of calcium
and magnesium may cause interferences. Ferrous iron and dissolved sulfide can also interfere with ammonia
determinations using this method.
Samples collected for nitrogen speciation should be preserved by keeping cold (4°C) and measured within specified
holding times, generally < 48 h. Longer holding times are possible by acidifying samples with sulfuric acid and
keeping them cold. Before analysis the samples should be brought back to room temperature and neutralized by-
adding base.
Alkalinity is a measure of the acid neutralizing capacity of dissolved solutes in a water sample and is reported as
equivalents per liter or as mg/L of CaCO,. Alkalinity is not typically included as a redox parameter. The group
included this parameter because alkalinity production often accompanies oxidation-reduction reactions, e.g., in
sulfatc or iron reduction. The alkalinity concentration consists of the sum of titratablc carbonate and titratablc non-
carbonate species in a water sample, including proton-accepting organic compounds. Proton-accepting species
include CO32~. HCO,", OH* (and metal-OH complexes), HS~, PO43~, ammonium, silicate, and borate. The alkalinity
measurement is important for evaluating charge balance of a solution and is a critical component of modeling
approaches using gcochcmical speciation computer packages. Because particulatc materials can be an important
sink for acid, alkalinity determinations should be made in filtered or low-turbidity waters (turbidity- < 5 NTU).
The most accurate determinations of alkalinity are made in the field at the time of sample collection. Fixed end-
point titrations to pH 4.3 are suitable for ground waters with negligible concentrations of dissolved organic matter
or titratable species other than inorganic carbon. For more complex ground waters, especially those rich in
dissolved organic material, fixed end-point titrations will give inaccurate results (see Hcmond, 1990). Total
alkalinity should then be determined by measuring pH as a function of titrated acid to a pH of 3.0. Using this
method, alkalinity is calculated from the volume of acid added to reach the end-point pH. The end-point is
determined graphically by plotting pH as a function of titrant volume and locating the inflection point. The
importance of protecting the sample from oxidation was also discussed. In high Fc(II) waters, the oxidation of
Fe(II) to Fe(III) can produce sufficient acid to bias the titration, resulting in an underestimation of actual alkalinity.
Rapid analysis is recommended, and protection of the sample from aeration is needed for accurate measurements of
alkalinity in iron-rich water.
4.4
A variety of measurement options are available for monitoring the most important redox parameters in ground
water. The core list of parameters identified by this group was Eh, dissolved oxygen, iron speciation, sulfur
speciation, and nitrogen speciation. Selection of the appropriate technique for a given parameter will depend on
specific project objectives. It was recognized that project objectives and principal data uses collected in monitoring
programs change with time as more site-specific information is obtained and conceptual models evolve. There was
general agreement that there exists a need for improved standards for routine field practices. There was strong
agreement among the group that sample collection practices and measurement techniques are among the greatest
challenges to developing reliable site conceptual models.
References
American Public Health Association (APHA), American Water Works Association, Water Environment Federation.
1999. Standard Methods for the Examination of Water and Wastewater, 2ffh Edition. American Public Health
Association, Washington, D.C.
American Society for Testing Materials (ASTM). 1992. Standard test methods for dissolved oxygen in water, ASTM
D 888-92. Philadelphia, Pennsylvania.
97
-------
American Society for Testing Materials (ASTM). 1995a. Standard test methods for low-level dissolved oxygen in
water. ASTM D 5543-94. Philadelphia, Pennsylvania.
American Society for Testing Materials (ASTM). 1995b. Nitrate-Nitrite in water, Test method B, ASTM D 3867-90.
Philadelphia, Pennsylvania.
Baedecker, M. J. and Back. W. 1979. Hydrogeological processes and chemical reactions at a landfill site.
Ground Water, v. 17. pp. 429-437.
Baedecker, M. J. and Cozzarelli. I. M'. 1992. The determination and fate of unstable constituents of contaminated
groundwater. In Groundwater contamination and analysis at hazardous waste sites, eds. S. Lesage and R.
Jackson, Marcel-Dekker, New York, pp. 425-461.
Cline. J. D. 1969. Spectrophotometric determination of hydrogen sulfide in natural waters. Limnology and
Oceanography, v. 14. pp. 454-458.
Champ, D. R., Gulens, J., and Jackson. R. E. 1979. Oxidation-reduction sequences in ground water flow systems.
Canadian Journal of Earth Sciences, v. 16, pp. 12-23.
Davison, W. 1991. The solubility of iron sulphides in synthetic and natural waters at ambient temperature.
Aquatic Sciences, v. 53. pp. 309-329.
Eary, L. E. and Schramke, J. A. 1990. Rates of inorganic oxidation reactions involving dissolved oxygen. In Chemical
Modeling of Aqueous Systems, II, Chapter 30, eds. D. C. Melchior and R. L. Bassett, American Chemical
Society, Washington, B.C., pp. 379-396.
Gibbs, C. R. 1976. Characterization and application of ferrozine iron reagent as a ferrous iron indicator.
Analytical Chemistry, v. 48, pp. 1197-1200.
Gilbert, T W., Behymer, T. D., and Castaneda, H. B. 1982. Determination of dissolved oxygen in natural and
wastewaters. American Laboratory, M. 14, March, pp. 119-134.
Hemond, H. F. 1990. Acid neutralizing capacity, alkalinity, and acid-base status of natural waters contaning organic-
acids. Environmental Science and Lechnology, v. 24, pp. 1486-1489.
Hitchman, M. L. 1978. Measurement of Dissolved Oxygen, John Wiley & Sons, New York, 255 p.
Jackson, R. E. and Patterson, R. J. 1982. Interpretation of Eh and pH in a fluvial-sand aquifer. Water Resources
Research,^. 18, pp. 1255-1268.
Lawrence, N. S., Davis, J., and Compton, R. G. 2000. Analytical strategies for the detection of sulfide: a review.
Talanla, v. 52, pp. 771-784.
Langmuir, D. 1971. Eh-pH determination. \nPmceduresinSedimenlaryPelrology, Chapter 26, ed. R. E. Carver,
Wiley-Interscience, New York, pp 597-635.
Light, T. S. 1972. Standard solution for redox potential measurements. Analytical Chemistry, v. 44, pp. 1.038-1039.
Lindberg, R. D. and Runnells, D. D. 1984. Ground water redox reactions: An analysis of equilibrium state applied
to Eh measurements and geochemical modeling. Science, v. 225, pp. 925-927.
Morris, J. C. and Stumm, W. 1967. Redox equilibria and measurements of potentials in the aquatic environment. In
Equilibrium Concepts in Natural Water Systems, ed. R. F. Gould, American Chemical Society. Washington, D.C..
pp.270-285
Moses, C. O.. Nordstrom. D. K.. and Mills, A. L. 1984. Sampling and analyzing mixtures of sulphate, sulphite.
thiosulphate, and polythionate. 'lalanta, v. 31, pp. 331-339.
Nordstrom, D. K. 1977. Thermochemical redox equilibria of Zobell's solution. Geochimica et Cosmochimica Acta,
v. 41, pp. 1835-1841.
Nordstrom. D. K.. Jenne, E.A.. and Ball, J. W. 1979. Redox equilibria of iron in acid mine waters. \& ACS Symposium
Series 93, Chapter 3, ed. E. A. Jenne. American Chemical Society, Washington, D.C., pp. 51-79.
Reynolds, J. F. 1969. Comparison studies ofWinklcr vs oxygen sensor. Journal Water Pollution Federation, v. 41,
" pp. 2002-2009.
98
-------
Stookey, L. L. 1970. Ferrozine-A new spectrophotometric reagent for iron. Analytical Chemistry, v. 42, pp. 779-781.
Stumm, W. 1966. Redox potential as an environmental parameter; conceptual significance and operational limitation.
In 3rd International Conference on Water Pollution Research, Section .1, No, 13, Water Pollution Control Federation,
Washington D.C., pp. 1-16.
Tamura, H., Goto, K., Yotsuyanagi, T., and Nagayama, M. 1974. Spectrophotometric determination of iron (II) with
1,10 phenanthroline in the presence of large amounts of iron (III). Talanta. v. 21, pp. 314-318.
Thorstenson, B.C. 1984. Tlie concept of electron activity and its relation to redox potentials in aqueous geochemical
systems. U.S. Geological Survey Open-File Report 84-072, 45 p.
To, T. B., Nordstrom, D. K., Cunningham, K. M., Ball, J. W., and McCleskey, R. B. 1999. New method for the direct
determination of dissolved Fe(III) concentration in acid mine waters. Environmental Science and Technology.
v. 33,pp.807-813.
U.S. Geological Survey. 1997. National Field Manual for the Collection of Water-Quality Data. U.S. Geological
Survey Techniques of Water-Resources Investigations, book 9. [http://water.usgs.gov/owq/FieldlVI.anual/
mastererrata.html]
Walton-Day, K., Macalady, D. L., Brooks, M. H., and Tate, V. T. 1990. Field methods for measurement of ground
water redox chemical parameters. Ground Water Monitoring Review^. 10, pp. 81-89.
White, A. F., Peterson, M. L., and Solbau, R. D. 1990. Measurement and interpretation of low levels of dissolved
oxygen in ground water. Ground Water, v. 28, pp. 584-590.
Wilkin, R. T, McNeil, M. S., Adair, C. J., and Wilson, J. T. 2001. Field Measurement of Dissolved Oxygen: A
Comparison of Methods. Ground Water Monitoring and Remediation, v. 21, pp. 124-132.
Whitfield. M. 1974. Thermodynamic limitations on the use of the platinum electrode in Eh measurements. Limnology
and Oceanography, v. 19, pp. 857-865.
99
-------
100
-------
5
of
Donald Kampbell and Frank Chapel le
5.1
A dissolved gas refers to a chemical element in a gaseous phase at ambient temperature. The dissolved gas is
absorbed by a liquid, in this case water, without a chemical interaction occurring. Dissolved gases such as oxygen.
hydrogen, and aliphatic hydrocarbons such as methane, are important components to determine the status of
biodegradation processes in ground water. For example, when dissolved oxygen is present, microorganisms will
preferentially use oxygen as the primary electron acceptor and aerobic metabolism will tend to dominate the
system. When dissolved oxygen is not present, however, the microbial population must use alternative electron
acceptors such as nitrate, ferric iron, sulfate, carbon dioxide, or, in some contaminated aquifers, chlorinated ethenes
in order to carry on respiration. When carbon dioxide serves as the primary electron acceptor (methanogenic
respiration) methane is produced. Thus, concentrations of dissolved oxygen and methane can be useful in
determining the kinds of microbial metabolism occurring in a ground-water system.
In general, decreasing concentrations of oxygen along ground-water flow paths are indicative of aerobic metabo-
lism. Similarly, increasing concentrations of methane along ground-water flow paths indicate the presence of
active methanogenesis in an aquifer. Frequently, the anaerobic process becomes functional after oxygen becomes
depleted. Then nitrates and sulfates can become electron acceptors. Finally, methane can become a major
component characteristic of an anaerobic medium. Although the assay of dissolved gases is important to
characterize biodegradation, other conditions need to be favorable for a given xenobiotic such as the presence of
compatible microorganisms and favorable environmental settings. Plume ground-water that has a decrease in
oxygen and an increase in methane is an important characteristic that should be monitored.
