United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/625/R-01/011a
November 2001
US EPA Office of Rfiwarc?! 3rd Dovebpmcn t
Summary of the
Phytoremediation
State of the Science
Conference
Boston, Massachusetts
May 1-2, 2000
-------
-------
625/R-01/011a
November 2001
Summary of the Phytoremediation
State of the Science Conference
Boston, Massachusetts
May 1-2, 2000
United States Environmental Protection Agency
Office of Research .and Development
National Risk Management Research Laboratory
Cincinnati, OH
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free.
-------
NOTICE
The U.S. Environmental Protection Agency through its Office of Research and
Development funded and managed the research described here under Contract 68-D7-
0001 to Eastern Research Group. It has been subjected to Agency review 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 threaten 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.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
-------
ACKNOWLEDGEMENTS
Many people participated in planning and organizing the Phytoremediation: State of the Science
Conference which was held May 1-2," 2000. The primary contributors for the conference are listed
below:
Steven Rock, U.S. Environmental Protection Agency (U.S. EPA), National Risk Management
Research Laboratory (NRMRL), Cincinnati. OH
Steven McCutcheon, U.S. EPA, National Exposure Research Laboratory (NERL), Athens, GA
Norman Kulujian, U.S. EPA, Office of Research and Development (ORD), Philadelphia, PA
Joan Colson, U.S. EPA, NRMRL, Cincinnati, OH
Mitch Lasat, U.S. EPA, Office of Solid Waste and Emergency Response (OSWER), Washington,
DC
Steven Hirsch, U.S. EPA Region 3, Philadelphia, PA
Judy Canova, Bureau of Land and Waste Management, South Carolina Department of Health and
Environmental Control
Jeff Heimerman, U.S. EPA, OSWER, Washington, DC
Linda Fiedler, U.S. EPA, OSWER, Washington, DC
The success of the conference and the preparation of this document are also due to the efforts of the
many speakers and poster presenters.
Eastern Research Group provided logistical support both in advance of and during the conference
as well as for the development of this document under Contract 68-D7-0001.
IV
-------
CONTENTS
Notice ii
Foreword iii
Acknowledgments iv
Session I: Introduction and Plenary
Welcome and Introductions 1
Norm Kulujian •
EPA Policy Overview 1
Stephen Luftig
The Science and Practice of Phytoremediation 3
Steven McCutcheon
Interstate Technical Regulatory Coorperation (ITRC): Making it Easier for Regulators ... 4
Robert Mueller
International Perspective on the Cleanup of Metals arid Other Contaminants 6
Terry Mclntyre
Looking Forward on Phytotechnologies 7
Steven Rock
Speaker Panel and Audience Discussion 8
Session II: Fundamental Processes of Plants and Soils
Transport of Contaminants in Plant and Soil Systems 9
Larry Erickson
Enzymatic Processes Used by Plants to Degrade Organic Compounds 10
Nelson Lee Wolfe
Biosystem Treatment of Recalcitrant Soil Contaminants 11
John Fletcher
-------
Speaker Panel and Audience Discussion 12
Session IIIA: Brownfields Applications and Beneficial Use Of Land
Goals for Brownfields Pilots 13
John Podgurski
Hartford Brownfields Site: Site Description and Summary of
Phytoremediation Project 14
Jeanne Webb
Hartford Brownfields Site: Public Health Perspective 15
Jennifer Kertanis
Integrating Remediation into Landscape Design .16
Niall Kirkwood
Speaker Panel and Audience Discussion 17
Session EBB: Radionuclides
Summary of DOE Projects and Recommendations Offered at a Recent
DOE Workshop 18
Scott McMullin
Capturing a "Mixed" Contaminant Plume: Tritium Phytoevaporation at
Argonne National Laboratory-East 19
M. Cristina Negri
Phytoremediation Application for Radionuclide Removal at
Argonne National Laboratory-West 20
Scott Lee
Speaker Panel and Audience Discussion 21
Session IVA: The Fate of Chlorinated Solvents that Disappear from
Planted Systems
Phytoremediation of Solvents 22
Milton Gordon
The Case for Phytovolatization 23
William Doucette
vi
-------
Phytotransformation Pathways and Mass Balance for Chlorinated
Alkanes and Alkenes 25
Valentine Nzengung
Speaker Panel and Audience Discussion 27
Session IVB: Innovative Solutions for Metals Removal
Phytoextraction of Metals from Contaminated and Mineralized Soils Using
Hyperaccumulator Plants 27
Rufus Chaney
Phytoextraction: Commercial Considerations 29
Michael Blaylock
Zinc Hyperaccumulation in Plants: The Case of Zinc Hyperaccumulation in
T. Caerulescens 30
Mitch Lasat ;
Speaker Panel and Audience Discussion 31
Session V: Simulations and Forecasts
Chasing Subsurface Contaminants 32
Joel G. Burken
Effect of Woody Plants on Groundwater Hydrology and Contaminant Fate 33
James Landmeyer
Modeling Plume Capture at Argonne National Laboratory-East (ANL-E) 35
John Quinn
Phytoremediation Potential of a Chlorinated: Solvents Plume in
Central Florida 36
Stacy Lewis Hutchinson
Speaker Panel and Audience Discussion 37
Session VI: Plume Control—On-the-Ground Experience
Phytoremediation at Aberdeen Proving Ground, Maryland: O&M,
Monitoring, and Modeling 37
Steven Hirsh
VII
-------
Phytoremediation Systems Designed to Control Contaminant Migration 39
Ari Ferro
Deep Planting 40
Edward Gatliff
Transpiration: Measurements and Forecasts 41
James Vose
Speaker Panel and Audience Discussion 42
Session VH: Vegetative Covers
Monitoring Alternative Covers 43
Craig Benson
Growing a Thousand-Year Landfill Cover 44
William Jody Waugh
Tree Covers for Containment and Leachate Recirculation 45
Eric Aitchison
EPA Draft Guidance on Landfill Covers 47
Andrea McLaughlin and Ken Skahn
Activities at an EPA Region 3 Site 48
Donna McCartney
Speaker Panel and Audience Discussion 48
Session VET: Degradation of Organic Compounds in Soils
Remediation of Petroleum Contaminants Using Plants 49
M. Katherine Banks
Phytoremediation of Explosives 50
Phillip L. Thompson
Case Study: Union Pacific Railroad 52
Felix Flechas
Phytoremediation in Alaska and Korea 53
Charles (Mike) Reynolds
VIII
-------
Speaker Panel and Audience Discussion 54
Closing Remarks 55
IX
-------
-------
SUMMARY OF THE PHYTOKEMEDIATION
STATE OF THE SCIENCE CONFERENCE
Omni Parker House Hotel
Boston, Massachusetts
May 1-2, 2000
SESSION I: INTRODUCTION AND PLENARY
Welcome and Introductions
Norm Kulujian, EPA, Office of Science and Policy
On behalf of the U.S. Environmental Protection Agency (EPA), Norm Kulujian welcomed speakers (see
Appendix B), poster presenters (see Appendix C), and meeting attendees (see Appendix D). Five years ago,
he said, many people were unfamiliar with the concepts that underlie phytoremediation. Today, the field is
burgeoning with interest, and many site owners are asking for permission to implement phytoremediation at
their sites. Regulators, Kulujian said, are eager to obtains better understanding of phytoremediation so that
they will know where it is likely to be successful, when to dismiss it, and when to perform preliminary tests
to determine whether it is appropriate for a particular site. Of the proposals that have been submitted,
Kulujian said, most have focused on using phytoremediation as a containment technology. This does not
mean that enough data have been collected to prove conclusively that phytoremediation is successful when
applied this way, nor does it mean that phytoremediation has no potential to reduce contaminant
concentrations. Phytoremediation's optimal application, said Kulujian, will differ across sites and will be
determined by what contaminants, climatological conditions, and geological conditions are present. Thus,
phytoremediation will be useful as a containment technology at some sites and as a destruction approach at
others; at some sites, it will probably serve as part of a treatment train.
Kulujian said that sites in arid or semi-arid climates might be excellent candidates for phytoremediation,
noting that plants might be able to extract enough moisture to prevent leachate from forming at these sites.
In addition, he said, phytoremediation might be beneficial to use at sites that have widespread contamination
and concentrations that are close to cleanup levels. Also, he said, sites that require ecosystem restoration
might obtain significant benefits by applying phytoremediation.
EPA Policy Overview
Stephen Luftig, EPA, Office of Solid Waste and Environmental Response
Stephen Luftig provided an overview of the Superfund program and explained how it views
phytoremediation technologies. He said that this year marks the 20th anniversary of the Superfund program,
and that the program is operating well. Over the last few.years, the program has (1) released guidance on
future land uses, (2) established a national remedy board, and (3) revisited hundreds of Records of Decision
that were written in the 1980s. The latter was done, he said, to determine whether sites would benefit from
the use of innovative technologies that have been developed over the last 10 years. Luftig provided statistics
to summarize the accomplishments of the Superfund program made since its inception. About 685 National
Priorities List (NPL) sites have been completed, he said, noting that the number is expected to rise to 750
before the end of the fiscal year and to 900 before the end of 2001. Of the 43,000 Comprehensive
Environmental Responsibility, Compensation, and Liability Act (CERCLA) sites that have been inventoried,
Views expressed are those of the participants, not necessarily EPA.
1 '•
-------
he continued, about 40,000 have been addressed and 32,000 have been designated as no further action
(NFA) sites. In addition, the Superfund program has completed about 6,000 removal actions. Luftig said that
the Superfund program must remain viable for many years to come, because there are still several
established sites that must be addressed and about 40 new sites are added to the NPL every year. In addition,
he said, the federal Superfund program is important because it prompts site owners to participate in state
voluntary cleanup programs, state Superfund programs, and the Brownfields program. Many site owners, he
explained, clean up their sites through these programs as alternatives to being listed on the NPL.
Luftig said that the Superfund program is interested in phytoremediation, a technology that uses living plants
to reduce the in situ risk of contaminated media through contaminant removal, degradation, or containment.
EPA's Environmental Response Team (ERT) has already started working on nine different
phytoremediation projects. (These sites are distributed throughout the country. Locations include New
Hampshire, Rhode Island, Maryland, Colorado, Utah, Wyoming, and Oregon.) Luftig said that the program
program is excited about phytoremediation technologies because they are less expensive than many
technologies, effective, driven by solar power, aesthetically pleasing, and capable of restoring properties for
future use. He said that phytoremediation might be a viable approach to use at a variety of Superfund,
Resource Conservation and Recovery Act (RCRA), brownfields, and leaking underground storage tank sites.
He envisioned it being particularly useful at sites that have shallow contaminants that are present at low
levels. In addition, Luftig said phytoremediation could be useful to include in treatment trains.
Despite its impressive benefits, Luftig said, phytoremediation does have some limitations. For example, the
technology may not be appropriate for all locations, because it is heavily influenced by seasonal, geographic,
and climatological conditions. Also, phytoremediation may not be very useful at sites that have deep
contaminants—roots might not be able to reach these—or at sites that must meet cleanup goals rapidly. In
addition, phytoremediation (like almost all technologies) may only be effective for certain types of
contaminants and at certain contaminant concentrations. Luftig said that some regulatory concerns have
been expressed about phytoremediation. For example, will it introduce subsurface contaminants into the
food chain? Restricting access to the site is an issue that must be done to protect the public. Potential
damage (e.g., fires, drought, or storms) may impact phytoremedial systems.
Luftig listed some of the things that should be taken into account if phytoremediation is being considered at
a site. First, he said, site-specific testing should be conducted for site characterization purposes and for
information to be used in the design phase. (He said that EPA's ERT could help with this.) In addition, he
said, according to Executive Orders, native plants should be used if possible, and invasive alien species
should be managed. If phytoremediation is going to be used at a federal Superfund site, Luftig said, the
technology must comply with the National Contingency Plan and meet the nine criteria that drive remedy
selection. He described two of the criteria: (1) protect human health and the environment, and (2) meet
Applicable or Relevant and Appropriate Requirements (ARARs).
Luftig said that phytoremediation references are available at http://www.clu-in.org. He recommended
reading .4 Citizen's Guide to Phytoremediation and the Phytoremediation Resource Guide. In the future, he
said, EPA will release guidance on evapotranspiration covers. Also, the Interstate Technology Regulatory
Cooperation (ITRC) Work Group will write a document that addresses regulatory impediments at the state
level.
Views expressed are those of the participants, not necessarily EPA.
2
-------
The Science and Practice of Phytoremediation
Steven McCntcheon, EPA, National Exposure Research Laboratory
Steve McCutcheon summarized what is currently known about phytoremediation. The term, he said, was
introduced in 1991, in a proposal by Ilya Raskin which was submitted to the Superfund program. An article
published in 1995 in Environmental Science and Technology recognized plants as agents that could degrade
organic chemicals and expanded the definition of phytoremediation. This article's release stimulated a
cascade of interest and research projects. Data collected from these efforts suggest that plants can be used to
address a wide variety of contaminants, such as heavy metals (lead, nickel, zinc, and chromium); other
inorganics (perchlorate, selenium, arsenic, and radionuclides); chlorinated aliphatics (e.g., tfichloroethylene
[TCE] and tetrachloroethylene [PCE]); benzene, toluene, ethylbenzene, and xylene (BTEX); total petroleum
hydrocarbons [TPH]; polycyclic aromatic hydrocarbons (PAHs); polychlorinated biphenyls (PCBs);
phosphorus-based pesticides and nerve agents; munitions (2,4,6-trinitrotoluene [TNT], hexhydro-1,3,5-
trinitrotoluene [RDX], and octahydro-l,3,5,7-tetranitro-l,3,5,7-tetrazine [HMX]); and other
nitroaromatics. In addition, studies are being performed to evaluate whether plants can be used to remediate
atrazine, cyanided compounds, DDT, methyl bromide, nitrites, phenols, and methyl terra butyl ether
(MTBE). ;
McCutcheon provided a brief overview of the limitations and the obstacles that phytoremediation faces. He
said that the technology may only be practical at sites that have shallow contaminants that are present at low
to moderate levels. Also, he said, the technology has mass transport limitations, and this slows down
remedial processes. In addition, regulatory obstacles still remain. Efforts must be made to address the
questions of regulators about phytoremediation and gather information that addresses bioaccumulation and
product toxicity concerns.
McCutcheon said that there are six types of phytoremediation. He described these and summarized what is
known about them:
Types of Phytoremediation/State of the Science/Proof of Concept
Phytotransformation or phytodegradation: Contaminants are metabolized or broken down within plants.
State of the Science:
• Researchers have isolated and are starting to forecast some of the plant enzymes that degrade organic contaminants.
• Researchers have shown that some contaminants are completely degraded in plants. (This was shown in mass balance and
pathway analyses.)
• Researchers have used axenic tissues to show that some plant enzymatic processes are powerful remediation agents.
• The following biotechnology tools are being used: (1) monoclonal antibodies, (2) ELISA field kits, and (3)
immunofluorescence.
Proof of Concept:
Field study at the Iowa Army Ammunition Plant (AAP) shows that plants metabolize TNT and RDX.
Feasibility study at Joliet AAP shows that natural attenuation by plants and microbes can be used to clean up munitions.
Georgia Tech laboratory study shows that plants clean up industrial waste waters that are contaminated with munitions.
Laboratory studies show that plants degrade chlorinated solvents.
Field studies performed at Hill Air Force Base (AFB), Carswell AFB, and the Cape Canaveral Air Station show that
phytoremedial systems degrade chlorinated solvents. Studies performed by the University of Washington also support this
finding.
Views expressed are those of the participants, not necessarily EPA.
3
-------
Types of Phytoremediation/State of the Science/Proof of Concept
Rhizosphere bioremediation, phytostimulation, or plant-assisted bioremediation: Plants promote remediation by
enhancing microbial activity in the rhizosphere.
State of the Science:
• Researchers have isolated microbes that degrade petroleum hydrocarbons and PAHs.
• Investigators have identified advanced field management techniques for determining what plants to use and whether to
fertilize and irrigate.
• Precise and accurate analytical methods, have been developed; these are used to evaluate treatment efficacy.
Proof of Concept:
• Field studies performed at Craney Island and the Texas Gulf Coast show that plant-based systems degrade petroleum
hydrocarbons and PAHs.
• Field study at Milan AAP shows that phytoremediation systems degrade TNT and RDX.
Phytostabilization: Uses plants to change soil conditions or to stabilize or hold contaminants in place.
Proof of Concept:
• Stabilization has been demonstrated at a Canadian mining site and at zinc smelter sites.
• Lead, nickel, zinc, and cadmium have been stabilized at some industrial sites.
Phytovolatilization: Contaminants are taken up by plants and transpired to the atmosphere.
State of the Science:
• Researchers discovered that grasses and plants speciate and transpire selenium.
• Transgenic Arabidopsis is immune to mercury (Hg) poisoning and transpires Hg(0).
Proof of Concept:
• Selenium uptake has been documented in the Central Valley of California.
• Wetlands have been shown to remove selenium from oil refinery wastes.
Rhizofiltration: Plants extract contaminants from flowing water.
Phytoaccumulation, phytoextraction, or hyperaccumulation: Plants extract metals and organics from the subsurface for
storage in plant leaves and shoots.
State of the Science:
• More than 400 species of hyperaccumulaters have been cataloged worldwide.
• Field test kits have been developed to identify plants that are rich in metals.
Proof of Concept:
• Plants have removed lead at the NJ Magic Marker plots and at a site in Boston, Massachusetts.
• Plants have removed heavy metals and radionuclides at Department of Energy (DOE) facilities.
• Sunflowers have been evaluated to remove radionuclides from ponds at Chernobyl.
• Plants have been evaluated to remove uranium from groundwater at Ashtabula. .
In summary, McCutcheon said that phytoremediation has been shown to (1) control and accumulate lead and
nickel, (2) treat water that is contaminated with munitions, (3) treat petroleum hydrocarbons and PAHs in
contaminated soils, (4) control and treat shallow chlorinated solvent plumes, (5), control selenium in soils
and wetlands, (6) control and treat, radionuclides in soil, and (7) achieve water balance and leachate control
at landfills. Engineers are investigating a broad variety of other applications.
Interstate Technical Regulatory Cooperation (ITRC): Making It Easier for Regulators
Robert Mueller, New Jersey Department of Environmental Protection
Robert Mueller said that ITRC, a state-led organization, promotes innovative technologies by reducing
regulatory barriers, improving state permitting processes, and facilitating technology deployment. He said
that ITRC is currently composed of representatives from 31 states, as well as members from industry, the
Views expressed are those of the participants, not necessarily EPA.
4
-------
Western Governors Association, the Southern States Energy Board, EPA, DOE, the U.S. Department of
-Defense (DOD), and the Environmental Coalition of States. Working together, ITRC's members have
delivered training courses, established information-sharing avenues, served as technology advocates, and
have released case studies, regulatory and technical guidelines, and technology overviews. Mueller
described the benefits that ITRC members receive by participating. State representatives, he said, learn
about demonstrations that are conducted in other jurisdictions, obtain advance knowledge about innovative
technologies, and are able to use ITRC as a sounding board for problems. Industry members gain insight into
the regulatory process and this helps them understand regulatory requirements. Federal agencies have a
forum in which to exchange ideas. DOD and DOE benefit because the ITRC addresses many of the
contaminants that are of concern to them.
Mueller said that 12 work groups have formed within ITRC; each focuses on a particular technology or
technical issue. One is focusing on phytoremediation, he said, noting that this group released a 40-page
decision tree in December 1999 and that it planned to release a detailed regulatory and technical guidance
document in July 2000. He said that ITRC's 11 other work groups have also generated a number of products,
some of which might be of great interest to meeting attendees. Therefore, he summarized the products and
success stories of each of ITRC's work groups:
Status and Activities of ITRC Work Groups
Accelerated Site Characterization Work Group
• Products:
4 Two technology overviews
4 Two guidelines on technical requirements:
- Site characterization and analysis systems
(SCAPS), laser induced fluorometry (LIF)
- SCAPS-VOCs
• Success stories: SCAPS-LIF document helped with
decision to use LIF at an EPA Superfund creosote
site.
Enhanced In Situ: Bioremediation Work Group
• Products:
•4 Technology overview
4 Four guidelines, including Natural Attenuation of
Chlorinated Solvents in Groundwater—Principles and
Practices
4 Case studies
4 Natural attenuation courses (delivered in 1998)
• Success stories: More than 900 regulators and 500
consultants attended the training courses.
Permeable Reactive Barriers (PRBs) Work Group
• Products:
4 Regulatory guidance for chlorinated solvents
4 Regulatory guidance for inorganics and
radionuclides
4 Design guidance for chlorinated solvents
4 Training course (delivered in 1999 and 2000)
• Success stories: At a site in New Jersey, it took less
than four months to design and install a PRB system.
(Decreased permitting time by about four months.)
Metals in Soil Work Group
• Products: :
4 Overviews of three emerging technologies:
- Phytoremediation
- Electrokinetics
— In situ stabilization .
4 Soil Washing Guideline issued in 1997 and updated in
1999
• Success stories: Facilitated community acceptance of soil
washing and phytoremediation at Fort Dix.
Low Temperature Thermal Desorption Work Group
• Products:
4 Three guidelines, addressing:
- Petroleum/coal tar/gas plant wastes
- Chlorinated organics
- Mixed waste and/or mercury
• Success stories: Contributed to $100/ton savings for
treatment in New York.
Verification Work Group
• Products:
4 A matrix of data (provided by 16 states) on the elements
that verification program should have
• Success stories:. Technology verification programs are
incorporating states' verification needs into their programs.
This makes it easier for states to approve technologies.
Views expressed are those of the participants, not necessarily EPA.
5
-------
Status and Activities of ITRC Work Groups
Plasma Technologies Work Group
• Products:
+ Technology overview
Dense Nonaqueous Phase Liquids (DNAPLs) Work
Group
• Planned Products:
+ Overview of the technologies that can be used to
characterize and treat DNAPLs.
Unexploded Ordnance (UXO) Work Group
• Planned Products:
+ Case studies examining how to remove barriers to
using innovative UXO remediation technologies
Enhanced In Situ Biodentrification Work Group
• Planned Products:
4 Technology overview
Radionuclides Work Group
• Planned Products:
4 A catalog of state, federal, and international radionuclide
organizations and their activities
4 A glossary of radionuclide terms
Phytoremediation Work Group
See text above this table.
In closing, Mueller encouraged attendees to visit http://www.itrcweb.org or to contact the ITRC's co-chairs
(Brian Sogorka and Roger Kennett) or project manager (Rick Tomlinson) for additional information.