Hydrogen is an important intermediate product of anaerobic microbial metabolism. Under anaerobic conditions.
the oxidation of organic matter requires food chains of symbiotic microorganisms. First, complex organic matter
is partially oxidized by fermentative bacteria with the production of intermediate compounds such as acetate and
molecular hydrogen. Secondly, these fermentative products are fully oxidized by rcspirativc microorganisms
coupled to reduction of mineral electron acceptors such as nitrate, sulfate, ferric iron, and carbon dioxide.
Importantly, chlorinated ethenes such as perchloroethene (PCE) and trichloroethene (TCE) can also serve as
electron acceptors during the metabolism of hydrogen. Concentrations of dissolved hydrogen can be an indication
of the efficiency of reductive dcchlorination in ground-water systems. Generally, high concentrations of dissolved
hydrogen favor efficient reductive dechlorination. The rate of reductive dechlorination at particular field sites is
useful in assessing the viability of employing monitored natural attenuation. Because of this, concentrations of
hydrogen can be a useful indicator parameter.
The complete reductive dechlorination of PCE and TCE results in the formation of ethene. Furthermore, ethene can
then be further reduced to ethane under strongly reducing methanogenic conditions. Thus, the presence of ethene
and ethane can be indicators that reductive dechlorination of PCE and TCE is occurring in a ground-water system.
For these reasons, it is important to measure concentrations of dissolved ethene and ethane at sites where reductive
dechlorination may be occurring.
Abiotic and biological transformations arc important mechanisms controlling natural attenuation as a plume
management strategy. Plume management requires monitoring of the transformation processes to confirm
protection of the environment and human health. Thus, measuring concentrations of oxygen, H2, ethene, ethane,
and methane provide important information about the kinds of microbial processes occurring in ground-water
systems.
101
-------
5.2
The measurement of dissolved gases in ground water can aid in understanding what biochemical processes are
active, or have been active, in ground-water systems. If oxygen has been depleted, then most likely anaerobic
processes have predominated. The presence of hydrogen is useful to stimulate dechlorination of chlorinated
hydrocarbon pollutants.
Gases such as oxygen, hydrogen, and methane are slightly soluble in water. Thus, their concentrations are directly
proportional to their partial pressure as formalized by Henry's Law. Because the Henry's Law constant for each gas
is highly temperature-dependent, the solubility of each may be used to calculate the amount present at saturation for
given temperatures. The solubility varies greatly with temperature. For example, the solubility of oxygen in fresh
water at 1 arm. pressure ranges from 14.6 mg/L and 7 mg/L at 0°C and 35°C, respectively. The solubility of H2 and
methane are similarly variable with temperature. At high temperature conditions where oxygen is least soluble.
biological oxidation increases along with oxygen demand. These critical conditions can limit the purification
capacity to remove waste pollutants.
Because oxygen, hydrogen, and methane are so slightly soluble in water, sampling methods must rigorously
exclude exposure to the air. This is particularly problematic with oxygen, since air is about 21 percent oxygen.
Measurements of dissolved oxygen must always be done in a closed air-free environment, and measurements
should be made as soon after sample collection as possible. In general, all measurements of dissolved oxygen
should be made in the field at the time of sample collection. Similarly, hydrogen measurements are most accurate
when made in the field at the time of sample collection. However, storage methods are now available for
transporting hydrogen samples to the laboratory. Because methane is relatively stable, and because methane is not
a major component of air, methane samples are relatively easy to collect and store for later analysis in the
laboratory. Care must be used when collecting water samples for dissolved gases, especially oxygen and hydrogen.
Exposure to the air will lead to a potential loss or gain of these components. Water sampling and analyses should
be done in a closed air-free system and as soon as possible. Because air is about 21 percent oxygen, it can be a
critical factor of contamination when collecting low level oxygen water samples. Hydrogen and methane are
constituents of air, but in the ppm range. Generally, exchange of these components with air can be overcome by
care during sampling. The expected concentration ranges of dissolved methane and hydrogen in contaminated
ground water are in the ppm range.
The primary factor for quality assurance, besides careful sampling to avoid sample contamination by air, is
instrument calibration. Pressurized cylinders containing known hydrogen or methane, ethane, and ethene can be
used. Concentrations of oxygen in water can be obtained from standard oxygen solubility versus temperature
tables. Generally, 10 percent of the collected samples should be duplicated to measure precision.
5.3 Hydrogen
A routine technique developed to analyze water samples for dissolved hydrogen has been described by Mclnnes
and Kampbell (2000). This is an adaptation of a method originally developed by Chapelle et al. (1997) and is called
the "bubble strip" technique. This method is an application of Henry's Law. After the sample is collected in a
20 mL bubble inside a 250 m.L gas sample vessel, a 2 mL aliquot of the bubble is removed for analysis using a
reduction gas analyzer/chromograph. A typical calibration range for hydrogen is 0.5 to 10.0 ppm.
5.4 Oxygen
Several decades ago, the dominant procedure for determining dissolved oxygen in water was the Winkler, Full
Bottle technique (U.S. EPA, 1979b; Method 360.2). The test depended on oxygen oxidizing divalent manganese to
a higher state of valence under alkaline conditions. The higher valence state manganese would then oxidize iodine
to free iodine in an acid solution. The free iodine released was equivalent to dissolved oxygen. Standard sodium
thiosulfatc then measured the free iodine. Interferences from nitrates were reduced by use of sodium azidc.
Dissolved oxygen can also be analyzed by electrode probe (U.S. EPA, 1979a; Method 360.1. see Appendix A). On
the end of the probe is a thin organic film over two metallic electrodes in a media of liquid electrolyte. The oxygen
that diffuses through the film is reduced electrochemically. A current is generated proportional to oxygen in the
sample. These electrodes are well suited for analyses in water because their portability, ease of operation, and
maintenance make them convenient for field applications. They are less amenable to interferences than the
Winkler test. The principal disadvantage of electrodes for measuring dissolved oxygen relates to the diffusion-
dependent mode of operation. When concentrations of dissolved oxygen are low, oxygen diffusion is very low and
102
-------
the current generated by the electrode is correspondingly low. It is thus difficult for electrodes to accurately
measure dissolved oxygen concentrations below 0.5 mg/L. Because many ground-water systems contain low
concentrations of dissolved oxygen, the use of electrodes can be problematic in some systems (see Session 4
Summary).
More recent chemical methods for measuring dissolved oxygen concentrations were developed by CHEMetrics.
Inc. One method, the indigo carmine method (ASTM D 888-87), is used for dissolved oxygen concentrations in the
1-10 ppm range. The chemistry of this method relies on the reaction of reduced indigo carmine with oxygen to
produce a blue product. Another method, the rhodazinc D method (ASTM D 5543-94). is used to measure
dissolved oxygen concentrations in the 0.1 to 1.0 ppm range. The chemistry of this method relies on the reaction
of rhodazine D with oxygen to form a bright pink reaction product. Neither of these methods is subject to
interferences from temperature, salinity, or dissolved gases, which often do interfere with dissolved oxygen
electrodes. For both the high range (indigo carmine) and low range (rhodazine D), the reagents are scaled in
ampules designed to fit into a plastic beaker. The beaker is attached to a stream of water pumped from the well
(sealed from the air) and is allowed to overflow. After a brief period of flushing (~30 seconds), an ampule is
inserted into the stream of water and the tip cracked. This draws water into the ampule where it reacts with the
reagents. The intensity of the color development is directly proportional to the dissolved oxygen concentration.
which can be compared visually to standards supplied with the reagents. This method is fast, reproducible, and
minimizes sample exposure to atmospheric oxygen.
5.5
Glass serum bottles are filled with ground-water samples at the field site, preserved with diluted sulfuric acid, and
capped with air-tight septa. Ten percent of the water is then removed to create a gaseous headspace. After
equilibrium by shaking the bottles, an aliquot of the headspace is removed by a gas-tight syringe for analysis by gas
chromatography. Dissolved methane, ethane, and ethene in the water sample can then be completed using Henry's
Law as explained by Kampbell and Vandegrift (1998). Quantitation limits for methane, ethane, and ethene by this
method were 0.001, 0.002, and 0.003 mg/L, respectively.
5.6
Methods used for field site characterization studies to determine the status of biodegradation processes in ground
water as related to dissolved gases were discussed. Emphasis was directed toward dissolved oxygen, hydrogen, and
aliphatic hydrocarbons such as methane, ethane, and ethene. A method for each of these parameters is listed that
has been used over a number of years to aid in identifying processes of natural attenuation.
References
Chapelle, F. H., Vroblesky, D. A., Woodward, J. C., and Lovley, D. R. 1997. Practical considerations for measuring
hydrogen concentrations in groundwater. Environmental Science and Technology, v. 31, pp. 2873-2877.
Kampbell. D. H. and Vandegrift, S. A. 1998. Analysis of methane, ethane, and ethylene in ground-water by a
standard gas chromatographic technique. Journal of Chmmatographic Science, v. 36. pp. 253-256.
Mclnnes, D. M. and Kampbell, D. H. 2000. The bubble stripping method for determining dissolved hydrogen (FL,)
in well water. Field Analytical Chemistry and Technology, v. 4, pp. 283-296.
U.S. Environmental Protection Agency. 1979a. Dissolved Oxygen by the Membrane Electrode Technique. Methods
for Chemical Analysis of Water and Wastes, EPA Method 360.1, EPA 600/4-79-020. U.S. EPA, Government
Printing Office, Washington, DC.
U.S. Environmental Protection Agency. 1979b. Dissolved Oxygen by the Modified Winkler Full-Bottle Technique.
Methods for Chemical Analysis of Water and Wastes, EPA Method 360.2, EPA 600/4-79-020. U.S. EPA,
Government Printing Office, Washington, DC.
103
-------
104
-------
6
Robert G. Ford
6.1
Monitoring oxidation-reduction (redox) processes for ground-water remediation must be applied both to the liquid
and solid phases in the remediation zone. Many of the redox sensitive elements that participate in chemical
reactions during subsurface remedial processes are cycled between these two phases during the reaction lifecycle.
Examples of important redox sensitive elements include carbon, iron, nitrogen, and sulfur. Site characterization
and monitoring may require the development of a site-specific database for the composition of the aquifer solid
matrix and the capacity of the matrix with respect to remedial performance measures.
The level of effort for solid phase characterization will depend on the specific remedial technology employed. Two
examples include in-situ redox manipulation of the aquifer matrix versus emplacement of a new solid matrix to
serve as a reactive barrier. Both treatment technologies require knowledge of the mechanisms and capacity of
specific redox reactions to achieve the treatment goals. For the first example, one must identify the mass and
distribution of the reactive components naturally occurring within the contaminated aquifer. This may require an
extensive level of effort to characterize both the abundance and distribution of redox-sensitive phases as well as
their reactivity during the remedial process. For the second example, characteristics of the reactive solid
components are typically known prior to barrier installation. However, it may be necessary to cam- out field
measurements to characterize changes in the reactive phase(s) during remedial monitoring activities.
6.2
The overall goal of mis breakout session was to identify' the key solid phase parameters relevant to redox-controlled
processes active during in-situ remediation of ground water. As part of this discussion, the group was to identify:
1) a list of solid phase parameters that addressed a variety of remedial technologies, 2) appropriate methods of
sample collection specific to characterization needs, and 3) available analytical methods for estimating solid phase
parameters. Throughout this discussion, the group was tasked with identifying characterization issues with the
greatest level of uncertainty and making recommendations concerning technical areas where the U. S. Environmental
Protection Agency should focus investment of resources.
6.3
A remedial case study was evaluated to provide a specific frame of reference for discussion of solid phase
characterization analyses that may be required as part of remedial design and monitoring. The case study was the
remedial effort implemented by the U. S. Department of Energy to treat chromate-contaminated oxic ground water
via in-situ redox manipulation of aquifer sediments at the Hanford Site in southeastern Washington (Williams ct al.,
2000). The goal of this remedial effort was to reduce naturally occurring ferric iron, Fe(III). in aquifer sediments
to soluble and solid forms of ferrous iron via injection of a buffered solution of a strong reductant, dithionite. The
ferrous iron, Fe(II). subsequently reacted with dissolved chromium; reducing Cr(VI) to insoluble Cr(III)-bearing
precipitate phases (Amoncttc, 2000). While the reaction of Fc(II) with Cr(VI) was the specific process controlling
chromium stabilization, the overall success of the remedial process was dependent on the efficiency of the initial
reduction of native Fe(III) via dithionite injection. Thus, solid phase characterization was focused towards
evaluating the production and longevity of Fe(II) within the treatment zone.
Design of the remedial process required analysis of the mass and distribution of reactive iron in aquifer sediments
and the rates of iron reduction/oxidation and non-beneficial reactions involving dithionite. Non-beneficial
reactions included those that consumed dithionite without production of Fe(IT) and the release of potentially toxic
elements during dissolution of iron oxide minerals. Sediment cores were collected within the proposed treatment
zone for mineralogical characterization and to evaluate rates and products for beneficial and non-beneficial
reactions. All tests were carried out using the <4 mm sediment fraction, with the assumption that the largest
105
-------
proportion of reactive iron would be contained in this size fraction. Following is a list with test description and
purpose used for site characterization and remedial performance evaluation.
Procedures that were performed included establishing the uniformity of site geology through well-site boring logs
and characterization of physical properties on split tube samples, including porosity, bulk density, and particle size
distribution. In addition, prc- and post-dithionitc injection hydraulic tests were performed to establish treatment
design criteria and to evaluate potential changes in formation hydraulic properties following dithionite injection.
The distribution of iron in sediment minerals was examined for the <4 mm size fraction for pre- and post-treatment
samples. The predominant phases identified for pre-treatment samples included iron adsorbed to mineral surfaces.
amorphous and crystalline Fe(III) oxides, Fe(II) carbonate, and Fe-bearing clay minerals. Post-treatment samples
had a decreased fraction of amorphous Fe(III) oxide and increased fractions of adsorbed Fe(II) and Fe(II)
carbonate. Mass balance calculations suggested that dithionite reduction primarily converted amorphous Fc(III)
oxides to adsorbed Fe(II). Separate experiments indicated that adsorbed Fe(II) provided the greatest reactivity for
Cr(VI) reduction. Thus, the observed changes in iron mineralogy provided verification of the success of dithionite
reduction.
Ferric iron reduction in the presence of dithionite (S,O42~) occurs via a chemical reaction between Fe(III) and a
sulfoxyl free radical (SO,"*) that exists in equilibrium with S2O42~ (Amonette, 2000). However, dithionite is also
subject to pH-dependent disproportionation reactions resulting in other dissolved forms of sulfur (S2O32~ and SO32")
that reduce Fe(ITI) at a much slower rate. These latter reactions are non-beneficial and occur in solution or may be
catalyzed at mineral surfaces. Tests were carried out to evaluate the optimal reaction pH and the rate at which
aquifer sediments catalyzed dithionite disproportionation. These parameters were required to establish the extent
to which the injection solution pH had to be buffered and the optimal rate of injection into the aquifer in order to
minimize the impact of dithionite disproportionation.
Oxidation capacity was defined for mis study as the degree of dithionite consumption during the reduction of
Fe(III) in aquifer sediments. These results were corrected for disproportionation reactions that resulted in
dithionite oxidation without concurrent Fe(III) reduction. An operational definition for oxidation capacity was
required, since not all Fc(III) was accessible to reduction due to physical constraints or slow rates of reduction for
certain mineral phases. This result highlights the need to tailor characterization methods for consistency with the
remedial technology.
Sediment reduction capacity for post-treatment sediment samples was characterized by exposure of reduced
sediments to oxygen-saturated synthetic ground water. Rates of Fe(II) oxidation in 1-D column studies, calculated
from oxygen consumption, were used to estimate the lifetime of Fe(II) within the in-situ reaction zone. These tests
were run on sediments collected prior to and following dithionite injection. Reduction capacity for pre-treatment
sediments was evaluated in the laboratory following dithionite treatment. Comparable reduction capacities were
measured for pre- and post-treatment sediment samples, providing a performance measure for success of the
reduction stage of the remedial effort. The rate of Fe(II) oxidation was also used to provide an estimate of in-situ
reaction zone longevity.
The stability of reduced chromium was evaluated during sediment re-oxidation with oxygen-saturated water to
simulate field conditions (Fruchter et al., 2000; Williams et al.. 2000). No leaching of chromium occurred as a
result of re-oxidation. In addition, the potential mobilization of native trace metals in the aquifer sediments that are
sensitive to changes in ground-water redox was examined during reduction with dithionite. Release of metals such
as arsenic did occur, but soluble concentrations were elevated only in reduced sediments. Soluble metal
concentrations decreased significantly during sediment re-oxidation. These results indicated that sediment
reduction effectively stabilized chromium and did not negatively impact water quality due to concurrent trace metal
leaching. In order to minimize soluble trace metal concentrations in the treated aquifer, the spent dithionite
106
-------
injection reagent and several additional injection volumes of ground water were withdrawn following the
predetermined iron reduction period.
This example illustrates the rationale and procedures that may be a necessary component of remedial evaluation
and monitoring. Procedures such as assessment of site geology may be common to a wide range of remedial
strategies. Assessment of aquifer sediment capacity parameters may also be common, but may require tailored
procedures that are more consistent with remedial objectives. In summary', it is clear that solid phase characteriza-
tion procedures will likely be required at all stages of remedial design, implementation and monitoring. In order to
provide a broader scope to encompass other remedial technologies, the discussion panel developed a more
universal list of potential characterization procedures and evaluated their utility for a wider range of remedial
technologies. These observations are discussed in the following section.
6.4 of Information
The discussion group developed a list of potential solid phase measurements to be applied as part of site assessment
and monitoring. Some measurements would be required as part of the initial remedial assessment while other
measurements would primarily be implemented as part of remedial monitoring. Measurement of the listed
parameters may require field- and/or laboratory-based tests. A general description is provided in Table 6.1 for each
of the discussed parameters. There may be additional parameters that require characterization for specialized
remedial technologies, but the list in Table 6.1 is generally comprehensive.
The list in Table 6.1 was then used as a reference point for evaluating data requirements for assessment, design, and
monitoring of a select number of remedial technologies. The outcome of this activity is illustrated in Table 6.2.
The remedial technologies mat were evaluated included oxidant injection (permanganate), reductant injection
(dithionite), reactive barrier installation (zero valent iron), enhanced biological degradation, and monitored natural
attenuation (MNA). Several general trends were observed:
1) The reactive barrier technology requires the least amount of solid phase characterization, since it was
assumed that properties of the barrier material would be fully characterized. This also assumes
sufficient knowledge of the influence of ground-water geochemical parameters on the (bio)chemical
reactions controlling contaminant remediation within the barrier. However, some characterization of
the native aquifer material may be required to address potential deleterious effects from alteration of
the ground-water redox state.
2) Injection (oxidant/reductant), enhanced biological, and MNA technologies require more extensive
solid phase characterization, since the aquifer matrix (mineralogy and microbial population) is an
integral reactant in the remedial technology.
3) Use of MNA as a remedial technology for an inorganic contaminant will necessitate the greatest extent
of solid phase characterization. This observation is due to the importance of inorganic contaminant
immobilization onto aquifer solids as the remedial endpoint.
Analysis of this select group of remedial technologies indicates that a single list of required solid phase
characterization parameters will not apply for all technologies. Of the listed parameters, only solubilized non-
target contaminant and aquifer permeability were considered universal for all technologies. In addition, a
determination must be made whether solid phase characterization is performed on the entire sample or a specific
size fraction. Parameters that pertain to a reaction capacity, such as acid neutralizing capacity, should be assessed
forthe entire sample. However, determinations of mineralogy may only be feasible following initial isolation of a
specific size fraction, e.g., determination of the clay mineral content. Thus, it was generally recommended that a
preliminary case-specific assessment should be made to identify solid phase parameter measurements critical to
successful implementation of the proposed remedial technology.
6.5 Collection
Recommendations for collection and preservation of solid phase material and its derived components were solicited
from the discussion group. In general, collection and preservation methods were grouped based on the desired
parameter to be characterized. The recommendations listed below were specific to the retrieval of subsurface
material from the saturated zone. This type of sample will generally display the greatest instability during above-
ground processing.
107
-------
Mineralogy. The greatest concerns for preservation of mineralogy in aquifer materials were the prevention of
contact with oxygen and removal of pore water. For example, iron- and manganese-bearing minerals may exist in
a reduced state within the saturated zone. Minerals that are stable under a reducing environment are subject to
significant alteration upon exposure to oxygen. In addition, solid phase structural and chemical transformations are
commonly mediated or facilitated by the pore water. Thus, removal of pore water may act to retard or impede
transformation. If dewatering is required to minimize sample alteration, it was recommended that non-aggressive
techniques, such as draining by gravity, be employed for removal of pore water.
General methods for sample handling were grouped into those employed in the field and laboratory. In the field.
sample cores should be immediately capped and frozen. Sample freezing can be accomplished either by-
submersing in liquid nitrogen or placement in a portable freezer located in the field. Following transport to a
laboratory setting, frozen materials should be thawed under an oxygen-free or inert atmosphere, e.g. within a glove
box. Residual pore water may then be displaced using anon-reactive solvent such as acetone. However, while this
treatment will leave the mineralogy intact, it will likely alter the solid organic matter. Drying at ambient
temperature in air or an inert atmosphere should be employed in order to minimize alteration of amorphous
minerals. Significant changes in reactivity may result from changes in mineral structure and surface area due to
drying at even slightly elevated temperatures (e.g., Stanjck and Wcidlcr, 1992). Some of the group participants
have had success storing materials in de-oxygenated water. However, extended periods of storage should be
avoided, since sample mineralogy will alter with time during approach to a new equilibrium with the fresh de-
oxygenated water.
Microbiology. The method used to preserve the microbial community within the core depends on the goal of
characterization. If characterization requires quantification of bacterial counts, then the solids should be
maintained at a reduced temperature, e.g. 4 °C. but not frozen. However, if there is no need to maintain a viable
microbial population, then samples can be frozen in the field. No other recommendations were made with regard
to microbial characterization, since the general group consensus w7as that this is a less problematic issue than
mineralogical preservation.
Pore Water. It was recommended that pore water be isolated from the core material in the field prior to
preservation of core solids. This may be accomplished for coarse-grained material by allowing the pore water to
drain into a collection container. Pore water may be extracted from fine-grained material by manual pressure
filtration using a syringe. Alternatively, pore water may be isolated by ccntrifugation if the location for sample
collection is close to laboratory facilities. Pore water should not be isolated in the laboratory after the core has been
subjected to freezing and thawing. Preservation of isolated pore water should be consistent with the desired
analytical measurements.
Insufficient time was available to make recommendations for methods to be employed to characterize all of the
solid phase parameters listed in Table 6.1. Discussion was thus directed towards a select number of parameters. In
all cases, the subsurface material bulk density should be measured so that the chemical parameters listed in
Table 6.1 can be reported on a mass basis. In addition, any testing protocol must include evaluation of the
performance of the analytical procedures. Methods referenced in Tables 6.3 and 6.4 should be initially evaluated
to assess their appropriateness for site materials. The capacities of materials derived from impacted environments
may exceed the analytical range for which the methods were originally developed.