International Perspective on the Cleanup of Metals and Other Contaminants
Terry Mclntyre, Environment Canada
Terry Mclntyre said that phytoremediation, which has attracted interest internationally, is the most exciting
area of research that he has worked on over the last 30 years. Mclntyre's presentation summarized research
efforts that are being performed in Canada as well as other countries outside the United States. He said that
researchers around the world are working to improve their understanding of the interactions that occur
between contaminants, plant exudates, microorganisms, and a variety of other abiotic and biotic processes.
For example, he said, researchers in Canada are evaluating whether plant exudates can remediate inorganic
and organic contaminants. They are also working diligently to elucidate the role that rhizosphere microbes
play in plant-based remediation. Once more information is available, Mclntyre said, researchers will have a
better understanding of phytoremediation's true potential.
In Canada, Mclntyre said, the majority of phytoremediation research has focused on using plants to address
metals. The decision to focus on metals was made because: (1) many sites in Canada are contaminated with
them; (2) regulators are becoming increasingly interested in cleaning up lead, mercury, and cadmium; and
(3) available physical, chemical, and thermal remediation technologies have had limited success in cleaning
up inorganics. Mclntyre said that Canadian researchers are examining various plants to determine whether
they can mobilize metals. For example, hemlock bark is being used to mobilize metals at a military site in
British Columbia. Also, seaweeds are being evaluated to determine how well they accumulate a variety of
substances (e.g., arsenic, vanadium, iodine, and uranium). Researchers in Australia and New Zealand have
also taken an active interest in plant-based systems that accumulate metals and they have launched efforts to
understand the mechanisms that are involved.
Mclntyre said that Environment Canada believes that it is important to generate a comprehensive list of
plants that have remedial properties. Toward this end, two inventories—Phytorem and Phytopet—have been
developed. Mclntyre said that Phytorem lists about 810 plants that accumulate, tolerate, or hyperaccumulate
metals. (This list was put together from the collective input of 39 countries.) The other database, Phytopet,
lists about 110 grasses, 25 bacteria, and 35 fungi that are known to degrade petroleum hydrocarbons.
Mclntyre said that both databases are available on CD-ROM, and that they would be given to all meeting
Views expressed are those of the participants, not necessarily EPA.
6
-------
attendees at no charge. He said that some of the plants ithat are listed might raise regulatory concerns if they
were chosen for a phytoremediation project in North America. This is because some are not native to the
continent, some are poisonous, some are considered noxious weeds, some are commercial crop species, and
others are on the rare or endangered species list. In addition, he said, some are not available commercially;
thus, some regulators are concerned that the plants might be robbed from relatively pristine wild areas if
approved for use in phytoremediation projects. Mclntyre said that plant selection is not the only aspect of
phytoremediation that has raised regulatory concerns. For example, he said, regulators want to know how
contaminated plants will be disposed of. They are also wrestling with the following issues: (1) is it
acceptable to leave contaminants in the ground if they are not bioavailable? and (2) is Environment
Canada's regulatory oversight on the use and production of plant exudates adequate?
Mclntyre summarized some of the other activities that are currently ongoing in Canada. Strides are being
made to gain a better understanding of agronomics, silvaculture, and ecological factors that must be
considered when choosing which plants to use in a phytoremediation system. Mclntyre said that efforts are
underway to identify the characteristics that make an ecosystem susceptible to plant invasion. Researchers in
Canada are also performing work in the area of biotechnology; some believe that the remedial mechanisms
of plants can be enhanced in the future.
Before closing, Mclntyre reiterated his first point: phytoremediation has attracted international interest. He
said demonstration projects will be initiated in China, Cuba, and the Ukraine before the end of the year.
Also, representatives from Japan have expressed interest in learning more about plant-based technologies. In
addition, phytoremediation conferences are being held around the world and attracting international
audiences. In March 2000, Environment Canada, EPA, and the United Nations Environment Program
(UNEP) met to discuss the possibility of implementing phytoremediation in developing countries. In June
2000, the North Atlantic Treaty Organization plannned to discuss phytoremediation at a meeting in Prague.
In addition, a meeting was planned in Canada in June 2001. Lastly, Environment Canada, EPA, UNEP, and
the United Nations Educational, Scientific, and Cultural Organization (UNESCO) hope to hold a
phytoremediation workshop in Prague sometime over the next few years.
Looking Forward on Phytotechnologies
Steven Rock, EPA, National Risk Management Research Laboratory
Steven Rock expressed great interest in phytotechnologics, saying that plant-based systems have a bright
future and are likely to grow in popularity. So far, he said, about 22 patents have been awarded, and research
efforts have been initiated by a number of academic institutions and other organizations. Rock said that he
envisions plants being used across a wide variety of applications and believes that interdisciplinary research
efforts will lead to innovative uses. He described a number of scenarios in which plant-based systems could
be of use. For example, they can be used to address contaminated soil and groundwater, he said, and have a
promising future in the areas of waste-water treatment, erosion control, mine site reclamation, and landfill
management. The latter application has already attracted significant attention; several site owners have
expressed interest in installing evapotranspiration covers. At many sites, Rock said, phytotechnologies will
probably be used as part of a treatment train. He described one site, which has shallow contaminated
groundwater, where a plant-based system is being used in combination with other treatment systems. As part
of the site's original remediation plan, Rock said, a barrier was installed and a pump-and-treat system was
established. Not surprisingly, site owners found operating a pump-and-treat system to be very costly. In an
effort to defray costs, a tree plantation has been installed at the site. Once fully established, the trees should
serve as an adequate pump during the growing season. This will allow site managers to shut down their
pump-and-treat system for three to five months of the year, which will result in tremendous cost savings.
Views expressed are those of the participants, not necessarily EPA.
7
-------
(The costs incurred by implementing the plant-based system will be compensated during the first year for
which site managers shut their pumps down.)
At some sites, Rock said, phytotechnologies can be used to address a number of environmental problems
simultaneously. For example, at a Region III site, managers decided that installing a plant-based system
would revegetate sterile land, get rid of wastes that are high in fertility, and prevent methane from forming.
As part of this project, wastes from concentrated animal feeding operations were distributed over a sterile
mining site that was covered with fly ash. Once the site was fertilized, plants were installed to clean it up. If
aminal waste had remained confined, it would have become anaerobic and produced methane—a greenhouse
gas.
Rock said that phytotechnologies must be designed on a site-specific basis, noting that researchers should
obtain a thorough understanding of a site's contaminants, subsurface conditions, and climate before
designing a plant-based system. Because phytotechnologies are so new, Rock said, researchers must
incorporate a number of treatability and greenhouse studies in their predesign activities. These tests, which
are costly, raise the overall price tag of phytoremedial approaches. In the future, Rock said, as a broader
body of knowledge is accumulated, researchers may not have to run as many predesign tests. Instead, they
might be able to choose a design by reviewing the results obtained at other sites that had similar conditions.
Rock said that he is glad that developed countries are working toward expanding the knowledge base, noting
that this information will benefit developing countries, which have limited research funds but a great need
for phytotechnologies.
Before closing, Rock sent a special thanks to Tom Wilson, who was not able to attend the conference. He
also recommended two books: Natural Capitalism: Creating the Next Industrial Revolution, by Paul
Hawkens, Amory B. Lovins, and L. Hunter Lovins, and The Lorax, by Dr. Seuss.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
* Plant disposal.One attendee asked if regulators are concerned about contaminants accumulating in plant
shoots, leaves, and fruits. Rock said that the likelihood of contaminants accumulating v/ill differ across
sites, depending on the contaminants present and the plants used. For example, he said, accumulation is
likely at sites that have radionuclides, but not at those that have organic contaminants. McCutcheon said
that plants that accumulate metals are typically harvested and disposed of in hazardous waste landfills. It
may be possible, however, to recover some of these metals. For example, McCutcheon said, the Bureau
of Mines developed a method to recycle nickel and other metals from plants. Mclntyre said that
Environment Canada is: (1) evaluating disposal and recovery techniques, and (2) analyzing the impact
that contaminated biomass has on herbivores and omnivores.
• Gaining regulatory approval to use plants at sites that are contaminated with metals and radionuclides.
One meeting attendee asked whether U.S. regulators will allow site managers to use phytotechnologies at
sites that have metal and radionuclide contamination. Rock said that demonstration projects have already
been initiated at sites that fit this profile. Thus, approval has already been granted, at least to some
degree. Rock said that proposals must adequately address media transfer issues if they are to be
approved.
Views expressed are those of the participants, not necessarily EPA.
8
-------
• Stewards for the environment andphytotechnology. A meeting participant asked the attendees to think
about how they can act as stewards for the environment. Rock said that stewards come in many forms:
not only regulators, consultants, academics, and activists, but also industry representatives. Many of the
latter, he said, are sincerely interested in adopting better pollution prevention and cleanup practices.
SESSION II: FUNDAMENTAL PROCESSES OF PLANTS AND SOIL
Transport of Contaminants in Plant and Soil Systems
Larry Erickson, Kansas State University
Larry Erickson said that laboratory studies have been performed to determine how plants impact
contaminant fate and transport. He described the laboratory setup that is used to track these processes.
Planted systems are enclosed within a chamber, he said, and contaminated water is pumped into the bottom
to establish a saturated zone and a vadose zone. Then, measurements are taken to determine how much water
and contaminant moves into the plant, how much is converted to gas, and how much exits the chamber.
Erickson said that researchers from Kansas State University (KSU) evaluated the fate and transport of
toluene and phenol in mass balance studies. Both of these contaminants virtually disappeared when exposed
to plants, he said, and neither was detected in plant transpirate. Researchers used these findings to formulate
the following hypothesis: toluene is drawn into the rhizosphere and degraded by aerobic microbes. Erickson
said that phenol is probably degraded in a similar fashion, but that anaerobic degradation might also play a
role in degrading it. ;
Erickson said that several research teams have performed experiments to evaluate TCE fate and transport.
Some studies suggest that TCE diffuses through plant roots and is carried to aboveground plant tissues. In
large bald cypress trees, he said, TCE has been detected in tree trunks. (Concentrations were highest near
ground surface and decreased with trunk height.) Studies also show, Erickson said, that TCE diffuses out of
tree walls. He cited TCE diffusivity values that have been measured in bald Cyprus (3 x 10~6 square
centimeters per second [cm2/s]), poplars (1 x 10~6 cm2/s to 3 x 10"6 cm2/s), and willows (1.5 x 10"6 cm2/s to 3
x 10~6 cm2/s). These diffusivity values, he said, are just one order of magnitude lower than that measured for
TCE in water (1 x 10~5 cm2/s). Studies also suggest that plants release TCE to the atmosphere. Under
confined conditions, concentrations in air may be fairly significant. However, when TCE and water
evaporate together, there is a dramatic reduction in the concentration of TCE that is detected in the gas
phase. Thus, in the field, the amount of TCE released to the atmosphere is rapidly dissipated. (More
information on TCE fate and transport, said Erickson, can be found in an article that was recently released in
Environmental Progress.)
Erickson said that KSU used a six-chamber study design to evaluate MTBE transport through alfalfa plants.
(One chamber was unplanted, but the other five were planted. Some chambers were inoculated with
microbes, but others were not. Some chambers were air sparged, but others were not.) From this study,
Erickson said, investigators learned that MTBE behaves differently than TCE. MTBE does get taken up by
plants, he said, but it does not diffuse out as readily as TCE does. Erickson said that KSU calculated
diffusivity for MTBE under uniform and nonUniform boundary conditions. All of the values recorded were
lower than those measured for TCE. (Diffusivity estimates for MTBE ranged from 5 x 10~8 cm2/s to 1.11 x
10~7 cm2/s under uniform boundary conditions and from 8.57 x 10~8 cm2/s to 1.62 x 10"7 cm2/s under
nonuniform boundary conditions.) Erickson said that KSU's results indicate that plants transpire a
significant amount of water, but a much smaller amount of MTBE. According to mass balance studies,
Views expressed are those of the participants, not necessarily EPA.
9
-------
Erickson said, only a fraction of the MTBE that is present in groundwater is taken up into alfalfa plants.
Evidence suggests that some of the contaminant is biodegraded in the rhizosphere; KSU estimated
biodegradation rates (under aerobic conditions) to be about 2 to 5 milligrams (mg) per kilogram (kg) of soil
per day.
Erickson closed by providing references—he advised attendees to peruse Hazardous Substance Research,
which is available online at http://www.engg.ksu.edu/HSRC—and expressing great enthusiasm for
phytoremediation's future. He said that several universities are interested in this area of research, and noted
that many sites have already started using phytoremediation.
Enzymatic Processes Used by Plants to Degrade Organic Compounds
Nelson Lee Wolfe, EPA, National Exposure Research Laboratory
Nelson Lee Wolfe said that efforts are underway to study the role that plant enzymes play in degrading
recalcitrant contaminants. These enzymes do not necessarily mineralize contaminants, he said, but they do
break them down into forms that are more susceptible to microbial degradation and hydrolysis. Wolfe said
that many researchers are evaluating plant redox and hydrolysis reactions to determine whether these play a
role in plant-mediated contaminant breakdown. Before summarizing some of the studies that have been
performed, Wolfe described laboratory techniques that are used to isolate plant enzymes. First, he said,
researchers make sure that plant samples are sterile. That way, they can be confident that the enzymes
identified are not from microorganisms or fungi that have attached themselves to the plants. Wolfe said that
surface sterilization, which involves washing and plating plants, is most commonly used, but that
sterilization by gamma radiation is also used on occasion. Once researchers are certain that plants are sterile,
he said, plants are ground up, an extraction solution is added, and enzymes are separated using
chromatography and sequencing techniques. As an alternative to the process just described, Wolfe said,
enzymes can be isolated from seedlings.
Wolfe said that he and others have studied plant nitroreductases, noting that researchers believe that these
enzymes play an important role in the degradation of munitions and pesticides. Results collected to date
support this idea. In one study, Wolfe said, researchers found that whole plants are able to degrade TNT and
RDX by reducing nitro groups to amino groups. In another study, plants were placed in TNT-contaminated
water and constituent concentrations were measured over time. The contaminant degraded rapidly, with a
half-life of only 70 minutes. Wolfe said that the processes involved in TNT degradation are not completely
known, but that they are complicated, involving the transfer of 18 electrons. Wolfe said that different plant
organs appear to have more nitoreductase activity than others. Studies suggest that activity is highest in roots
and lowest in stems. Efforts have been made to identify specific nitoreductase enzymes in several plants,
such as spinach and Elodea. In the latter, Wolfe said, five enzymes from the nitroreductase class have
already been detected; all of them have different molecular weights.
Wolfe said that several experiments have been performed to evaluate dehalogenase activities in plants. For
example, he said, Dr. Valentine Nzengung found that spirogyra breaks down hexachloroethane (HCA)
rapidly, forming some pentachloroethane, trichloroacetic acid, TCE, and PCE in the process. In another
study, Wolfe said, TCE was shown to degrade quickly when exposed to spirogyra. The degradation rate,
Wolfe said, was influenced by the condition of the spirogyra. When exposed to live spirogyra, TCE had a
half-life of about 22 hours. Rates were much higher when the TCE was exposed to dead spirogyra. Wolfe
said that studies have been performed to determine whether plant leaves have high dehalogenase activity. In
one study, Dr. Peter Jeffries collected leaves from 29 different types of plants, placed them in separate-
containers, and spiked the surrounding air with methylene bromide. In all cases, the contaminant decreased,
Views expressed are those of the participants, not necessarily EPA.
10
_
-------
with disappearance rates proportional to leaf surface area. Wolfe said that Jeffries also experimented with
chlorinated compounds, finding that HCA disappears in the presence of holly trees and that some PCE and
chloroform form in the process. Wolfe said that studies have been performed to determine how fast TCE
degrades in leaves. For this effort, he said, cottonwood leaves from Carswell AFB were ground up and
spiked with TCE. Significant degradation was observed; the average half-life recorded was 14.1 hours.
Wolfe said that research has also been performed on plant phosphatases to determine whether they can
degrade organophosphate pesticides. Researchers in Japan isolated an alkaline phosphatase from duckweed,
he said, but it did not exhibit degradation activity towards organophosphate pesticides. Wolfe's group has
isolated an acid phosphatase, that has been shown to break down three organophosphate pesticides that
differ in structure.
Wolfe said that efforts have recently been initiated to evaluate plant peroxidases, enzymes that are
considered fairly stable and therefore capable of persisting on the ground long enough to percolate into soil.
To date, he said, several peroxidases have been isolated from grasses. In addition, Wolfe said, new efforts
are just getting underway to evaluate laccases and nitrilases for phytoremediation.
Biosystem Treatment of Recalcitrant Soil Contaminants
John Fletcher, University of Oklahoma
John Fletcher said that the University of Oklahoma is optimistic that plant-based systems can be used to
treat sites that are contaminated with PCBs and PAHs. He said that these recalcitrant contaminants are
highly immobile. They have low water solubilities and high log Kow. Within the last 150 years, Fletcher said,
PAH contaminants have been disposed of at about 18,000 industrial installations. He could not estimate how
many PCB-contaminated sites exist, but did say that some may have been contaminated as far back as 70
years ago. Fletcher said that he has turned to nature in, an effort to find solutions to environmental problems,
and that he is optimistic that contaminated sites can be remediated using biosystem treatments that
incorporate evapotranspiration, rhizosphere degradation, and natural attenuation.
Fletcher said that plants produce polyaromatics (e.g., tannins and flavanoids), that can break these
compounds down. (Researchers know this because the amount of natural polyaromatics detected in soil is
much less than the amount that has been produced by plants over extended time periods.) Fletcher said that
these natural events serve as the underpinning for a hypothesis that he formulated about 10 years ago. The
hypothesis is: roots of some plant species enhance the degradation of recalcitrant, organic soil contaminants
(e.g., PCBs and PAHs) by releasing co-metabolites (e.g., flavanoids) and facilitating soil aeration as fine
roots turn over. Fletcher said that research has been conducted to test this hypothesis and to gain a better
understanding of the mechanistic basis for rhizosphere degradation. The results obtained support the
hypothesis. For example, in studies performed on mulberry plants, Fletcher said, researchers learned that
natural polyaromatics reside in fine roots (less than 5 millimeters in diameter). When roots die, these
polyaromatics are released to the surrounding soil, where they promote the growth of rhizosphere microbes.
Fletcher said that he suspects that these microbes, which proliferate in the presence of the natural
polyaromatics, can also degrade compounds with similar structures, such as recalcitrant PCBs and PAHs.
Assuming that this is true, he said, contaminants that are located near dying root hairs would be degraded.
As each year passes, Fletcher continued, roots expand their network, contact more contaminants, facilitate
oxygen diffusion, and sustain an active community of PAH-degrading microbes. Fletcher asked participants
to take note of four points: (1) fine roots grow to contaminants over time, (2) roots turn over each year, (3)
plants must be able to produce polyaromatics that are structurally similar to anthropogenic contaminants,
and (4) plants must grow for several years before reductions in contaminant concentrations will be
Views expressed are those of the participants, not necessarily EPA.
11
-------
statistically significant. Fletcher stressed the latter point, noting that plant-based systems will be unfairly set
up for failure if their efficacy is measured based on short-term studies that simply consider end results (i.e.,
the disappearance of contaminants). During the first five years of a phytoremediation study, Fletcher said, it
is more meaningful to study the mechanisms that are responsible for degradation, such as root growth, root
depth, and the presence and activity of PAH- and PCB-degrading microbes. He said that the University of
Oklahoma is working toward identifying useful tools for assessing these parameters.
With the help of Union Carbide, Fletcher said, the University of Oklahoma studied the impact that plants
had at a former sludge lagoon. This site, Fletcher said, was invaded by plants in 1983, one year after the
lagoon was drained. Since that time, he continued, mulberry trees and 50 other plant species have become
firmly established at the site. Fletcher said that researchers collected about 60 cores from the lagoon, over
three different depths, to determine whether the 17-year-old plant-based system had caused a reduction in
PAH concentrations. It had. While total PAHs in the parent sludge were present at about 17,000 parts per
million (ppm), concentrations were only 1,121 ppm in the site's first foot of soil, 2,654 ppm in the second
foot, and 9,190 ppm at the bottom of the root zone. Fletcher said that the same pattern was observed with
individual compounds. For example, the concentrations of benzo(a)pyrene (a water soluble contaminant)
were 135 ppm in the parent sludge, 21 ppm in the first foot of soil, 34 ppm in the second foot, and 70 ppm at
the bottom of the root zone. Fletcher said that researchers collected a number of soil samples from the site
and isolated microorganisms. About 250 different isolates that have the ability to degrade PAHs were
identified. In-depth studies were conducted on three of these microbes; all three were shown to grow on
polyaromatics (e.g., morusin and catechin) that are present in mulberry and oak systems.
Fletcher addressed some of the concerns that have been voiced about plant-based systems. First, he said,
some people claim that it takes too long to determine whether plants are effectively remediating a site.
Fletcher said that this view is based upon using contaminant concentrations as the only measure of success.
If the mechanisms of degradation are monitored instead, Fletcher said, it will be possible to show, relatively
quickly, whether a plant-based system is working. Another view is that phytoremediation does not work
quickly enough, saying that it takes too long to achieve end results. Fletcher said that he did not think the
time periods required were unreasonable, given that some sites have already been contaminated for many
years. Lastly, Fletcher noted, some people say that it is better to use impermeable caps instead of plant-based
systems, because using the former has already been demonstrated as a proven solution. Fletcher pointed out
that the decision to use impermeable caps was made in the 1970s, long before information had been gathered
on natural attenuation or phytoremediation. Also, he said encapsulating waste retains it for future
generations.
In closing, Fletcher said that he is a proponent of biosystem treatment, a holistic approach that involves
evapotranspiration, rhizosphere degradation, and natural attenuation. He advised establishing systems that
perform long-term sustained cleanups which in his view, natural systems can accomplish.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
• Biodiversity. Mclntyre asked Fletcher to comment on biodiversity at the Union Carbide site. Fletcher
said that there was much variety at this site, and said that more specific information will be presented in
three upcoming papers. (One of the papers will be called "Ecological Recovery with Phytoremediation
Overtones.") Fletcher said that it is not uncommon for contaminated sites to exhibit as much diversity as
Views expressed are those of the participants, not necessarily EPA.
12
-------
uncontaminated sites do. The types of plants that are represented at contaminated and uncontaminated
sites, however, are not necessarily the same. •
Biological versus physical processes. McMillan asked Fletcher how he could be sure that biological
processes were responsible for contaminant reductions at the Union Carbide site. Wasn't it possible, he
asked, that the contaminants simply migrated downward over time? Fletcher said that he is confident that
biological processes were responsible. He noted that sludge is still present below the site's root zone. If
contaminants had simply migrated downward, he said, a contaminated hot zone would have been
detected at the border between the root zone and the sludge. No such zone was detected. Also, he said,
many of the contaminants at the site are immobile. Thus, it is unlikely that they would have migrated
downward at any significant rate.