Oxidation Capacity. Methods for the determination of an 'oxidation capacity' for solids are listed in Table 6.3. The
most aggressive approach involves digestion of the solid with a strong chemical reductant, i.e., aqueous Cr(II).
This approach may overestimate the effective oxidation capacity of the solid material. A more targeted approach
may be achieved via dithionite titration, but this may only be useful for evaluation of dithionite injection as a
remedial technology. Likewise, methods designed to target iron oxides will only be valid for aquifer sediments in
which the oxidation capacity is dominated by these minerals. Development of a tailored method may prove most
useful, but this will necessitate additional research and development.
Reduction Capacity. Several methods are documented for the determination of a 'reduction capacity' for the solid
material. A selection of these methods is listed in Table 6.4. Methods for determining chemical oxygen demand
(COD) and organic matter content were not developed for the purpose of collecting data relevant towards
implementation of a remedial technology. The reducing capacity may be under- or overestimated relative to the
remedial target. However, these methods are well documented and have been commonly employed for solid phase
characterization in soil and waste treatment studies. The method for measuring dissolved oxygen demand
108
-------
developed at the Pacific Northwest National Laboratory was designed to provide data specific to the remedial
technology employed (Williams et al, 2000). This method involves measuring oxygen demand in a continuous
Table 6.1 Solid Phase Parameters Requiring Measurement as Part of Remedial Assessment, Design, and Monitoring
Solid Phase Parameter
Oxidation Capacity
Reduction Capacity
Neutralization Capacity
Contaminant Sorption
Capacity
Ion Exchange Capacity
Sorbed Contaminant
Solubilized Non-Target
Contaminant
Reagent Stability
Mineralogy (Bulk and Trace)
Non-Target Solid Phase
Contaminant
Extractable Fe/Al/Mn
Extractable Sulfide
Total Organic Carbon
Total Inorganic Carbon
Reduction/Oxidation Reaction
Rate
Microbial Activity or
Physiology
Microbial Population
Aquifer Permeability
Parameter Description
Capacity of sediment to oxidize a reduced chemical (contaminant or
introduced remedial reagent)
Capacity of sediment to reduce an oxidized chemical (contaminant or
introduced remedial reagent)
Capacity of solid phase to buffer change in pH (acid or alkaline)
Total mass of contaminant that can be partitioned to solid phase by
various mechanisms
Total mass of charged ions that can be partitioned to solid phase via
an electrostatic mechanism
Mass of contaminant that is partitioned to solid phase
Mass of non-target contaminant associated with solid phase that may
be solubilized as a result of remediation
Identification of undesirable by-products produced during reaction
between remedial reagent and contaminant/ solid phase
Identity of mineral phases present in various size fractions of solid
phase
Non-target contaminant that may negatively interact with remedial
technology
Mass of Fe/Al/Mn extracted from solid phase using reagents designed
to attack specific mineral phases
Mass of sulfide extracted from solid phase using reagents designed to
attack specific sulfidic mineral fractions (see Appendix A)
Mass of carbon associated with organic solid phases in sediment (see
Appendix A)
Mass of carbon associated with inorganic solid phases in sediment
The rate at which solid phase will reduce/oxidize (consume)
oxidizing/reducing reagent
Characterization of the microbial processes/characteristics controlling
contaminant degradation/stabilization
Identification of the species of microbes that inhabit contaminated
solid phase
The ability of aquifer material to transmit ground water based on
measurement of porosity
109
-------
Table 6.2 Solid Phase Measurements Applied as Part of Site Assessment and Monitoring for a Selected Number
of Remedial Technologies.
Measured
Parameter
Oxidation Capacity
Reduction Capacity
Neutralization
Capacity
Contaminant
Sorption Capacity
Ion Exchange
Capacity
Sorbed Contaminant
Solubilized Non-
Target Contaminant
Reagent Stability
Mineralogy (Bulk
and Trace)
Solid Phase
Contaminant
Extractable Fe/Al/Mn
Extractable Sulfide
Total Organic Carbon
Total Inorganic
Carbon
Reduction/Oxidation
Reaction Rate
Microbial Activity or
Physiology
Microbial Population
Aquifer Permeability
Remedial Technology
Oxidant
Injection
(Permanganate)
X
X
X
X
X
X
X
X
X
Reductant
Injection
(Dithionite)
X
X
X
X
X
X
X
X
(X)
X
X
Reactive
Barrier
(TNI)
X
(X)
X
[X]
X
Enhanced
Biological
Degradation
X
X
X
(X)
X
X
X
X
X
[X]
X
MNA
Inorganic
Contaminant
X
X
X
X
X
X
X
X
X
X
X
[X]
[X]
X
110
-------
Table 6.3 Methods That May Be Employed for Estimating an Oxidation Capacity for Solid Materials
Method
Cr(II)
Digestion with Ti(III)-EDTA
Titration/digestion with dithionite
solution
Exposure to treatment reagent
Estimate based on mineralogy
Source
(Barcelona and Holm, 1991a)
(Barcelona and Holm, 199 Ib)
(Barcelona and Holm, 1992)
(Ryan and Gschwend, 1991)
(Loeppert and Inskeep, 1996)
(Williams etal., 2000)
Not Documented
Not Documented
Comments
Most aggressive but oxygen-free
atmosphere recommended; high
estimate
Developed for extraction of Fe
oxides; applicability limited to iron
oxide dominated solids
Valid for only specific remedial
technology; targets Fe oxides
Case specific evaluation; requires
preliminary research and
development effort
No consensus values for reference
minerals; unreliable value
Table 6.4 List of Methods That May Be Employed for Estimating a Reduction Capacity for Solid Materials
Method
Chemical Oxygen Demand by
digestion with acid dichromate
Digestion in hydrogen peroxide
solution
Dissolved oxygen consumption in
air-saturated water
Exposure to treatment reagent
Estimate based on mineralogy
Source
(U.S. EPA, 1979)
(Barcelona and Holm, 1991a)
(Barcelona and Holm, 1991b)
(Barcelona and Holm, 1992)
(Nelson and Sommers, 1996)
(Williams etal., 2000)
Not Documented
Not Documented
Comments
Precipitate coatings if pH not
buffered; high estimate
Developed for quantifying organic
matter content
Dynamic column test with
mathematical simulation; test design
must minimize gas diffusion from
external sources; time-consuming,
but realistic
Case specific evaluation; requires
preliminary research and
development effort
No consensus values for reference
minerals; unreliable value
111
-------
flow column configuration. Variation of column flow parameters and mathematical simulation may be required to
arrive at a reliable estimate. Implementation of this method would strongly depend on the level of heterogeneity of
the aquifer solids within the treatment zone and the relative value of the data for successful implementation of the
remedial technology. As with recommendations for oxidizing capacity measurements, method development or
modification may be required to optimize the usefulness of the derived data.
Acid-Base Neutralizing Capacity. The ability of the aquifer solids to buffer changes in ground water pH is a critical
parameter for the success of a remedial technology dependent on redox reactions. Many redox reactions involve
the production/consumption of protons/liydroxyl. and the reaction rates may be optimal only within a restrictive pH
range. Thus, the acid-base neutralizing capacity of the aquifer material must be known in order to evaluate the
efficiency of the desired redox reaction(s). It may prove necessary to supplement the aquifer neutralizing capacity
depending on the remedial technology that is employed (e.g., Williams et al., 2000). Determination of the acid-
base neutralizing capacity may be readily achieved through use of a standardized acid/base solution. The primary
decision with regard to quantification of the capacity is the time for equilibration at the desired pH endpoint, since
part of the neutralizing capacity of a soil is derived from rate-dependent mineral dissolution reactions. A reaction-
based determination consistent with anticipated site conditions provides the most reliable estimate. It is
inappropriate to base neutralizing capacity estimates on bulk mincralogical content (e.g., total carbonate analysis),
since a fraction of the mineral content of the aquifer may not be reactive under site conditions.
Microbially Reducible Iron. The recognition that degradation of organic contaminants in ground water can be
coupled to the microbial reduction of ferric iron in aquifer sediment minerals has prompted the development of
screening methods for the estimation of solid phase microbially reducible iron (e.g., Lovley and Phillips, 1987).
Quantification of the reducible ferric iron fraction by this type of method typically involves the determination of
dissolved iron. The inherent assumption behind such a test is that solid phase ferric iron that is easily reduced under
mild conditions by a chemical reductant can be correlated to the fraction of microbially reducible iron. However,
there was no consensus within the group concerning the iron mineralogy associated with this type of extraction. In
addition, there is evidence that iron bound in clay minerals may serve as a terminal electron acceptor in the absence
of iron dissolution (Ernstsen et al., 1998). Thus, estimates from this type of extraction should be viewed with
caution. There was group consensus that observed changes in the 'reducible' solid phase iron fraction within an
aquifer sediment may be used as an indicator of iron serving as a terminal electron acceptor. However, this
conclusion is only valid if the absence of abiotic iron reduction reactions can be verified. There is also evidence for
reduction of solid phase iron in soils through biologically controlled reactions for which no terminal electron
acceptor process is active (Schwertmann, 1991).
6.6
It was clear from this discussion that a single, clear-cut approach towards characterization of solid phase redox
parameters is not feasible. While technical consensus may be reached for a select number of analytical approaches,
a complete characterization effort will require a high degree of technical expertise. However, there was consensus
that this type of analysis is critical towards the successful implementation of many in-situ remedial technologies.
Thus, the discussion group arrived at the following overall conclusions:
1. Solid phase characterization will play a significant role for some remedial technologies.
2. More solid phase characterization will be required for technologies that rely on the inherent or
enhanced treatment capacity of the aquifer matrix versus installation of a barrier technology.
3. Evaluation of several solid phase characteristics may be of common importance to a variety of
remedial technologies, but specificity based on site, contaminant, or remediation technique prevents a
universal approach.
4. Recommendations can be made for best practices relative to characterization of solid phase redox
parameters, but it is not technically feasible to issue specific (restrictive) requirements.
5. Synthesis and documentation of existing methods and development of new methods is required to
fulfill data needs.
Solid phase characterization issues have generally been avoided from the standpoint of regulatory guidance due to
the complexity of the technical issues. However, the need for a consistent framework for evaluating solid phase
redox processes is significant with increased implementation of remedial technologies that manipulate or monitor
solid phase processes.
112
-------
References
Amonette. J. E. 2000. Iron Redox Chemistry of Clays and Oxides: Environmental Applications. In Electrochemistry
of Clays, ed. A. Fitch, Clay Minerals Society, Aurora. Colorado.
Barcelona, M. J. and Holm, T. R. 1991a. Oxidation-reduction capacities of aquifer solids. Environmental Science
and Technology, v. 25, pp. 1565-1572.
Barcelona, M. J. and Holm. T. R. 1991b. Oxidation-reduction capacities of aquifer solids. Environmental Science
and Technology, v. 25, 2540.
Barcelona, M. J. and Holm, T. R., 1992. Oxidation-reduction capacities of aquifer solids. Environmental Science
and Technology, v. 26, pp. 2538-2539.
Ernstsen, V., Gates, W. P. and Stucki, J. W. 1998. Microbial reduction of structural iron in clays - A renewable source
of reduction capacity. Journal of Environmental Quality, v. 27, pp. 761-766.
Fruchter, J. S., Cole, C. R., Williams, M. D., Vermeul, V. R., Amonette, J. E,, Szecsody, J. E,, Istok, J. D., and M. D.