Perspectives on the length of time it takes for plant-based systems to achieve end results. An attendee
noted that plant-based systems may require as long as 20 years to reduce chemical concentrations to
acceptable levels. She asked whether regulators are willing to wait that long for results. Mclntyre said
that regulators from Environment Canada would probably not regard this as an unreasonable time frame.
Rock said that he did not think it would be a huge stumbling block among EPA regulators either, noting
that many conventional treatments (e.g., pump-andTtreat) require at least that much time to achieve their
goals. Harry Compton agreed, noting that regulators are often flexible with the amount of time it takes to
clean up sites, as along as the site poses no immediate threat to human health or the environment. Judy
Canova said that the same holds true at the state level, at least in South Carolina. EPA representatives,
said that regulators might not be willing to wait for 20 years unless they could have some assurances
along the way that the system was working. Responding to this last statement, Fletcher said that he hopes
to find ways to give these assurances. He reiterated a point he made earlier: it is time to start monitoring
the processes that cause degradation rather than just focusing on contaminant concentrations. If
researchers can prove that the appropriate degradation mechanisms are operating at a site, Fletcher said,
that should serve as sufficient proof that plant-based systems are working. He said that much research
remains to be done to identify ways to monitor the mechanisms, but said that this research really must be
done. Fletcher said that it is time that phytoremediation gets the funding and the prominence it deserves.
Meeting attendees responded in a chorus of agreement.
Life cycle studies. One meeting attendee said that he was glad to hear Fletcher and Wolfe address the
role of plant life cycles.
SESSION IIIA: BROWNFIELDS APPLICATIONS AND BENEFICIAL USE OF LAND
Goals for Brownfields Pilots •
John Podgurski, EPA, Region 1 ',
John Podgurski opened his presentation by providing general information about brownfields sites, noting
that some estimates suggest that there are 500,000 located across the country. He defined these sites as
vacant or underutilized industrial/commercial properties where redevelopment is complicated by real or
perceived contamination. Brownfields are located, he said, in small communities as well as large urban
centers. In the latter areas, most of the sites are relatively small and have structures on top of them.
Podgurski presented data collected from Lawrence, Massachusetts, to demonstrate his point. In this city,
60% of brownfields sites are less than 1.5 acres in size, 60% have one building, and 10% have two or more
buildings.
Views expressed are those of the participants, not necessarily EPA.
13
-------
Podgurski described the steps that developers take and the issues they encounter when deciding whether to
redevelop brownfields sites. They must assess the local property value of the site, he said, along with
existing infrastructure (e.g., utilities and roadways), security, and safety. Also, developers must address
several technical issues. For example, if the extent of site contamination is not known, developers must
invest in assessment activities. Also, if a developer finds out that a remediation effort will be required, he or
she will consider the following when choosing a cleanup technology: cost, time frame, efficacy, regulatory
acceptance, space constraints, and the potential that community concerns will delay remedial efforts.
Podgurski said that phytoremediation has the potential to be useful at many, but not all, urban brownfields
sites. Poor candidates include sites located on very small parcels of land, sites whose surfaces are largely
covered by structures or pavement, and sites that have difficult growing conditions. (Podgurski said that
growing conditions may be less than ideal at brownfields sites, especially if they are shaded by tall buildings
or if their soils are littered with debris.) In addition, Podgurski said, phytoremediation is not appropriate for
sites that require quick clean up. At privately controlled sites, he said, it is likely that developers have
already made reuse plans. If so, they typically require cleanup quicker than phytoremediation can offer.
Other issues that must be addressed when determining whether phytotechnologies are appropriate include
security, vandalism, equipment access, and the public's willingness to have an innovative remediation
technology used for an area that many people have access to.
Despite the obstacles listed above, Podgurski said, he does believe that there is great potential for
phytoremediation at a large number of urban brownfields sites, particularly those that are publicly
controlled. The time frame for remediation may not be an important issue at these sites. If this is the case,
city managers might use phytoremediation as a cleanup strategy, because it is a more cost-effective way to
remediate and stabilize sites in preparation for future redevelopment. In addition, he said, city managers will
be attracted to phytoremediation because it could reduce contaminant exposures to the public and improve
an area's aesthetics,. Podgurski said that city managers might find the startup costs of phytoremediation to
be high, but said that equipment and other startup costs could be spread out if managers decided to use
phytoremediation at several sites. Podgurski said that Operation and Maintenance (O&M) costs would
probably not be very significant particularly if they can be maintained by the local government itself.
Hartford Brownfields Site: Site Description and Summary of Phytoremediation Project
Jeanne Webb, City of Hartford
Jeannie Webb provided a description of the Chestnut and Edwards Street site, a brownfields site in Hartford,
Connecticut, and explained how phytoremediation is being used to remediate it. The site is about 1.7 acres
in size and is located in a transitional neighborhood. A soup kitchen is located next door, Webb said, and
many people use the site as a pathway to and from the kitchen. Webb said that there are also shelters,
residential buildings, and a furniture store in close proximity.
A few years ago, Webb said, neighborhood representatives asked whether the site could be used as a
recreational area and community garden. She said that this idea stimulated significant interest, and a team of
organizations banded together to determine whether this wish could become reality. Some of the major
players on the team included the City of Hartford, Trinity College, a local middle school, House of Bread,
neighborhood revitalization committees, and the University of Connecticut Master Gardeners Program.
Webb said that a Phase II site investigation was performed at the site to determine whether it could be used
as a community garden. The results indicated that lead was scattered throughout the site's surface soil at
concentrations as high as 2,800 ppm in some areas. Thus, it was clear that remedial action had to be taken,
Views expressed are those of the participants, not necessarily EPA.
14
-------
and several different approaches were assessed. For example, soil removal was considered, but city
managers decided that they did not have enough funds for this. Also, Webb said, some consideration was
given to covering the soil with mulch and creating raised-bed gardens. In the end, the decision was made to
fence off the most contaminated portion of the site and establish a phytoremediation project on it. Field
activities for this effort were initiated in the summer of 1999. Trinity College, which proposed using
phytoremediation in the first place, worked with Edenspace Systems Corporation to establish crops of
mustard, sudan grassland sunflowers. Webb said that the crop was planted twice, During the first planting,
it was difficult for seeds to penetrate the ground, because the site was littered with so many bricks and other
kinds of litter. Before the second planting effort was initiated, the site was rototilled, and the plants grew
much better. Webb said that the city recently identified additional funds that can be used for the site. As a
result, plans were made to install a plant-based system at another portion of the site during summer 2000.
Webb said that communication has played a key role in propelling this phytoremediation project, noting that
substantial efforts have been made to educate community members and municipal workers. The latter, she
said, will perform O&M at the site. Webb said that all of the interested parties sat down together to discuss
options for the site. As a result, people felt that their opinions and concerns had been heard. In the end, the
community became very excited about the prospect of phytoremediation.
t
Hartford Brownfields Site: Public Health Perspective
Jennifer Kertanis, Connecticut Department of Public Health
Jennifer Kertanis, an epidemiologist and public health official, said that she was also involved with the
Chestnut and Edwards Street site. (This site was described in the previous presentation.) Kertanis said that
her department performed a thorough exposure pathway assessment for the site. This involved evaluating the
source and type of contamination, the media impacted, potential exposure routes (i.e., ingestion, dermal
contact, or inhalation), and property usage patterns. Several public health concerns were identified during
the assessment process. For example, Kertanis said, there was concern that people could be contacting the
site's contaminated soil directly or inhaling airborne soil particles, because there was unrestricted access to
the site, passive recreational usage, and limited ground cover over contaminated soils. Thus, she said, before
voicing approval for a remedial approach, she had to be assured that the technology chosen would address
existing public health hazards without creating brand new ones.
Kertanis said that community members and public health officials both thought it was important to evaluate
phytoremediation to make sure it would not create nevy exposure routes. Some concerns were identified, she
said, but solutions were found to address each of them. For example, there was some concern that rototilling
could create airborne dust, but soils were wetted beforehand to reduce the chance of this happening. Also,
she said, concern was expressed about soils becoming'airborne in the winter, when plants die and soils are
bare. She said that these concerns can be addressed by placing ground mats on top of the soils or adding
mulch to the site. Also, Kertanis said, public health officials were concerned that professors and students,
who were involved with plant installation, could be exposed to contaminants. Thus, these populations were
given instructions on how to reduce the potential for contact. Kertanis said that community members
expressed concern about people having access to the project's crops. By listening to the community, public
health officials learned that members of certain local ethnic groups eat a lot of mustard, one of the crops
used in the phytoremedial system. To prevent people from eating the crops, Kertanis said, a fence was
installed and information was distributed advising community members to stay away from the plants.
Views expressed are those of the participants, not necessarily EPA.
15
-------
Kertanis stressed the importance of including local public health departments in decision-making processes.
She said that public health officials can do the following to assist phytoremediation projects: (1) assess and
address community concerns, (2) sort out real versus perceived health risks, (3) highlight issues that are not
covered by EPA or local departments of environmental protection, (4) provide input and support, and (5)
communicate effectively with the public. Kertanis said that she believes that the most successful projects are
those in which all involved parties—site owners, community members, environmental agencies, public
health organizations, and a variety of other stakeholders—sit down together to discuss issues and solutions.
When collaborative efforts are fostered, it becomes clear that economic development, site reuse, public
health issues, environmental issues, and innovative technologies do not have to be competing interests.
Integrating Remediation into Landscape Design
Niall Kirhvood, Harvard University, Graduate School of Design
Niall Kirkwood said that landscape architects are interested in applying phytoremediation and hope that
additional research will be performed to support the use of such technologies. He said that landscape
architects and phytoremediation designers have much in common: both have extensive knowledge about
plants, and both are interested in using plants to improve quality of life. Furthermore, they use similar
processes to address sites. He described the steps that landscape designers go through: (1) site analysis, (2)
conceptual design, (3) schematic design, (4) documentation and bidding, (5) implementation, and (6)
maintenance and post-occupancy evaluation. Kirkwood said that there is potential for great synergy between
the two disciplines, and he is optimistic that objectives will merge in the future. If so, plant-based systems
could provide multiple benefits simultaneously. For example, he said, systems could be designed so that they
simultaneously provide remedial benefits, aesthetics and visual screening, microclimate control, spatial
enclosure and form, and a "green" buffer that attracts wildlife.
Kirkwood showed slides of a variety of plant-based projects. At one site along the Colorado River, he said,
about 30,000 cottonwoods have been planted in an effort to stabilize soil. Outside of London, he said, a
landscape design project has been initiated to implement infrastructure. He noted that fuel oil is present in
the groundwater under this site, and wondered whether the design could have been set up to address this and
implement infrastructure at the same time. At another site, located in Germany, landscape design was used
to convert a former steel smelting site into a public park with beautiful fields and gardens. Kirkwood said
that trees are being used to take up groundwater. Thus, the site is also serving a remedial purpose, although
it is not advertised as a phytoremediation project.
Kirkwood said that he envisions several areas where phytoremediation could have potential: (1) regional
parks, city parks, and community recreational open spaces; (2) commercial/industrial parks and
biotechnology centers; (3) assisted and private housing areas; (4) locations that require "green"
infrastructure, such as roads and rail corridors; (5) landfills that are being converted into golf courses; and
(6) areas that are being developed as urban arboretums or environmental education centers. He expressed «
much enthusiasm for the latter, noting that consideration is currently being given to creating an
arboretum/botanical garden at a former zinc smelting site in Mexico.
Kirkwood spent the remainder of his presentation talking about brownfields sites and how phytoremediation
could be used to address them. Within the brownfields community, he said, four issues must be addressed:
(1) sustainable economics, (2) social issues, (3) environmental educational opportunities, and (4)
innovations in assessment and remediation technologies. He said that phytoremediation could be beneficial
at many brownfields sites, but said that some misconceptions must be corrected before it is seriously
considered as a viable approach. For example, he said, it must be made clear that phytoremediation has
Views expressed are those of the participants, not necessarily EPA.
16
-------
applications at the urban scale, not only at the agricultural scale. To determine whether phytoremediation is
appropriate for a specific brownfields site, he said, the following issues must be assessed: urban context,
existing site conditions, other engineering activities, plant growth concerns, adjacent community concerns,
proposed reuse program, implementation controls, disposal methods, and time lines. Kirkwood said that he
believes that phytoremediation will be able to address a wide variety of issues that arise at brownfields sites,
including environmental protection, public health, economic development, and municipal planning for
development. If phytoremediation is adopted as a popular approach at brownfields sites, he said, it will
allow researchers to explore how the technology works across a variety of scales and contaminant profiles. It
will also allow researchers to test a variety of plant species all over the country.
In closing, Kirkwood encouraged attendees to think of ways in which phytoremediation could offer multiple
benefits at brownfields sites. For example, he said, it could be used to create sustainable open space or be
combined with biomass fuel projects or "Brightsfield" solar projects. Also, he said, phytoremediation could
be used to improve aesthetics and public perception, promote smart growth, and restore habitats.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
* Accumulation of contaminants and plant disposal issues. A question was asked whether lead was
detected in the Chestnut and Edwards Street site mustard plants, and at what concentrations. Webb and
Kertanis did not know. Another meeting attendee asked how the mustard plants were disposed of. Webb
said that they were harvested, burned at the Trinity College biology laboratory, and then disposed of as
contaminated materials. One meeting attendee asked Kirkwood a question about the landscape design
project that was performed in Germany: are tree leaves harvested and disposed of off site? Kirkwood
said that they are raked up during normal maintenance activities and burned in the open air.
• Brownfields site near Boston, Massachusetts. It was suggested by an attendee, that the Boston Public
Health Commission might initiate a phytoremediation project at a site near Boston.
• Treatment trains. Erickson said that some researchers use phytoremediation to reduce lead
concentrations, and then use stabilization techniques to convert the remainder to lead pyromorphite, a
benign compound. He asked whether a similar approach will be used at the Chestnut and Edwards Street
site. Kertanis said this will probably not be necessary. She is optimistic that phytoremedial efforts will
reduce lead concentrations below 500 ppm, the Connecticut residential cleanup standard. If this is
achieved, she said, no more remedial activities will be required, and the site will be considered safe for
community gardening.
• Weighing public health risks. One meeting attendee asked Kertanis to explain how assessors determine
when the public health hazards posed by phytoremediation are so significant that a project cannot be
allowed to proceed. Kertanis said that the decision is site-specific. From a public health perspective, she
said, it would have been easier to excavate and remove soils from the Chestnut and Edwards Street site,
because doing so would have eliminated the potential for any exposure. However, she said, at this site,
the community was educated about phytoremediation and very excited about the demonstration. Thus,
the decision was made to use phytoremediation, and to identify solutions to the potential risks that the
technology could pose. :
Views expressed are those of the participants, not necessarily EPA.
17
-------
SESSION IHB: RADIONUCLEDES
Summary of DOE Projects and Recommendations Offered at a Recent DOE Workshop
Scott McMullin, DOE
Scott McMullin said that several DOE sites are contaminated with heavy metals and radionuclides. In the
past, DOE officials thought that excavation and removal would be used to address the vast majority of this
contamination. More recently, however, it has become clear that in situ technologies will play a large role in
the remediation of these inorganic wastes; in fact, one DOE official has predicted that as much as three-
quarters of the wastes will be addressed in situ. Thus, DOE has started evaluating stabilization, containment,
and a variety of biologically driven technologies, such as phytoremediation and natural attenuation.
McMullin said that DOE's science program has already issued 17 grants for projects that involve
phytoremediation or biological processes. For example, he said, plant-based systems are being used as
cleanup technologies for tritium-contaminated groundwater at Savannah River and Argonne National
Laboratory-East. In addition, at a site in Poland, phytoremediation is being used to extract heavy metals
from surflcial soils.
In November 1999, McMullin said, DOE hosted a workshop to discuss what is currently known about using
plant-based systems to address inorganics. About 75 people attended; representatives from consulting
companies, federal agencies, state agencies, and a number of international groups were present. Four
breakout groups formed, each of which discussed a specific phytoremedial application. The following table
lists the groups, along with the key issues and suggestions that each identified during the meeting.
Focus Group
Key Issues and Recommendations
Using plants to
address groundwater
More information is needed on rhizosphere activities.
More demonstration sites are needed.
A plant database is needed.
Phytoextraction
More demonstration sites are needed.
Cost data need to be generated and a life-cycle accounting system should be developed.
Basic research on plant physiology should be conducted.
Ecological risks must be addressed. .
Phytostabilization
Efforts should be made to improve the understanding of:
— Long-term bioavailability.
— Rhizosphere biochemistry.
- Toxicity relative to the food chain.
— Predictive models.
- The physiological processes involved with sequestration.
Using plants as a
monitoring tool
Field tests should be conducted.
Plant characterization tools should be developed.
Monitoring elements should be defined.
A monitoring framework should be defined.
McMullin said that DOE is evaluating the recommendations that each breakout group offered. In 2001,
another workshop will be held to match up some of these recommendations with programmatic directives.
McMullin was optimistic that phytoremediation will continue to grow under DOE.
Views expressed are those of the participants, not necessarily EPA.
18
-------
Capturing a "Mixed" Contaminant Plume: Tritium Phytoevaporation at Argonne
National Laboratory-East
M. Cristina Negri, Argonne National Laboratory
Cristina Negri said that a phytoremediation project has been initiated at the 317/319 area at Argonne
National Laboratory-East (ANL-E). This area received wastes between 1940 and 1960, Negri said, noting
that solvents were poured down a french drain and a variety of wastes were disposed of in a landfill. As a
result of these practices, soils are contaminated with volatile organic compounds (VOCs) and groundwater is
contaminated with VOCs and tritium. (Tritium is a low-energy beta emitter that has a half-life of 12.6 years;
decays to helium-3; shares hydrogen's chemical and physical properties; can pose human health hazards
when absorbed, ingested, or inhaled; and has an average biological half-life of 7.5 to 9.5 days within the
human body.) Negri said that the site's glacial subsurface is very complex and heterogenous, consisting of
boulders, gravel, sand lenses, clay lenses, and silts. Water-bearing intervals are interconnected sand and
gravel zones.
Negri said that groundwater at the site is about 20 to 30 feet deep and is migrating toward property boundary
lines. Extraction wells have been installed to capture some of the plume; the water collected is forwarded to
Argonne's treatment plant before being released. Althpugh the extraction wells do capture a portion of the
plume, Negri said, site managers agreed that more had to be done to address the site. They decided not to use
DOE's "baseline" technology, which would have involved installing an asphalt cap and adding additional
extraction wells, because the site's complicated subsurface makes it too difficult to predict where to drill
wells so that they have a significant zone of influence. Instead, site managers decided to install plants,
because they have the ability to grow and extend roots towards water. Negri said that using
phytoremediation made sense, since the following had already been shown or hypothesized: (1) tritium can
be directly incorporated into water and biological tissues; (2) plants transpire tritiated water vapor; (3)
tritium can accumulate in plant tissues, but the contaminant mean residence time is only about 4 to 37 days;
and (4) contaminated groundwater can be controlled using engineered plant-based systems.
To gain regulatory approval for the phytoremediation project, Negri said, site managers had to prove that the
plants would not release hazardous quantities of airborne tritium. They did so using EPA's CAP-88PC
exposure equations to calculate tritium emissions. When putting parameters into the equation, she said, site
managers assumed Worst-case scenarios. That is, they identified the highest tritium concentrations that had
ever been detected in the groundwater, assumed that plants would be exposed to that level at all times, and
assumed that all of the tritium would be transpired. Emission levels were then calculated for an entire
mature plantation, based on the assumption that transpiration rates would range from 2 to 50 gallons per day
per tree, depending on the season. The results, said Negri, indicated that emission rates would be several
orders of magnitude lower than what is allowed under National Emission Standards For Hazardous Air
Pollutants.
In 1999, Negri said, a phytoremediation project was initiated at ANL-E. It has the following components: (1)
a herbaceous cover that is designed to minimize water infiltration and soil erosion; (2) deep-planted unlined
TreeMediation® hybrid willows, which have been installed to address VOC source areas; and (3) deep-
planted Tree Well® hybrid poplars, which are expected to achieve hydraulic groundwater control. The latter
were planted in lined caissons that extended to depths of 30 feet and were backfilled with materials that
promote root development. Thus, the Tree Well® roots are cut off from all shallow sources of water and
have been forced to grow downward in search of groundwater. Negri said that plume velocity and winter
dormancy were considered when the number of trees to include in the phytoremediation design was
determined. She said that the phytoremediation system is currently being monitored by two entities:
Views expressed are those of the participants, not necessarily EPA.
-------
Argonne National Laboratory and the EPA Superfund Innovative Technologies Evaluation (SITE) program.
Site managers hope that results will show that: (1) hydraulic control can be obtained within four years, (2)
tritium can be transpired without causing significant hazardous emissions, and (3) existing extraction wells
can be shut off.
Negri closed her presentation by citing the potential benefits that could be realized using phytoremediation
instead of DOE's "baseline" technology for this site. First, she said, installation costs associated with
phytoremediation were about 33% less and O&M costs are projected to provide a cost savings of more than
30%. Also, she said, using in situ technologies: (1) minimizes the amount of contact that remediation
workers have with contaminants and (2) eliminates the need to transport hazardous materials off site. In
addition, Negri said, phytoremediation has the potential to accelerate the cleanup time of VOCs, restore the
site's rhizosphere, prepare the site for native prairie species, and provide protection against unknown source
areas that might be located within the site.
Phytoremediation Application for Radionuclide Removal at Argonne National Laboratory-West
Scott Lee, Argonne National Laboratory
Scott Lee said that a phytoremediation project has been initiated at Argonne National Laboratory-West
(ANL-W). The laboratory, located in Idaho, was listed on the NPL in 1991. Several contaminated sites have
been identified within the property's boundaries, many of which are contaminated with heavy metals and
radionuclides. Cesium-137 (Cs-137), a radionuclide with a half-life of 30 years, is present at concentrations
that could pose potential human health hazards. Lee described efforts that are being made to address Cs-137
contamination at ANL-Ws industrial waste pond, interceptor canal, and interceptor canal mound. Cs-137 is
detected at elevated levels in all three of these areas, but the site's groundwater, which is about 650 feet
deep, has not been contaminated. Lee said that ANL-W evaluated a number of technologies (e.g., excavation
and capping) before deciding to use phytoremediation. Before gaining approval for this technology, he said,
ANL-W representatives had to address the following concerns:
" Ecological risks. There is a wide variety of wildlife at ANL-W. For example, burrowing animals, owls,
antelope, and rabbits frequent the site and use the industrial waste pond as a watering hole. Community
members were concerned that Cs-137 could be introduced to the food chain if plants were used to extract
the constituent from the ground. Lee said that this concern did not pose an obstacle to the
phytoremediation project, because Cs-137 was not identified as an major ecological risk at the site.