Humphrey. 2000. Creation of a subsurface permeable treatment zone for aqueous chromate contamination using
in situ redox manipulation. Ground Water Monitoring & Remediation, v. 20, pp. 66-77.
Loeppert, R. H. and Inskeep, W. P. 1996. Iron. In Methods of Soil Analysis: Part 3 - Chemical Methods, ed. D. L.
Sparks, Soil Science Society of America Book Series, Soil Science Society of America. Inc., Madison, Wisconsin.
pp.639-664.
Lovley, D. R. and Phillips, E. J. P. 1987. Rapid assay for microbially reducible ferric iron in aquatic sediments.
Applied and Environmental Microbiology, v. 58, pp. 1536-1540.
Nelson. D. W. and Sommers. L. E. 1996. Total Carbon. Organic Carbon, and Organic Matter. In Methods of
Soil Analysis: Part 3 - Chemical Methods, ed. D. L. Sparks, Soil Science Society of America Book Scries, Soil
Science Society of America, Inc., Madison, Wisconsin, pp. 961-1010.
Ryan. J. N. and Gschwend. P. M. 1991. Extraction of iron oxides from sediments using reductive dissolution by
titanium(III). Clays and Clay Minerals, v. 39, pp. 509-518.
Schwertmann, U. 1991. Solubility and dissolution of iron oxides. Plant and Soil. v. 130, pp. 1-25.
Stanjek, H. and Weidler. P. G. 1992. The effect of dry heating on the chemistry, surface area, and oxalate solubility
of synthetic 2-line and 6-line ferrihydrites. Clay Minerals, v. 27, pp. 397-412.
U.S. Environmental Protection Agency. 1979. Chemical Oxygen Demand. Methods for Chemical Analysis ofWater
and Wastes, EPA Method 410.4, EPA 600/4-79-020. U.S. EPA, Government Printing Office, Washington, DC.
Williams, M. D., Vermeul, V. R., Szecsody, J. E. and Fmchter, J. S. 2000. 100-D Area In Situ Redox Treatability Test
for Chromate-Contaminated Groundwater, PNNL-13349, Pacific Northwest National Laboratory.
Richland, Washington.
113
-------
114
-------
A
A list of EPA documented analytical methods for redox sensitive elements is included in Table Al. Abbreviated
descriptions for these methods are provided in Table A2. The full references in Table Al should be consulted for
details on these methods. There are no EPA documented methods for the determination of ferrous iron or
oxidation-reduction potential in aqueous solutions. Summary' Chapter 4 should be consulted for these methods.
Included in Tables Al and A2 are methods for the determination of total organic carbon, total inorganic carbon and
acid-volatile sulfur in solids. There are no EPA documented methods for total sulfur and acid-digestible iron in
solids. Multiple procedures listed in Tables Al and A2 for a given element are due to either differences in
analytical range or method of detection. The most appropriate method will be based on the total analytc mass and
specific matrix characteristics of the sample. The listed methods were developed for implementation in a
laboratory setting, though many of the colorimetric and potentiometric methods for aqueous solutions have been
adapted to field-based procedures by several commercial vendors.
115
-------
Table Al. EPA Methods for Selected Redox Parameters
or
Number
EPA
Alkalinity - Colorimetric, Automated
Alkalinity - Titrimetric, pH 4.5
Ammonia in Estuarine by Gas Segmented CF/CA
Ammonia-N / automated colorimetry / prelim distill
Ammonia-N /ion-selective electrode potentiometry
Ammonium in Wet Deposition by Electrometric
Ammonium in Wet Deposition/Automated Colorimetric
Bomb Preparation Method for Solid Waste/Metallic
Carbon & Nitrogen in sediments & particulates
Extractable Sulfides
Inorganic Anions by Ion Chromatography
Inorganic Anions by Ion Chromatography
Inorganic Anions by Ion Chromatography
Inorganic Anions in Aqueous Matrices - CIE
Nitrate - Ion-Selective Electrode
Nitrate - Ion-Selective Electrode
Nitrate & Nitrite by Gas Segmented CF/CA
Nitrate/nitrite-N in biosolids - auto photometry
Nitrate/nitrite-N in biosolids/ manual colorimetry
Nitrate-N by 1C
Nitrate-Nitrate by Automated Colorimetry
Nitrate-Nitrite by Automated Colorimetric
Nitrite by 1C
Nitrogen, Ammonia - Colorimetric
Nitrogen, Ammonia - Colorimetric, Titrimetric
Nitrogen, Ammonia - Potentiometric
Nitrogen, Kjeldahl, Total - Colorimetric
Nitrogen, Kjeldahl, Total - Colorimetric/Automated
Nitrogen, Kjeldahl, Total - Colorimetric/Titrimetric
Nitrogen, Kjeldahl, Total - Potentiometric
Nitrogen, Nitrate - Colorimetric, Brucine
Nitrogen, Nitrate-Nitrite- Colorimetric/Cadmium
Nitrogen, Nitrate-Nitrite - Colorimetric/Hydrazine
Nitrogen, Nitrate-Nitrite - Manual Cadmium
Nitrogen, Nitrite - Spectrophotometric
Organic Carbon, Total - Combustion or Oxidation
Organic Carbon, Total - UV Promoted
Oxygen, Dissolved - Membrane Electrode (Probe)
Oxygen, Dissolved - Modified Winkler
Sulfate
Sulfate - Colorimetric, Automated, Chloranilate
Sulfate - Colorimetric, Automated, Chloranilate
Sulfate - Colorimetric, Automated, Methylthymol
Sulfate - Colorimetric, Automated, Methylthymol
Sulfate - Colorimetry, Automated
Sulfate - Gravimetric
Sulfate - Turbidimetric
Sulfate - Turbidimetric
Sulfide - Colorimetric, Methylene Blue
Sulfide - Ion-Selective Electrode
Sulfide - Titrimetric, Iodine
Sulfides, Acid-Soluble & Insoluble - Titrimetric
Sulfides, Acid-Soluble & Insoluble -Distillation
Total Kjeldahl Nitrogen - Semi-Automated
Total Kjeldahl nitrogen / preliminary
Total Kjeldahl nitrogen / preliminary
Total Organic Carbon
0310.2
0310.1
0349.0
1690
1689
0350.6
0350.7
5050
0440.0
9031
0300.0
9056
9056A
6500
9210
9210A
0353.4
1685
1686
0300.0
0353.2
0353.6
0300.0
0350.1
0350.2
0350.3
0351.2
0351.1
0351.3
0351.4
0352.1
0353.2
0353.1
0353.3
0354.1
0415.1
0415.2
0360.1
0360.2
0300.0
0375.1
9035
0375.2
9036
0375.2
0375.3
0375.4
9038
0376.2
9215
0376.1
9034
9030B
0351.2
1687
1688
9060
600/4-79-020
600/4-79-020
600/R-97-072
600/4-86-024
600/4-86-024
SW-846 Ch 5
600/R-97-072
SW-846 Ch 5
600/R-93-100
SW-846 Ch 5
SW-846 Ch 5
SW-846 Ch 3.3
SW-846 Ch 5
SW-846 Ch 5
600/R-97-072
600/4-79-020
600/R-93-100
600/4-86-024
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
600/4-79-020
SW-846 Ch 5
600/4-79-020
SW-846 Ch 5
600/R-93-100
600/4-79-020
600/4-79-020
SW-846 Ch 5
600/4-79-020
SW-846 Ch 5
600/4-79-020
SW-846 Ch 5
SW-846 Ch 5
600/R-93-100
SW-846 Ch 5
09/01/1997
02/01/1999
02/01/1999
03/01/1986
03/01/1986
09/01/1994
09/01/1997
07/01/1992
08/01/1993
09/01/1994
01/01/1998
12/01/1996
09/01/1997
02/01/1999
02/01/1999
08/01/1993
03/01/1986
09/01/1986
09/01/1986
08/01/1993
09/01/1986
12/01/1996
12/01/1996
12/01/1996
08/01/1993
02/01/1999
02/01/1999
09/01/1986
116
-------
OF EPA
Index to EPA Test Methods, May 2000 Revised Edition, U.S. EPA Region 1 Library at http://www.epa.gov/regionO I/
oarm/testmeth.pdf
EPA 530/SW-846 Test Methods for Evaluating Solid Waste: Physical/Chemical Methods 3rd ed - 4 vols. November
1986. (Subscription from GPO / SN955-001-00000-1; one-time print purchase from NTIS; also on CD ROM
from NTIS and Solutions) All methods including Update IVA and Update IVB and other pre-releases online at
www.epa.gov/epaoswer/hazwaste/test/sw846.htm
EPA 600/R-93-100 Methods for the Determination of Inorganic Substances in Environmental Samples. August 1993
(NTIS/PB94-120821)
EPA 600/R-97-072 Methods for the Determination of Chemical Substances in Marine and Estuarinc Environmental
Matrices. 2nd edition. September 1997. (NSCEP or NTIS /PB97-127326 or www.epa.gov:80/nerlcwww/
ordmeth.htm or CD ROM)
EPA 600/4-79-020 Methods for Chemical Analysis ofWater and Wastes. Revised March 1983 (NTIS/PB84-128677
for paper; CD ROM; NEPIS/ http://www.epa.gov/cincl)
EPA 600/4-86-024 Development of Standard Methods for the Collection and Analysis of Precipitation. May 1986
(NTIS/PB86-201365)
117
-------
Table Al. EPA Methods for Selected Redox Parameters
Analyte
Alkalinity
Alkalinity
Carbon,
Total Organic
Method
310.1
310.2
415.1
Detection Limits
In a single laboratory
using surface water
samples at an average
concentration of 122 mg
CaCO3 /L , the standard
deviation was 3.
The applicable range is
10 to 200 mg/L as
CaCO3. In a single
laboratory, using surface
water samples at
concentrations of 1 5, 57,
154, and 193 mg/L as
CaCO3 the standard
deviation was 0.5.
The method is most
applicable to
measurement of organic
carbon above 1 mg/L.
Preservation
The sample should be refrigerated at
4°C and run as soon as practical. Do
not open sample bottle before analysis.
The sample must not be filtered,
diluted, concentrated, or altered in any
way.
Sample should be refrigerated at 4°C
and run as soon as practical. Do not
open sample bottle before analysis.
Sampling and storage of samples in
glass bottles is preferable. Sampling
and storage in plastic bottles such as
conventional polyethylene and
cubitainers is permissible if it is
established that the containers do not
contribute contaminating organics to
the samples. NOTE 1: A brief study
performed in the EPA Laboratory
indicated that distilled water stored in
new, one quart cubitainers did not show
any increase in organic carbon after
two weeks exposure. Because of the
possibility of oxidation or bacterial
decomposition of some components of
aqueous samples, the lapse of time
between collection of samples and start
of analysis should be kept to a
minimum. Also, samples should be kept
cool (4°C) and protected from sunlight
and atmospheric oxygen. In instances
where analysis cannot be performed
within two hours (2 hours) from time of
sampling, the sample is acidified to
pH<2)with HCIorH2SO4.
Description
Titrimetric: An unaltered sample
is titrated to an electrometrically
determined end point of pH 4.5.
Automated ColorimetricTitration
Methyl Orange indicator
Combustion or oxidation. This
method includes the
measurement of organic carbon
in drinking, surface and saline
waters, domestic and industrial
wastes.