(Chromium*3, mercury, selenium, silver, and zinc have been identified as ecological risk drivers.)
• Public concerns. Lee said that radionuclides generate great fear among the public. ANL-W
representatives addressed public health concerns.
• Leaching of contaminants. Lee said that irrigation is included in ANL-W's phytoremediation design; the
water frees up the Cs-137 so that it can be taken up by plant roots. Lee said that some people were
concerned that Cs-137 would leach downward. To assure community members that this would not occur,
soil moisture reflectometers were installed. These are stacked on top of each other, he said; the one on
top is connected to the irrigation system and signals it to turn off when soil moisture is high. The one on
the bottom is used to determine whether moisture is migrating downward.
• Noxious weeds. ANL-W proposed using Russian thistle, but the state would not allow it because that
plant was considered a noxious weed. ANL-W then proposed using kochia weed. This choice was
approved after ANL-W agreed to control growth by harvesting plants before they flower, establish a
Views expressed are those of the participants, not necessarily EPA.
20
-------
clear zone around the site, and apply seeds with a paper mulch so that the seeds would not be dispersed
by the wind. ,
Lee said that the phytoremediation project was initiated in 1999. ANL-W did the following before planting:
(1) installed irrigation lines, pressure regulators, risers, heads and moisture probes; (2) added organic matter
to the site; (3) prepared the soil using a plow, ripper, and rototiller; and (4) collected real-time gamma
emissions using a global positioning radiometric scanrier (GPRS). Then, kochia weed was
hydroseeded—seeds were distributed about 4 inches apart from each other—and allowed to grow. The weed
was harvested after about 110 days, Lee said, using potato harvesting farming equipment. He said that this
equipment was used because it harvests roots as well as aboveground portions of plants. After the plants
were harvested, a separator was used to remove attached soils, and the plants were baled and taken to an
incinerator. Once the site was cleared, gamma emissions were measured again using the GPRS. Lee said that
problems were encountered using the detectors. Thus, it was difficult to determine whether Cs-137 levels
had decreased over the course of the first growing season. Over the next several years, Lee said, several
additional plants will be seeded and harvested at the site. He believes that it will take about five growing
seasons to meet cleanup goals.
Lee closed his presentation by listing some of the upcqming activities that are planned for this site. First, he
said, the efficacy of three amaranth species will be compared against that of the kochia weed to determine
which plants are best to use for Cs-137 cleanup. Second, he said, ANL-W will perform studies to determine
whether efficacy rates differ between stressed and unstressed plants. Lastly, he said, ISSOX, a directional
sodium germanium detector, will be used prior to planting, harvesting, raking, and bailing.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
• Bioaccumulation of radionuclides. One attendee asked whether Cs-137 bioaccumulates in kochia weeds.
Lee said that the bioaccumulation ratio is low; for example, at soil concentrations of 30 picocuries per
gram (pCi/g), only about 1 pCi/g is accumulated in plants. The attendee said that higher bioaccumulation
rates have been reported by researchers who have experimented with other plants.
• Cs-137 accumulation in roots versus shoots. One attendee asked whether roots and shoots accumulate
different amounts of Cs-137. Lee said that they do. As shown in a greenhouse study, he said, Cs-137
accumulation rates in kochia weed roots and shoots are about 10 to 1.
• Ways to optimize treatment. One attendee asked whether soil amendments could be applied to increase
Cs-137 uptake by plants. Negri said that a number of treatability studies were performed on ANL-W's
soils to determine whether removal rates could be improved by adding ammonium salts, potassium salts,
EDTA, or citric acid. No significant improvements:were detected. Mclntyre asked whether any efforts
are being made to enhance the remedial capabilities of plants: have efforts been made to identify the
kochia weed's mode of action and to incorporate it into plants that produce more biomass? Mclntyre also
asked whether other plants have been studied to evaluate their potential to uptake Cs-137. Lee said that
Amaranthus retroflexus, A. bicolor, and A. paniculatum will be evaluated for their potential to uptake
Cs-137. Negri said that it might be possible to identify plants that uptake Cs-137 more vigorously than
kochia weed, but said that she thinks it is more useful to evaluate soil amendments at the ANL-W site.
Views expressed are those of the participants, not necessarily EPA.
2l
-------
• Costs. One attendee asked Lee how much farming equipment cost. Lee said that $10,000 of equipment
was purchased for the ANL-W site. Another attendee said that O&M procedures at ANL-W are
intensive. He asked whether site managers had considered designing a system that required less labor
and irrigation. Lee said that lower^maintenance approaches will be considered for ANL-W after cleanup
goals are met.
• Using soil moisture reflectometers to determine whether leaching is occurring. Glendon Gee noted that
soil moisture reflectometers measure water content, but not downward water flux. He said that
information on subsurface hydraulic properties must be combined with reflectometer data to make
meaningful conclusions about leaching potentials.
SESSION IVA: THE FATE OF CHLORINATED SOLVENTS THAT DISAPPEAR FROM
PLANTED SYSTEMS
Phytoremediation of Solvents
Milton Gordon, University of Washington
Milton Gordon described studies that the University of Washington has performed to elucidate the role that
poplars play in remediating chlorinated solvents. He said that poplars were chosen for study because the
University had already collected about 20 years of data on poplar breeding and metabolism. Gordon
launched his discussion by describing laboratory studies that have been performed to determine whether
poplars can remediate TCE. In one study, Gordon said, poplar tumor cells were shown to break down TCE.
Trichloroethanol, trichloroacetic acid, and dichloroacetic acid, all oxidative metabolites of TCE, were
detected, as well as carbon dioxide. Gordon said that mass balance studies have also been conducted on a
number of chlorinated solvents to determine how poplar cuttings impact these contaminants. Gordon
described the experimental apparatus that the University of Washington used to perform mass balance
studies, noting that efforts were made to replicate outdoor conditions. For example, a steady flow of air was
passed through the test chamber and intense light was shined on the plants. Even using the strongest light
available, however, researchers were only able to provide about one-fifth of the intensity of normal sunlight.
In the laboratory studies, Gordon said, a large proportion of the contaminants that were introduced simply
transpired through the plant. For example, he said, about 70% of the TCE was detected as a gas in the upper
portions of the study chamber. Gordon said that good mass recovery was observed when the polar cuttings
were grown hydroponically, but that rates were only about 60% to 70% when the plants were grown in soil.
Gordon also described experiments that the University of Washington has performed in the field. Working
with Occidental Chemical Corporation and the Department of Ecology, Gordon said, the University set up a
field study. In this study, a series of crypts were designed, each of which had poplar cuttings planted on top
of them. The crypts were irrigated with TCE-contaminated water and all liquid effluent leaving the system
was captured by an extraction well. (Safety mechanisms were in place to make sure that the site did not
become contaminated during the study.) Over a three-year period, Gordon said, hundreds of analyses were
performed to compare TCE concentrations in the irrigation stream and in the extraction well. Only about 1%
to 2% of the TCE that was added over this period was detected in the extraction well. Gordon said that
breakthrough only occurred at the end of growing seasons, coinciding with the time that leaves fell from
trees. As each year passed, and the trees matured, TCE concentrations in the extraction well decreased.
Gordon said that investigators wanted to know what was causing the TCE to disappear in the field. Wood
from the poplars was analyzed, but only small quantities of TCE were detected in the samples. To determine
whether TCE was being transpired, leaves were enclosed in Teflon bags; air was pulled through them at a
rate of 1 to 2 liters per minute; gases were collected, sent through a carbon filter, and heated; and TCE was
Views expressed are those of the participants, not necessarily EPA.
22
-------
measured. Gordon said that TCE was detected in the gas, but results indicated that only about 5% of the
TCE that was added over the three-year study was transpiring through the trees. (This quantity is less than
investigators expected based on the results of the mass balance laboratory study.) After additional
investigations, it became clear that many TCE degradation products were present in the rhizosphere. During
soil sampling, large quantities of chloride were detected near root systems. Investigators concluded that TCE
is degraded in poplar roots and excess chloride is excreted into surrounding soils. To confirm that rhizobial
microbes were not responsible for the degradation, at least at this particular field site, investigators irrigated
the site with more water than the tree roots could absorb. When they did this, they found that TCE levels in
the extraction well increased. Thus, it appears that TCE-contarninated water must enter poplar roots for
degradation to occur. '
Gordon said that a variety of other studies were performed at the field site as well. For example, a student at
Clemson University evaluated the toxicity of poplar detritus by feeding it to pill bugs. The insects did not
suffer adverse effects. In addition, Gordon said, the site was used to study the effect that poplars have on
carbon tetrachloride. The study showed that this chemical is also remediated by poplar trees, but that the
mechanisms involved differ somewhat from those observed for TCE. That is, none of the carbon
tetrachloride was released from the leaves as a gas. In general, Gordon said, these studies (as well as others
that have been performed on chlorinated solvents) suggest the following: poplars are able to uptake
contaminants, metabolize chlorinated solvents in plant tissues, and break down the solvents into chloride and
carbon dioxide. ;
Before closing, Gordon brought up a controversial subject: using transgenic plants to remediate
contaminants. In mammalian systems, he said, the cytochrome P450 enzyme plays a key role in TCE
degradation. In one study, he said, researchers incorporated a mammalian P450 enzyme into the genome of a
tobacco plant. They found that the oxidation rate of TCE was increased about 600 times when this
modification was made. Gordon said that this raises the question of whether it would be wise to tailor-make
plants that have an increased activity against chlorinated solvents. He said that phytoremediation might
travel down this path in the future, but acknowledged that using such an approach could raise some
controversial issues.
The Case for Phytovolatilization
William Doucette, Utah State University
William Doucette noted that a variety of mechanisms (e.g., enhanced rhizosphere microbial degradation,
sorption, plant uptake, plant metabolism, and phytovolatilization) are proposed to play a role in plant-
mediated remediation. He said that he would discuss the role that uptake and phytovolatilization (also called
transpiration) play in remediating TCE. Doucette said that there are three major pathways for organics to
enter plants: (1) root uptake, (2) atmospheric deposition, and (3) diffusive transport through air spaces into
roots and shoots. He said that his presentation would address the first route, in which TCE is dissolved in
subsurface waters or vapors and then passively diffuses through root membranes. He said that he would also
examine how readily TCE translocates from roots to shoots, and whether, once TCE reaches the
aboveground portions of a plant, it is released to the atmosphere through transpiration. Before proceeding,
he defined a term: transpiration stream concentration factor (TSCF), which defines how readily compounds
move from roots to shoots. In 1982, he said, some researchers suggested that a contaminant's log Kow
determined its TSCF value, and that contaminants with a log Kow of 2.0 had the highest TSCF values. After
collecting more data, however, scientists learned that compounds with the same log Kow do not always have
the same TSCF values. Thus, the relationship is more complicated than originally proposed.
Views expressed are those of the participants, not necessarily EPA.
23
-------
Doucette provided a brief literature review, summarizing the major conclusions that have been made over
the last six years about plant uptake and phytovolatilization of TCE. Regarding uptake, he said, Schnoor et
al. reported little root uptake and Schnabel et al. detected low (1% to 2%) uptake in shoots. As for
phytovolatilization, Doucette continued, several researchers have detected TCE in plant transpirate, but
reported values differ across the studies. For example, Newman et al reported transpiration as "measurable"
in one study and about 5% in another one, but a 1998 study by Burken and Schnoor cited values of 21%.
Doucette said that many researchers have measured TSCF values for TCE as well, and that reported values
range between 0.02 to 0.75. Doucette acknowledged that the data presented in the literature are quite
scattered. He described some of the reasons why this might be. First, he said, some of the data were
generated in the laboratory and some were generated in the field. Also, he said, experimenters did not all use
the same TCE concentrations, exposure durations, plant species, or growing conditions. For example, some
plants were grown in soil and some were grown hydroponically. Also, the age of the plants differed between
experiments. In addition, Doucette said, a variety of different experimental approaches have been used, and
this might contribute to the variation in results that is observed across studies. While some laboratory
experiments were performed in static chambers, he said, others were conducted in flow-through chambers.
Also, some researchers established a barrier between plant root zones and foliar portions, but others did not.
Doucette said that there are a variety of factors that must be considered when performing laboratory
experiments, including (1) humidity and temperature build-up, (2) unnatural plant conditions, (3) foliar
deposition versus root uptake, (4) leaks and low recovery, and (5) trapping mechanisms. The extent to which
researchers address these, he said, affects results. In the field, Doucette continued, researchers have used a
wide variety of approaches to measure phytovolatilization, including using (1) open bags, (2) sealed bags,
(3) flow-through bags or chambers, and (4) open path Fourier Transform Infrared (FTIR).
Doucette described laboratory studies that Utah State University (USU) has performed to quantify
phytovolatilization. He described the experimental design used in these studies, noting that a dual-chamber
flow-through system was used for all of the investigations and that researchers made great efforts to mimic
outdoor environments. Mass recovery in three of the studies was greater than 92%, he said, and the studies
revealed that: (1) TCE and the resulting oxidative metabolite (e.g., trichloroacetic acid and dichloroacetic
acid) concentrations were highest in roots, lower in leaves, and lowest in stems, (2) little to no
phytovolatilization occurred, (3) TSCF values ranged between 0.02 and 0.21, and (4) TSCF values were not
influenced by exposure concentrations (a range of 1 ppm to 70 ppm was tested) or exposure duration (12 to
26 days). Doucette described another study, which used a slightly modified experimental apparatus and was
performed over a 43-day period; in that study, mass recovery was about 97%, and results indicated that
about 70% to 75% of the TCE was present in plant roots, 10% to 13% was stored in leaves, 5% to 6% was
stored in stems, and about 4% to 7% had been volatilized out of the plant. He noted a point of interest: in
short-term studies, TCE concentrations in the stems were greater than those in the leaves, but the opposite
was true in longer-term studies.
Doucette said that USU has also evaluated the fate of chlorinated solvents at two different field sites. At
both sites, he said, mature trees had established themselves naturally in contaminated areas. Doucette
described the sites and the evaluations performed:
• Cape Canaveral Air Station (CCAS), Site 1381. Doucette said that plants have been growing for 10 to 20
years at this site, and that TCE (1 to 10 milligrams per liter, or mg/L) has been detected in an aquifer that
is located 2 to 4 feet below ground surface (bgs). Doucette said that USU took measurements from three
different tree species, as well as from the media that surrounded them. Results indicated that (1) TCE
was present in groundwater and soil samples but its metabolites were not typically found; (2) no TCE
was transpired through the plant; (2) TCE and its metabolites were present in plant roots, shoots, and
Views expressed are those of the participants, not necessarily EPA.
24
-------
leaves; (3) TCE and dichloroethene (DCE) were present in soil gas, but there was no measurable surface
emission flux; and (4) 75% of the roots were located in the top 2 feet of soil.
• HillAFB. Doucette said that this site had many things in common with the CCAS site. For example,
TCE concentrations were similar, and plants had established roots in contaminated areas many years
before USU initiated field studies. Doucette said that plant tissues were analyzed and that PCE, TCE,
and resulting metabolites were detected in stems and leaves. The ratio of PCE to TCE in the plants, he
said, was similar to that found in the site's groundwater. He also said that the TCE concentrations
detected in plant tissues were several orders of magnitude greater than those detected in the CCAS
plants. In addition, he said, contaminants were detected in plant transpirate at fairly high concentrations;
these reached 0.04 mg/L for PCE and 0.8 mg/L for TCE. Doucette did wonder, however, whether the
transpiration rates were a bit skewed, noting that'investigators had problems controlling humidity in the
measurement chambers that were installed around leaves.
Doucette noted that the results obtained at CCAS and Hill AFB differed quite a bit. More TCE accumulated
in plant tissues and transpired at the latter. He said that differences in precipitation rates may be responsible
for the different results. (Average annual rainfalls are 50 inches and 15 inches at CCAS and Hill AFB,
respectively.) Plants in areas with a lot of precipitation may not grow as deep, or tap into groundwater
sources as efficiently, as plants that are growing in areas where water is scarce.
Phytotransformation Pathways and Mass Balance for Chlorinated Alkanes and Alkenes
Valentine Nzengung, University of Georgia
Valentine Nzengung described studies that have been performed to determine how willows and cottonwoods
impact chlorinated alkenes and alkanes. The following table summarizes the studies and the results:
Views expressed are those of the participants, not necessarily EPA.
: 25
-------
Laboratory and Greenhouse Studies
Willows were dosed with 60 ppm of TCE. Over 50 days, TCE concentrations in solution decreased to less than 10 ppm.
Willows were dosed with PCE. At the end of the study, significant amounts of PCE were still detected in the rhizosphere. PCE
was also detected in aboveground portions of the plant, but concentrations decreased with plant height. The following
oxidative metabolites were detected during the study: trichloroethanol, trichloroacetic acid, and dichloroacetic acid.
• Cottonwoods were dosed with PCE. TCE concentrations increased as PCE concentrations decreased. Nzengung said that this
result was observed after researchers made some adjustments to experimental reactors. He said that he would like to know if
other researchers could duplicate these results. He acknowledged that the results could surprise some people, because they
suggest that reductive dechlorination processes are involved in plant degradation.
• Cottonwoods and willows were exposed to PCE. The following were detected in the rhizosphere head space: PCE, TCE, trans
1,2-DCE, ethene, ethane, and methane.
Mass balance studies were performed. Total mass recovery was poor (about 45% for TCE and 74% for PCE), but investigators
hope to improve these rates by modifying test reactors. The following was learned: (1) a significant portion of contaminant
was volatilized, (2) carbon dioxide was recovered in traps when plants were exposed to TCE, and (3) carbon dioxide was
mostly in the solution phase when plants were exposed to PCE.
• Aquatic plants were dosed with radiolabeled carbon tetrachloride. The results showed that: (1) 10% to 12% of the contaminant
was transformed to chloroform, (2) 80% to 85% was bound residue, and (3) 45% of the bound fraction was assimilated.
• Aquatic plants were dosed with HCA. A portion of the sequestered HCA became irreversibly bound residue. Several
contaminant breakdown products were detected, including PCE, TCE, pentachloroethane, trichloroacetic acid, and
dichloroacetic acid.
Studies Conducted on the Leaves and Roots of Cottonwoods That Are Growing at the Carswell AFB Site
Nzengung said that Cottonwoods were planted at the Carswell site in April 1996 in an effort to clean up groundwater that is
contaminated with PCE. Some of the trees were planted as whips and others as "5-gallon bucket" size cuttings. Researchers
collected leaf and root samples from the Cottonwoods in-1998 (after the second growing season) and in 1999 (after the third
growing season. Results (expressed in mg/kg) indicated the following:
September 1998 Results January 1999 Results
Trichloroethanol Trichloroacetic Acid TCE Trichloroethanol Trichloroacetic Acid TCE
Whip, leaf
Whip, root
5-gallon, leaf
5-gallon, root
0.161 ±0.016
0.640 ± 0.057
1.059 ±0.127
0.211 ±0.024
2.54 ± 0.003
ND
1.30 ±0.066
ND
ND
Trace
ND
Trace
1.566 ±0.056
0.623 ± 0.025
0.257 ±0.125
0.252 ± 0.074
ND
28.44
>30
31.34 ±0.72
ND
ND
ND
ND
Nzengung noted that values differed between the whips and the cuttings. He said that this probably reflects the fact that one
reached groundwater before the other. After evaluating the results, researchers returned to Carswell AFB to collect leaves from
mature Cottonwoods that had been established at the site before the phytoremediation demonstration project was started. The
results indicated that no detectable TCE or trichloroethanol was measured in the leaf extract, but that trichloroacetic acid and
dichloroacetic acid concentrations were very high. Nzengung hypothesized that the age of a plant influences phytotransformation
pathways.
Basing his findings on these and other studies, Nzengung proposed the following as a possible conclusion:
there are three phyto-processes that play a primary role in remediating volatile chlorinated aliphatics. These
are phytovolatilization, rhizodegradation, and phytodegradation. (Nzengung defined the latter as including
only those processes that occur in the aboveground portions of plants.) He also concluded that: (1)
sequestration and assimilation are important processes for some plants, especially aquatic plants; (2)
phytotransformation of volatile chlorinated aliphatics involves multiple pathways; (3) reductive
dechlorination and mineralization of TCE and PCE occurs mainly in the rhizosphere of woody plants; and
(4) oxidation and reductive transformation occurs in the tissues of aquatic and terrestrial plants. Nzengung
also said that data collected suggest that woody plants volatilize a significant portion of TCE and PCE under
greenhouse conditions.
Views expressed are those of the participants, not necessarily EPA.
26
-------
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
• Will phytovolatilization cause significant air contamination*? Rock asked panelists whether trees should
be considered significant sources of air contamination. Gordon said that phytovolatilization was minimal
in the studies he performed, but did say that rates might vary across sites. Thus, he said, tests should be
performed to measure phytovolatilization at every site that is proposed for phytoremediation. Doucette
agreed that phytovolatilization rates are site-specific, noting that this was confirmed in the studies
performed at CCAS and Hill AFB. Also, he said, it may be premature to make definitive conclusions
about the significance of phytovolatilization, since few efforts have been made to determine how much
contaminant will be transpired over an entire canopy. Nzengung said that he still thinks that
volatilization rates could prove to be significant in some cases, especially if a plant system has a
significant amount of reductive dechlorination (where parent compounds are converted to highly volatile
products, such as DCE) occurring in the rhizosphere. One participant said that he did not think more
than 10 to 15 pounds of contaminant would be released per acre over the course of a growing season, but
said that he could not predict whether this type of release would cause concern among regulators. One
attendee said that this amount is negligible compared to the total amount of organic compounds that trees
release naturally each year. Building on this point, Doucette said that many naturally occurring
hydrocarbons are released from trees, and that emissions could be as much as 2,000 times greater than
the rates of TCE release measured in phytovolatilization studies. Before closing on this topic, Harry
Compton advised field investigators to monitor air at all phytoremediation sites. Regardless of what is
found in the laboratory, he said, regulators will be satisfied that no dangerous air emissions pathways
have been created if field data back up this conclusion.
• Tips on how to perform laboratory experiments. Gordon said that investigators could be led to believe
that reductive dechlorination reactions are taking place in the rhizosphere if anaerobic conditions are
inadvertently established in the laboratory. In early studies performed by the University of Washington,
Gordon said, dissolved TCE and ethanol were pumped in to study chambers, driving out oxygen. As a
result, cis 1,2-DCE, a product of reductive dehalogenation, formed. When investigators started injecting
TCE and water rather than TCE and ethanol, he said, dissolved oxygen rates increased and cis 1,2-DCE
concentrations decreased. Gordon also advised investigators to check the purity of TCE that is provided
by chemical supply houses before using it in laboratory studies. He also said that air should be filtered
before it flows through a laboratory chamber. Doucette agreed, noting that he has found TCE in "clean"
bottled air.