Organic carbon in a sample is
converted to carbon dioxide
(CO2) by catalytic combustion or
wet chemical oxidation. The CO2
formed can be measured directly
by an infrared detector or
converted to methane (CH4) and
measured by a flame ionization
detector.
118
-------
Table Al. continued
Analyte
Carbon, Total
Organic
Carbon (&
Nitrogen)
Carbon, Total
Organic
Method
415.2
440.0
9060
Detection Limits
The applicable range is
from approximately
50|jg/L to 10mg/L.
Higher concentrations
may be determined by
sample dilution.
In a single laboratory,
using raw river water,
centrifuged river water,
drinking water, and the
effluent from a carbon
column which had
concentrations of 3.11,
3.10, 1.79, and
0.07mg/L total organic
carbon respectively, the
standard deviations from
ten replicates were 0.13,
0.03, 0.02, and
0.02mg/L, respectively.
Method detection limits
(MDLs)of10.5|jg/Land
62.3 |jg/L for Particulate
Nitrogen and Particulate
Carbon, respectively,
were obtained for a
200mL sample volume.
Sediment MDLs of PN
and PC are 84 mg/kg
and 1300mg/kg, respec-
tively, for a sediment
sample weight of
10.00mg. The method
has been determined to
be linear to 4800|jg of C
and 700|jg of N in a
sample.
Precision and accuracy
data are available in
Method 415.1 of
Methods for Chemical
Analysis of Water and
Wastes.
Preservation
Sampling and storage of samples must
be done in glass bottles. Caution: Do
not leave any headspace in the sample
bottle as this may contribute to loss of
purgeable organics. Because of the
possibility of oxidation or bacterial
decomposition of some components of
aqueous samples, the lapse of time
between collection of samples and start
of analysis should be kept to a
minimum. Also, samples should be kept
cool (4°C) and protected from sunlight
and atmospheric oxygen. When
analysis cannot be performed within
two hours from time of sampling, the
sample should be acidified to pH < 2
with H2SO4. Note: HCI should not be
used because it is converted to chlorine
during the analysis. This causes
damage to the instrument.
Whenever possible, immediately filter
the samples. Store the filtered sample
pads by freezing at -20°C or storing in a
desiccator after drying at 1 03-1 05°C for
24 hr. No significant difference has
been noted in comparing the two
storage procedures for a time period of
up to 100 days. If storage of the water
sample is necessary, place the sample
into a clean amber glass bottle and
store at 4°C until filtration is done.
Because of the possibility of oxidation
or bacterial decomposition of some
components of aqueous samples, the
time between sample collection and the
start of analysis should be minimized.
Also, samples should be kept cool (4°C)
and protected from sunlight and
atmospheric oxygen. In instances
where analysis cannot be performed
within 2 hr from time of sampling, the
sample is acidified (pH < 2) with HCI or
H2SO4. Samples should be collected in
glass bottles.
Description
UV-promoted oxidation. This
method covers the determination
of total organic carbon in
drinking water and other waters.
This instrument is designed for a
two-step operation to distinguish
between purgeable and
nonpurgeable organic carbon.
These separate values are not
pertinent to this method.
This method is applicable only to
the carbonaceous matter which
is either soluble or has a particle
size of 0.2|im or less.
Analysis of particulates by
combustion at 975°C using an
elemental analyzer. The
combustion products are passed
over a copper reduction tube to
convert the oxides of N into
molecular N. Carbon dioxide,
water vapor and N are
homogeneously mixed at a
known volume, temperature and
pressure. The mixture is
released to a series of thermal
conductivity detectors/traps,
measuring in turn by difference,
hydrogen (as water vapor), C (as
carbon dioxide) and N (as N ).
Inorganic and organic C may be
determined by two methods
which are also presented.
Method 9060 is used to
determine the concentration of
organic carbon in ground water,
surface and saline waters, and
domestic and industrial wastes.
Method 9060 is most applicable
to measurement of organic
carbon above 1 mg/L.
119
-------
Table Al. continued
Analyte
Nitrogen
(Ammonia)
Nitrogen
(Ammonia)
Nitrogen
(Ammonia)
Nitrogen
(Ammonia)
Method
349.0
350.1
350.2
350.3
Detection Limits
A method detection limit
(MDL) of 0.3|jg N/Lhas
been determined by one
laboratory from spiked
surface water of three
different salinities.
Range of 0.01 to
2.0mg/L NH3 as N.
In a single laboratory,
using surface water
samples at
concentrations of 1.41,
0.77, 0.59 and 0.43mg
NH3 -N/L, the standard
deviation was 0.005.
Range from about 0.05
to1.0mgNH3-N/Lforthe
colorimetric procedure,
from 1.0 to 25mg/L for
the titrimetric procedure,
and from 0.05 to
1400mg/L for the
electrode method.
Range from 0.03 to
1400 mgNH3 -N/L. In a
single laboratory, using
surface water samples
at concentrations of
1.00, 0.77, 0.19, and
0.13mg NH3 -N/L,
standard deviations
were 0.038, 0.017,
0.007, and 0.003,
respectively.
Preservation
After collection and filtration or
centrifugation, samples should be
analyzed as soon as possible. If
samples will be analyzed within 3 hours
then keep refrigerated in tightly sealed,
glass or high density polyethylene
bottles in the dark at 4°C until the
analysis can be performed.
At low concentrations of ammonia (< 20
|jg N/L), no preservation technique is
satisfactory. Samples must be analyzed
within 3 hours of collection. At
moderate and high concentrations of
ammonia (> 20 |jg N/L) samples can be
preserved by the addition of 2 mL of
chloroform per liter of sample and
refrigerated in the dark at 4°C. Samples
can be stored in either glass or high
density polyethylene bottles. A
maximum holding time for preserved
estuarine and coastal water samples
with moderate to high concentrations of
ammonia is two weeks.
Preservation by addition of cone. H2SO4
to pH < 2 and refrigeration at 4°C.
Preservation by addition of cone. H2SO4
to pH < 2 and refrigeration at 4°C.
Preservation by addition of cone. H2SO4
to pH < 2 and refrigeration at 4°C.
Description
Gas segmented CF/CA. The
method is based upon the
indophenol reaction, here
adapted to automated gas-
segmented continuous flow
analysis.
Colorimetric, Automated
Phenate.
This method covers the
determination of ammonia in
drinking, surface and saline
waters, domestic and industrial
wastes.
Colorimetric Titration,
Potentiometric Distillation
Procedure.
This method covers the
determination of ammonia in
drinking, surface and saline
waters, domestic and industrial
wastes.
Potentiometric, Ion Selective
Electrode.
This method covers the
determination of ammonia in
drinking, surface and saline
waters, domestic and industrial
wastes.
120
-------
Table Al. continued
Analyte
Nitrogen
(Ammonium)
Nitrogen
(Ammonium)
Nitrogen
(Ammonia)
Nitrogen
(Ammonia)
Nitrogen (Nitrate,
Nitrite), Sulfate
Method
350.6
350.7
1689
1690
300.0
Detection Limits
MDL = 0.05 mg/ L
Range: 0.05-2.00mg/L
N as NH4+
MDL = 0.03 mg/ L
Range: 0.03-2.00mg/L
N as NH4+
Method detection limits
and minimum levels for
ammonia-N have not
been formally
established for this draft
method. These values
will be determined
during the validation of
the method.
Method detection limits
and minimum levels for
ammonia-N have not
been formally
established for this draft
method. These values
will be determined
during the validation of
the method.
MDLs (mg/L):
Nitrite-N 0.004
Nitrate-N 0.002
Sulfate 0.02
Ranges (mg/L)
Nitrate-N 0.42-14.0
Nitrite-N 0.36-12.0
Sulfate 2.85 - 95.0
Preservation
Collect in HOPE containers that have
been thoroughly rinsed with ASTM type
II water. Filtration through pre-rinsed
0.45 urn filters, followed by refrigeration
at4°C. Holding time one week or less.
See 350.6
See 1685
See 1685
Samples should be collected in
scrupulously clean glass or
polyethylene bottles. Sample
preservation and holding times for the
anions that can be determined by this
method are as follows:
Nitrate-N, cool to 4°C, 28 days
Nitrite-N, cool to 4°C, 48 hours
Sulfate, cool to 4°C, 28 days
Description
Electrometric detection of
ammonia using a gas-sensing
ion-selective electrode.
This method is applicable to the
measurement of ammonium in
wet deposition samples: rain,
snow, dew, sleet, and hail.
This colorimetric method
(phenate) is applicable to the
measurement of ammonium in
wet deposition samples: rain,
snow, dew, sleet, and hail.
Ion selective electrode poten-
tiometry.
This method describes
procedures for the determination
of ammonia as N in drinking,
ground, and surface water;
domestic and industrial waste;
and biosolids (municipal sewage
sludge). Distillation of the
ammonia from the sludge is
followed by analysis using an
ion-selective probe. This method
is based on U.S. Environmental
Protection Agency (EPA)
Methods 350.2.
This automated colorimetric
method describes procedures for
the determination of ammonia-N
in drinking, ground, and surface
water; domestic and industrial
waste; and biosolids (municipal
sewage sludge). Distillation of
ammonia from the sample is
followed by analysis using an
automated colorimetric analyzer.
This method is based on 350.1
and 350.2.
Determination of Inorganic
Anions in Water by Ion
Chromatography
121
-------
Table Al. continued
Analyte
Nitrogen (Nitrate)
Nitrogen (Nitrate,
Nitrite)
Nitrogen (Nitrate,
Nitrite)
Nitrogen (Nitrate,
Nitrite)
Nitrogen (Nitrate,
Nitrite)
Method
352.1
353.1
353.2
353.3
353.4
Detection Limits
The applicable range of
concentrations is 0.1 to
2 mg NO3 -N/liter.
± std. dev., mg N/L
0.16 ±0.92
0.19 ±0.083
1.08 ±0.245
1.24 ±0.214
The applicable range of
this method is 0.01-
10mg/L nitrate-nitrite
nitrogen. In a single
laboratory using drinking
water, surface water and
industrial waste at
concentrations of 0.39,
1.15, 1.76 and 4.75ug
NO3 -N/L, the standard
deviations were 0.02,
0.01, 0.02 and 0.03,
respectively.
The applicable range of
this method is 0.05 to
10.0 mg/L nitrate-nitrite
nitrogen. The range may
be extended with
sample dilution.
The applicable range of
this method is 0.01 to
1 .0 mg/L nitrate-nitrite
nitrogen. The range may
be extended with
sample dilution.
A statistically deter-
mined method detection
limit (MDL) of 0.075|jg
N/L has been
determined by one
laboratory in seawaters
of five different salinities.
The method is linear to
5.0 mg N/L using a Flow
Solution System
(Alpkem, Wilsonville,
Oregon) with respect to
analyte concentration.
Preservation
Analysis should be made as soon as
possible. If analysis can be made within
24 hours, the sample should be
preserved by refrigeration at4°C. When
samples must be stored for more than
24 hours, they should be preserved by
acidification to pH < 2 with sulfuric acid
and refrigeration.
Analysis should be made as soon as
possible. If analysis can be made within
24 hours, samples should be preserved
by refrigeration at 4°C. When samples
must be stored for more than 24 hours,
they should be preserved by
acidification to pH < 2 with sulfuric acid
(H2SO4) and refrigerated at 4°C.
Analysis should be made as soon as
possible. If analysis can be made within
24 hours, the sample should be
preserved by refrigeration at4°C. When
samples must be stored for more than
24 hours, they should be preserved
with sulfuric acid to pH < 2 and
refrigeration at 4°C. CAUTION:
Samples for reduction column must not
be preserved w/ mercuric chloride.