SESSION IVB: INNOVATIVE SOLUTIONS FOR METALS REMOVAL
Phytoextraction of Metals from Contaminated and Mineralized Soils Using Hyperaccumulator Plants
Rufus Chaney, U.S. Department of Agriculture
Rufus Chaney said that phytostabilization and phytoextraction can be used to address soils that are
contaminated with metals. The former converts metals into less toxic or less bioavailable forms in situ. The
latter, Chaney continued, extracts metals from the root zone and accumulates the contaminants within plant
shoots. In many cases, Chaney said, metal-contaminated sites have very low pH and are deficient in
nutrients. As a result, these sites are often barren and unable to support plant growth. To achieve
phytostabilization of such toxic soils, Chaney said, researchers must act as agronomists, improving soil
Views expressed are those of the participants, not necessarily EPA.
27
-------
conditions and converting metals to non-bioavailable forms. Chaney said that much exciting research has
been conducted to determine ways to improve soil conditions so that plants can grow in highly toxic soil. In
recent years, investigators have starting mixing fertilizers with byproducts, such as biosolids, paper wastes,
and composts, and applying these to poor soils. This approach has achieved the desired phytostabilization
results in many cases. Thus, a beneficial use is being identified for byproducts that are not always
considered desirable. Chaney provided some specific examples of metal-contaminated sites that were barren
before having their soils amended, including the Bunker Hill site; Aberdeen Proving Ground; a zinc-
smelting site in Palmerton, Pennsylvania; a nickel-smelting site in Canada; and a smelter slag waste site in
Poland.
Chaney described the concepts that underlie phytoextraction, a technology that removes metals that are
present in the rhizosphere at low costs. Most plants, he said, cannot withstand high metal concentrations.
Most die when exposed to zinc concentrations of 500 ppm, cadmium concentrations of 10 to 20 ppm, copper
concentrations greater than 25 ppm, or nickel concentration greater than 100 ppm. There are some plants,
however, that exhibit great tolerance for metals, and actually accumulate them in very large quantities within
their tissues. Chaney said that the ability to accumulate metals may have originally evolved as a defense
against plant-eating insects. Today, these plants, called hyperaccumulators, are recognized to have great
remedial potential. Species that naturally hyperaccumulate selenium, zinc, copper, cadmium, and nickel
have been identified, but none have yet been found that will hyperaccumulate lead or chromium.
Chaney said that the U.S. Department of Agriculture (USDA) has patented a phytoextraction procedure. The
patent involves planting hyperaccumulators on contaminated sites, harvesting the plants, incinerating them,
recycling extracted metals from ash, and then selling the recycled metals for profit. He described two plants
that have been identified as hyperaccumulators. Alysswn murale can be used to hyperaccumulate nickel and
cobalt, which both have fairly high market value in the United States. Thlaspi caerulescens, Chaney said,
can be used to remove cadmium and zinc. He said that neither of these metals are worth much in the United
States; thus, site managers who have cadmium and zinc contamination may not jump at the chance to use
phytoextraction unless they are mandated to perform a cleanup. T. caerulescens might prove to be an
extraordinarily important remedial tool, however, in developing countries, where mines and smelters have
contaminated rice paddies and consumption of the rice has caused cadmium poisoning in subsistence
farming families. Peasant farmers in these countries could probably make a profit on recycled metals,
Chaney said, while simultaneously reducing the potential for human cadmium toxicity.
Chaney described the process that USDA uses to promote the development of hyperaccumulators. First, he
said, researchers identify species that are worth investing in. Once this is accomplished, germplasm is
collected and genetic diversity is evaluated. The genotypes are tested under uniform conditions using
contaminated soil, and efforts are made to identify the agronomic needs of the plant. Also, researchers
evaluate the fertility needs of hyperaccumulators and learn how to breed the plants. In addition, researchers
evaluate the economic benefits that the plants offer, address concerns about ecological risks, and test the
plants in the field.
In closing, Chaney said that hyperaccumulators have great potential. If developed and applied properly, he
said, they can provide economic benefits and address difficult environmental problems simultaneously.
Views expressed are those of the participants, not necessarily EPA.
28
-------
Phytoextraction: Commercial Considerations
Michael Blaylock, Edenspace Systems Corporation
Michael Blaylock said that Edenspace Systems Corporation (Edenspace) promotes the use of
phytoextraction, a technology that can be used to remediate metal-contaminated sites. He did note, however,
that there are some obstacles that must be hurdled before phytoextraction will be used on a wide-scale
commercial basis. The first of these is inertia, Blaylock said: many site managers have no motivation to
remediate sites until a government agency forces them to do so. Another issue that must be addressed, he
said, is determining a way to measure the success of a technology. Also, he said, efforts must be made to
answer the following questions:
• Can phytorextraction serve any benefit if contaminants are located deeper than 2 or 3 feet? Blaylock
said that efforts are underway to determine whether phytoextraction and electrokinetic transport can be
used in combination. Also, he said, at some sites, deep soils are being excavated, spread out in ex situ
treatment cells, and planted with species that extract metals.
• Will phytoextraction cause ecological toxicity? Blaylock said that some studies are underway to evaluate
this concern. For example, at the Aberdeen Proving Grounds, researchers are evaluating whether plants
that have been used to remediate uranium could pose a risk to the surrounding ecosystem.
• Will phytoextraction increase the solubility of some contaminants and cause leaching? Blaylock said
that lysimeters can be used to address these concerns. Also, if the issue is of great concern at a particular
site, phytoextraction could be used ex situ rather than in situ.
• Will phytoextraction serve any benefit at sites that have insoluble contaminants? Blaylock said that
researchers are experimenting with soil amendments to determine whether metals can be made more
soluble and absorbed through plant roots. \
Blaylock said that it is often best to integrate conventional remediation techniques with innovative ones.
When combinations are used, the strengths of multiple technologies can be used to overcome site-specific
challenges. Blaylock said that soil washing, particle size separation, excavation, electrokinetics, and
stabilization can all be used with phytoextraction. In fact, phytoextraction and phytostabilization approaches
are both being used at the Simsbury, Connecticut, site, a site that Edenspace is currently remediating. Before
initiating remedial activities at this site, Edenspace measured total lead concentrations in surficial soils at
concentrations as high as 6,000 ppm. Average SPLP leachable lead values were 0.85 mg/L. Blaylock said
that phytoextraction—using Brassica juncea—will be used for two years in an effort to remediate total lead.
After that time, the site will be reassessed, and phytostabilization will be used to address any remaining
problems. Blaylock said that the B. juncea has already been harvested once, and that average concentrations
from all the crops exceeded 1,000 ppm. Also, he said,: soils have been sampled to track progress to date.
Already, total lead concentrations have decreased dramatically and SPLP leachable lead values have
decreased to 0.08 mg/L.
Blaylock said that Edenspace has completed phytoextraction projects at several sites. For example, he said,
the technology was applied at a DaimlerChrysler site that had lead contaminants below root zones and
required cleanup activities to be completed over a one-year period. Edenspace identified a creative remedial
strategy to use at this site. First, subsurface soils were excavated; spread out over a lined, unused, paved
area; and mapped. Then, contaminant hot spots were identified and excavated for disposal. The remaining
soil, which was spread over about 2 acres, was planted with B. juncea in April and replanted with
Views expressed are those of the participants, not necessarily EPA.
29
-------
sunflowers in June. (Blaylock said that the sunflowers do not extract lead as readily as B. juncea, but they do
produce more biomass and extend roots deeper.) After harvesting and removing all the plants, Blaylock said,
Edenspace sampled soils to identify whether lead was still present above regulatory standards. Only about 2
cubic yards of soil still had too much lead; this soil was properly disposed of off site. Blaylock said that the
site owner was very happy with the results, and estimated that $ 1 million was saved by using
phytoextraction instead of a conventional approach. Blaylock said that Edenspace received an award for this
project.
Blaylock said that there are several exciting research efforts ongoing in the area of phytoextraction. For
example, the University of Florida is working with a fern that might be capable of hyperaccumulating
arsenic. Also, he said, Edenspace has started evaluating turf grasses, a low-maintenance crop, to determine
whether they can be induced to uptake lead. Initial laboratory results are promising, he said, noting that this
low-maintenance crop might have the potential to serve as a long-term sustainable approach for addressing
roadside lead contamination.
Zinc Hyperaccumulation in Plants: The Case of Zinc Hyperaccumulation in T. Caerulescens
Mitch Lasat, EPA
Mitch Lasat explained why there is great need to develop phytoextraction, a cost-effective technology that
removes metals from the soil. First, he said, there are a number of sites that would benefit from using it,
noting that elevated metal concentrations have been detected at many NPL sites and brownfields sites.
Second, phytoextraction is an environment-friendly approach that appeals to the public. Most importantly,
Lasat said, the technology is able to succeed where others have failed: it extracts metals from large surface
areas in a cost-effective fashion.
Lasat said that some metals are considered essential plant micronutrients. Plants have developed
mechanisms to extract these metals from the soil, absorb the metals into their roots, and translocate the
metals to aboveground plant tissues. While most plants accumulate only small quantities of metals,
hyperaccumulators have the ability to accumulate enormous quantities. For example, Lasat said, while a
regular plant accumulates about 50 ppm of nickel or zinc in its shoots, a hyperaccumulator can store more
than 20,000 ppm. Lasat said that much interest has been generated in using hyperaccumulators as remedial
agents, but that scientists have identified one problem: the hyperaccumulators are typically small, with a low
potential for biomass production. This problem could be ameliorated, Lasat said, if researchers identify,
clone, and transfer the genes that cause hyperaccumulation into plants that yield higher biomass.
Lasat said the remainder of his presentation would describe what scientists have learned about the
mechanisms that underlie zinc hyperaccumulation in T. caerulescens. He said that this plant was chosen for
study because it is an excellent hyperaccumulator. Some studies indicate that the plant suffers no toxic
effects even when more than 3% of its shoot weight consists of metals. To gain an understanding of what
causes hyperaccumulation, Lasat said, the physiological and molecular processes of T. caerulescens were
compared against those of T. arvense. The latter is closely related to T. caerulescens, but cannot survive in
heavily contaminated soils or hyperaccumulate metals. Lasat said that radiotracer flux studies were
performed to determine the physiology of zinc transport in the two species. The studies showed that T.
caerulescens uptakes zinc into its roots more readily than T. arvense. (In a three-hour study, the root
concentration of zinc was about two times higher in T. caerulescens.) Over longer periods, however, zinc
concentrations were higher in the T. arvense roots; compartmentalization studies indicated that zinc
concentrations in the root cell wall and cytoplasm were similar between the two species, but that vacuolar
zinc concentrations were significantly higher in T. arvense. In the shoots and leaves, zinc concentrations in
* Views expressed are those of the participants, not necessarily EPA.
30
-------
T. arvense were significantly lower than those measured in T. caerulescens. Based on this information, the
following hypothesis was generated: zinc is sequestered in T. arvense root vacuoles and made unavailable
for translocation to shoots, but this mechanism is disabled in T. caerulescens. Lasat said that information
about the mechanisms of zinc transport in the abovegrqund portions of T. caerulescens was also revealed
during the radiotracer studies. For example, investigators found that zinc tends to associate with vascular
tissues, but that it diffuses into intervenous spaces when concentrations are very high.
After obtaining information on the mechanisms that underlie zinc transport in T. caerulescens, Lasat said,
efforts were made to identify and clone the responsible genes. To do so, he said, the following steps were
taken: (1) a T. caerulescens cDNA library was constructed in a yeast expression vector pFL61; (2) zhy3, a
yeast mutant defective in zinc uptake, was complemented with T. caerulescens genes from the library; and
(3) the library was screened for clones that are capable of growing on restrictive zinc levels. Lasat said that
seven colonies were identified, five of which were discovered to represent the ZNT1 gene. He said that a
Northern analysis was performed to evaluate ZNT1 expression in T. caerulescens and T. arvense shoots and
roots. These analyses showed that T. arvense stimulates the expression ofZNTl under zinc-deficient
conditions. In T. caerulescens, however, the gene is expressed strongly even when there is plenty of zinc.
available in growth media. This suggests that the T. caerulescens does not downregulate zinc uptake once it
has accumulated enough zinc to meet its metabolic needs.
I
Lasat said that some investigators are confident that they have identified the gene that is responsible for zinc
hyperaccumulation in T. caerulescens. As a next step, he said, they will need to identify the gene that allows
this plant to tolerate highly toxic metal concentrations. Once this is determined, he said, researchers might
be able to bioengineer a hyperaccumulator that yields: high biomass, tolerates toxic metal concentrations, and
extracts them from the subsurface at high rates.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
• Harvesting plant roots. An attendee asked whether roots should be harvested along with aboveground
plant materials. Specifically, she asked whether investigators were concerned about animals that could
contact contaminated roots. Blaylock said that Edenspace has evaluated metal concentrations in the roots
and shoots of many plants. In general, he said, concentrations in shoots are about 5 to 10 times greater
than in roots. He said that an ecotoxicity study, performed at the Aberdeen Proving Ground, suggests
that earthworms and soil-dwelling insects do not experience adverse effects when exposed to uranium-
contaminated roots. Chaney said that some of the hyperaccumulators are perennials and that it would not
be economical to harvest their roots each season. . ,
• Costs. One participant asked for information on how much it costs to apply phytoextraction technologies.
Blaylock said that cost varies from site to site, but that $50,000 an acre is a rough average. At the
DaimlerChrysler site, he said, it cost between about $100,000 to $200,000 to clean up a 3-acre area.
Chaney reminded participants that the costs of application can be defrayed if site managers choose to
recycle the metals that they extract. It may even be possible to profit from the activity.
• The use of the term "hyperaccumulator." Chaney said that he defines hyperaccumulators as plants that
have evolved, through natural processes, to uptake large quantities of metals. Some plants can be
induced to act as hyperaccumulators if EDTA is applied, he said, but these are not "true"
hyperaccumulators. (Chaney said that B. juncea and sunflower are not hyperaccumulators.)
Views expressed are those of the participants, not necessarily EPA.
31
-------
" Incineration and methods used to extract metals. A participant asked whether metals could be released to
the air when plant materials are burned at incinerators. Chaney said that this was unlikely, as long as the
materials were sent to a permitted incinerator that was up to date on emission control technologies.
Blaylock said that several researchers are evaluating whether there are other ways to handle harvested
biomass. For example, a group in the Soviet Union has developed a technology that extracts metals from
biomass without requiring them to be incinerated first.
• Using local plants in phytostabilization projects. One participant asked whether efforts are being made
to use local plants in phytostabilization projects. Chaney said that this is the end goal, but other species
often have to be planted first to improve the quality of soils. At many sites, sterile plants are used with
mixtures of byproducts to hold the amendments on the soil and improve conditions before establishing
native crops that can proliferate.
• Using aquatic plants to extract metals. One participant asked whether plants can be used to extract
metals from contaminated water. Blaylock said that Rutgers University has experimented with
rhizofiltration, which involves exposing contaminated water to plant roots. Chaney said that work has
also been performed using constructed wetlands.
SESSION V: SIMULATIONS AND FORECASTS
Chasing Subsurface Contaminants
Joel G. Burken, University of Missouri
Joel Burken summarized mechanisms that are involved with the phytoremediation of organic compounds.
He said that subsurface contaminants can be either bound to soils, mineralized, or taken up by plants. Once
inside plants, contaminants can be metabolized, bound in plant materials, or released through
phytovolatilization. Burken said that these mechanisms have been documented, but the importance of each is
still being debated.
Burken described laboratory studies in which he has been involved, noting that the study results offer
interesting insights, but cannot necessarily be extrapolated directly to the field. In the early 1990s, he said,
researchers evaluated whether poplars can clean up atrazine. Several interesting findings were recorded.
First, investigators learned that poplars do not increase mineralization rates in the subsurface, but they do
reduce contaminant concentrations. Also, poplars were shown to take up the atrazine. Once the contaminant
was inside the plant, some of it degraded (about eight breakdown products were detected), some became
bound, and some transpired. In addition, researchers learned that subsurface soil conditions have a large
impact on transpiration rates. Uptake is not as vigorous when components of the subsurface sorb or compete
with contaminants. (When poplars were planted in silica sand, all of the atrazine was taken up by plants over
a 10-day period. Only 30% was taken up over an 80-day period when a silt-loam soil was used.)
In recent years, Burken said, the University of Missouri has performed several studies to evaluate how
poplar cuttings impact a variety of VOCs. In all of the studies, he said, a steady flow of air was injected into
a reactor, aerial and subsurface regions were compartmentalized, and significant efforts were made to mimic
outdoor conditions. Plants were grown hydroponically in some of the studies, but were rooted in soil in
others. When soils were used, investigators were able to create two subsurface zones: a saturated layer
(where contaminated water was injected) and a vadose zone. Burken described what was learned when
benzene was exposed to poplar cuttings:
Views expressed are those of the participants, not necessarily EPA.
32
-------
• Subsurface benzene concentrations were reduced. Benzene was injected into implanted and planted -
reactors, and investigators measured the effluent that came out of each. After six days passed, benzene
was no longer detected in the planted reactor effluent. It persisted in the unplanted reactor effluent,
however, for the entire course of the experiment.
• Subsurface mineralization processes were enhanced. Carbon dioxide was measured in the rhizosphere
headspace of unplanted and planted reactors. Very; little was detected in the former; significant quantities
were found in the latter.
• Very little benzene was stored in plants. Negligible benzene concentrations were detected in plant
tissues.
• Benzene was transpired, but the amounts recorded differed across experiments. Significant amounts of
benzene were transpired by hydroponically grown plants, but those that were grown in soil released
much less benzene. This proves that differences in experimental design have profound impacts on
results.
• Subsurface contaminant migration, water table height, and subsurface oxygen levels were impacted.
Very little benzene migrated to the upper soil layers in unplanted reactors because soil pore spaces wfere
mostly filled with water. In the planted reactors, contaminants moved more freely throughout the
subsurface. (Roots removed water, and this opened up pore spaces.) The studies also showed that poplars
have a profound effect on rhizosphere oxygen levels. The rhizosphere in planted reactors was aerobic,
but soils in the unplanted reactors proved to be oxygen deprived. Aerobic environments attract microbes
that help to degrade organics. Thus, it is clear that plants exert a combination of beneficial effects
concurrently. :
Burken identified areas that require additional research. Efforts should be made to understand plants better,
he said, noting that researchers should assess the phytoremedial properties of a wide variety of plants. In
addition, he said, researchers should ask: can plant systems be altered to improve their phytoremedial
properties? Burken said that some scientists propose using genetically enhanced plants or genetically
engineered microbes (GEM) in the rhizosphere. He described some work that has been performed on the
latter. Toluene o monooxygenase has been incorporated into three different rhizosphere microbes, Burken
said, and experiments have been performed to (1) evaluate how well GEM compete in wheat, bean, barley,
and poplar rhizospheres; (2) determine the degradation properties of GEM; and (3) determine whether GEM
increase TCE reduction rates when they are used in a poplar rhizosphere. The results of the latter, Burken
said, suggest that adding GEM to plant systems increase degradation rates dramatically.
Effect of Woody Plants on Groundwater Hydrology and Contaminant Fate
James Landmeyer, U.S. Geological Survey . .
James Landmeyer said that a study is being performed at a site in Charleston, South Carolina, to evaluate the
impact that poplars have on groundwater hydrology. The site is located in an urbanized setting, near a
residential area and an aquarium. Groundwater at the site, which is about 2 to 4 feet bgs, is polluted with
PAH and BTEX. It is also impacted by a coastal water body; thus portions of the site have saline
groundwater and are influenced by the tides. In December 1998, Landmeyer said, 2-foot-tallpoplar whips
were planted in an effort to address the site's groundwater contamination. Many of the trees prospered even
though the site experienced a harsh drought shortly after planting. By September 1999, several of the trees
had reached about 14 feet in height. Landmeyer said that excavations were performed to evaluate root
Views expressed are those of the participants, not necessarily EPA.
33
-------
profiles. From this activity, investigators learned that the poplar roots are growing vertically, into the
capillary fringe of the shallow water table. Landmeyer said that researchers were relieved to find this,
because tree roots typically grow horizontally, rather than vertically, in this part of the country.
Landmeyer said that the Penman equation was used to estimate how much water the poplars extract from the
site's groundwater and surrounding soil. A weather station was installed to collect several of the parameters
(e.g., wind speed and direction, air temperature, precipitation, solar radiation, relative humidity, barometric
pressure) that are required for the equation. In addition, Landmeyer said, information on leaf temperature
and wetness was also collected and entered into the equation. The equation's output indicated that the
potential evapotranspiration (PET) is about 0.5 inches per day. Extrapolating this over the entire plantation,
Landmeyer said, researchers concluded that 2,000 gallons of soil moisture and groundwater would be
removed from the site each day. He said that this value exceeds the amount of water (1,500 gallons) that is
thought to enter the study area each day. Thus, according to the modeled data, the poplar plantation should
cause a depression in the water table. Landmeyer said that extensive groundwater-level sampling would
have to be conducted to determine whether depressions are actually resulting.
Landmeyer said that some efforts have been made to measure groundwater fluctuation at the site. He
presented a series of graphs, which showed how groundwater levels changed during three different study
periods (August 6-11, 1999; August 24-September 2, 1999; and April 5-9, 2000) and at three different
locations across the site (in the center of the site, 1,000 feet from the tidal body, and 2,000 feet from the tidal
body). These data showed that: (1) rainfall causes the groundwater table to rise, (2) there is no evidence of
hydraulic capture yet, (3) fluctuation is more significant in locations that are closer to the tidal body, and (4)
groundwater levels dip slightly around noon—the time when evapotranspiration rates are expected to be
highest. Using the data obtained, Landmeyer said, investigators determined that each one-year-old poplar at
the site is taking up about 0.08 gallons of groundwater and soil moisture each day. This translates to an
uptake rate of about 0.0002 gallons per minute. After more data are collected, Landmeyer said, he will use a
modeling program to estimate how the trees will perform as they get older.
Landmeyer said that some efforts have been made to determine whether the poplars are impacting
subsurface geochemistry. For example, a dissolved oxygen probe has been installed in one well to determine
whether poplars have caused dissolved oxygen levels to increase in groundwater. To date, he said,
fluctuations in this parameter seem only to correlate with rainfall occurrences. He said that some preliminary
efforts have been made to determine the fate of the groundwater PAH and BTEX. These contaminants have
been detected in tree leaves, but it is unclear whether these chemicals originated in the groundwater or came
from some other source.