Analysis should be made as soon as
possible. If analysis can be made within
24 hours, the sample should be
preserved by refrigeration at4°C. When
samples must be stored for more than
24 hours, they should be preserved
with sulfuric acid to pH < 2 and
refrigeration to 4°C. CAUTION:
Samples for reduction column must not
be preserved w/ mercuric chloride.
Turbid samples should be filtered as
soon as possible after collection. After
collection and filtration, samples should
be analyzed as soon as possible. If
samples will be analyzed within 3 hours
then keep refrigerated in tightly sealed,
high density polyethylene bottles in the
dark at 4°C until analysis can be
performed.
Description
This colorimetric method
(brucine) is applicable to the
analysis of drinking, surface and
saline waters, domestic and
industrial wastes. Modification
can be made to remove or
correct for turbidity, color,
salinity, or dissolved organic
compounds in the
sample.
This colorimetric (hydrazine)
method is applicable to drinking
and surface water, and domestic
and industrial wastes.
This automated colorimetric
method pertains to the
determination of nitrite singly, or
nitrite and nitrate combined in
surface and saline waters, and
domestic and industrial wastes.
Manual, copper-coated cadmium
reduction column.
This method is applicable to the
determination of nitrite singly, or
nitrite and nitrate combined in
drinking, surface and saline
waters, domestic and industrial
wastes.
Automated gas segmented
continuous flow colorimetric
method for the analysis of
nitrate. Copper-coated cadmium
reduction column.
122
-------
Table Al. continued
Analyte
Nitrogen, Nitrate
Nitrogen (Nitrite)
Nitrogen (Nitrate,
Nitrite)
Nitrogen (Nitrate,
Nitrite)
Nitrogen (Nitrate,
Nitrite), Sulfate
Method
353.6
354.1
1685
1686
5050
Detection Limits
MDL = 0.02 mg/L
Range: 0.02-5.00mg/L
N as NH4+
The method is applic-
able in the range from
0.01 to1.0mgNO2-N/L
MDL not available
Method detection limits
and minimum levels for
nitrate/nitrite-N have not
been formally
established for this draft
method. These values
will be determined
during the validation of
the method.
This is a draft method,
and is currently
undergoing validation.
Method performance
criteria will be set
following the validation
of the method.
The method detection
limit (MDL) varies for
anions as a function of
sample size and the
conductivity scale used.
Generally, minimum
detectable concen-
trations are in the range
of 0.05mg/L for F and
0. 1 mg/L for Br,CI,N03
, N02 , P04-3 , and SO4 2.
Preservation
Collect in HOPE containers that have
been thoroughly rinsed with ASTM type
II water. Filtration through pre-rinsed
0.45 urn filters, followed by refrigeration
at4°C. Holding time one week or less.
Samples should be analyzed as soon
as possible. They may be stored for 24
to 48 hours at 4°C.
Samples should be collected in wide
mouth glass jars with a minimum of air
space above the sample. Minimize
exposure of samples to air and intense
light as much as possible. Nitrate can
be formed or lost during storage due to
biological activity or lost by oxidation. If
the sample contains free water or is
aqueous: samples should be checked
for residual chlorine and treated with
sodium thiosulfate, if necessary, during
collection. Samples should be
preserved by acidification to pH < 2
with concentrated H2SO4 and cooled to
4°C as soon as possible aftercollection.
The holding time for samples should
not exceed 28 days from sampling. If
the sample contains no free water or is
solid, the sample should be cooled to
4°C as soon as possible aftercollection.
The holding time for samples should
not exceed 28 days from sampling.
See 1685
Analyze the samples as soon as
possible after collection. Preserve by
refrigeration at 4°C. Samples may be
collected in teflon, polyethylene, or
glass containers.
Description
This colorimetric method is
applicable to the measurement
of nitrate in wet deposition
samples: rain, snow, dew, sleet,
and hail.
This spectrophotometric method
is applicable to the deter-
mination of nitrite in drinking,
surface and saline waters,
domestic and industrial wastes.
This method describes pro-
cedures for the determination of
nitrate/nitrite-nitrogen, or
oxidized nitrogen, in drinking,
ground, and surface water;
domestic and industrial waste;
and biosolids (treated municipal
sewage sludge). Automated
photometry is used to determine
the nitrate/nitrite concentration of
the sample. This method is
based on U.S. Environmental
Protection Agency (EPA)
Method 353.2.
This method describes pro-
cedures for the determination of
nitrate/nitrite-nitrogen, or
oxidized nitrogen, in drinking,
ground, and surface water;
domestic and industrial waste;
and biosolids (municipal sewage
sludge). Manual colorimetry is
used to determine the
nitrate/nitrite-N concentration.
This method is based on U.S.
Environmental Protection
Agency (EPA) Method 353.3.
Sequential determination by ion
chromatography of the anions
chloride, fluoride, bromide,
nitrate, nitrite, phosphate, and
sulfate in the collection solutions
from the bomb combustion of
solid waste samples, as well as
all water samples.
123
-------
Table Al. continued
Analyte
Nitrogen (Nitrate,
Nitrite), Sulfate
Nitrogen (Nitrate,
Nitrite), Sulfate
Nitrogen (Nitrate)
Nitrogen, Total
Kjeldahl (TKN)
Nitrogen, Total
Kjeldahl
Method
6500
9056
9210
9210A
351.1
351.2
Detection Limits
Nitrate, Nitrite and,
Sulfate in the concen-
tration range of 0.1 to
50mg/L
Creek water, mg/L
nitrite <0.072
nitrate 3.00 ± 0.30
sulfate 0.37 ±0.1 9
Generally, minimum
detectable concen-
trations are in the range
of 0.05mg/L for F and
0.1 mg/L for Br, Cl , NO3
N02 , P04 3, and SO4 2-
The method detection
limit is 2.0mg/L Nitrate
concentrations from 0.2
to 1,OOOmg/L may be
measured. However,
using a linear
calibration, results less
than 2mg/L may be
biased up to
approximately 420
percent high; results
greater than 400 mg/L
may be biased up to
approximately 50
percent low.
Range is 0.05 to 2.0 mg
N/L.
Std. Dev.
Kjeldahl-N mg N/L
1.89 ±0.54
2. 18 ±0.61
5.09 ± 1.25
5.81 ± 1.85
The applicable range of
this method is 0.1 to
20mg/L TKN. The range
may be extended with
sample dilution. In a
single laboratory, using
sewage samples of
concentrations of 1 .2,
2.6, and 1.7mg N/L, the
precision was ±0.07,
±0.03 and ±0.15,
respectively.
Preservation
Analyze samples as soon as possible
after collection. For nitrite, nitrate, and
phosphate refrigerate the sample at 4°C
after collection and warm to room
temperature before dilution and
analysis. Determine nitrite and nitrate
within 48 hours.
Analyze the samples as soon as
possible after collection. Preserve by
refrigeration at4°C. Collect samples in
teflon, polyethylene, or glass
containers.
In the presence of a reducing agent
(e.g., organic matter), bacteria can
utilize nitrate as an oxidant, causing a
slow decrease in the nitrate
concentration. This potential
interference can be obviated by using a
preservative. Therefore, samples must
be preserved by adding 1mL of 1M
boric acid solution per 100mL of
sample. Samples should be stored at
4°C and contained in teflon,
polyethylene, or glass bottles.
Samples may be preserved by
acidification to pH < 2 with H2SO4 and
refrigeration at 4°C. Even when
preserved in this manner, conversion of
organic nitrogen to ammonia may
occur. Therefore, samples should be
analyzed as soon as possible.
See Method 351.1.
Description
Capillary ion electrophoresis
with indirect UV detection. This
test method is applicable for
determination of the dissolved
inorganic anions; fluoride,
bromide, chloride, nitrite, nitrate,
ortho-phosphate, and sulfate in
aqueous matrices.
Sequential determination by ion
chromatography of the anions
chloride, fluoride, bromide,
nitrate, nitrite, phosphate, and
sulfate in the collection solutions
from the bomb combustion of
solid waste samples, as well as
all water samples.
Total solubilized nitrate is
determined potentiometrically
using a nitrate ion-selective
electrode (ISE).
Colorimetric, Automated
Phenate Surface and saline
waters
Semi-automated colorimetric
determination of total Kjeldahl
nitrogen in drinking and surface
waters, domestic and industrial
wastes.
124
-------
Table Al. continued
Analyte
Nitrogen, Total
Kjeldahl
Nitrogen, Total
Kjeldahl
Nitrogen, Total
Kjeldahl
Nitrogen, Total
Kjeldahl
Method
351.3
351.4
1687
1688
Detection Limits
Method applicable to the
range 0.05 to 1400mg/L
± std. dev.
0.20 ±0.197
0.31 ± 0.247
4.10 ± 1.056
4.61 ± 1.191
This method covers the
range from 0.03 to 25
mg TKN/L.
Method detection limits
and minimum levels for
TKN have not been
formally established for
this draft method. These
values will be
determined during the
validation of the method.
Method detection limits
and minimum levels for
TKN have not been
formally established for
this draft method. These
values will be
determined during the
validation of the method.
Preservation
See Method 351.1.
See Method 35 1.1.
Ammonia can be formed or lost during
storage due to biological activity or lost
by volatilization or oxidation. The
following preservation procedures will
help prevent significant changes in the
analyte concentration. Samples should
be collected in wide mouth jars with a
minimum of air space. If the sample
contains free water or is aqueous:
samples should be checked for residual
chlorine and treated with sodium
thiosulfate, if necessary, during
collection. Samples should be
preserved by acidification to pH < 2
with concentrated H2SO4 and cooled to
4°C as soon as possible after collection.
The holding time for samples should
not exceed 28 days from sampling. If
the sample contains no free water or is
solid, the sample should be cooled to
4°C as soon as possible after collection.
The holding time for samples should
not exceed 28 days from sampling.
See 1687.
Description
This colorimetric method covers
the determination of total
Kjeldahl nitrogen in drinking,
surface and saline waters,
domestic and industrial wastes.
This potentiometric method
covers the determination of total
Kjeldahl nitrogen in drinking,
surface and saline waters,
domestic and industrial wastes.
This method describes
procedures forthe determination
of total kjeldahl nitrogen (TKN)
and/or organic nitrogen in
drinking, ground, and surface
water; domestic and industrial
waste; and biosolids (municipal
sewage sludge).
Digestion/distillation of the TKN
from the sample is followed by
analysis using automated
colorimetry. This method is
based on U.S. Environmental
Protection Agency (EPA)
Method 35 1.1.
This method describes
procedures for the determination
of total kjeldahl nitrogen (TKN)
and/or organic nitrogen in
drinking, ground, and surface
water; domestic and industrial
waste; and biosolids (municipal
sewage sludge). Automatic
digestion of the sample is
followed by analysis using
automated colorimetry. This
method is based on U.S.
Environmental Protection
Agency (EPA) Method 351 .2.
125
-------
Table Al. continued
Analyte
Oxygen,
Dissolved
Oxygen,
Dissolved
Sulfate
Sulfate
Method
360.1
360.2
375.1
375.2
Detection Limits
Manufacturer's spec-
ification claim 0.1mg/L
repeatability with 1%
accuracy.
Exact data are unavail-
able on the precision
and accuracy of this
technique; however,
reproducibility is
approximately 0.2mg/L
of DO at the 7.5mg/L
level due to equipment
tolerances and uncom-
pensated displacement
errors.
Range of 10 to 400 mg
SO4/L. In a single
laboratory, using surface
water samples at
concentrations of 39,
111, 188 and 294mg
SO4/L, the standard
deviations were 0.6, 1 .0,
2.2, 4, and 0.8,
respectively.
Samples in the range of
3 to SOOmg SO4/L can
be analyzed. The
sensitivity of the method
can be increased by a
minor modification to
analyze samples in the
range of 0.5 to 30mg
S04/L.
Preservation
Special precautions are required to
avoid entrainment or solution of
atmospheric oxygen or loss of
dissolved oxygen. Care must be taken
to prevent turbulence and the formation
of bubbles when filling bottle. At time of
sampling, the sample temperature
should be recorded as precisely as
required. If samples must be preserved
the method below may be employed.
Add 2mL of manganous sulfate solution
and then 2mL of alkaline iodide-azide
solution to the sample contained in the
BOD bottle. Both reagents must be
added well below the surface of the
liquid. Stopper the bottle immediately
and mix the contents thoroughly. The
sample should be stored at the
temperature of the collection water, or
water sealed and kept at a temperature
of 10 to 20°C, in the dark. Complete
the procedure by adding 2mL H2SO4 at
time of analysis.
See 360.1 above.
Refrigeration at 4°C.
Refrigeration at 4°C.
Description
The membrane electrode
method for dissolved oxygen is
recommended forthose samples
containing materials which
interfere with the modified
Winkler procedure such as
sulfite, thiosulfate, polythionate,
mercaptans, free chlorine or
hypochlorite, organic substances
readily hydrolyzed in alkaline
solutions, free iodine, intense
color or turbidity and biological
floes. The probe method is
recommended as a substitute for
the modified Winkler procedure
in monitoring of streams, lakes,
outfalls, etc., where it is desired
to obtain a continuous record of
the dissolved oxygen content of
the water under observation.
Modified Winkler
This method is applicable for
use with most wastewaters and
streams that contain nitrate
nitrogen and not more than
1 mg/L of ferrous iron. Other
reducing or oxidizing materials
should be absent.
This automated colorimetric
(chloranilate) method is
applicable to drinking and
surface waters, domestic and
industrial wastes.
Automated colorimetric, methyl
thymol. This automated
colorimetric method (methyl
thymol) is applicable to drinking
and surface waters domestic
and industrial wastes.
126
-------
Table Al. continued
Analyte
Sulfate
Sulfate
Sulfate
Sulfate
Sulfate
Sulfate
Method
375.3
375.4
5050
6500
9035
9036
Detection Limits
This method is the most
accurate method for
sulfate concentrations
above 1 0 m g / L .
Therefore, it should be
used whenever results
of the greatest accuracy
are required.
A synthetic unknown
sample containing
259mg/L sulfate,
108mg/L Ca, 82mg/L
Mg, 3.1mg/L K,
19.9mg/L Na, 241mg/L
chloride, 250 Fg/L nitrite
N, 1.1mg/L nitrate Nand
42.5mg/L alkalinity
(contributed by
NaHCO3), was
analyzed in 32
laboratories by the
gravimetric method, with
a relative standard
deviation of 4.7% and a
relative error of 1 .9%.
The method is suitable
for all concentration
ranges of sulfate;
however, in order to
obtain reliable readings,
use a sample aliquot
containing not more than
40mg SO4/L. The
minimum detectable limit
is approximately 1mg/L
sulfate.
Applicable to ground
water, drinking and
surface waters, and
domestic and industrial
wastes containing 10 to
400mg SO4 2 /liter.
Preservation
Refrigeration at 4°C.
Refrigeration at 4°C.
Refrigerate at4°C. Samples should be
centrifuged or filtered before analysis.
Teflon, polyethylene, or glass bottles
may be used.
Refrigerate at 4°C. Samples with pH
below 2 should be neutralized because
high acid concentrations elute cations
from the ion-exchange resin. Turbid
samples should be filtered or
centrifuged.
Description
This gravimetric method is
applicable to drinking, surface
and saline water, domestic and
industrial wastes.
This turbidimetric method is
applicable to drinking and
surface waters, domestic and
industrial wastes.
Bomb preparation
Inorganic anions in aqueous
matrices, capillary ion
electrophoresis
Automated colorimetric (acid
chloranilate ion)
This automated colorimetric
method using methylthymol blue
is applicable to ground water,
drinking and surface waters, and
domestic and industrial wastes.
127
-------
Table Al. continued
Analyte
Method
Detection Limits
Preservation
Description
Sulfate
9038
The minimum detectable
limit is approximately
1 mg/L of SO4 2.
Use a sample aliquot
containing not more than
40mg/L of SO4 2.
Preserve by refrigerating at 4°C.
Turbidimetric method by
nephelometer, filter photometer,
or spectrophotometer applicable
to ground water, drinking and
surface waters, and domestic
and industrial wastes.
Sulfate
9056
Ion chromatography
Sulfide
376.1
This method is suitable
for the measurement of
sulfide in concentrations
above 1mg/L
Precision and accuracy
for this method have not
been determined.
If the sample is not preserved with zinc
acetate and NaOH, the analysis must
be started immediately. Similarly, the
measurement of dissolved sulfides
must also be commenced immediately.
This iodimetric method is
applicable to the measurement
of total and dissolved sulfides in
drinking, surface and saline
waters, domestic and industrial
wastes. Acid insoluble sulfides
are not measured by the use of
this test. (Copper sulfide is the
only common sulfide in this
class).
Sulfide
376.2
The method is suitable
for the measurement of
sulfide in concentrations
up to 20mg/L The
precision has not been
determined. The
accuracy is about 10%.
Samples must be taken with a minimum
of aeration. Sulfide may be volatilized
by aeration and any oxygen
inadvertently added to the sample may
convert the sulfide to an unmeasurable
form. Dissolved oxygen should not be
present in any water used to dilute
standards. The analysis must be
started immediately.
This colorimetric method
(methylene blue) is applicable to
the measurement of total and
dissolved sulfides in drinking,
surface and saline waters,
domestic and industrial wastes.
Acid insoluble sulfides are not
measured by this method.
Copper sulfide is the only
common sulfide in this class.
Sulfide, Acid
Volatilized
9030B
Detection Limit - The
detection limit was
determined by analyzing
seven replicates at 0.45
and 4.5mg/L The detec-
tion limit was calculated
as the standard de-
viation times the
student's t-value for a
one-tailed test with n-1
degrees of freedom at
99% confidence level.
The detection limit for a
clean matrix (H2O) was
found to be between 0.2
and 0.4mg/L Method
9030 is suitable for
measuring sulfide
concentrations in
samples which contain
between 0.2 and
50mg/kg of sulfide.
All aqueous samples and effluents must
be preserved with zinc acetate and
sodium hydroxide. Use four drops of 2N
zinc acetate solution per 100ml_ of
sample. Adjust the pH to greater than 9
with 6N sodium hydroxide solution. Fill
the sample bottle completely and
stopper with a minimum of aeration.
The treated sample is relatively stable
and can be held for up to seven days.
If high concentrations of sulfide are
expected to be in the sample, continue
adding zinc acetate until all the sulfide
has precipitated. For solid samples, fill
the surface of the solid with 2N zinc
acetate until moistened. Samples must
be cooled to 4°C and stored head
space free.
The distillation procedure
described in this method is
designed forthe determination of
sulfides in aqueous, solid waste
materials, or effluents. For acid-
soluble sulfide samples,
separation of sulfide from the
sample matrix is accomplished
by the addition of sulfuric acid to
the sample. The sample is
heated to 70°C and the
hydrogen sulfide (H2S) which is
formed is distilled under acidic
conditions and carried by a
nitrogen stream into zinc acetate
gas scrubbing bottles, where it is
precipitated as zinc sulfide.
128
-------
Table Al. continued
Analyte
Sulfide, Acid
Volatilized
(Extractable)
Sulfide, Acid
Volatilized
Method
9031
9034
Detection Limits
Method 9031 is suitable
for measuring sulfide in
solid samples at
concentrations above
1mg/kg.
Method 9034 is suitable
for measuring sulfide
concentrations in
samples which contain
0.2 mg/kg to 50 mg/kg of
sulfide.
Preservation
All samples must be preserved with
zinc acetate and sodium hydroxide.
Use four drops of 2N zinc acetate
solution per 100 ml_ of aqueous or
multiphasic sample. Adjust the pH to
greater than 9.0 with 50% NaOH. Fill
the sample bottle completely and
stopper with a minimum of aeration. For
solid samples, fill the surface of solid
with 2N zinc acetate until moistened.
Samples must be cooled to 4°C during
storage.
Distillates that are not analyzed
immediately should be stored in a
sealed flask at 4°C.
Description
Separation of sulfide from the
sample matrix is accomplished
by the addition of sulfuric acid to
the sample. The sample is
heated to 70°C and the
hydrogen sulfide which is formed
is distilled under acidic
conditions and carried by a
nitrogen stream into zinc acetate
gas scrubbing bottles where it is
precipitated as zinc sulfide. The
sulfide in the zinc sulfide
precipitate is oxidized to sulfur
with a known amount of excess
iodine. Then the excess iodine is
determined by titration with a
standard solution of
phenylarsine oxide (PAO) or
sodium thiosulfate until the blue
iodine starch complex
disappears. The use of standard
sulfide solutions is not possible
because of their instability to
oxidative degradation. Therefore
quantitation is based on the PAO
or sodium thiosulfate.
This procedure may be used as
a determinative step for acid-
soluble and acid-insoluble
sulfides following distillation of
the sample by SW-846 Method
9030. Sulfide is extracted from
the sample by a preliminary
distillation procedure (see
Method 9030) and precipitated in
a zinc acetate scrubber as zinc
sulfide. The sulfide is oxidized to
sulfur by adding a known excess
amount of iodine. The excess
iodine is determined by titration
with a standard solution of
phenyl arsine oxide (PAO) or
sodium thiosulfate until the blue
iodine starch complex
disappears. As the use of
standard sulfide solutions is not
possible because of oxidative
degradation, quantitation is
based on the PAO or sodium
thiosulfate.
129
-------
Table Al. continued
Analyte
Sulfide
Method
9215
Detection Limits
The method detection
limit is 1.0mg/L Sulfide
concentrations from 0.1
to 12,OOOmg/L may be
measured. However,
when a linear calibration
is used, results less than
1 mg/L may be biased up
to approximately 90
percent low.
Preservation
All original, undistilled samples should
be handled as described in Section 6 of
Methods 9030 and/or 9031. Samples
should be stored at 4°C prior to
analysis. The distilled samples should
be analyzed with the ISE immediately
following distillation, distillates that are
not analyzed immediately should be
stored in a sealed flask at 4°C.
Description
This method can be used for
measuring total sulfide in a
distilled sample. The method is
meant to be used as an alternate
determinative step following the
distillation in either SW-846
Methods 9030 or 9031. The
distillations in Methods 9030
and/or 9031 are performed,
except that the scrubber solution
is sulfide anti-oxidant buffer
(SAOB), with ascorbic and
salicylic acids added as oxygen
scavengers. The distillates are
then analyzed potentiometrically
using a sulfide ion-selective
electrode (ISE) in conjunction
with a double-junction reference
electrode and a pH meter with
an expanded millivolt scale or an
ISE meter capable of being
calibrated directly in terms of
sulfide concentration.
130
-------
&EPA
United States
Environmental Protection
Agency
National Risk Management
Research Laboratory
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HEREEH;
detach, or copy this cover, and return to the address in the
upper ieft-hand corner.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/R-02/002
January 2002
------- |