Before closing, Landmeyer presented information on another site, located in Beaufort, South Carolina,
which has a BTEX and MTBE plume emanating from an underground storage tank. He said that oak trees
established themselves at the site about 40 years ago. In 1997, researchers initiated evaluations to determine
whether these oaks were effecting groundwater contaminants. They concluded that hydraulic capture has not
been achieved because groundwater flow rates are too fast and portions of the plume are too deep. However,
they did find strong evidence that contaminants are being taken up into the oak trees. Landmeyer said that
the trees were cored and evaluated using gas chromatography/mass spectrometry, and that BTEX, MTBE
and trimethylbenzene isomers were detected in those trees that had extended roots into the groundwater
table. Landmeyer said that the oaks might be transpiring a significant amount of MTBE, which could cause
some people to express concern about cross-media contamination. He said that it may be more effective to
decrease MTBE concentrations by using plants that release MTBE to the air (where it has a half-life on the
Views expressed are those of the participants, not necessarily EPA.
34
-------
order of hours or days) than to allow MTBE to be transported in anaerobic aquifer systems (where its half-
life is on the order of years).
i
Modeling Plume Capture at Argonne National Laboratory-East (ANL-E)
John Quinn, Argonne National Laboratory
John Quinn described some of the modeling activities that were performed for the phytoremediation project
at ANL-E's 317/319 area. He noted that this site was described earlier in the conference, but offered the
following information as a brief summary: (1) the groiindwater is contaminated with VOCs and tritium, (2)
13 extraction wells were installed as an interim measure to control groundwater, (3) the contaminated
groundwater plume is about 25 to 30 feet deep, (4) the site's subsurface stratigraphy is extremely complex,
(5) the aquifer is fairly thin—about 5 feet thick—but varies widely in texture and permeability, (6)
groundwater flows towards the Waterfall Glen Forest Preserve and a recreational path, (7) contaminants
have appeared in downgradient seeps, (8) an engineered phytoremediation system has been installed to
control site contamination, and (9) fences have been installed to prevent deer from contacting the planted
system.
Quinn described the phytoremediation system that has been installed at the site. He said that a herbaceous
cover has been installed to minimize water infiltration and soil erosion. In addition, trees have been installed
near the site's french drain in an effort to enhance the remediation of subsurface VOCs. (Quinn said that the
downward migration of tree roots should not be hindered by subsurface heterogeneities, because the area
around the french drain was homogenized by deep soil mixing a couple of years earlier.) Also, Quinn said,
Tree Wells® have been installed in an effort to achieve hydraulic control. He said that these trees were
planted in 30-foot-deep, lined, large-diameter boreholes. Thus, roots are cut off from any shallow water
sources and forced to grow vertically in search of deeper groundwater sources.
Quinn described modeling efforts that have been conducted on the aquifer that lies 25 to 30 feet below the
site. He said that MODFLOW was used to model transient groundwater flow under pre- and post-
phytoremediation conditions. For the latter, he said, TreeWell® evapotranspiration estimates were provided
by Ray Hinchman and Edward Gatliff. Quinn showed computer simulations for the pre- and post-
conditions, and told attendees that these could be viewed more leisurely by visiting
http://web.ead.anl.gov/phyto. In both simulations, Quinn pointed out, water levels pulsate, with the highest
levels predicted in spring, during snow melt events, and the lowest levels predicted during late summer and
early fall. Quinn said that he also performed particle tracing studies to evaluate groundwater flow
throughout the plantation. Summarizing the results of ;his studies and simulations, Quinn made the following
conclusions about the ANL-E site: (1) mature plantations will achieve groundwater containment, even
during the dormant winter months; (2) the site's 13 extraction wells can be phased out in the future; (3)
groundwater that is pulled into microbially active rhizosphere areas will have a residence time of 5 to 17
months; and (4) de-watering of the perched aquifer could occur.
In closing, Quinn listed some of the activities that are ongoing or planned at the ANL-E site. These include
collecting groundwater samples, making water-level measurements, collecting weather data, sampling tree
tissues, evaluating transpirate, analyzing subsurface soils, and installing viewing tubes so that miniature
cameras can be used to view root development.
Views expressed are those of the participants, not necessarily EPA.
35
-------
Phytoremediation Potential of a Chlorinated Solvents Plume in Central Florida
Stacy Lewis Hutchinson, EPA, National Exposure Research Laboratory
Stacy Lewis Hutchinson said that models can be used to predict whether a plant-based system will be able to
control groundwater at a site. She said that water balance can be determined by evaluating groundwater flow
patterns and the amount of water that plants withdraw from aquifers. She presented the water balance
equation—a calculation that is incorporated into predictive models—and listed parameters included in it:
hydraulic conductivity, aquifer thickness, hydraulic head, sources and sinks of water, precipitation,
irrigation, surface runoff, and evapotranspiration. The latter two, she said, are often estimated, and'therefore
tend to have more error associated with them.
Hutchinson talked briefly about the evapotranspiration parameter, noting that plants that exhibit high values
may be able to reduce average net recharge to subsurface aquifers, prevent plume diving, and create an
upward flux of groundwater. A plantation's capability of achieving these results, she said, is influenced by
geography and climate. For example, she said, evidence suggests that plants are more likely to achieve
hydraulic control in the Southwest than in the Southeast. (Both areas are hot and exhibit fairly equivalent
evaporation energies, but precipitation rates differ between the two. In the wet Southeast, plant roots are
more likely to use precipitation as a water source. In the dry Southwest, roots are forced to search for deeper
groundwater aquifers.) Hutchinson described the methods that are used to determine evapotranspiration
values. She said that many researchers use the Penman-Montieth equation to generate estimates, but
cautioned that this equation may provide overinflated evapotranspiration values. Recently, Hutchinson said,
researchers have started using sap flow collars and probes to obtain a more direct measurement of
evapotranspiration.
Hutchinson described modeling efforts that were conducted at a site in Orlando, Florida. This site's
groundwater, which is contaminated with PCE and TCE, flows under a paved area, to a ditch, a grassy area,
and then a more densely vegetated area before discharging to a lake. Groundwater analyses, she said, reveal
that there is plume diving at the site. Thus, groundwater is too deep for existing trees to tap into. In an effort
to fix the plume diving problem, Hutchinson said, investigators used models to determine (1) why the plume
diving is occurring and (2) how the problem could be fixed. She said that a one-dimensional analytical
recharge model was used to evaluate the local patterns of aquifer recharge. Using this, researchers were able
to mimic the pattern of plume diving that was observed at the site. (The model was not perfect, she said,
noting that it did not precisely predict how deep the plume would dive.) Once researchers had a better
understanding of recharge, a numerical model was developed for the site, using the MODFLOW code and a
U.S. Geological Survey (USGS) regional calibration of the same model. Two layers of differing
conductivity were incorporated into the model, and fine scale layering was incorporated to define lake
bathymetry. Hutchinson said that addressing vertical discretization is important for evaluating three-
dimensional flow features. She said that the model was run using three different evapotranspiration
estimates: 50 cm/year (a value cited in a USGS report), 80 cm/year (a value cited in a report about the actual
site), and 130 cm/year (a value cited by the Florida Agricultural Extension Service). All of these values, she
noted, are less than 140 cm/year—the average annual precipitation estimate.
After the model was used to simulate existing conditions, Hutchinson said, it was used to predict the impact
that different engineering solutions might have. For example, she said, simulations were performed to
determine whether plume diving could be reversed by (1) diverting runoff from the ditch, and (2) planting
some trees in the paved area. She noted that the latter area has no recharge, because water cannot infiltrate
pavement. Thus, by dispersing some trees in this area, site managers could be assured of causing a net loss
of groundwater over this area. The modeling results indicated that combining the two approaches would
Views expressed are those of the participants, not necessarily EPA.
36
-------
reverse much of the diving that is currently observed at the site. Assuming that recharge in the paved area is
zero and evapotranspiration rates are 80 cm/year, she said, vegetation would be able to capture about 77% of
the plume. The percentage is even higher (88%), she said, when evapotranspiration rates are assumed to be
130 cm/year. In closing, Hutchinson stressed two important points: (1) vertical characterization is required
to delineate plumes, and (2) the localized recharge distribution should be carefully assessed.
Speaker Panel and Audience Discussion ;
Audience members asked questions or provided comments about the following topics:
• The role models play in obtaining regulatory approval. Hirsh said that regulators will want to review
modeling and forecasting data when trying to determine whether to approve a phytoremediation project.
Having these data available, he said, will help build convincing arguments in support of
phytoremediation.
• Using evergreens. A participant asked whether evapotranspiration rates at the ANL-E site could have
been enhanced by including evergreens in the phytoremediation design. Quinn said that this was a good
question, but that he did not have an answer. Quinn said that evergreens do transpire water for a longer
portion of the year, but pointed out that their evapotranspiration rates are lower than those exhibited by
poplars and willows. McMillan acknowledged that the rates are lower, but thought it might be beneficial
to include evergreens in a tree mix, since they actively transpire during early spring and late fall—two
critical recharge periods that occur when poplars and willows are not transpiring optimally.
• Public perception on genetic enhancements. One meeting attendee said that the public might express
concern about genetically enhanced plants and GEM. She asked Burken if he had any recommendations
on how to handle this. He suggested hiring someone with public relations experience to address the
public's concerns. He said that the public would need to be educated on these topics. For example, he
said, it is important for the public to understand that the GEM that are being created have suicide
implantations.
• GEM. Phil Sayre advised researchers who are creating GEM to be careful if they are working with the
Burkholderia cepacia complex, which has been reported to cause death in people with cystic fibrosis.
Sayre also advised investigators to evaluate GEM for resistance to antibiotics.
SESSION VI: PLUME CONTROL—ON-THE-GROUND EXPERIENCE
Phytoremediation at Aberdeen Proving Ground, Maryland: O&M, Monitoring, and Modeling
Steven Hirsh, EPA Region 3
Steven Hirsh described a phytoremediation project that has been initiated at the Aberdeen Proving Ground, a
site that has conducted chemical warfare research since 1917. Prior to the 1960s, Hirsh said, munitions,
laboratory wastes, and solvents were dumped at the site's burn pits and ignited. A contaminated groundwater
plume has formed below the pits, and is discharging to a fresh-water marsh and the Chesapeake Bay. Hirsh
said that contaminant concentrations are very high. For example, 1,1,2,2,-tetrachloroethane has been
detected in groundwater at 390,000 ppm and TCE is present at 93,000 ppm. Hirsh said that the surficial
aquifer is about 30 to 40 feet thick, has a thick confining unit, consists mostly of silty sands, and has a
conductivity of about 1 foot per day.
Views expressed are those of the participants, not necessarily EPA.
37
-------
Hirsh said that two objectives were identified for managing the plume—reducing VOC mass and achieving
hydraulic control—and that a variety of remedial technologies were evaluated as potential solutions,
including hydrogen release compounds (HRC™), in situ chemical reactions, monitored natural attenuation,
recirculating wells, and phytoremediation. Significant interest was generated in the latter; therefore, about
200 hybrid poplars have been planted in a horseshoe shape at the site. Before the trees were planted, Hirsh
said, an agronomic assessment was performed, a drainage system was installed to divert rainwater off of the
burn pits, and monitoring wells were installed. The trees were installed by Applied Natural Sciences, Inc., he
said, using techniques that promote vertical root growth. Boreholes were augered, plastic sleeves were
installed, two-year-old trees were planted, and boreholes were backfilled. Hirsh said that excavation
activities were performed last year to evaluate root growth. For the most part, very little horizontal growth
was observed, and the roots appeared to be bound within the borehole. Hirsh said that the trees are growing
well at the site, noting that they were about 30 feet tall in 1999. The site was hit by a hurricane in the fall of
1999 and this caused about 30% of the trees to lean over, Hirsh said, but site managers have straightened
them.
Hirsh said that groundwater samples are being collected, but that it is too early to determine whether the
trees are reducing contaminant concentrations at this site. He said that a number of other parameters are also
being monitored to determine how much groundwater the tree plantation is removing from the plume. For
example, data are being collected on leaf area, tree diameter, land area, sap flow rates, and a variety of
climatological parameters, such as temperature, relative humidity, solar radiation, and wind speed. These
data are then used, Hirsh said, to calculate PET and crop coefficients. Based on the calculated values,
researchers estimate that each tree is currently taking up about 8 gallons of water per day, and that this value
will grow to 12 gallons per day once the canopy closes. Extrapolating, Hirsh said, this suggests1 that the
mature plantation will pump about 2,000 gallons per day. He said that this is only about one-fifth of the
pumping rate that is needed to achieve hydraulic control at this particular site.
Hirsh said that transpiration gas is being measured at the site, and that 1,1,2,2-tetrachloroethane and TCE
have been detected as offgassing products in at least a couple of trees. (Gases were measured by placing
bags over leaves and sealing them. Cold traps were added to some of the bags, but investigators found that
these devices had minimal impact on gas measurements.) Hirsh said that flux chambers have also been
established to determine whether contaminants are offgassing from soils. Some TCE has been detected in
the chambers. Fourier transform infrared (FTIR), another air sampling technique, is also being used to
determine whether the plantation, as a whole, is emitting VOCs to the environment at dangerous levels.
Results collected to date indicate that this is not the case.
Hirsh said that site managers at the Aberdeen Proving Ground know that phytoremediation systems do raise
some ecological concerns. Therefore, soap has been hung on the poplars to deter herbivores from feeding on
them. In addition, efforts are being made to determine whether soil functioning is adversely impacted when
contaminated leaves degrade on the ground. For example, Hirsh said, nematode populations are being
evaluated. Honeybee hives are also being monitored to determine if volatiles are accumulating within them.
Hirsh closed his presentation by talking about modeling efforts that have been performed for the site. He
said that these were run under the assumption that phytoremediation, recirculating wells, and natural
attenuation would all be working in concert. According to MODFLOW, he said, phytoremediation should
have a significant effect on the top portions of the aquifer, but the impact will decrease in deeper layers. He
said that three-dimensional modeling was performed so that researchers could estimate how much
contaminant would be removed by each technology. Results indicated that 400 pounds would be degraded
Views expressed are those of the participants, not necessarily EPA.
38
-------
through phytoremediation, 1,500 pounds by microbial degradation, and 1,950 pounds removed by
recirculating wells.
Phytoremediation Systems Designed to Control Contaminant Migration
Art Ferro, Phytokinetics !
l
Ari Ferro said that researchers believe that tree plantations can be used to prevent subsurface contaminants
from migrating. In fact, he said, this theory is currently being tested at a site in Ogden, Utah. The site,
Chevron's Light Petroleum Products Terminal, has a dissolved-phase TPH plume located about 6 to 8 feet
bgs. (Concentrations have been detected between about 5,000 parts per billion [ppb] and 10,000 ppb.) In
April of 1996, Ferro said, a dense stand of hybrid poplar trees was planted perpendicular to groundwater
flow. (The design has three rows of poplars. There are about 6.5 feet between rows and about 7.5 feet
between trees.) Ferro described the planting method that was used at the site: 8-foot deep boreholes were
drilled, long poplar cuttings were placed into the holes, and the holes were backfilled with sand and peat.
Ferro said that the trees were planted directly into the saturated zone; thus, from the start, the trees used
subsurface groundwater as their water source rather than relying on surface irrigation. Growth has been
robust, he said, and the trees are approaching canopy closure. Ferro said that the root structure for one tree
was analyzed at the end of the third growing season. A dense proliferation of roots, many extending into the
saturated zone, was observed in and around the borehole. Interestingly, the deepest roots did not come from
the tree's original pole, which was dead; instead, the deepest roots originated from highly branched roots
that were located nearer to the surface.
Ferro said that researchers have generated estimates on the amount of groundwater that is being used by the
tree plantation. He said that each tree pumped about 0.25 gallons per day during the second growing season,
that this value rose to 11.9 by the fourth growing season, and that it is expected to increase to 17.7 by the
end of the fifth season. Ferro said that these values represent the volumetric water used by trees (Vt) minus
the amount of precipitation that falls during a growing season. Both parameters have been measured directly
at the site; precipitation is measured at an onsite weather station, and V, is quantified using thermal
dissipation probes that measure sap flow. Vt values can also be determined indirectly, Ferro said, using
calculations that have the following inputs: PET, crop coefficient, leaf area index, and tree stand area. Ferro
said that investigators did use calculations to estimate Vt, and that the results obtained were similar to those
identified through direct measurements.
Ferro said that five piezometers have been installed at the Ogden, Utah, site. Some are located upgradient of
the trees, some are in the middle of the plantation, and others are downgradient. These were sampled several
times between May 1998 and August 1999, and TPH, BTEX, and water levels were measured. The results
indicate that contaminant concentrations decrease as groundwater moves through the tree stand. As for
water-level measurements, Ferro said, groundwater was not shown to dip during the growing season and this
surprised site investigators. (Based on their calculations, researchers expected to find that the trees
transpired a volume of water equivalent to an 11-foot thickness of the saturated zone. Using estimated V,,
crop coefficient, and leaf area index values for the fourth growing season, researchers determined that the
tree stand would transpire about 480 gallons of groundwater per day. To determine whether transpirational
water use by the trees would be significant relative to the total flow of groundwater, researchers used
Darcy's Law to calculate the approximate volume of water that flows beneath a vertical cross-section of the
site. The rate of groundwater flux through a 1-foot thick vertical cross-section was calculated to be about 44
gallons per day.) Ferro said that he is not sure why water-level measurements failed to show a dip; efforts
are underway to identify plausible explanations.
Views expressed are those of the participants, not necessarily EPA.
39
-------
Before closing, Ferro described a phytoremediation project that has been proposed for the Bofors-Nobel
Superfund site in Muskegon, Michigan. This site, placed on the NPL in 1989, produced industrial chemicals
between the 1960s and 1980s and disposed of waste sludge into 10 lagoons. The lagoons have been drained,
Ferro said, but the sludge that remains is highly contaminated with 1,2,4-trichlorobenzene, 3,3,-
dichlorobenzidine, 2-chloroanaline, and a variety of other contaminants. Ferro said that site managers want
to plant deep-rooted trees in the lagoons in an effort to reduce leachate formation, stabilize contaminants,
and promote rhizosphere degradation. The proposal also recommends planting trees and native grasses in the
areas between the lagoons in an attempt to minimize rain-water infiltration. In addition, a poplar stand is
proposed in another portion of the site to intercept groundwater flow, and revegetation has been proposed
for a 9-acre segment of the site. Ferro said that predesign studies are being performed, noting that a small
field study was initiated in 1999 to evaluate which planting methods would be best to use and which plants
are most tolerant of the site's toxic sludges. He said that a greenhouse study will be initiated in the near
future.
Deep Planting
Edward Gatliff, Applied Natural Sciences, Inc.
At most sites, Edward Gatliff said, 80% to 90% of tree roots develop within the first 3 feet of soil. If enough
precipitation falls upon a site, he said, roots will not have an incentive to grow much deeper. Thus, Applied
Natural Sciences, Inc., has developed "deep planting" techniques to coax plant roots to grow deeper and to
tap into aquifers or horizon soils that lie deeper than 3 feet. By using this approach, Gatliff said, site
managers can be assured that roots will come into contact with targeted subsurface areas. Gatliff said that
Applied Natural Sciences, Inc., developed its first "deep planting" technique in 1990, while working on a
site in New Jersey, which had a contaminated aquifer that was located about 16 to 20 feet bgs. Investigators
considered using alfalfa to tackle the deep contamination at this site, but soon realized that the site's soil had
too much clay for alfalfa roots to penetrate. So, they started experimenting with trees. Since that time,
Gatliff said, Applied Natural Sciences, Inc., has identified several types of "deep planting" techniques that
can be used. Some of these have been patented.
Gatliff described the process involved with "deep planting." If the top 5 to 10 feet of subsurface are targeted
in a phytoremediation project, he said, trees can be installed using a trenching technique. For deeper depths,
however, boreholes are drilled, trees are placed inside, and the boreholes are backfilled. Gatliff said that the
borehole diameter and depth dictates the type of drill rig that must be used. For example, a three-point auger
or a tractor can be used to install boreholes to depths less than 5 feet, but a skid steer with an auger
extension is needed to install boreholes up to 10 feet deep. For holes that are drilled between 10 and 20 feet
deep, a medium-sized drill rig with an 8-foot stroke can be used, but anything deeper requires a caisson rig.
Gatliff said that casings are usually installed in the boreholes, noting that these help to limit preferential
shallow root development and surface-water short-circuiting. He said that Applied Natural Sciences, Inc.,
has experimented with three different kinds—ADS and metal casings, sonotubes, and plastic casings—and
found that each has pros and cons. In general, Gatliff prefers plastic casings, because these are cheap and
easy to use. He said that casings do not have to be installed in the boreholes, but said that roots have been
shown to develop in surficial areas when casings were not included. He said that this does not hold true for
all sites, noting that roots grow deep in uncased boreholes that are drilled in areas that have high clay
content. Gatliff said that Applied Natural Sciences, Inc., has not yet determined the optimal diameter for
boreholes, but believes that they should be greater than 16 inches. At many sites, good results have been
obtained with 18-inch boreholes. Gatliff said that trees planted in boreholes can be planted with the rootballs
either near the surface or deeper in the subsurface. If the latter approach is used, roots reach targeted depths
Views expressed are those of the participants, not necessarily EPA.
40
-------
more quickly and trees are less likely to require irrigation, because their roots tap into groundwater almost
immediately.
Gatliff said that trees that have been installed with "deep planting" techniques have been shown to remediate
contaminants. For example, at a site in Finley, Ohio, TCE concentrations decreased significantly after a
plantation was installed. Trees at this site are clearly tapping into groundwater sources, he said, noting that
the trees were unaffected by a serious drought that recently impacted the region.
Gatliff said that he does not know what the depth limitations are for "deep planting" techniques. At some
sites, trees have been installed in boreholes that reach 35 feet deep. At the ANL-E site's 317/319 area, he
said, Treewells® have been planted in 30-foot-deep boreholes. The Treewell® design, Gatliff said, uses
elongated roots, which are pregrown and attached to rootballs. At ANL-E, rootballs were placed about 5 to
10 feet bgs and elongated roots were suspended another 2 to 8 feet. Gatliff said that it may be possible to
design systems that penetrate even deeper. For example, he said, acacia trees have been known to extend
roots as deep as 100 feet bgs. Roots that grow in sewers have been reported to reach lengths greater than 200
feet.
Gatliff provided rough estimates of the costs that are associated with "deep planting" techniques:
System Description
Target depth of ^ 10 feet
Target depth of 10 to 20 feet
Target depth of > 20 feet
Costs Per Tree ;
$100 to $300 '.. - : •
$250 to $500 :
$500 to $1,500
Costs Per Acre
$20,000 to $60,000
$50,000 to $100,000
$100,000 to $300,000
He said that costs can be controlled by using the right methodologies and equipment for a particular site.
Transpiration: Measurements and Forecasts
James Vose, U.S. Forest Service ;
James Vose opened his presentation by clarifying the definitions of "transpiration" and
"evapotranspiration," two terms, he said, that are often used interchangeably. Vose said that transpiration
represents the amount of water that trees take up and release to the atmosphere. Evapotranspiration
represents transpiration plus evaporation (i.e., the free water in soils and on the surface that undergoes a
phase transformation to water vapor). Vose said that the evaporation component is small compared to the
transpiration component. Therefore, he focused on transpiration. He described the controlling factors.
Climate plays a large role: temperature, solar radiation,; vapor pressure, wind speed, and precipitation affect
how much water a plant transpires. Also, he said, plant physiology, including stomatal conductance, xylem
anatomy, rooting depth, sap wood quantities, and sap wood flow rates, dictate how much water is used. To
demonstrate his point, Vose presented a graph that showed that transpiration rates generally increase with
increasing sap wood quantities. He asked attendees to note, however, that plant species that have the same
quantities of sap wood do not all have the same transpiration rates. This is due, in part, to the fact that plants
have different anatomies, which affects how water flows through their sapwood. Vose said that leaf area
index and leaf distribution also exert great influence on transpiration rates. In addition, site characteristics
play an important role, Vose said, noting that a plant's capability to draw water into its roots is influenced
by the accessibility of groundwater sources, the site rooting volume, and the soil water holding capacity.
Views expressed are those of the participants, not necessarily EPA.
41
-------
Vose described the methods that can be used to estimate transpiration rates. At the tree level, he said,
transpiration can be measured using weighing lysimeters, leaf-level measurements, and sap flow probes or
collars. Once transpiration rates are established for an individual plant, Vose said, these rates are used to
extrapolate rates for entire tree stands. He warned, however, that several assumptions must be made to
perform the extrapolations, and said that this could compromise the accuracy of predictions. Vose said that
there are ways to measure tree stand transpiration rates directly, such as the eddy flux or gauged water shed
methods, but said that these techniques do have limitations. For example, both techniques measure whole-
system response; thus, it is difficult to determine how much transpiration is coming from trees versus
herbaceous cover. Vose said that a variety of models, ranging from simple empirical models to complex
mechanistic ones, are also available to help investigators predict transpiration rates.
Vose described one site, located in Fort Worth, Texas, that initiated a phytoremediation project in 1996.
Whips and one-year-old eastern cottonwoods were planted at this site in an effort to intercept a TCE plume.
Starting in 1997, Vose said, data have been collected for a variety of parameters, such as leaf water
potential, stomatal conductance, soil moisture release, and climatic conditions. In addition, collars and
probes were installed to collect monthly sap flow measurements from about 16 cottonwoods. (More than one
probe was used on each tree.) Vose said that the data collected were used to calculate transpiration, and he
presented what has been learned over the last few years. He said that transpiration rates in the one-year-old
trees were higher than those measured in the whips during the first year of growth, because the former were
larger. However, he said, when investigators standardized for tree size, they found that the whips actually
had higher sap flow rates. By the end of the second year, Vose said, transpiration rates in the one-year-old
trees and the whips were about 6 gallons/day/tree and 1.6 gallons/day/tree, respectively. Vose said that
investigators also used PROSPER, a model that has been in existence for about 25 years, to estimate
transpiration rates at the site. Initially, he said, the model appeared to do a good job of simulating
transpiration; modeled data were in good agreement with measured data. By the second year, however,
correlations were not as strong, and modeled estimates diverged significantly from measured values during
the driest months of the year. Over the next year, Vose said, he hopes to improve the PROSPER model.
Vose said that extrapolations were made on the measured and modeled data to predict how much water
might be transpired across the Fort Worth, Texas, site in future years. Predictions ranged from 500,000 to
2,000,000 gallons/hectare/year. The lower bound, he said, represents what PROSPER estimated as the
transpiration rate at a 15-year-old plantation. The upper bound, he said, represents projected transpiration
rates in a mature open forest. The rates would be closer to about 1,000,000 gallons/hectare/year, Vose said,
if investigators accounted for the shading that will result in a closed canopy situation. Vose said that it is
useful to make predictions using a wide variety of methods, because it helps investigators determine the
outer boundaries for possible transpiration rates. He is hopeful that models will be improved significantly in
the future so that more accurate predictions can be made.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
• Red cedar juniper trees. Fletcher noted that red cedar junipers, which are evergreens, are growing at the
Fort Worth, Texas, site. He said that he thinks these trees might be desirable to use in phytoremediation
designs, and asked Vose whether he plans to model performance of these trees. Vose said that he had not
planned to do that for this site, but did agree that sites in the Southwest might benefit from having a
mixture of conifers and hardwood in their remedial designs.
Views expressed are those of the participants, not necessarily EPA.
42
-------
The suitability of using poplars and-willows over -wide geographical ranges. J.G. Isebrand, who serves
on the International Willow and Poplar Commission, said that it is his job to ensure that willow and
polar cultures are exchanged responsibly. He said that he was delighted to learn about the different ways
in which these trees are being used in phytoremediation projects. He did caution, however, that trees can
fail if they are planted in inhospitable soils or in regions where the trees are prone to disease. He asked
meeting attendees to keep this in mind, noting that all efforts should be made to avoid failures—even
just a little negative publicity could set back the acceptance of phytoremediation.
Biochemical parameters that impact transpiration rates. McMillan noted that Vose's presentation
focused on physical parameters that control transpiration rates. He asked whether biochemical
parameters, which have been discussed in the European literature, might also have an influence. Vose
said that this could be, but that he was not aware of any that played a significant role. In general, he said,
water uptake processes are really dominated by physical processes {e.g., resistances and differences in
potentials), as well as anatomical features (e.g., number of stomata, distribution of stomata, amount of
sap wood). He said that biophysical processes do cdntrol how water is pumped at the cellular level, but
speculated that the influence on overall water usage; is only minimal.
Poplar tree costs. One meeting attendee asked howniuch poplar trees cost. Gatliff said that trees cost
about $10 to $20, but that whips can be obtained for about $1 to $3.
SESSION VII: VEGETATIVE COVERS
Monitoring Alternative Covers \
Craig Benson, University of Wisconsin
Craig Benson said that many people have expressed interest in using vegetated covers as an alternative to
prescriptive RCRA-style landfill covers. He said that this is because the latter, which do not always perform
well, typically cost a great deal of money. For example, compacted clay caps, which are listed under RCRA
Subtitle D, cost about $125,000 per acre and are prone to desiccation cracking, frost damage, settlement, and
root intrusion. Other prescribed covers, such as composite-type caps, do perform well, but these cost
between $ 175,000 to $200,000 per acre. Benson said that investigators are confident that effective
vegetative covers can be designed for far less money. He said that vegetative covers, like any landfill cover,
must (1) prevent physical contact with underlying waste, (2) prevent harmful gas production, and (3) keep
water from percolating downward toward groundwater^ tables. The latter objective, he said, is of paramount
importance; he described how vegetative covers achieve it. The covers act like sponges: they have thick soil
layers that hold water during dormant seasons, and areisucked dry during the growing season, when plant
roots extract stored water. If a vegetative cover is designed so that soil storage capacity is never exceeded,
Benson said, water will not leak into underlying wastes. There are two basic types of vegetative covers:
monolithic barriers and capillary barriers. While a cap pf the former type simply consists of a thick layer of
dirt with vegetation on top of it, Benson said, a cap of the latter type has fine textured soils laid over coarser
materials. The contrast in texture in the latter design buoys water up in the top layers, making the water
more accessible to plant roots. :
Benson said that several test covers have been installed under the Alternative Cover Assessment Program
(ACAP). One of the main goals of ACAP is to determine whether vegetative covers perform as well as
prescriptive designs. This must be determined, Benson said, because RCRA Subtitle D states that
percolation from an alternative cover must be less than or equal to percolation from prescriptive covers. At
Views expressed are those of the participants, not necessarily EPA.
43
-------
many sites, Benson said, ACAP has installed vegetative and prescriptive covers side by side, believing that
this is truly the best way to compare the percolation rates of the two. At some sites, it was not possible to set
up side-by-side comparisons, and ACAP was forced to use default equivalency values to determine whether
vegetative caps are equivalent to prescribed covers. For composite caps, he said, these values have been
defined as 30 millimeters of percolation per year in humid environments and 10 millimeters per year in arid
areas. With composite covers, he said, percolation default values are 3 millimeters per year in both humid
and arid environments. (Regulatory documents do not list the amount of percolation that is allowed for
prescriptive covers. Thus, ACAP had to define these default values.)
Benson said that ACAP is using an elaborate monitoring system at the demonstration sites. Test sections
have been set up with lysimeters so that any drainage that leaks through a cover can be measured directly.
The lysimeters are like big pans; each sits on a compacted base that is about 10 meters by 20 meters in size.
The bottom layer of each lysimeter consists of an impervious geomembrane that is made of a low-density
polyethylene. A geocomposite drainage layer lies over the geomembrane and carries any infiltrating water to
a collection sump, which in turn shuttles the water to tipping buckets and siphons so that drainage can be
measured. An interim soil cover is placed over the drainage layer, Benson said, and a root barrier is placed
on top of that. (The barrier has an herbicide; roots that come in contact are redirected, but not killed.) The
test cover sits on top of the root barrier. Benson said that careful quality control is performed when
lysimeters are installed; for example, geomembrane seams are examined and test systems are filled with
water to determine whether leaks are present. Aside from the lysimeters, Benson said, a variety of other
monitoring systems are established at ACAP sites. For example, weather stations are set up to collect
meteorological data, and a variety of water content reflectometers, thermocouples, and heat dissipation
probes are installed. All of the monitoring systems, he said, are wired to solar-powered data loggers, which
transmit data to the Desert Research Institute in a near-real-time fashion. These data are collected,
organized, and made available to ACAP members via the Internet. Users can select specific parameters that
interest them, Benson said, and graph the parameters over specified time periods. He said that access to the
data is currently limited to those who have passwords, but the data will eventually be made public in
published reports.
Growing a Thousand-Year Landfill Cover
William Jody Waugh, MacTec-ERS
Jody Waugh said that the U.S. Department of Energy (DOE) wants to design covers that perform adequately
for 1,000 years or more. These kinds of covers are needed to isolate radioactive wastes (e.g., uranium mill
tailings and fission products). About 20 years ago, Waugh said, DOE started capping sites under the
Uranium Mill Tailings Radiation Control Act (UMTRCA) project. As part of this effort, manyisites v/ere
covered with compacted soil layers, a fine sand layer, and rock rip-rap. (Caps like this were installed in Tuba
City, Arizona; Rifle, Colorado; Mexican Hat, Utah; and Lowman, Idaho.) Some vegetative caps were also
installed, Waugh said. Vegetative caps of various kinds were installed at sites in Pennsylvania, Utah, and
Colorado. Waugh said that DOE is assessing the lessons that have been learned on the UMTRCA covers,
and will take these into account when designing the next generation of DOE covers, which are to be
installed at DOE weapon sites. Waugh said that DOE will compile what it has learned in a guidance
document on how to design covers for long-term performance. This document will help end users, he said,
and will hopefully be embraced by the regulatory community.
Waugh said that all covers are subjected to dynamic ecosystems, and that a site's ecology changes over the
long term. Thus, it is not realistic to use data from a short field study to predict cover performance in future
centuries. More accurate predictions can be made, he said, if monitoring efforts, modeling efforts, and
Views expressed are those of the participants, not necessarily EPA.
44
-------
analog studies are combined to make these estimates. The latter involves evaluating natural or
archaeological settings. By looking at analogs, Waugh said, researchers reconstruct the past and collect clues
that can be used to predict what will happen in the future. He said that researchers can use analogs to gain an
understanding of how pedogenesis (i.e., soil development), ecological change, and climatic change could
impact covers. With this information in hand, DOE might be able to design covers that endure changing
conditions over the centuries. If so, this may save DOE billions of dollars in stewardship costs.
Waugh said that pedogenesis could affect the physical and hydraulic properties of a cover over time. He said
that designers can obtain an idea of how their engineered caps might develop if they evaluate natural or
archeological settings that have soil profiles similar to the engineered soils. For example, he said,
researchers were able to predict how soil would develop in a cover that was installed in Monticello, Utah, by
analyzing how soils developed in Anasazi pit houses that were abandoned about 800 to 1,000 years ago.
Also, he said, by evaluating a natural capillary barrier that formed in the state of Washington, researchers
learned about the water-holding capacity of a certain soil that was used at the Hanford DOE site. This
natural barrier, which has a soil profile that is about 13^000 years old, has fine materials deposited over a
coarse layer. Waugh said that carbonates, which serve as tracers for water movement, were found on the
coarse materials. This means that the thin layer of fine soil was not thick enough to prevent downward
percolation. Thus, if this material were used in a vegetative cover design, the layer would have to be made
thicker. ' , !
Waugh said that ecological changes, such as plant succession and biological intrusion, can have dramatic
impacts on a cover's evapotranspiration rate. He said that analogs were used to predict the impact of
preferential flow on a cover that was installed in Pennsylvania. This cover has a clay layer that is overlain by
a layer of sand and a layer of rock. Plants have penetrated the top two layers and have started establishing in
the clay layer. In situ saturated hydraulic conductivity was measured in areas that had plants, as well as areas
that had not been invaded. Values were about 10"7 cm/sec in the plant-free areas, but were two orders of
magnitude higher near Japanese knotweed plants. Waugh said that the values detected near the plants were
close to what was measured at an analog site that also had plants established. He also said studying natural
environments has helped researchers at the Hanford site determine what potential vegetation patterns could
emerge if area soils were incorporated into cover design. In addition, studies have been performed at an
UMTRCA site near Lakeview, Oregon, to determine how leaf area index values change when plants invade
an area.
Waugh said that climatic changes also occur over the long term, and that changes in meteorological
parameters could have dramatic impacts on cover performance. Thus, when designing a cover to perform
over the long term, researchers should evaluate natural paleoclimate analogs to obtain reasonable estimates
of how climate could change in a certain area in the future. That way, the range of climate changes in the
past can be entered as bounds in design models. Waugh described how climatic conditions in the Four
Corners area were reconstructed. He said that pollen cores, packrat middens, and other proxy climate data
were used to reconstruct past plant populations, and this helped to determine what past climates were like.
Tree Covers for Containment and Leachate Recirculation
Eric Aitchison, Ecolotree, Inc. \
Eric Aitchison described the Ecolotree® Cap, a patented containment system designed to achieve hydraulic
control. This cap, which consists of densely planted hybrid poplars and a grass understory, acts like a
sponge, holding moisture during dormant seasons and then drying out during the growing season. Aitchison
said that poplars are used in the Ecolotree® Cap because these trees grow fast, tolerate a variety of
Views expressed are those of the participants, not necessarily EPA.
45
-------
environmental and chemical stressors, develop deep and dense root systems, are relatively easy to plant and
maintain, can be grown from cut stumps, and can be used as a cash crop. Aitchison said that Ecolotree®
Caps have been installed at several sites across the county. He described six of them:
Ecolotree® Cap Projects
Lakeside Reclamation Landfill in Beaverton, Oregon: The first Ecolotree® Cap, designed to cover a 3-acre area with
11,000 hybrid poplars, was installed at this site in 1990. About 90% of the trees survived, and trees grow about 5 to 8
feet each year. Tree roots have grown through a 4-foot-thick soil layer and have penetrated the site's wastes. No
contaminants have been detected in monitoring wells since the cap was installed. The site owner is happy with the
system, and recently received approval to extend the cap over the entire site. The Ecolotree® Cap has also improved
aesthetics and has attracted wildlife. The site owner has brought sheep in to graze among the trees.
The Bluestem Landfill in Cedar Rapids, Iowa: At this site, an Ecolotree® Cap was compared to a prescriptive cover
that the Iowa Department of Natural Resources approved. Soil moisture was measured with reflectometers between
November 1995 and October 1996. At all times, soil moisture was lower in the Ecolotree® Cap than in the prescribed
cover. (Soils did not exceed the water-holding capacity in either cap.) In addition to achieving environmental objectives,
the Ecolotree® Cap served a secondary purpose: the trees trapped and prevented litter from blowing off site.
PAH-contaminated site in Tennessee: Before planting, compost was spread over this site to improve soil fertility and
water-holding capacity. The Ecolotree® Cap is expected to serve three purposes at this site: (1) achieve hydraulic
control, (2) stabilize soil, and (3) enhance rhizosphere degradation.
Landfill in Seattle, Washington: The Ecolotree® Cap will not be able to achieve complete hydraulic control at this site
because it receives too much winter rain. Modeling has been performed; the results suggest that the cap could reduce
leachate production by 50%. For this site, regulators decided that this partial reduction is sufficient. So, the cap was
installed over 13 acres in April 2000. The project cost about $600,000. Site owners believe that a geomembrane would
have cost about $3,000,000 to install at this site.
Military base in Georgia: A side-by-side comparison is being performed between the Ecolotree® Cap and a prescribed
cover. This work is being performed under ACAP.
Landfill in Michigan: Site owners told Ecolotree, Inc., representatives that the site's soils were suitable for plant
growth. Thus, an Ecolotree® Cap was installed over a chemical waste landfill. When only 30% of the trees survived, a
more detailed soil and groundwater sampling effort was conducted. Results showed that the site has high salinity and pH
values. Greenhouse studies were performed to determine whether amendments could make the soils fertile. Results are
)romisingjbut site owners have not yet decided how to move forward
Aitchison also described the Ecolotree® Buffer, which was patented in summer 2000. In this system, he said,
plants are exposed to contaminated water in a flow-through fashion. For example, waste water or leachate is
used as irrigation for Ecolotree® Buffers. Aitchison described three sites where Ecolotree® Buffers have
been used:
Views expressed are those of the participants, not necessarily EPA.
46
-------
Ecolotree® Buffer Projects
Riverbend Landfill, McMinnville, Oregon: About 35,000 hybrid poplar trees were planted over 17 acres in 1992 and
1993. Ammonium-rich leachate is shuttled from a collection pond to the tree plantation and used as irrigation water. In
the fourth growing season, about 860,000 gallons of leachate were applied to each acre of the plantation, and about 460
pounds of nitrogen was added per acre. By the end of the growing season, only about 30 pounds of nitrogen were found
in the soils. (Investigators know that the nitrogen is not simply being flushed out of the system because Time Domain
Refiectometry probes have been installed to determine whether water is moving downward.) Roots were shown to extend
about 7 feet bgs, and leaf matter was also detected at this depth. It is believed that worms are dragging the organic
material down, and that this will improve soil fertility and water-holding capacity. Ecolotree, Inc., and CH2M Hill
received an award for this project.
City of Woodburn Wastewater Treatment Plant: About 17,000 hybrid poplars were planted over 10 acres in 1995. In
1999, the project was extended over a 90-acre area. The goal was to remove thermal energy and ammonium from treated
wastewater. Wastewater was used to irrigate the site during the summer months. Ecolotree, Inc., and CH2M Hill received
an award for this project. •
GRRWA Landfill, Fort Madison, Iowa: About 6,800 hybrid poplar trees were planted over 6 acres in 1997 and 1998.
Site managers installed the system in an effort to treat leachate in a cost-effective manner. Before installing the
Ecolotree® Buffer, the managers paid about $150,000 annually to dispose of 7,000,000 gallons of leachate. After
installing the plantation, site owners only had to dispose of 50,000 gallons. The trees transpired or absorbed the
remainder. Leachate was sprayed right onto tree leaves. This burned the leaves, so site managers started irrigating at
night so that materials would not get baked onto the leaves. Stanley Consultants and Ecolotree, Inc., received an award
for this project. __^_
EPA Draft Guidance on Landfill Covers
Andrea Mclaughlin and Ken Skahn, EPA, Office of Emergency and Remedial Response
Andrea McLaughlin described regulatory frameworks, and explained what site managers must do to obtain
approval to use alternative covers. Under the EPA Liquids Management Strategy, she said, landfill owners
are expected to detect, collect, and remove any leachate that is generated in their landfills. In addition,
owners are expected to prevent leachate generation by preventing liquids from percolating through waste
materials. \
McLaughlin described what is expected of landfills that are closed under RCRA Subtitle D. First, she said,
permeability of bottom layers must be greater than or equal to those of top layers. Also, permeability rates
may not be greater than 1 x 10"5 cm/sec. McLaughlin said that federal regulations explicitly indicate that
state officials can approve alternative covers as long asithe cap is able to achieve equivalent reductions in
infiltration (i.e., permeability must not exceed a rate of 1 x 10'5 cm/sec). McLaughlin said that some states
may have even stricter performance standards. For example, in Illinois, covers must be designed so that they
do not exceed infiltration rates of 1 x 10'7cm/sec. McLaughlin said that covers that are selected for
CERCLA-mandated landfills are expected to meet ARARs—standards or requirements that are specified
under federal laws or promulgated under state environmental laws. According to federal regulations, she
said, alternative covers may be used at CERCLA sites in states that already have provisions for alternative
covers written into state law. If no such provision exists, alternative covers can still be considered as a
potential remedial approach if ARAR waivers are obtained. These waivers can be obtained, she said, if an
alternative cover is shown to perform at least as well as prescribed covers. To prove this, site owners must
show that the alternative cover infiltration rate does not exceed the minimum permeability rate that is
defined under RCRA Subtitle D. McLaughlin stressed that alternative covers must be able to meet the
minimum permeability standards at all times, rather than over an averaged period. If an ARAR waiver is
Views expressed are those of the participants, not necessarily EPA.
47
-------
obtained, and an alternative cover meets the nine criteria of the National Contingency Plan, then the cover
will be considered a viable remedial approach.
McLaughlin said that two EPA guidance documents are being developed under the EPA Liquid Management
Strategy. One will address the use of alternative covers at CERCLA municipal landfills, and the other will
provide comprehensive technical guidance on RCRA/CERCLA final covers. Ken Skahn is leading the effort
to develop the latter, McLaughlin said. She turned the remainder of the presentation over to him.
Skahn said that EPA's technical guidance on RCRA/CERCLA final covers will be released in about 18
months. It will serve as an update to a previous guidance document that was written in 1991. He said that an
update is needed because existing RCRA guidance documents do not discuss landfill gas managment,
performance monitoring, or long-term maintenance. In addition, existing documents do not discuss
alternative covers or list cover materials (e.g., geocomposite clay liners and new drainage materials) that
have become available over the last decade. Skahn said that it is important to discuss new materials because
some state regulatory agencies are reluctant to use new materials until the materials are officially
acknowledged by EPA. Skahn said that the revised guidance document will cover the following topics:
regulatory requirements, design considerations, alternative designs, water balance models, geotechnical
analysis and design, lessons learned, and long-term maintenance. He said that the document will explain that
alternative covers can be used if the covers demonstrate equivalency, and that this can be proven either with
predictive models or side-by-side demonstrations. Skahn said that the document will explain that covers can
be designed to last for long periods if designers select appropriate materials and address slope stability,
erosion, long-term maintenance, and flow capacities for internal drainage systems. In addition, the document
will encourage designers to take the following steps: (1) determine if gas collection is necessary, (2) identify
critical infiltration events, (3) calculate minimum storage capacity, (4) characterize soil properties, (5)
identify appropriate cover thickness, (6) consider amending surface soils and installing vegetation, and (7)
use predictive modeling to establish the adequacy of proposed designs.
Activities at an EPA Region 3 Site
Donna McCartney, EPA, Region 3
Donna McCartney described a site that received wastes from a chlorine manufacturer, a PCB manufacturer,
and a neighboring facility for more than a decade. These wastes were disposed in two disposal
impoundments. In February 2000, she said, approval was granted to test a vegetative cover as a potential
containment measure for this site. Site managers are hopeful that the cover will reduce infiltration, mitigate
erosion, eliminate direct contact with wastes, and promote contaminant degradation. The cover will be
analyzed over a three-year period; if proven effective, the cover may be considered as a viable remedial
approach during final remedy selection.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
• Measuring performance. One attendee called attention to one of the comments that McLaughlin made
during her presentation: alternative covers are expected to meet RCRA Subtitle D permeability standards
at all times. He thought this was excessively strict, and asked whether the regulations would permit
momentary lapses following extreme weather events, such as a 500-year rain. McLaughlin said that the
regulations imply that no exceedences are acceptable. She recommended designing covers so that they
are able to perform effectively during extreme events.
Views expressed are those of the participants, not necessarily EPA.
48
-------
Percolation rates versus amounts. Rock said that the regulations list percolation rates that cannot be
exceeded, but that these rates are not translated into actual amounts. He said that it is difficult to measure
rates in the field. Thus, many investigators are measuring drainage amounts instead, and performing side-
by-side comparisons to determine whether vegetative caps are equivalent to prescribed covers.
Guidance documents for alternative covers. Erickson asked whether guidance documents have been
produced on vegetative cover designs. Waugh said that DOE created a guidance document that describes
how to design UMTRCA covers. In addition, he said, ;DOE plans to release another guidance document
in about three or four years. Rock said that ACAP has not developed guidance documents yet. Both he
and Benson stressed that vegetative cover designs arejvery site-specific; one design cannot be applied to
all sites. Aitchison agreed that vegetative cover designs are site-specific, but said that it might be possible
to make some general design recommendations at this point. For example, he said, it might be realistic to
say that the top layers of a vegetative cover should never be less than 18 inches thick. If thinner layers are
used, he said, plants might be killed by high methane levels.
SESSION VIII: DEGRADATION OF ORGANIC COMPOUNDS IN SOILS
Remediation of Petroleum Contaminants Using Plants
M. Katherine Banks, Purdue University i
Kathy Banks said that grasses can be used to remediate petroleum contaminants by enhancing microbial
activity in the rhizosphere. She described studies that have been performed at three different field sites:
Craney Island, Virginia; Port Hueneme, California; and Bedford, Indiana.
Banks said that sediments at the Craney Island site, a large 16-acre landfarming facility, are contaminated
with diesel fuels. A portion of the site was used in a phytoremediation demonstration project, she said, and
four treatments were evaluated: tall fescue, Bermuda grass with annual rye, white clover, and nonvegetated
controls. Banks said that six replicates were included in the study design and that all plots were fertilized and
irrigated as needed. Over a two-year period, aboveground biomass quantities were assessed, plant tissue
samples were analyzed, and soil samples were analyzed. These analyses revealed that: (1) all of the plants
were able to grow in the site's contaminated soils, (2) contaminants did not accumulate within plant tissues,
and (3) plants did induce statistically significant contaminant reductions—the clover remediated more TPH
than the other plants. Banks said microbial analyses were performed on soil samples using a variety of
techniques. For example, researchers evaluated total plate counts to determine whether plants stimulated
enhanced microbial activities, performed BIOLOG analyses to assess the functional diversity of microbial
communities, and used most probable number (MPN) procedures to evaluate the proliferation of petroleum
degraders. The results showed that planted treatments exhibited a higher degree of functional diversity and
that MPN values were highest in clover treatments.
Banks said that a demonstration project was also performed at the Port Hueneme site, which had heavily
weathered soils that were contaminated with fuel oil. Three treatments were tested: (1) fescue and legume,
(2) California roadside mix, and (3) nonvegetated controls.-Banks said that four replicates were included in
the study design, all plots were fertilized and irrigated as needed, and samples were collected over a 30-
month period. Most of the analyses described for the Craney Island site were also performed at Port
Hueneme. In addition, efforts were made to assess ecological toxicity. Researchers found that: (1) plants
were able to grow in the site's contaminated soils; (2) plants did induce statistically significant contaminant
Views expressed are those of the participants, not necessarily EPA.
49 ,
-------
reductions (the fescue remediated PAHs more efficiently than the California roadside mix); (3) microbial
activity was enhanced in the planted treatments; (4) the vegetated plots had a higher number of
pseudomonads; (5) soils, which were slightly toxic at the beginning of the project, were considered nontoxic
after treatment; and (6) the different treatments exhibited no difference in lettuce germination rates, which
are used to measure toxicity, over the long term.
Banks said that a field study was recently initiated at a former manufactured gas plant in Bedford, Indiana.
PAH concentrations are very high at this site, she said, and the contamination is about 3 to 6 feet bgs.
Phytoremediation is being tested at this site, along with three other types of remedial systems: natural
attenuation, land farming, and composting. Banks said that poplars and a grass understory, which are being
irrigated and fertilized as needed, are being used in the phytoremediation plots. Researchers will determine
how effective the planted treatments are by comparing these treatments to the natural attenuation plots.
Banks said that the water table is high at this site, and that researchers hope the poplars will lower the water
table so that more of the subsurface can be oxygenated and treated via natural aerobic degradation. Banks
said that soils will be collected over a three-year period and analyzed for contaminants, microbial
characteristics, and toxicity. Some preliminary data do suggest that the soils in the planted plots are
becoming less toxic.
Banks said that phytoremediation sites must be managed, noting that contaminant reduction rates leveled off
when investigators stopped fertilizing, irrigating, and removing invading species at the Craney Island site. In
the greenhouse, Banks said, Stacy Lewis Hutchinson has shown that there is strong correlation between
contaminant degradation rates, fertilization rates, and irrigation. In one of Hutchinson's studies, four types of
irrigation—surface, continuous, cycling, and subsurface—were evaluated to determine whether using
different application methods impacts degradation rates. Hutchinson found that degradation rates were
highest when the subsurface irrigation was applied. Banks said that these are just preliminary results, and
said that she hopes Hutchinson will perform additional experiments to back up her findings.
In closing, Banks summarized some of the research that Purdue University will perform over the next five
years. Significant efforts will be made to evaluate microbial communities and identify the rhizosphere
conditions that foster degradation. Banks said that she plans to evaluate microbial diversity, and suspects that
researchers may be able to use this as an indicator to measure the remedial potential of a plant. In addition,
the University will evaluate: (1) how plants affect soil toxicity, (2) the fate of PAH in the rhizosphere, (3)
depth limitations for phytoremediation, and (4) using phytoremediation as a polishing tool.
Phytoremediation of Explosives
Phillip L. Thompson, Seattle University
Phillip Thompson said that several experiments have been performed to determine whether plants can be
used to remediate explosives. For example, in the early 1980s, researchers evaluated the capability of yellow
nutsedge to take up TNT. Uptake of this contaminant, as well as RDX, was also investigated in the late
1980s5 in studies that the Pacific Northwest Laboratory performed using bush bean, maize, and wheat. Also,
the University of Nebraska has evaluated the uptake of explosives in tall fescue and switchgrass. In recent
years, Thompson said, the University of Iowa and Iowa AAP worked together to determine whether poplars
can be used to remediate low concentrations of TNT and RDX.
Thompson provided brief chemical profiles for TNT and RDX. The former, which is a mutagen, has a Log
Kow of 1.9, a solubility of 100 mg/L at 25° C, and an electron-deficient ring. Aerobic degradation pathways
have been identified for TNT and are documented in bioremediation and microbial remediation literature.
Views expressed are those of the participants, not necessarily EPA.
50
-------
RDX, which has been shown to cause central nervous system disorders, has a Log Kow of 0.9, a solubility of
40 mg/L at 25° C, and a saturated ring. Unlike TNT, RDX does mot degrade readily under aerobic conditions.
Thompson said that treatability studies have been performed to determine how TNT and RDX impact poplar
growth. In the laboratory, RDX was shown to be nontoxic, but trees experienced adverse effects when
exposed to 5 mg/L of TNT. When the same experiment was performed in the greenhouse, with larger trees
and under more natural sunlight conditions, the poplars were able to withstand higher TNT concentrations.
Thompson said that mass balance studies have been performed to determine how poplars affect the fate and
transport of TNT. Excellent mass recovery rates (about 95%) were exhibited in these experiments. Results
indicated that 70% of TNT accumulates in plant roots, and that much smaller amounts accumulate in shoots
and leaves. Transformation products were measured, Thompson said, noting that about 50% were bound
residue and could not be extracted. Of the other 50%, he Said, only about 10% could be identified. Many of
the products identified in the roots were those that form under aerobic degradation. Most of the products in
the leaves were unknown polar compounds. These might: represent novel oxidative products, Thompson said;
if so, this suggests that plants create transformation products other than those identified in aerobic microbial-
mediated degradation pathways. •
Thompson said that mass balance studies were also performed using RDX, but that recovery rates were only
about 80%. This may have been because the RDX used in the study was not pure and/or because some of the
RDX was phytodegraded and transformed into formic acid within the leaves. Thompson was not sure
whether the latter is occurring, but said this has been suggested as a potential fate for RDX in preliminary
studies by other researchers. Thompson said that the fate'of RDX differs markedly from that of TNT. About
65% to 70% of the RDX that was used in mass balance studies accumulated in plant leaves. Thompson said
that this finding could raise some regulatory concern about high RDX concentrations. He did stress,
however, that results that are obtained in the laboratory do not necessarily indicate what will be observed in
the field. To demonstrate his point, he noted that pop lare have been established as part of a phytoremediation
effort at the Iowa Army Ammunition Plant. So far, he said, RDX has not been detected in tree leaves at the
site. He said that this may be because RDX concentrations in the soil are very low, or because RDX in the
leaves is being transformed to formic acid.
Thompson said that he has used models to predict how TNT and RDX would degrade if one contaminated
acre was treated with about 600 to 700 poplar trees. He said that the following parameters were used to
estimate uptake: TSCF, transpiration rates, and porewater concentration. (Thompson stressed the importance
of accounting for the latter, noting that this parameter is often overlooked.) Also, he said, the following
assumptions were made: contaminants are located in the top 3 feet of soil, bulk density is 1,500 kg/m3,
porosity is 0.3, contamination is homogeneous across the site, there is instantaneous desorption of
contaminants, 10 to 20 gallons of water is transpired per day by each tree, and microbial influences are
negligible. The results of this modeling effort, he said, suggest that TNT and RDX would have half-lives of
20 years and 5 years, respectively. ,
In closing, Thompson listed some issues that require further investigation. First, he said, researchers should
evaluate how natural conditions affect RDX concentrations in leaves. It would also be wise, he said, to
evaluate the toxicity of plant tissues by feeding leaves to snails and worms. In addition, he said, efforts
should be initiated to determine how microbial/mycorrhizal associations affect phytoremedial systems.
Thompson also recommended performing studies to determine how effective phytoremediation is at sites that
are contaminated with a variety of different contaminants. Lastly, he said, researchers must gain a better
understanding of the true time frames that are associated with phytoremedial systems.
Views expressed are those of the participants, not necessarily EPA;
-------
Case Study: Union Pacific Railroad
Felix Flechas, EPA, Region 8
Felix Flechas opened his presentation by summarizing regulatory issues that must be considered when
deciding whether phytoremediation is appropriate to use at a particular site. First, he said, the technology
should be evaluated against the nine criteria of the National Contingency Plan. Flechas described the nine
criteria, splitting them into two categories: selection criteria and balancing criteria. The former, he said, are
the most important. They state that selected remedies must: (1) protect human health and the environment,
(2) comply with applicable rules and regulations, (3) control sources of release, and (4) attain cleanup
standards. Flechas said that it is not always possible to identify a remedy that can meet the latter criteria. For
example, he said, if a site has NAPL concentrations or a karst formation, available technologies may not be
capable of cleaning the site to meet applicable standards. In these cases, he said, Technical Impracticability
Waivers can be obtained. Flechas also listed the five balancing criteria: (1) long-term effectiveness and
permanence; (2) reduction of toxicity, mobility, or volume with a preference towards treatment; (3) short-
term effectiveness; (4) implementabiliry; and (5) cost.
Flechas said that phytoremediation can be used as a treatment technology, an immobilization technology, or
a containment technology. Before presenting a proposed remedy to a regulator, he said, site managers should
make sure to clearly define the functionality of the remedy. In addition, he said, site managers should
identify an approach that can be used to track progress. He acknowledged that it may take several years for
plant-based systems to achieve cleanup goals, but said that regulators will probably want periodic assurances
that the remedy is working. At CERCLA and RCRA sites, he said, remedies will be evaluated every five
years. In order to show that statistically significant contaminant reductions are occurring, site managers will
need to collect extensive baseline data and fully characterize heterogeneities at the site. Before initiating
field studies, Flechas said, site managers should create detailed sampling and analytical plans. If they do not,
they have little hope of demonstrating statistically significant results. Flechas said that site managers must
also be prepared to address ecological concerns. He said that EPA has created a guidance document for
baseline ecological assessments; this document can be used to determine whether installing a
phytoremediation system will create new ecological hazards. Flechas said that site managers should also
make sure that they understand the needs of the community. If eager for green spaces to be created, the
public may be very accepting of phytoremediation. However, if community members are hoping to redevelop
a site for commercial use, they may want more aggressive technologies used so cleanup goals can be
achieved quickly.
Flechas described a phytoremediation demonstration project that has been proposed for the Union Pacific
Railroad site. This site, located in Laramie, Wyoming, was used as a railroad tie treatment plant for about
100 years. About 10 million gallons of creosote have been released over a 90-acre area, he said, and high
polynuclear aromatic hydrocarbon and pentachlorophenol concentrations have been detected at this site.
Flechas said that phytoremediation is just one of several technologies that are being used to clean up the site.
For example, about 2 miles of slurry wall have been installed around the site to achieve hydraulic isolation.
Also, a dual drain line system was installed to recover about 3.5 million gallons of creosote. Amendments
have also been used in an effort to remediate drained surface impoundment areas.
Flechas said that unsaturated soils at the Union Pacific Railroad site are highly contaminated. Some of these
areas, he said, are located near a proposed bike path area. Also, the soils occasionally travel to a nearby river
via overland flow. Thus, there is strong incentive to clean up the site soils. Risk analyses have been
performed on the soils. Results indicate that areas near the river posed cancer risks in the 10'6 or 10'7 range,
but other areas pose risks in the 10"4 range. Both areas will be addressed through phytoremediation, he said,
Views expressed are those of the participants, not necessarily EPA.
52
-------
but the latter will be covered with 18 inches of clean soil first. Flechas said that four different treatments will
be applied to the site: (1) cottonwood and willow trees, (2) hackberry bushes, (3) alfalfa, and (4) dryland
grass mixture. The demonstration project will be performed over 20 years and evaluated every 5 years. Site
managers have agreed to measure the following through the course of the demonstration project: plant
survival and growth, vegetative cover, groundwater extraction rates, soil oxygen levels, rooting depths and
density, soil organic matter, and contaminant concentrations. In addition, he said, site managers have agreed
to think about restoration goals and; beneficial site reuse.
Phytoremediation in Alaska and Korea . :
Charles (Mike) Reynolds, U.S. Army Corps of Engineers
Mike Reynolds said that the Cold Regions Research and Engineering Laboratory has established several
phytoremediation demonstrations on former military sites that are located in cold regions. Many of these
sites are contaminated with petroleum hydrocarbons, he said; remediating these cold-weather sites has been a
challenge, because the sites are located in remote areas and often have limited infrastructure. In recent years,
he said, researchers have started experimenting with phytoremediation at these sites. This is viewed as a
reasonable solution because the approach is cheap, relatively easy to install and maintain, and can be
installed over permafrost. It is not yet clear how these phytotechnologies will perform at these sites,
however. Reynolds said that it will likely take several years,to achieve cleanup goals, especially since the
growing season at these extreme northern sites is short. , ,
I
Reynolds presented contaminant degradation curves that are observed in phytoremediation projects. These
curves typically exhibit an initial lag, he said, in which ho contaminant concentration reductions are
detected. (He said that researchers are interested in finding ways to minimize the lag time.) Then,
contaminant concentrations decrease steadily for a while before reaching an asymptotic plateau. In some
cases, Reynolds said, this asymptotic level falls above site cleanup standard. If this is the case, some
regulators might regard phytoremediation to be a failure at a particular site. Reynolds questioned the
legitimacy of such a conclusion, .noting that he is not sure that contaminant concentrations are the most
appropriate endpoints for measuring success. He said that some researchers believe that ecological endpoints
should be used to define success instead. For example, he said, it might be niore meaningful to use microbial
diversity as a measure of success. He advised looking at ;the, health .of the overall community when trying to
determine whether phytoremediation has been a success.
Reynolds said that a phytoremediation project has been installed at the Farmers Loop site in Fairbanks,
Alaska. Soil samples have been collected from the site and microbes have been analyzed using different
methods. All of the results point to.the. same conclusion: there is more microbial diversity in planted
treatments than there are in unplanted treatments. Reynolds said that researchers believe that this diversity
indicates that the system is becoming "healthier," either from reduced contaminant-caused stress, reduced
bioavailability of the contaminant carbon, or both. '.
Reynolds said that efforts are underway to determine whether microbial diversity can be correlated with
reductions in contaminant concentration. He said that phytoremediation projects have been initiated at five
sites (three in Alaska and two in Korea), and that chemical and microbial data are being collected from these
sites. At all of the sites, he said, four treatments are being evaluated: unplanted/fertilized,
unplanted/unfertilized, vegetated/fertilized, and vegetated/unfertilized. Reynolds said that three of the sites
are being evaluated in conjunction with the Remediation Technologies Development Forum's (RTDF's)
TPH in Soil Subgroup, and chemical analyses are being performed,in accordance with the Subgroup's
protocol. He said that the percent of TPH degradation at these sites will be measured using a recalcitrant
Views expressed are those of the participants, not necessarily EPA.
53
-------
biomarker (i.e., hopane) to normalize data. Using this approach helps investigators obtain meaningful results'
at sites that exhibit much variability. Reynolds summarized some of the data that have been collected from
the sites in Alaska. At one site, he said, no contaminant concentration reductions were detected after the first
year of plant growth, but researchers did find that microbial degraders are more abundant in
unplanted/fertilized and planted/fertilized treatments. At two sites in Korea, Reynolds said, he observed no
plant or fertilization effects on hopane-normalized TPH depletion relative to the controls, but using Gas
Chromatography/Mass Spectography data, a clear vegetative effect was observed on the heavier PAH
fraction compounds at both sites. These data agree with observations seen in laboratory studies using field
soils, where degradation of heavy-end PAH was most robust in the planted/fertilized treatment. In general,
Reynolds said, he has found that phytoremediation exerts the greatest benefits on heavy, more recalcitrant
compounds, and that benefits are somewhat less pronounced for the lighter compounds that are relatively
easy to degrade.
Reynolds reiterated a point that Banks made during her presentation: fertilization may have profound
impacts on the efficacy of phytoremediation systems. At the sites that he has worked on so far, he said,
fertilization has been applied to meet agronomic needs, but little thought has been given to managing
fertilizer in an attempt to optimize degradation rates. He said that he plans to give this issue more attention in
the near future.
Speaker Panel and Audience Discussion
Audience members asked questions or provided comments about the following topics:
* Focusing on microbial diversity. Fletcher commended Banks and Reynolds for their interest in learning
more about microbial diversity in the rhizosphere. He reiterated that changes in microbial communities
may serve as good indicators that phytoremediation is working.
• Identifying microbes that are associated with specific plants. Erickson asked Banks whether she proposes
identifying the individual microbes that are associated with different plants. Banks said that she thought
this would be too research-intensive. She thinks it might be more useful to focus on whole microbial
communities. That is, she proposes determining which plants are able to create microbial communities
that promote degradation.
• Plant selection for cold-weather regions. One meeting attendee noted that some of the sites that
Reynolds is working on are located above the Arctic Circle. He asked how Reynolds selected species for
his phytoremediation projects, and whether it was difficult to find plants that perform well under Arctic
conditions. Reynolds said that the Alaskan sites are being evaluated under the RTDF TPH in Soil
Subgroup; thus, plants were chosen to fit in with the Subgroup's protocol. Mclntyre said that he recently
found a Ph.D. thesis that lists about 150 plants that grow well in the eastern, central, and western Arctic.
He said that he would share this paper with Reynolds.
" Guidance documents for plant selection. Erickson asked whether guidance documents have been
produced on plant selection. Banks said that she did not know of any, and that she thought it was too
early to create such a document. Reynolds said that it might be a good idea to start keeping a list of the
plants that do not work. Mclntyre reminded meeting attendees that plants are currently cataloged on two
databases: Phytopet and Phytorem.
Views expressed are those of the participants, not necessarily EPA.
54
-------
Remedy selection. Burken asked Flechas whether regulators consider the aesthetic and restorative
benefits of phytoremediation when choosing a remedial technology for a site. Specifically, he asked
whether these two benefits could sway a regulator to accept phytoremediation over a faster-acting
approach. Flechas said that these factors might be considered, but said that the weight they would be
given would differ from site to site.
CLOSING REMARKS
Joan Colson thanked everyone for attending the meeting, and gave special thanks to all of the speakers,
poster presenters, organizers, and everyone who helped plan the meeting.
Views expressed are those of the participants, not necessarily EPA.
5i5
-------
-------
-------
United States
Environmental Protection Agency/ORD
National Risk Management
Research Laboratory
Cincinnati, OH 45268
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HERE [3;
detach, or copy this cover, and return to the address in the
upper left-hand comer.
PRESORTED STANDARD |
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use
$300
EPA/625/R-01/011a
------- |